{ "best_metric": 0.7636064291000366, "best_model_checkpoint": "final_models/lait_hau_mpt/checkpoint-12750", "epoch": 6.0, "eval_steps": 500, "global_step": 12750, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00047058823529411766, "grad_norm": 101.49613592746898, "learning_rate": 1.5e-07, "loss": 22.75, "step": 1 }, { "epoch": 0.0009411764705882353, "grad_norm": 112.2388930526183, "learning_rate": 3e-07, "loss": 25.0, "step": 2 }, { "epoch": 0.001411764705882353, "grad_norm": 116.76067668284688, "learning_rate": 4.5e-07, "loss": 24.5, "step": 3 }, { "epoch": 0.0018823529411764706, "grad_norm": 115.84574253530985, "learning_rate": 6e-07, "loss": 25.875, "step": 4 }, { "epoch": 0.002352941176470588, "grad_norm": 132.33052016952934, "learning_rate": 7.499999999999999e-07, "loss": 29.0, "step": 5 }, { "epoch": 0.002823529411764706, "grad_norm": 108.81027753995384, "learning_rate": 9e-07, "loss": 25.0, "step": 6 }, { "epoch": 0.0032941176470588237, "grad_norm": 106.5293346660472, "learning_rate": 1.05e-06, "loss": 22.125, "step": 7 }, { "epoch": 0.0037647058823529413, "grad_norm": 86.36964709277535, "learning_rate": 1.2e-06, "loss": 19.5, "step": 8 }, { "epoch": 0.0042352941176470585, "grad_norm": 144.2823063298436, "learning_rate": 1.3499999999999998e-06, "loss": 28.25, "step": 9 }, { "epoch": 0.004705882352941176, "grad_norm": 145.50911756736946, "learning_rate": 1.4999999999999998e-06, "loss": 27.0, "step": 10 }, { "epoch": 0.0051764705882352945, "grad_norm": 145.28792000066682, "learning_rate": 1.6499999999999997e-06, "loss": 25.125, "step": 11 }, { "epoch": 0.005647058823529412, "grad_norm": 138.63678500827925, "learning_rate": 1.8e-06, "loss": 16.875, "step": 12 }, { "epoch": 0.00611764705882353, "grad_norm": 337.87181528956177, "learning_rate": 1.9499999999999995e-06, "loss": 14.625, "step": 13 }, { "epoch": 0.006588235294117647, "grad_norm": 117.88942577946123, "learning_rate": 2.1e-06, "loss": 13.125, "step": 14 }, { "epoch": 0.007058823529411765, "grad_norm": 196.16784031514246, "learning_rate": 2.2499999999999996e-06, "loss": 16.5, "step": 15 }, { "epoch": 0.0075294117647058826, "grad_norm": 72.3662821367274, "learning_rate": 2.4e-06, "loss": 6.8438, "step": 16 }, { "epoch": 0.008, "grad_norm": 36.88527078878357, "learning_rate": 2.55e-06, "loss": 6.625, "step": 17 }, { "epoch": 0.008470588235294117, "grad_norm": 147.17817025227978, "learning_rate": 2.6999999999999996e-06, "loss": 12.25, "step": 18 }, { "epoch": 0.008941176470588235, "grad_norm": 130.53389105525557, "learning_rate": 2.85e-06, "loss": 10.9375, "step": 19 }, { "epoch": 0.009411764705882352, "grad_norm": 104.5694695030932, "learning_rate": 2.9999999999999997e-06, "loss": 9.125, "step": 20 }, { "epoch": 0.00988235294117647, "grad_norm": 86.07862342767167, "learning_rate": 3.15e-06, "loss": 7.75, "step": 21 }, { "epoch": 0.010352941176470589, "grad_norm": 291.7854297417925, "learning_rate": 3.2999999999999993e-06, "loss": 9.6875, "step": 22 }, { "epoch": 0.010823529411764706, "grad_norm": 206.3907373016001, "learning_rate": 3.4499999999999996e-06, "loss": 9.625, "step": 23 }, { "epoch": 0.011294117647058824, "grad_norm": 115.29621377587144, "learning_rate": 3.6e-06, "loss": 7.5625, "step": 24 }, { "epoch": 0.011764705882352941, "grad_norm": 54.57082775236266, "learning_rate": 3.7499999999999997e-06, "loss": 7.0312, "step": 25 }, { "epoch": 0.01223529411764706, "grad_norm": 42.159147570431244, "learning_rate": 3.899999999999999e-06, "loss": 5.8125, "step": 26 }, { "epoch": 0.012705882352941176, "grad_norm": 41.20171868170702, "learning_rate": 4.049999999999999e-06, "loss": 6.3125, "step": 27 }, { "epoch": 0.013176470588235295, "grad_norm": 59.81392772288977, "learning_rate": 4.2e-06, "loss": 7.375, "step": 28 }, { "epoch": 0.013647058823529411, "grad_norm": 33.0559230526184, "learning_rate": 4.35e-06, "loss": 5.6562, "step": 29 }, { "epoch": 0.01411764705882353, "grad_norm": 34.605772343203995, "learning_rate": 4.499999999999999e-06, "loss": 5.75, "step": 30 }, { "epoch": 0.014588235294117647, "grad_norm": 44.39881589576797, "learning_rate": 4.6499999999999995e-06, "loss": 5.9688, "step": 31 }, { "epoch": 0.015058823529411765, "grad_norm": 41.50780482964435, "learning_rate": 4.8e-06, "loss": 6.0312, "step": 32 }, { "epoch": 0.015529411764705882, "grad_norm": 25.818247112284777, "learning_rate": 4.95e-06, "loss": 6.0312, "step": 33 }, { "epoch": 0.016, "grad_norm": 27.896607385160312, "learning_rate": 5.1e-06, "loss": 5.7188, "step": 34 }, { "epoch": 0.01647058823529412, "grad_norm": 41.46313957245083, "learning_rate": 5.25e-06, "loss": 5.5625, "step": 35 }, { "epoch": 0.016941176470588234, "grad_norm": 28.890172443371533, "learning_rate": 5.399999999999999e-06, "loss": 5.7188, "step": 36 }, { "epoch": 0.017411764705882352, "grad_norm": 20.64321553485829, "learning_rate": 5.549999999999999e-06, "loss": 5.5, "step": 37 }, { "epoch": 0.01788235294117647, "grad_norm": 15.407300067462652, "learning_rate": 5.7e-06, "loss": 5.2812, "step": 38 }, { "epoch": 0.01835294117647059, "grad_norm": 29.052431214602308, "learning_rate": 5.85e-06, "loss": 5.1875, "step": 39 }, { "epoch": 0.018823529411764704, "grad_norm": 26.442242807401243, "learning_rate": 5.999999999999999e-06, "loss": 5.625, "step": 40 }, { "epoch": 0.019294117647058823, "grad_norm": 18.377219720667565, "learning_rate": 6.1499999999999996e-06, "loss": 5.5312, "step": 41 }, { "epoch": 0.01976470588235294, "grad_norm": 16.930402874214824, "learning_rate": 6.3e-06, "loss": 5.0, "step": 42 }, { "epoch": 0.02023529411764706, "grad_norm": 19.859055803159734, "learning_rate": 6.449999999999999e-06, "loss": 5.25, "step": 43 }, { "epoch": 0.020705882352941178, "grad_norm": 30.64130227930262, "learning_rate": 6.599999999999999e-06, "loss": 5.0, "step": 44 }, { "epoch": 0.021176470588235293, "grad_norm": 21.325375318760546, "learning_rate": 6.749999999999999e-06, "loss": 5.25, "step": 45 }, { "epoch": 0.02164705882352941, "grad_norm": 19.302916478103924, "learning_rate": 6.899999999999999e-06, "loss": 5.1875, "step": 46 }, { "epoch": 0.02211764705882353, "grad_norm": 19.013965858247154, "learning_rate": 7.049999999999999e-06, "loss": 5.0938, "step": 47 }, { "epoch": 0.02258823529411765, "grad_norm": 20.5642346552464, "learning_rate": 7.2e-06, "loss": 4.75, "step": 48 }, { "epoch": 0.023058823529411764, "grad_norm": 18.57073878009592, "learning_rate": 7.35e-06, "loss": 4.8125, "step": 49 }, { "epoch": 0.023529411764705882, "grad_norm": 33.046521804321365, "learning_rate": 7.499999999999999e-06, "loss": 5.0938, "step": 50 }, { "epoch": 0.024, "grad_norm": 25.917378059734396, "learning_rate": 7.65e-06, "loss": 5.125, "step": 51 }, { "epoch": 0.02447058823529412, "grad_norm": 18.856474460276843, "learning_rate": 7.799999999999998e-06, "loss": 5.0312, "step": 52 }, { "epoch": 0.024941176470588234, "grad_norm": 29.144004463117664, "learning_rate": 7.949999999999998e-06, "loss": 4.8438, "step": 53 }, { "epoch": 0.025411764705882352, "grad_norm": 29.150680350987862, "learning_rate": 8.099999999999999e-06, "loss": 5.0625, "step": 54 }, { "epoch": 0.02588235294117647, "grad_norm": 32.769161581571474, "learning_rate": 8.249999999999999e-06, "loss": 4.7812, "step": 55 }, { "epoch": 0.02635294117647059, "grad_norm": 30.95125339751104, "learning_rate": 8.4e-06, "loss": 4.875, "step": 56 }, { "epoch": 0.026823529411764704, "grad_norm": 23.76598249187105, "learning_rate": 8.55e-06, "loss": 4.9688, "step": 57 }, { "epoch": 0.027294117647058823, "grad_norm": 23.908836790462598, "learning_rate": 8.7e-06, "loss": 4.75, "step": 58 }, { "epoch": 0.02776470588235294, "grad_norm": 38.28663717060975, "learning_rate": 8.849999999999998e-06, "loss": 4.9375, "step": 59 }, { "epoch": 0.02823529411764706, "grad_norm": 25.662046681188414, "learning_rate": 8.999999999999999e-06, "loss": 4.7812, "step": 60 }, { "epoch": 0.028705882352941175, "grad_norm": 23.875479314809386, "learning_rate": 9.149999999999999e-06, "loss": 4.5, "step": 61 }, { "epoch": 0.029176470588235293, "grad_norm": 16.670702389031845, "learning_rate": 9.299999999999999e-06, "loss": 4.6875, "step": 62 }, { "epoch": 0.029647058823529412, "grad_norm": 20.478243823618094, "learning_rate": 9.45e-06, "loss": 4.4062, "step": 63 }, { "epoch": 0.03011764705882353, "grad_norm": 25.239103559162466, "learning_rate": 9.6e-06, "loss": 4.2188, "step": 64 }, { "epoch": 0.03058823529411765, "grad_norm": 36.492718803507415, "learning_rate": 9.75e-06, "loss": 4.375, "step": 65 }, { "epoch": 0.031058823529411764, "grad_norm": 16.454813088249846, "learning_rate": 9.9e-06, "loss": 4.5, "step": 66 }, { "epoch": 0.03152941176470588, "grad_norm": 23.207748256083462, "learning_rate": 1.005e-05, "loss": 4.625, "step": 67 }, { "epoch": 0.032, "grad_norm": 21.149585742549043, "learning_rate": 1.02e-05, "loss": 3.9375, "step": 68 }, { "epoch": 0.03247058823529412, "grad_norm": 12.685374897669782, "learning_rate": 1.035e-05, "loss": 4.5938, "step": 69 }, { "epoch": 0.03294117647058824, "grad_norm": 28.22378135455897, "learning_rate": 1.05e-05, "loss": 4.1875, "step": 70 }, { "epoch": 0.033411764705882356, "grad_norm": 26.496476385096706, "learning_rate": 1.0649999999999998e-05, "loss": 4.5, "step": 71 }, { "epoch": 0.03388235294117647, "grad_norm": 33.31826882344378, "learning_rate": 1.0799999999999998e-05, "loss": 4.2812, "step": 72 }, { "epoch": 0.034352941176470586, "grad_norm": 21.772278844592897, "learning_rate": 1.0949999999999998e-05, "loss": 4.5625, "step": 73 }, { "epoch": 0.034823529411764705, "grad_norm": 22.264768894413663, "learning_rate": 1.1099999999999999e-05, "loss": 4.4062, "step": 74 }, { "epoch": 0.03529411764705882, "grad_norm": 19.230295773103034, "learning_rate": 1.1249999999999999e-05, "loss": 4.5625, "step": 75 }, { "epoch": 0.03576470588235294, "grad_norm": 32.59427863615375, "learning_rate": 1.14e-05, "loss": 4.2188, "step": 76 }, { "epoch": 0.03623529411764706, "grad_norm": 36.496187867844796, "learning_rate": 1.155e-05, "loss": 4.0312, "step": 77 }, { "epoch": 0.03670588235294118, "grad_norm": 22.96770553726211, "learning_rate": 1.17e-05, "loss": 3.9062, "step": 78 }, { "epoch": 0.0371764705882353, "grad_norm": 31.577803011236764, "learning_rate": 1.1849999999999998e-05, "loss": 3.9844, "step": 79 }, { "epoch": 0.03764705882352941, "grad_norm": 30.102933978807457, "learning_rate": 1.1999999999999999e-05, "loss": 4.0312, "step": 80 }, { "epoch": 0.03811764705882353, "grad_norm": 32.75592144805966, "learning_rate": 1.2149999999999999e-05, "loss": 3.5312, "step": 81 }, { "epoch": 0.038588235294117645, "grad_norm": 34.1045239410945, "learning_rate": 1.2299999999999999e-05, "loss": 3.9219, "step": 82 }, { "epoch": 0.039058823529411764, "grad_norm": 22.675455933949355, "learning_rate": 1.245e-05, "loss": 4.3125, "step": 83 }, { "epoch": 0.03952941176470588, "grad_norm": 17.50169898171744, "learning_rate": 1.26e-05, "loss": 4.0938, "step": 84 }, { "epoch": 0.04, "grad_norm": 20.386295201551015, "learning_rate": 1.275e-05, "loss": 4.125, "step": 85 }, { "epoch": 0.04047058823529412, "grad_norm": 29.90841689758957, "learning_rate": 1.2899999999999998e-05, "loss": 3.7656, "step": 86 }, { "epoch": 0.04094117647058824, "grad_norm": 35.33978732552655, "learning_rate": 1.3049999999999999e-05, "loss": 3.2031, "step": 87 }, { "epoch": 0.041411764705882356, "grad_norm": 31.062408946852734, "learning_rate": 1.3199999999999997e-05, "loss": 3.5312, "step": 88 }, { "epoch": 0.04188235294117647, "grad_norm": 23.17689972908053, "learning_rate": 1.3349999999999998e-05, "loss": 3.8281, "step": 89 }, { "epoch": 0.042352941176470586, "grad_norm": 36.190540225308325, "learning_rate": 1.3499999999999998e-05, "loss": 3.2812, "step": 90 }, { "epoch": 0.042823529411764705, "grad_norm": 36.42021929733866, "learning_rate": 1.3649999999999998e-05, "loss": 3.125, "step": 91 }, { "epoch": 0.04329411764705882, "grad_norm": 36.330491664222144, "learning_rate": 1.3799999999999998e-05, "loss": 3.25, "step": 92 }, { "epoch": 0.04376470588235294, "grad_norm": 26.02983060826279, "learning_rate": 1.3949999999999999e-05, "loss": 3.75, "step": 93 }, { "epoch": 0.04423529411764706, "grad_norm": 24.19972781587301, "learning_rate": 1.4099999999999999e-05, "loss": 3.8906, "step": 94 }, { "epoch": 0.04470588235294118, "grad_norm": 35.323969594711485, "learning_rate": 1.4249999999999999e-05, "loss": 2.75, "step": 95 }, { "epoch": 0.0451764705882353, "grad_norm": 30.48439805632442, "learning_rate": 1.44e-05, "loss": 3.3125, "step": 96 }, { "epoch": 0.04564705882352941, "grad_norm": 18.384623872021983, "learning_rate": 1.455e-05, "loss": 3.8438, "step": 97 }, { "epoch": 0.04611764705882353, "grad_norm": 19.56318145075929, "learning_rate": 1.47e-05, "loss": 3.5625, "step": 98 }, { "epoch": 0.046588235294117646, "grad_norm": 31.618451275852667, "learning_rate": 1.485e-05, "loss": 3.1719, "step": 99 }, { "epoch": 0.047058823529411764, "grad_norm": 37.57565888884787, "learning_rate": 1.4999999999999999e-05, "loss": 2.7656, "step": 100 }, { "epoch": 0.04752941176470588, "grad_norm": 29.29155435568703, "learning_rate": 1.5149999999999999e-05, "loss": 2.7969, "step": 101 }, { "epoch": 0.048, "grad_norm": 22.337022045309705, "learning_rate": 1.53e-05, "loss": 3.5156, "step": 102 }, { "epoch": 0.04847058823529412, "grad_norm": 34.7046163321294, "learning_rate": 1.545e-05, "loss": 2.3438, "step": 103 }, { "epoch": 0.04894117647058824, "grad_norm": 30.62739504319389, "learning_rate": 1.5599999999999996e-05, "loss": 2.6719, "step": 104 }, { "epoch": 0.04941176470588235, "grad_norm": 54.78520623580544, "learning_rate": 1.5749999999999997e-05, "loss": 2.7969, "step": 105 }, { "epoch": 0.04988235294117647, "grad_norm": 17.636959795875743, "learning_rate": 1.5899999999999997e-05, "loss": 3.3438, "step": 106 }, { "epoch": 0.050352941176470586, "grad_norm": 27.92949647985338, "learning_rate": 1.6049999999999997e-05, "loss": 2.5938, "step": 107 }, { "epoch": 0.050823529411764705, "grad_norm": 23.315477403505213, "learning_rate": 1.6199999999999997e-05, "loss": 2.7344, "step": 108 }, { "epoch": 0.05129411764705882, "grad_norm": 32.89685836242179, "learning_rate": 1.6349999999999998e-05, "loss": 2.2031, "step": 109 }, { "epoch": 0.05176470588235294, "grad_norm": 32.761208472865974, "learning_rate": 1.6499999999999998e-05, "loss": 1.8047, "step": 110 }, { "epoch": 0.05223529411764706, "grad_norm": 25.030594088061843, "learning_rate": 1.6649999999999998e-05, "loss": 2.6406, "step": 111 }, { "epoch": 0.05270588235294118, "grad_norm": 33.39501558931582, "learning_rate": 1.68e-05, "loss": 3.1406, "step": 112 }, { "epoch": 0.0531764705882353, "grad_norm": 30.70592064174349, "learning_rate": 1.695e-05, "loss": 1.9766, "step": 113 }, { "epoch": 0.05364705882352941, "grad_norm": 27.353793811462733, "learning_rate": 1.71e-05, "loss": 2.0, "step": 114 }, { "epoch": 0.05411764705882353, "grad_norm": 25.49220344295269, "learning_rate": 1.725e-05, "loss": 2.25, "step": 115 }, { "epoch": 0.054588235294117646, "grad_norm": 26.92355181253648, "learning_rate": 1.74e-05, "loss": 2.2656, "step": 116 }, { "epoch": 0.055058823529411764, "grad_norm": 33.09281442277799, "learning_rate": 1.755e-05, "loss": 2.8594, "step": 117 }, { "epoch": 0.05552941176470588, "grad_norm": 27.01206988401169, "learning_rate": 1.7699999999999997e-05, "loss": 1.7109, "step": 118 }, { "epoch": 0.056, "grad_norm": 27.8987884205982, "learning_rate": 1.7849999999999997e-05, "loss": 1.3984, "step": 119 }, { "epoch": 0.05647058823529412, "grad_norm": 21.932991170834192, "learning_rate": 1.7999999999999997e-05, "loss": 2.5312, "step": 120 }, { "epoch": 0.05694117647058824, "grad_norm": 25.691244203986567, "learning_rate": 1.8149999999999997e-05, "loss": 1.5547, "step": 121 }, { "epoch": 0.05741176470588235, "grad_norm": 19.972158258155726, "learning_rate": 1.8299999999999998e-05, "loss": 2.2344, "step": 122 }, { "epoch": 0.05788235294117647, "grad_norm": 25.805867579019992, "learning_rate": 1.8449999999999998e-05, "loss": 1.4062, "step": 123 }, { "epoch": 0.05835294117647059, "grad_norm": 18.13514866906183, "learning_rate": 1.8599999999999998e-05, "loss": 2.2344, "step": 124 }, { "epoch": 0.058823529411764705, "grad_norm": 17.836659337475243, "learning_rate": 1.875e-05, "loss": 2.0781, "step": 125 }, { "epoch": 0.059294117647058824, "grad_norm": 19.673429304680766, "learning_rate": 1.89e-05, "loss": 1.9844, "step": 126 }, { "epoch": 0.05976470588235294, "grad_norm": 17.803979825011382, "learning_rate": 1.905e-05, "loss": 1.6875, "step": 127 }, { "epoch": 0.06023529411764706, "grad_norm": 20.402557382421737, "learning_rate": 1.92e-05, "loss": 1.6562, "step": 128 }, { "epoch": 0.06070588235294118, "grad_norm": 18.851479594532623, "learning_rate": 1.935e-05, "loss": 1.7812, "step": 129 }, { "epoch": 0.0611764705882353, "grad_norm": 20.092447799096274, "learning_rate": 1.95e-05, "loss": 1.2734, "step": 130 }, { "epoch": 0.06164705882352941, "grad_norm": 17.46935207512688, "learning_rate": 1.965e-05, "loss": 1.4375, "step": 131 }, { "epoch": 0.06211764705882353, "grad_norm": 19.63336393854412, "learning_rate": 1.98e-05, "loss": 0.8867, "step": 132 }, { "epoch": 0.06258823529411765, "grad_norm": 12.058056259436794, "learning_rate": 1.995e-05, "loss": 2.4062, "step": 133 }, { "epoch": 0.06305882352941176, "grad_norm": 14.28786815282879, "learning_rate": 2.01e-05, "loss": 1.3672, "step": 134 }, { "epoch": 0.06352941176470588, "grad_norm": 15.925459032269789, "learning_rate": 2.025e-05, "loss": 0.7773, "step": 135 }, { "epoch": 0.064, "grad_norm": 14.725466343599697, "learning_rate": 2.04e-05, "loss": 0.5859, "step": 136 }, { "epoch": 0.06447058823529411, "grad_norm": 9.440969107166815, "learning_rate": 2.055e-05, "loss": 1.6562, "step": 137 }, { "epoch": 0.06494117647058824, "grad_norm": 13.27879142886898, "learning_rate": 2.07e-05, "loss": 0.7109, "step": 138 }, { "epoch": 0.06541176470588235, "grad_norm": 12.011691711291492, "learning_rate": 2.085e-05, "loss": 0.5078, "step": 139 }, { "epoch": 0.06588235294117648, "grad_norm": 8.862005498305646, "learning_rate": 2.1e-05, "loss": 1.7188, "step": 140 }, { "epoch": 0.06635294117647059, "grad_norm": 10.251822863986336, "learning_rate": 2.1149999999999996e-05, "loss": 1.0469, "step": 141 }, { "epoch": 0.06682352941176471, "grad_norm": 9.485232465996303, "learning_rate": 2.1299999999999996e-05, "loss": 0.6055, "step": 142 }, { "epoch": 0.06729411764705882, "grad_norm": 9.231507181743552, "learning_rate": 2.1449999999999996e-05, "loss": 0.2773, "step": 143 }, { "epoch": 0.06776470588235294, "grad_norm": 7.592277081154745, "learning_rate": 2.1599999999999996e-05, "loss": 1.6094, "step": 144 }, { "epoch": 0.06823529411764706, "grad_norm": 6.542620279148565, "learning_rate": 2.1749999999999997e-05, "loss": 1.6406, "step": 145 }, { "epoch": 0.06870588235294117, "grad_norm": 6.104841770147069, "learning_rate": 2.1899999999999997e-05, "loss": 1.0859, "step": 146 }, { "epoch": 0.0691764705882353, "grad_norm": 5.95474283428887, "learning_rate": 2.2049999999999997e-05, "loss": 1.1016, "step": 147 }, { "epoch": 0.06964705882352941, "grad_norm": 7.123747413551944, "learning_rate": 2.2199999999999998e-05, "loss": 0.375, "step": 148 }, { "epoch": 0.07011764705882353, "grad_norm": 5.680807899906948, "learning_rate": 2.2349999999999998e-05, "loss": 0.9883, "step": 149 }, { "epoch": 0.07058823529411765, "grad_norm": 5.90332902344141, "learning_rate": 2.2499999999999998e-05, "loss": 1.3594, "step": 150 }, { "epoch": 0.07105882352941177, "grad_norm": 6.7838739167243824, "learning_rate": 2.2649999999999998e-05, "loss": 0.8281, "step": 151 }, { "epoch": 0.07152941176470588, "grad_norm": 5.672275364248789, "learning_rate": 2.28e-05, "loss": 1.3594, "step": 152 }, { "epoch": 0.072, "grad_norm": 5.667614343428989, "learning_rate": 2.295e-05, "loss": 0.7656, "step": 153 }, { "epoch": 0.07247058823529412, "grad_norm": 5.460318785186627, "learning_rate": 2.31e-05, "loss": 1.3359, "step": 154 }, { "epoch": 0.07294117647058823, "grad_norm": 7.095681683310174, "learning_rate": 2.325e-05, "loss": 1.3359, "step": 155 }, { "epoch": 0.07341176470588236, "grad_norm": 4.837756217965252, "learning_rate": 2.34e-05, "loss": 0.3145, "step": 156 }, { "epoch": 0.07388235294117647, "grad_norm": 4.923123737942311, "learning_rate": 2.3549999999999996e-05, "loss": 1.9922, "step": 157 }, { "epoch": 0.0743529411764706, "grad_norm": 4.276827397925142, "learning_rate": 2.3699999999999997e-05, "loss": 1.1484, "step": 158 }, { "epoch": 0.0748235294117647, "grad_norm": 4.984012041046594, "learning_rate": 2.3849999999999997e-05, "loss": 1.2969, "step": 159 }, { "epoch": 0.07529411764705882, "grad_norm": 4.351734206423633, "learning_rate": 2.3999999999999997e-05, "loss": 0.4668, "step": 160 }, { "epoch": 0.07576470588235294, "grad_norm": 4.93649385257823, "learning_rate": 2.4149999999999997e-05, "loss": 1.6016, "step": 161 }, { "epoch": 0.07623529411764705, "grad_norm": 5.1788288848015895, "learning_rate": 2.4299999999999998e-05, "loss": 1.2812, "step": 162 }, { "epoch": 0.07670588235294118, "grad_norm": 3.9569039940747173, "learning_rate": 2.4449999999999998e-05, "loss": 1.0391, "step": 163 }, { "epoch": 0.07717647058823529, "grad_norm": 4.628925518368876, "learning_rate": 2.4599999999999998e-05, "loss": 0.5352, "step": 164 }, { "epoch": 0.07764705882352942, "grad_norm": 3.90744775109024, "learning_rate": 2.475e-05, "loss": 1.4141, "step": 165 }, { "epoch": 0.07811764705882353, "grad_norm": 3.899990950713678, "learning_rate": 2.49e-05, "loss": 0.8242, "step": 166 }, { "epoch": 0.07858823529411765, "grad_norm": 4.101256389460906, "learning_rate": 2.505e-05, "loss": 0.9883, "step": 167 }, { "epoch": 0.07905882352941176, "grad_norm": 3.8319862166598524, "learning_rate": 2.52e-05, "loss": 0.9258, "step": 168 }, { "epoch": 0.07952941176470588, "grad_norm": 3.858091629574467, "learning_rate": 2.535e-05, "loss": 0.6406, "step": 169 }, { "epoch": 0.08, "grad_norm": 3.6148140214060676, "learning_rate": 2.55e-05, "loss": 1.2188, "step": 170 }, { "epoch": 0.08047058823529411, "grad_norm": 4.597719079294744, "learning_rate": 2.565e-05, "loss": 2.125, "step": 171 }, { "epoch": 0.08094117647058824, "grad_norm": 3.3889060899272097, "learning_rate": 2.5799999999999997e-05, "loss": 1.4531, "step": 172 }, { "epoch": 0.08141176470588235, "grad_norm": 3.383815825819804, "learning_rate": 2.5949999999999997e-05, "loss": 0.6445, "step": 173 }, { "epoch": 0.08188235294117648, "grad_norm": 2.8198111179313314, "learning_rate": 2.6099999999999997e-05, "loss": 0.2051, "step": 174 }, { "epoch": 0.08235294117647059, "grad_norm": 2.4569552473331098, "learning_rate": 2.6249999999999998e-05, "loss": 0.1523, "step": 175 }, { "epoch": 0.08282352941176471, "grad_norm": 3.7371653034419374, "learning_rate": 2.6399999999999995e-05, "loss": 0.7109, "step": 176 }, { "epoch": 0.08329411764705882, "grad_norm": 3.5664434122110378, "learning_rate": 2.6549999999999995e-05, "loss": 1.3281, "step": 177 }, { "epoch": 0.08376470588235294, "grad_norm": 3.2065448879076497, "learning_rate": 2.6699999999999995e-05, "loss": 0.9766, "step": 178 }, { "epoch": 0.08423529411764706, "grad_norm": 2.823868875331691, "learning_rate": 2.6849999999999995e-05, "loss": 0.4277, "step": 179 }, { "epoch": 0.08470588235294117, "grad_norm": 2.5225890178165753, "learning_rate": 2.6999999999999996e-05, "loss": 0.4746, "step": 180 }, { "epoch": 0.0851764705882353, "grad_norm": 3.2810121189778125, "learning_rate": 2.7149999999999996e-05, "loss": 0.8438, "step": 181 }, { "epoch": 0.08564705882352941, "grad_norm": 2.8783085624825926, "learning_rate": 2.7299999999999996e-05, "loss": 1.3359, "step": 182 }, { "epoch": 0.08611764705882353, "grad_norm": 2.7316658492833534, "learning_rate": 2.7449999999999996e-05, "loss": 0.7188, "step": 183 }, { "epoch": 0.08658823529411765, "grad_norm": 3.0419372803141567, "learning_rate": 2.7599999999999997e-05, "loss": 1.375, "step": 184 }, { "epoch": 0.08705882352941176, "grad_norm": 2.9347108413184633, "learning_rate": 2.7749999999999997e-05, "loss": 0.9883, "step": 185 }, { "epoch": 0.08752941176470588, "grad_norm": 2.1177733289515537, "learning_rate": 2.7899999999999997e-05, "loss": 0.4668, "step": 186 }, { "epoch": 0.088, "grad_norm": 5.466340773884225, "learning_rate": 2.8049999999999997e-05, "loss": 2.125, "step": 187 }, { "epoch": 0.08847058823529412, "grad_norm": 2.5884957544223215, "learning_rate": 2.8199999999999998e-05, "loss": 0.7031, "step": 188 }, { "epoch": 0.08894117647058823, "grad_norm": 1.9642120584736031, "learning_rate": 2.8349999999999998e-05, "loss": 0.2139, "step": 189 }, { "epoch": 0.08941176470588236, "grad_norm": 4.098128669106752, "learning_rate": 2.8499999999999998e-05, "loss": 1.2656, "step": 190 }, { "epoch": 0.08988235294117647, "grad_norm": 4.041450143168705, "learning_rate": 2.865e-05, "loss": 1.6953, "step": 191 }, { "epoch": 0.0903529411764706, "grad_norm": 2.797561988353154, "learning_rate": 2.88e-05, "loss": 1.1875, "step": 192 }, { "epoch": 0.0908235294117647, "grad_norm": 2.9830354999474795, "learning_rate": 2.895e-05, "loss": 1.3594, "step": 193 }, { "epoch": 0.09129411764705882, "grad_norm": 3.0431325210980127, "learning_rate": 2.91e-05, "loss": 1.25, "step": 194 }, { "epoch": 0.09176470588235294, "grad_norm": 2.2138179691729274, "learning_rate": 2.925e-05, "loss": 0.6484, "step": 195 }, { "epoch": 0.09223529411764705, "grad_norm": 2.507434932740471, "learning_rate": 2.94e-05, "loss": 0.4082, "step": 196 }, { "epoch": 0.09270588235294118, "grad_norm": 2.7025753504049885, "learning_rate": 2.955e-05, "loss": 0.6445, "step": 197 }, { "epoch": 0.09317647058823529, "grad_norm": 2.8140446327606283, "learning_rate": 2.97e-05, "loss": 0.4219, "step": 198 }, { "epoch": 0.09364705882352942, "grad_norm": 3.4556992553192396, "learning_rate": 2.985e-05, "loss": 1.5156, "step": 199 }, { "epoch": 0.09411764705882353, "grad_norm": 2.537615950425544, "learning_rate": 2.9999999999999997e-05, "loss": 1.0078, "step": 200 }, { "epoch": 0.09458823529411765, "grad_norm": 9.897745703107537, "learning_rate": 3.0149999999999998e-05, "loss": 0.3633, "step": 201 }, { "epoch": 0.09505882352941177, "grad_norm": 15.74279068840129, "learning_rate": 3.0299999999999998e-05, "loss": 1.6094, "step": 202 }, { "epoch": 0.09552941176470588, "grad_norm": 1.8243101746618369, "learning_rate": 3.0449999999999998e-05, "loss": 0.2852, "step": 203 }, { "epoch": 0.096, "grad_norm": 6.006273106267926, "learning_rate": 3.06e-05, "loss": 0.4355, "step": 204 }, { "epoch": 0.09647058823529411, "grad_norm": 5.2128998366548425, "learning_rate": 3.0749999999999995e-05, "loss": 0.7188, "step": 205 }, { "epoch": 0.09694117647058824, "grad_norm": 3.811406630223774, "learning_rate": 3.09e-05, "loss": 1.5625, "step": 206 }, { "epoch": 0.09741176470588235, "grad_norm": 2.737507761205516, "learning_rate": 3.1049999999999996e-05, "loss": 1.1953, "step": 207 }, { "epoch": 0.09788235294117648, "grad_norm": 3.590165694924627, "learning_rate": 3.119999999999999e-05, "loss": 1.4453, "step": 208 }, { "epoch": 0.09835294117647059, "grad_norm": 2.131366149464595, "learning_rate": 3.1349999999999996e-05, "loss": 0.5312, "step": 209 }, { "epoch": 0.0988235294117647, "grad_norm": 4.101910536597718, "learning_rate": 3.149999999999999e-05, "loss": 1.0938, "step": 210 }, { "epoch": 0.09929411764705882, "grad_norm": 3.5805749911891906, "learning_rate": 3.165e-05, "loss": 1.0234, "step": 211 }, { "epoch": 0.09976470588235294, "grad_norm": 3.747967871546596, "learning_rate": 3.1799999999999994e-05, "loss": 1.1875, "step": 212 }, { "epoch": 0.10023529411764706, "grad_norm": 3.060654078771219, "learning_rate": 3.195e-05, "loss": 1.2969, "step": 213 }, { "epoch": 0.10070588235294117, "grad_norm": 1.5035227211932738, "learning_rate": 3.2099999999999994e-05, "loss": 0.0854, "step": 214 }, { "epoch": 0.1011764705882353, "grad_norm": 2.4002034819222744, "learning_rate": 3.225e-05, "loss": 1.0312, "step": 215 }, { "epoch": 0.10164705882352941, "grad_norm": 3.2025521749078427, "learning_rate": 3.2399999999999995e-05, "loss": 0.9453, "step": 216 }, { "epoch": 0.10211764705882354, "grad_norm": 4.718969140147516, "learning_rate": 3.255e-05, "loss": 0.9648, "step": 217 }, { "epoch": 0.10258823529411765, "grad_norm": 2.9876548096014157, "learning_rate": 3.2699999999999995e-05, "loss": 1.3828, "step": 218 }, { "epoch": 0.10305882352941176, "grad_norm": 2.712765639174939, "learning_rate": 3.285e-05, "loss": 0.9531, "step": 219 }, { "epoch": 0.10352941176470588, "grad_norm": 2.8715271320767757, "learning_rate": 3.2999999999999996e-05, "loss": 0.9219, "step": 220 }, { "epoch": 0.104, "grad_norm": 1.5329707780397739, "learning_rate": 3.315e-05, "loss": 0.332, "step": 221 }, { "epoch": 0.10447058823529412, "grad_norm": 1.7859644235054186, "learning_rate": 3.3299999999999996e-05, "loss": 0.5156, "step": 222 }, { "epoch": 0.10494117647058823, "grad_norm": 3.1594754106965013, "learning_rate": 3.345e-05, "loss": 1.3828, "step": 223 }, { "epoch": 0.10541176470588236, "grad_norm": 1.8265676462364886, "learning_rate": 3.36e-05, "loss": 0.4961, "step": 224 }, { "epoch": 0.10588235294117647, "grad_norm": 1.9212461316129499, "learning_rate": 3.375e-05, "loss": 0.4473, "step": 225 }, { "epoch": 0.1063529411764706, "grad_norm": 3.588115734649765, "learning_rate": 3.39e-05, "loss": 1.2891, "step": 226 }, { "epoch": 0.1068235294117647, "grad_norm": 3.25322524968237, "learning_rate": 3.405e-05, "loss": 1.3047, "step": 227 }, { "epoch": 0.10729411764705882, "grad_norm": 1.4458494510477518, "learning_rate": 3.42e-05, "loss": 0.3008, "step": 228 }, { "epoch": 0.10776470588235294, "grad_norm": 2.4875196872728154, "learning_rate": 3.435e-05, "loss": 0.9336, "step": 229 }, { "epoch": 0.10823529411764705, "grad_norm": 2.7288414854267127, "learning_rate": 3.45e-05, "loss": 1.4297, "step": 230 }, { "epoch": 0.10870588235294118, "grad_norm": 2.3793583970654, "learning_rate": 3.465e-05, "loss": 0.918, "step": 231 }, { "epoch": 0.10917647058823529, "grad_norm": 2.942744462297241, "learning_rate": 3.48e-05, "loss": 1.1953, "step": 232 }, { "epoch": 0.10964705882352942, "grad_norm": 2.1249695499913246, "learning_rate": 3.4949999999999996e-05, "loss": 0.4023, "step": 233 }, { "epoch": 0.11011764705882353, "grad_norm": 3.3883906755546853, "learning_rate": 3.51e-05, "loss": 1.5078, "step": 234 }, { "epoch": 0.11058823529411765, "grad_norm": 3.0484587331604542, "learning_rate": 3.5249999999999996e-05, "loss": 1.4375, "step": 235 }, { "epoch": 0.11105882352941177, "grad_norm": 3.1220281537875354, "learning_rate": 3.539999999999999e-05, "loss": 1.5469, "step": 236 }, { "epoch": 0.11152941176470588, "grad_norm": 1.4698777738570221, "learning_rate": 3.555e-05, "loss": 0.3379, "step": 237 }, { "epoch": 0.112, "grad_norm": 2.328692172667586, "learning_rate": 3.5699999999999994e-05, "loss": 0.9492, "step": 238 }, { "epoch": 0.11247058823529411, "grad_norm": 2.4546903391435797, "learning_rate": 3.585e-05, "loss": 1.1719, "step": 239 }, { "epoch": 0.11294117647058824, "grad_norm": 2.883815584389125, "learning_rate": 3.5999999999999994e-05, "loss": 1.2812, "step": 240 }, { "epoch": 0.11341176470588235, "grad_norm": 2.738639039329212, "learning_rate": 3.615e-05, "loss": 0.7773, "step": 241 }, { "epoch": 0.11388235294117648, "grad_norm": 2.4930017385985046, "learning_rate": 3.6299999999999995e-05, "loss": 1.2266, "step": 242 }, { "epoch": 0.11435294117647059, "grad_norm": 2.1230146208077367, "learning_rate": 3.645e-05, "loss": 1.1875, "step": 243 }, { "epoch": 0.1148235294117647, "grad_norm": 3.2569644465926384, "learning_rate": 3.6599999999999995e-05, "loss": 2.4062, "step": 244 }, { "epoch": 0.11529411764705882, "grad_norm": 2.42946018218758, "learning_rate": 3.675e-05, "loss": 0.8633, "step": 245 }, { "epoch": 0.11576470588235294, "grad_norm": 1.4269942689205244, "learning_rate": 3.6899999999999996e-05, "loss": 0.2305, "step": 246 }, { "epoch": 0.11623529411764706, "grad_norm": 2.854359921486424, "learning_rate": 3.705e-05, "loss": 1.0703, "step": 247 }, { "epoch": 0.11670588235294117, "grad_norm": 2.445219997163655, "learning_rate": 3.7199999999999996e-05, "loss": 0.8555, "step": 248 }, { "epoch": 0.1171764705882353, "grad_norm": 4.442391006050997, "learning_rate": 3.735e-05, "loss": 1.2266, "step": 249 }, { "epoch": 0.11764705882352941, "grad_norm": 2.992097363467336, "learning_rate": 3.75e-05, "loss": 1.7891, "step": 250 }, { "epoch": 0.11811764705882354, "grad_norm": 2.625327550746839, "learning_rate": 3.7649999999999994e-05, "loss": 0.9609, "step": 251 }, { "epoch": 0.11858823529411765, "grad_norm": 3.011437679507196, "learning_rate": 3.78e-05, "loss": 1.2812, "step": 252 }, { "epoch": 0.11905882352941176, "grad_norm": 1.1319960908564297, "learning_rate": 3.7949999999999994e-05, "loss": 0.252, "step": 253 }, { "epoch": 0.11952941176470588, "grad_norm": 1.7561583512386556, "learning_rate": 3.81e-05, "loss": 0.167, "step": 254 }, { "epoch": 0.12, "grad_norm": 3.1090967210890437, "learning_rate": 3.8249999999999995e-05, "loss": 1.2891, "step": 255 }, { "epoch": 0.12047058823529412, "grad_norm": 2.2821923122911834, "learning_rate": 3.84e-05, "loss": 1.2031, "step": 256 }, { "epoch": 0.12094117647058823, "grad_norm": 2.087449545459651, "learning_rate": 3.8549999999999995e-05, "loss": 1.1719, "step": 257 }, { "epoch": 0.12141176470588236, "grad_norm": 2.235388940782363, "learning_rate": 3.87e-05, "loss": 1.2422, "step": 258 }, { "epoch": 0.12188235294117647, "grad_norm": 2.6960598186490667, "learning_rate": 3.8849999999999996e-05, "loss": 1.2109, "step": 259 }, { "epoch": 0.1223529411764706, "grad_norm": 2.060933505519406, "learning_rate": 3.9e-05, "loss": 0.4668, "step": 260 }, { "epoch": 0.1228235294117647, "grad_norm": 2.17535582490466, "learning_rate": 3.9149999999999996e-05, "loss": 0.5742, "step": 261 }, { "epoch": 0.12329411764705882, "grad_norm": 1.6818164721295374, "learning_rate": 3.93e-05, "loss": 0.4746, "step": 262 }, { "epoch": 0.12376470588235294, "grad_norm": 3.4905596802047487, "learning_rate": 3.945e-05, "loss": 0.6953, "step": 263 }, { "epoch": 0.12423529411764705, "grad_norm": 1.2199905152038284, "learning_rate": 3.96e-05, "loss": 0.1709, "step": 264 }, { "epoch": 0.12470588235294118, "grad_norm": 3.115098250039668, "learning_rate": 3.975e-05, "loss": 1.2656, "step": 265 }, { "epoch": 0.1251764705882353, "grad_norm": 2.612240427593972, "learning_rate": 3.99e-05, "loss": 1.0469, "step": 266 }, { "epoch": 0.12564705882352942, "grad_norm": 1.8896759825529688, "learning_rate": 4.005e-05, "loss": 0.3887, "step": 267 }, { "epoch": 0.12611764705882353, "grad_norm": 3.1126529533394485, "learning_rate": 4.02e-05, "loss": 1.2734, "step": 268 }, { "epoch": 0.12658823529411764, "grad_norm": 1.7818690834907158, "learning_rate": 4.035e-05, "loss": 0.5273, "step": 269 }, { "epoch": 0.12705882352941175, "grad_norm": 1.9695925135488939, "learning_rate": 4.05e-05, "loss": 0.8203, "step": 270 }, { "epoch": 0.1275294117647059, "grad_norm": 2.773941317629207, "learning_rate": 4.065e-05, "loss": 0.4902, "step": 271 }, { "epoch": 0.128, "grad_norm": 2.1475149968941567, "learning_rate": 4.08e-05, "loss": 0.9414, "step": 272 }, { "epoch": 0.12847058823529411, "grad_norm": 3.189965478162992, "learning_rate": 4.095e-05, "loss": 1.1094, "step": 273 }, { "epoch": 0.12894117647058823, "grad_norm": 1.8447180925663897, "learning_rate": 4.11e-05, "loss": 0.5078, "step": 274 }, { "epoch": 0.12941176470588237, "grad_norm": 2.3453312719416504, "learning_rate": 4.125e-05, "loss": 0.8906, "step": 275 }, { "epoch": 0.12988235294117648, "grad_norm": 1.8683326407962295, "learning_rate": 4.14e-05, "loss": 1.1484, "step": 276 }, { "epoch": 0.1303529411764706, "grad_norm": 1.5260036882694452, "learning_rate": 4.155e-05, "loss": 0.3047, "step": 277 }, { "epoch": 0.1308235294117647, "grad_norm": 3.1487362571008433, "learning_rate": 4.17e-05, "loss": 1.7891, "step": 278 }, { "epoch": 0.1312941176470588, "grad_norm": 2.654896453588714, "learning_rate": 4.185e-05, "loss": 0.5625, "step": 279 }, { "epoch": 0.13176470588235295, "grad_norm": 2.1717768018964705, "learning_rate": 4.2e-05, "loss": 0.7617, "step": 280 }, { "epoch": 0.13223529411764706, "grad_norm": 2.000790306134065, "learning_rate": 4.215e-05, "loss": 0.2207, "step": 281 }, { "epoch": 0.13270588235294117, "grad_norm": 3.8223364504834763, "learning_rate": 4.229999999999999e-05, "loss": 0.8711, "step": 282 }, { "epoch": 0.13317647058823529, "grad_norm": 1.8160475186326681, "learning_rate": 4.2449999999999995e-05, "loss": 0.9844, "step": 283 }, { "epoch": 0.13364705882352942, "grad_norm": 1.96919364366986, "learning_rate": 4.259999999999999e-05, "loss": 0.7266, "step": 284 }, { "epoch": 0.13411764705882354, "grad_norm": 2.3820439897535772, "learning_rate": 4.2749999999999996e-05, "loss": 0.9297, "step": 285 }, { "epoch": 0.13458823529411765, "grad_norm": 2.8641858736053316, "learning_rate": 4.289999999999999e-05, "loss": 1.5234, "step": 286 }, { "epoch": 0.13505882352941176, "grad_norm": 1.2012603070637584, "learning_rate": 4.3049999999999996e-05, "loss": 0.2812, "step": 287 }, { "epoch": 0.13552941176470587, "grad_norm": 1.5894718182852974, "learning_rate": 4.319999999999999e-05, "loss": 0.5742, "step": 288 }, { "epoch": 0.136, "grad_norm": 1.8966947993341616, "learning_rate": 4.334999999999999e-05, "loss": 0.6602, "step": 289 }, { "epoch": 0.13647058823529412, "grad_norm": 4.837148630874499, "learning_rate": 4.3499999999999993e-05, "loss": 1.5938, "step": 290 }, { "epoch": 0.13694117647058823, "grad_norm": 2.26799245631818, "learning_rate": 4.364999999999999e-05, "loss": 1.125, "step": 291 }, { "epoch": 0.13741176470588234, "grad_norm": 0.7728096927818335, "learning_rate": 4.3799999999999994e-05, "loss": 0.0552, "step": 292 }, { "epoch": 0.13788235294117648, "grad_norm": 4.213567383479722, "learning_rate": 4.394999999999999e-05, "loss": 1.4688, "step": 293 }, { "epoch": 0.1383529411764706, "grad_norm": 2.848981512521546, "learning_rate": 4.4099999999999995e-05, "loss": 1.0156, "step": 294 }, { "epoch": 0.1388235294117647, "grad_norm": 2.0423719893501486, "learning_rate": 4.424999999999999e-05, "loss": 1.0156, "step": 295 }, { "epoch": 0.13929411764705882, "grad_norm": 2.5402984032103526, "learning_rate": 4.4399999999999995e-05, "loss": 1.1172, "step": 296 }, { "epoch": 0.13976470588235293, "grad_norm": 0.6584136052328181, "learning_rate": 4.454999999999999e-05, "loss": 0.0864, "step": 297 }, { "epoch": 0.14023529411764707, "grad_norm": 1.8006502632023358, "learning_rate": 4.4699999999999996e-05, "loss": 0.7773, "step": 298 }, { "epoch": 0.14070588235294118, "grad_norm": 2.2245536504490184, "learning_rate": 4.484999999999999e-05, "loss": 1.0312, "step": 299 }, { "epoch": 0.1411764705882353, "grad_norm": 3.2695351717983576, "learning_rate": 4.4999999999999996e-05, "loss": 1.7812, "step": 300 }, { "epoch": 0.1416470588235294, "grad_norm": 2.0805657015415484, "learning_rate": 4.514999999999999e-05, "loss": 1.0625, "step": 301 }, { "epoch": 0.14211764705882354, "grad_norm": 2.1572593693529063, "learning_rate": 4.5299999999999997e-05, "loss": 0.8047, "step": 302 }, { "epoch": 0.14258823529411765, "grad_norm": 2.2372936329368542, "learning_rate": 4.5449999999999993e-05, "loss": 0.8242, "step": 303 }, { "epoch": 0.14305882352941177, "grad_norm": 2.1718337667740237, "learning_rate": 4.56e-05, "loss": 1.3359, "step": 304 }, { "epoch": 0.14352941176470588, "grad_norm": 2.0020036045980545, "learning_rate": 4.5749999999999994e-05, "loss": 0.9023, "step": 305 }, { "epoch": 0.144, "grad_norm": 1.4177838454176748, "learning_rate": 4.59e-05, "loss": 0.1357, "step": 306 }, { "epoch": 0.14447058823529413, "grad_norm": 2.0484352349510657, "learning_rate": 4.6049999999999994e-05, "loss": 0.6445, "step": 307 }, { "epoch": 0.14494117647058824, "grad_norm": 1.8695392382568854, "learning_rate": 4.62e-05, "loss": 0.8203, "step": 308 }, { "epoch": 0.14541176470588235, "grad_norm": 1.2675306462833387, "learning_rate": 4.6349999999999995e-05, "loss": 0.1816, "step": 309 }, { "epoch": 0.14588235294117646, "grad_norm": 1.6107852966306957, "learning_rate": 4.65e-05, "loss": 0.3164, "step": 310 }, { "epoch": 0.14635294117647057, "grad_norm": 1.6808614356300777, "learning_rate": 4.6649999999999996e-05, "loss": 0.6602, "step": 311 }, { "epoch": 0.1468235294117647, "grad_norm": 2.5085658978847154, "learning_rate": 4.68e-05, "loss": 1.4141, "step": 312 }, { "epoch": 0.14729411764705883, "grad_norm": 2.178717924412675, "learning_rate": 4.6949999999999996e-05, "loss": 0.8086, "step": 313 }, { "epoch": 0.14776470588235294, "grad_norm": 2.065668243780936, "learning_rate": 4.709999999999999e-05, "loss": 1.2969, "step": 314 }, { "epoch": 0.14823529411764705, "grad_norm": 1.5024360788510707, "learning_rate": 4.7249999999999997e-05, "loss": 0.2197, "step": 315 }, { "epoch": 0.1487058823529412, "grad_norm": 2.047112489430929, "learning_rate": 4.7399999999999993e-05, "loss": 1.25, "step": 316 }, { "epoch": 0.1491764705882353, "grad_norm": 1.8443086492973062, "learning_rate": 4.755e-05, "loss": 0.6797, "step": 317 }, { "epoch": 0.1496470588235294, "grad_norm": 0.8166306287436869, "learning_rate": 4.7699999999999994e-05, "loss": 0.0859, "step": 318 }, { "epoch": 0.15011764705882352, "grad_norm": 4.05039814472271, "learning_rate": 4.785e-05, "loss": 1.4531, "step": 319 }, { "epoch": 0.15058823529411763, "grad_norm": 2.512186636483786, "learning_rate": 4.7999999999999994e-05, "loss": 0.957, "step": 320 }, { "epoch": 0.15105882352941177, "grad_norm": 2.1764378878579467, "learning_rate": 4.815e-05, "loss": 0.9062, "step": 321 }, { "epoch": 0.15152941176470588, "grad_norm": 2.3631120248197828, "learning_rate": 4.8299999999999995e-05, "loss": 1.0859, "step": 322 }, { "epoch": 0.152, "grad_norm": 1.0313282602370724, "learning_rate": 4.845e-05, "loss": 0.1079, "step": 323 }, { "epoch": 0.1524705882352941, "grad_norm": 2.5872902286732096, "learning_rate": 4.8599999999999995e-05, "loss": 1.1875, "step": 324 }, { "epoch": 0.15294117647058825, "grad_norm": 4.73445867231754, "learning_rate": 4.875e-05, "loss": 1.5156, "step": 325 }, { "epoch": 0.15341176470588236, "grad_norm": 1.2569979172926453, "learning_rate": 4.8899999999999996e-05, "loss": 0.3496, "step": 326 }, { "epoch": 0.15388235294117647, "grad_norm": 4.4627810853812, "learning_rate": 4.905e-05, "loss": 0.6055, "step": 327 }, { "epoch": 0.15435294117647058, "grad_norm": 1.630419965271155, "learning_rate": 4.9199999999999997e-05, "loss": 0.3867, "step": 328 }, { "epoch": 0.1548235294117647, "grad_norm": 2.998099187670558, "learning_rate": 4.935e-05, "loss": 0.9961, "step": 329 }, { "epoch": 0.15529411764705883, "grad_norm": 2.082842193677949, "learning_rate": 4.95e-05, "loss": 1.1016, "step": 330 }, { "epoch": 0.15576470588235294, "grad_norm": 1.0912609018521289, "learning_rate": 4.965e-05, "loss": 0.2393, "step": 331 }, { "epoch": 0.15623529411764706, "grad_norm": 2.369871712466468, "learning_rate": 4.98e-05, "loss": 0.5977, "step": 332 }, { "epoch": 0.15670588235294117, "grad_norm": 3.8215487118211957, "learning_rate": 4.995e-05, "loss": 1.0703, "step": 333 }, { "epoch": 0.1571764705882353, "grad_norm": 3.4283227367783016, "learning_rate": 5.01e-05, "loss": 2.3125, "step": 334 }, { "epoch": 0.15764705882352942, "grad_norm": 2.174047650144301, "learning_rate": 5.025e-05, "loss": 0.4727, "step": 335 }, { "epoch": 0.15811764705882353, "grad_norm": 1.7630068354439452, "learning_rate": 5.04e-05, "loss": 0.3027, "step": 336 }, { "epoch": 0.15858823529411764, "grad_norm": 2.249266443242867, "learning_rate": 5.055e-05, "loss": 0.7031, "step": 337 }, { "epoch": 0.15905882352941175, "grad_norm": 2.615010238128941, "learning_rate": 5.07e-05, "loss": 0.5898, "step": 338 }, { "epoch": 0.1595294117647059, "grad_norm": 2.4483345322820975, "learning_rate": 5.0849999999999996e-05, "loss": 1.1094, "step": 339 }, { "epoch": 0.16, "grad_norm": 3.8051237454920868, "learning_rate": 5.1e-05, "loss": 1.75, "step": 340 }, { "epoch": 0.16047058823529411, "grad_norm": 2.4141280429963037, "learning_rate": 5.1149999999999996e-05, "loss": 1.2422, "step": 341 }, { "epoch": 0.16094117647058823, "grad_norm": 2.934789148222637, "learning_rate": 5.13e-05, "loss": 1.9609, "step": 342 }, { "epoch": 0.16141176470588237, "grad_norm": 2.082952669047601, "learning_rate": 5.145e-05, "loss": 1.2734, "step": 343 }, { "epoch": 0.16188235294117648, "grad_norm": 2.3275935806164196, "learning_rate": 5.1599999999999994e-05, "loss": 1.0625, "step": 344 }, { "epoch": 0.1623529411764706, "grad_norm": 1.9530342658659814, "learning_rate": 5.174999999999999e-05, "loss": 1.0625, "step": 345 }, { "epoch": 0.1628235294117647, "grad_norm": 1.9102128565534644, "learning_rate": 5.1899999999999994e-05, "loss": 0.7969, "step": 346 }, { "epoch": 0.1632941176470588, "grad_norm": 1.3030202600017686, "learning_rate": 5.204999999999999e-05, "loss": 0.2637, "step": 347 }, { "epoch": 0.16376470588235295, "grad_norm": 3.616138860109933, "learning_rate": 5.2199999999999995e-05, "loss": 2.1094, "step": 348 }, { "epoch": 0.16423529411764706, "grad_norm": 2.863558437008838, "learning_rate": 5.234999999999999e-05, "loss": 1.7422, "step": 349 }, { "epoch": 0.16470588235294117, "grad_norm": 1.613370728604923, "learning_rate": 5.2499999999999995e-05, "loss": 0.2051, "step": 350 }, { "epoch": 0.16517647058823529, "grad_norm": 1.6283791826014704, "learning_rate": 5.264999999999999e-05, "loss": 0.4141, "step": 351 }, { "epoch": 0.16564705882352943, "grad_norm": 2.2875467203849595, "learning_rate": 5.279999999999999e-05, "loss": 1.0859, "step": 352 }, { "epoch": 0.16611764705882354, "grad_norm": 2.5658667964525206, "learning_rate": 5.294999999999999e-05, "loss": 1.8594, "step": 353 }, { "epoch": 0.16658823529411765, "grad_norm": 1.3179934643093218, "learning_rate": 5.309999999999999e-05, "loss": 0.7695, "step": 354 }, { "epoch": 0.16705882352941176, "grad_norm": 2.178901641764621, "learning_rate": 5.324999999999999e-05, "loss": 0.9922, "step": 355 }, { "epoch": 0.16752941176470587, "grad_norm": 2.0231347768555032, "learning_rate": 5.339999999999999e-05, "loss": 0.7773, "step": 356 }, { "epoch": 0.168, "grad_norm": 1.4427512473893422, "learning_rate": 5.3549999999999994e-05, "loss": 0.1011, "step": 357 }, { "epoch": 0.16847058823529412, "grad_norm": 2.1859315325493265, "learning_rate": 5.369999999999999e-05, "loss": 1.4766, "step": 358 }, { "epoch": 0.16894117647058823, "grad_norm": 1.1310566721786202, "learning_rate": 5.3849999999999994e-05, "loss": 0.1504, "step": 359 }, { "epoch": 0.16941176470588235, "grad_norm": 2.2273301591896915, "learning_rate": 5.399999999999999e-05, "loss": 1.1875, "step": 360 }, { "epoch": 0.16988235294117648, "grad_norm": 2.594180207029895, "learning_rate": 5.4149999999999995e-05, "loss": 1.25, "step": 361 }, { "epoch": 0.1703529411764706, "grad_norm": 1.4032488300540513, "learning_rate": 5.429999999999999e-05, "loss": 0.6406, "step": 362 }, { "epoch": 0.1708235294117647, "grad_norm": 1.9608035155938763, "learning_rate": 5.4449999999999995e-05, "loss": 0.793, "step": 363 }, { "epoch": 0.17129411764705882, "grad_norm": 1.8375488481130156, "learning_rate": 5.459999999999999e-05, "loss": 0.8789, "step": 364 }, { "epoch": 0.17176470588235293, "grad_norm": 0.9627216611766073, "learning_rate": 5.4749999999999996e-05, "loss": 0.1084, "step": 365 }, { "epoch": 0.17223529411764707, "grad_norm": 1.4501883819670796, "learning_rate": 5.489999999999999e-05, "loss": 0.334, "step": 366 }, { "epoch": 0.17270588235294118, "grad_norm": 0.943273781484865, "learning_rate": 5.5049999999999996e-05, "loss": 0.2236, "step": 367 }, { "epoch": 0.1731764705882353, "grad_norm": 1.1843528130040526, "learning_rate": 5.519999999999999e-05, "loss": 0.2324, "step": 368 }, { "epoch": 0.1736470588235294, "grad_norm": 3.127882997426753, "learning_rate": 5.535e-05, "loss": 1.6719, "step": 369 }, { "epoch": 0.17411764705882352, "grad_norm": 2.2366627781618256, "learning_rate": 5.5499999999999994e-05, "loss": 1.5156, "step": 370 }, { "epoch": 0.17458823529411766, "grad_norm": 3.7353281405130754, "learning_rate": 5.565e-05, "loss": 2.0469, "step": 371 }, { "epoch": 0.17505882352941177, "grad_norm": 1.4682039583234394, "learning_rate": 5.5799999999999994e-05, "loss": 0.6367, "step": 372 }, { "epoch": 0.17552941176470588, "grad_norm": 1.1785587492485496, "learning_rate": 5.595e-05, "loss": 0.1885, "step": 373 }, { "epoch": 0.176, "grad_norm": 2.3981290505770474, "learning_rate": 5.6099999999999995e-05, "loss": 0.7344, "step": 374 }, { "epoch": 0.17647058823529413, "grad_norm": 2.546407917229536, "learning_rate": 5.625e-05, "loss": 0.1807, "step": 375 }, { "epoch": 0.17694117647058824, "grad_norm": 1.1253733092210205, "learning_rate": 5.6399999999999995e-05, "loss": 0.1309, "step": 376 }, { "epoch": 0.17741176470588235, "grad_norm": 3.0888840285008614, "learning_rate": 5.654999999999999e-05, "loss": 1.1875, "step": 377 }, { "epoch": 0.17788235294117646, "grad_norm": 1.8370345995794368, "learning_rate": 5.6699999999999996e-05, "loss": 1.0625, "step": 378 }, { "epoch": 0.17835294117647058, "grad_norm": 3.262469863059897, "learning_rate": 5.684999999999999e-05, "loss": 1.2031, "step": 379 }, { "epoch": 0.17882352941176471, "grad_norm": 1.6349180626419697, "learning_rate": 5.6999999999999996e-05, "loss": 0.7539, "step": 380 }, { "epoch": 0.17929411764705883, "grad_norm": 1.6301143084725096, "learning_rate": 5.714999999999999e-05, "loss": 0.2197, "step": 381 }, { "epoch": 0.17976470588235294, "grad_norm": 3.7918691915420015, "learning_rate": 5.73e-05, "loss": 2.3281, "step": 382 }, { "epoch": 0.18023529411764705, "grad_norm": 1.551978292086268, "learning_rate": 5.7449999999999994e-05, "loss": 0.8203, "step": 383 }, { "epoch": 0.1807058823529412, "grad_norm": 1.475584309423767, "learning_rate": 5.76e-05, "loss": 0.8008, "step": 384 }, { "epoch": 0.1811764705882353, "grad_norm": 31.717527443393383, "learning_rate": 5.7749999999999994e-05, "loss": 0.2676, "step": 385 }, { "epoch": 0.1816470588235294, "grad_norm": 13.372256148147589, "learning_rate": 5.79e-05, "loss": 0.8438, "step": 386 }, { "epoch": 0.18211764705882352, "grad_norm": 3.1947182386740076, "learning_rate": 5.8049999999999995e-05, "loss": 1.3672, "step": 387 }, { "epoch": 0.18258823529411763, "grad_norm": 167.14538862327439, "learning_rate": 5.82e-05, "loss": 2.1094, "step": 388 }, { "epoch": 0.18305882352941177, "grad_norm": 55.87413482387034, "learning_rate": 5.8349999999999995e-05, "loss": 7.0625, "step": 389 }, { "epoch": 0.18352941176470589, "grad_norm": 3.742080080457822, "learning_rate": 5.85e-05, "loss": 1.9219, "step": 390 }, { "epoch": 0.184, "grad_norm": 14.09336634604679, "learning_rate": 5.8649999999999996e-05, "loss": 1.0703, "step": 391 }, { "epoch": 0.1844705882352941, "grad_norm": 7.714310873476905, "learning_rate": 5.88e-05, "loss": 0.9453, "step": 392 }, { "epoch": 0.18494117647058825, "grad_norm": 46.33023995767413, "learning_rate": 5.8949999999999996e-05, "loss": 2.625, "step": 393 }, { "epoch": 0.18541176470588236, "grad_norm": 2.8642396717529373, "learning_rate": 5.91e-05, "loss": 0.8984, "step": 394 }, { "epoch": 0.18588235294117647, "grad_norm": 6.587573382785803, "learning_rate": 5.925e-05, "loss": 1.0781, "step": 395 }, { "epoch": 0.18635294117647058, "grad_norm": 3.0432426398122057, "learning_rate": 5.94e-05, "loss": 1.5156, "step": 396 }, { "epoch": 0.1868235294117647, "grad_norm": 5.691127683146497, "learning_rate": 5.955e-05, "loss": 1.7266, "step": 397 }, { "epoch": 0.18729411764705883, "grad_norm": 2.9715233288028133, "learning_rate": 5.97e-05, "loss": 1.3672, "step": 398 }, { "epoch": 0.18776470588235294, "grad_norm": 204.7610481320438, "learning_rate": 5.985e-05, "loss": 2.2344, "step": 399 }, { "epoch": 0.18823529411764706, "grad_norm": 3.368437813354467, "learning_rate": 5.9999999999999995e-05, "loss": 0.6992, "step": 400 }, { "epoch": 0.18870588235294117, "grad_norm": 462.89393539594863, "learning_rate": 6.015e-05, "loss": 16.625, "step": 401 }, { "epoch": 0.1891764705882353, "grad_norm": 31.31686257665466, "learning_rate": 6.0299999999999995e-05, "loss": 0.5156, "step": 402 }, { "epoch": 0.18964705882352942, "grad_norm": 2.580151856063713, "learning_rate": 6.045e-05, "loss": 0.7969, "step": 403 }, { "epoch": 0.19011764705882353, "grad_norm": 7.119458343825339, "learning_rate": 6.0599999999999996e-05, "loss": 3.2656, "step": 404 }, { "epoch": 0.19058823529411764, "grad_norm": 3.680946753279621, "learning_rate": 6.075e-05, "loss": 2.3594, "step": 405 }, { "epoch": 0.19105882352941175, "grad_norm": 9.085773077183179, "learning_rate": 6.0899999999999996e-05, "loss": 4.9062, "step": 406 }, { "epoch": 0.1915294117647059, "grad_norm": 4.464561519551958, "learning_rate": 6.104999999999999e-05, "loss": 2.8281, "step": 407 }, { "epoch": 0.192, "grad_norm": 3.0647279723527014, "learning_rate": 6.12e-05, "loss": 2.5469, "step": 408 }, { "epoch": 0.19247058823529412, "grad_norm": 1.6218244133451922, "learning_rate": 6.134999999999999e-05, "loss": 0.8008, "step": 409 }, { "epoch": 0.19294117647058823, "grad_norm": 4.816737286970588, "learning_rate": 6.149999999999999e-05, "loss": 2.75, "step": 410 }, { "epoch": 0.19341176470588237, "grad_norm": 2.027956761133797, "learning_rate": 6.165e-05, "loss": 1.3672, "step": 411 }, { "epoch": 0.19388235294117648, "grad_norm": 1.386166766890797, "learning_rate": 6.18e-05, "loss": 0.8477, "step": 412 }, { "epoch": 0.1943529411764706, "grad_norm": 3.116913505803336, "learning_rate": 6.194999999999999e-05, "loss": 1.2109, "step": 413 }, { "epoch": 0.1948235294117647, "grad_norm": 1.913989769221226, "learning_rate": 6.209999999999999e-05, "loss": 1.4297, "step": 414 }, { "epoch": 0.1952941176470588, "grad_norm": 3.98831975877608, "learning_rate": 6.225e-05, "loss": 2.2188, "step": 415 }, { "epoch": 0.19576470588235295, "grad_norm": 0.9573248512125144, "learning_rate": 6.239999999999999e-05, "loss": 0.1875, "step": 416 }, { "epoch": 0.19623529411764706, "grad_norm": 2.762180183539294, "learning_rate": 6.254999999999999e-05, "loss": 2.9062, "step": 417 }, { "epoch": 0.19670588235294117, "grad_norm": 3.227554006732618, "learning_rate": 6.269999999999999e-05, "loss": 2.9062, "step": 418 }, { "epoch": 0.1971764705882353, "grad_norm": 1.7836326193659289, "learning_rate": 6.285e-05, "loss": 1.2578, "step": 419 }, { "epoch": 0.1976470588235294, "grad_norm": 1.1351935994424045, "learning_rate": 6.299999999999999e-05, "loss": 0.957, "step": 420 }, { "epoch": 0.19811764705882354, "grad_norm": 4.909668911651264, "learning_rate": 6.314999999999999e-05, "loss": 0.3281, "step": 421 }, { "epoch": 0.19858823529411765, "grad_norm": 1.93637825510381, "learning_rate": 6.33e-05, "loss": 0.6953, "step": 422 }, { "epoch": 0.19905882352941176, "grad_norm": 3.3321261988275204, "learning_rate": 6.345e-05, "loss": 1.9609, "step": 423 }, { "epoch": 0.19952941176470587, "grad_norm": 3.0364844853577506, "learning_rate": 6.359999999999999e-05, "loss": 2.25, "step": 424 }, { "epoch": 0.2, "grad_norm": 3.3817941993450717, "learning_rate": 6.374999999999999e-05, "loss": 2.5938, "step": 425 }, { "epoch": 0.20047058823529412, "grad_norm": 2.141408233089959, "learning_rate": 6.39e-05, "loss": 1.5078, "step": 426 }, { "epoch": 0.20094117647058823, "grad_norm": 2.6987019486294033, "learning_rate": 6.405e-05, "loss": 1.4375, "step": 427 }, { "epoch": 0.20141176470588235, "grad_norm": 4.721244793636129, "learning_rate": 6.419999999999999e-05, "loss": 2.25, "step": 428 }, { "epoch": 0.20188235294117646, "grad_norm": 1.950777179509344, "learning_rate": 6.434999999999999e-05, "loss": 1.4141, "step": 429 }, { "epoch": 0.2023529411764706, "grad_norm": 5.617342166821382, "learning_rate": 6.45e-05, "loss": 1.8281, "step": 430 }, { "epoch": 0.2028235294117647, "grad_norm": 1.3770959232449782, "learning_rate": 6.465e-05, "loss": 1.1016, "step": 431 }, { "epoch": 0.20329411764705882, "grad_norm": 4.222815588844367, "learning_rate": 6.479999999999999e-05, "loss": 2.2656, "step": 432 }, { "epoch": 0.20376470588235293, "grad_norm": 2.533221783526971, "learning_rate": 6.494999999999999e-05, "loss": 1.9844, "step": 433 }, { "epoch": 0.20423529411764707, "grad_norm": 2.7439406996653144, "learning_rate": 6.51e-05, "loss": 1.5469, "step": 434 }, { "epoch": 0.20470588235294118, "grad_norm": 1.3516241517031073, "learning_rate": 6.525e-05, "loss": 0.6719, "step": 435 }, { "epoch": 0.2051764705882353, "grad_norm": 2.454511119660775, "learning_rate": 6.539999999999999e-05, "loss": 1.5078, "step": 436 }, { "epoch": 0.2056470588235294, "grad_norm": 2.7991741854646355, "learning_rate": 6.555e-05, "loss": 1.6797, "step": 437 }, { "epoch": 0.20611764705882352, "grad_norm": 1.9037577939079204, "learning_rate": 6.57e-05, "loss": 1.0859, "step": 438 }, { "epoch": 0.20658823529411766, "grad_norm": 1.6179537960225938, "learning_rate": 6.584999999999999e-05, "loss": 0.9961, "step": 439 }, { "epoch": 0.20705882352941177, "grad_norm": 2.4569381791323557, "learning_rate": 6.599999999999999e-05, "loss": 1.0938, "step": 440 }, { "epoch": 0.20752941176470588, "grad_norm": 1.9916446706896354, "learning_rate": 6.615e-05, "loss": 1.5391, "step": 441 }, { "epoch": 0.208, "grad_norm": 1.8645054422859628, "learning_rate": 6.63e-05, "loss": 1.4453, "step": 442 }, { "epoch": 0.20847058823529413, "grad_norm": 1.0836431137271858, "learning_rate": 6.644999999999999e-05, "loss": 0.4941, "step": 443 }, { "epoch": 0.20894117647058824, "grad_norm": 1.1462845667576904, "learning_rate": 6.659999999999999e-05, "loss": 0.6094, "step": 444 }, { "epoch": 0.20941176470588235, "grad_norm": 12.306946646251378, "learning_rate": 6.675e-05, "loss": 0.9141, "step": 445 }, { "epoch": 0.20988235294117646, "grad_norm": 3.329870009738836, "learning_rate": 6.69e-05, "loss": 2.6406, "step": 446 }, { "epoch": 0.21035294117647058, "grad_norm": 1.529472378914684, "learning_rate": 6.704999999999999e-05, "loss": 0.9023, "step": 447 }, { "epoch": 0.21082352941176472, "grad_norm": 3.5680818672267214, "learning_rate": 6.72e-05, "loss": 0.9805, "step": 448 }, { "epoch": 0.21129411764705883, "grad_norm": 9.10742342212134, "learning_rate": 6.735e-05, "loss": 1.875, "step": 449 }, { "epoch": 0.21176470588235294, "grad_norm": 5.702529802037804, "learning_rate": 6.75e-05, "loss": 1.2891, "step": 450 }, { "epoch": 0.21223529411764705, "grad_norm": 1.620003860325722, "learning_rate": 6.764999999999999e-05, "loss": 0.75, "step": 451 }, { "epoch": 0.2127058823529412, "grad_norm": 3.4112311406509326, "learning_rate": 6.78e-05, "loss": 2.0469, "step": 452 }, { "epoch": 0.2131764705882353, "grad_norm": 1.456810035981253, "learning_rate": 6.795e-05, "loss": 0.5781, "step": 453 }, { "epoch": 0.2136470588235294, "grad_norm": 3.404681307446091, "learning_rate": 6.81e-05, "loss": 2.5781, "step": 454 }, { "epoch": 0.21411764705882352, "grad_norm": 9.858744033076427, "learning_rate": 6.824999999999999e-05, "loss": 1.625, "step": 455 }, { "epoch": 0.21458823529411764, "grad_norm": 2.665372015343191, "learning_rate": 6.84e-05, "loss": 2.125, "step": 456 }, { "epoch": 0.21505882352941177, "grad_norm": 1.7153374191519077, "learning_rate": 6.855e-05, "loss": 1.2969, "step": 457 }, { "epoch": 0.21552941176470589, "grad_norm": 1.261061282367109, "learning_rate": 6.87e-05, "loss": 0.957, "step": 458 }, { "epoch": 0.216, "grad_norm": 1.2580178692076815, "learning_rate": 6.884999999999999e-05, "loss": 1.0, "step": 459 }, { "epoch": 0.2164705882352941, "grad_norm": 4.961814191779544, "learning_rate": 6.9e-05, "loss": 3.3906, "step": 460 }, { "epoch": 0.21694117647058825, "grad_norm": 1.8034987692600177, "learning_rate": 6.915e-05, "loss": 1.3672, "step": 461 }, { "epoch": 0.21741176470588236, "grad_norm": 5.452664735679394, "learning_rate": 6.93e-05, "loss": 3.2188, "step": 462 }, { "epoch": 0.21788235294117647, "grad_norm": 2.941769128025957, "learning_rate": 6.945e-05, "loss": 2.9375, "step": 463 }, { "epoch": 0.21835294117647058, "grad_norm": 3.060018565495794, "learning_rate": 6.96e-05, "loss": 2.5312, "step": 464 }, { "epoch": 0.2188235294117647, "grad_norm": 0.9536954541727757, "learning_rate": 6.975e-05, "loss": 0.2197, "step": 465 }, { "epoch": 0.21929411764705883, "grad_norm": 3.517363712704531, "learning_rate": 6.989999999999999e-05, "loss": 2.25, "step": 466 }, { "epoch": 0.21976470588235295, "grad_norm": 2.2033182445727175, "learning_rate": 7.005e-05, "loss": 1.125, "step": 467 }, { "epoch": 0.22023529411764706, "grad_norm": 1.87864794265696, "learning_rate": 7.02e-05, "loss": 1.1016, "step": 468 }, { "epoch": 0.22070588235294117, "grad_norm": 4.005276035254549, "learning_rate": 7.034999999999999e-05, "loss": 2.0, "step": 469 }, { "epoch": 0.2211764705882353, "grad_norm": 4.969239812742473, "learning_rate": 7.049999999999999e-05, "loss": 2.4688, "step": 470 }, { "epoch": 0.22164705882352942, "grad_norm": 1.4228107741311549, "learning_rate": 7.065e-05, "loss": 1.2969, "step": 471 }, { "epoch": 0.22211764705882353, "grad_norm": 3.480388381796721, "learning_rate": 7.079999999999999e-05, "loss": 2.3906, "step": 472 }, { "epoch": 0.22258823529411764, "grad_norm": 3.1285057746854665, "learning_rate": 7.094999999999999e-05, "loss": 1.5781, "step": 473 }, { "epoch": 0.22305882352941175, "grad_norm": 4.735819272708921, "learning_rate": 7.11e-05, "loss": 3.4062, "step": 474 }, { "epoch": 0.2235294117647059, "grad_norm": 2.205147605633135, "learning_rate": 7.125e-05, "loss": 1.875, "step": 475 }, { "epoch": 0.224, "grad_norm": 2.153486862480025, "learning_rate": 7.139999999999999e-05, "loss": 1.0312, "step": 476 }, { "epoch": 0.22447058823529412, "grad_norm": 1.3039365093594875, "learning_rate": 7.154999999999999e-05, "loss": 0.5586, "step": 477 }, { "epoch": 0.22494117647058823, "grad_norm": 3.704321378412122, "learning_rate": 7.17e-05, "loss": 1.5234, "step": 478 }, { "epoch": 0.22541176470588234, "grad_norm": 2.1691707286701143, "learning_rate": 7.184999999999998e-05, "loss": 1.7344, "step": 479 }, { "epoch": 0.22588235294117648, "grad_norm": 2.8029159740378087, "learning_rate": 7.199999999999999e-05, "loss": 1.4531, "step": 480 }, { "epoch": 0.2263529411764706, "grad_norm": 3.2325307325262203, "learning_rate": 7.214999999999999e-05, "loss": 1.5078, "step": 481 }, { "epoch": 0.2268235294117647, "grad_norm": 2.2458938435478384, "learning_rate": 7.23e-05, "loss": 1.5312, "step": 482 }, { "epoch": 0.2272941176470588, "grad_norm": 1.982112800561155, "learning_rate": 7.244999999999999e-05, "loss": 1.2734, "step": 483 }, { "epoch": 0.22776470588235295, "grad_norm": 1.8667958268719982, "learning_rate": 7.259999999999999e-05, "loss": 1.0234, "step": 484 }, { "epoch": 0.22823529411764706, "grad_norm": 1.3811944400943856, "learning_rate": 7.274999999999999e-05, "loss": 0.7188, "step": 485 }, { "epoch": 0.22870588235294118, "grad_norm": 1.9381917785271925, "learning_rate": 7.29e-05, "loss": 1.2188, "step": 486 }, { "epoch": 0.2291764705882353, "grad_norm": 2.6488432351982922, "learning_rate": 7.304999999999999e-05, "loss": 2.2344, "step": 487 }, { "epoch": 0.2296470588235294, "grad_norm": 2.6226116376980895, "learning_rate": 7.319999999999999e-05, "loss": 1.6406, "step": 488 }, { "epoch": 0.23011764705882354, "grad_norm": 2.1301192088584555, "learning_rate": 7.335e-05, "loss": 1.5, "step": 489 }, { "epoch": 0.23058823529411765, "grad_norm": 2.2271928334079845, "learning_rate": 7.35e-05, "loss": 1.9453, "step": 490 }, { "epoch": 0.23105882352941176, "grad_norm": 1.6589520735039427, "learning_rate": 7.364999999999999e-05, "loss": 1.6797, "step": 491 }, { "epoch": 0.23152941176470587, "grad_norm": 1.8641356113897063, "learning_rate": 7.379999999999999e-05, "loss": 1.4609, "step": 492 }, { "epoch": 0.232, "grad_norm": 4.22899589379286, "learning_rate": 7.395e-05, "loss": 2.5625, "step": 493 }, { "epoch": 0.23247058823529412, "grad_norm": 1.440922635175555, "learning_rate": 7.41e-05, "loss": 0.9062, "step": 494 }, { "epoch": 0.23294117647058823, "grad_norm": 3.0133023840334103, "learning_rate": 7.424999999999999e-05, "loss": 2.7188, "step": 495 }, { "epoch": 0.23341176470588235, "grad_norm": 111.16126940847008, "learning_rate": 7.439999999999999e-05, "loss": 1.4062, "step": 496 }, { "epoch": 0.23388235294117646, "grad_norm": 3.625559016694518, "learning_rate": 7.455e-05, "loss": 1.4062, "step": 497 }, { "epoch": 0.2343529411764706, "grad_norm": 48.730411232095925, "learning_rate": 7.47e-05, "loss": 4.375, "step": 498 }, { "epoch": 0.2348235294117647, "grad_norm": 4.981768855734861, "learning_rate": 7.484999999999999e-05, "loss": 2.2188, "step": 499 }, { "epoch": 0.23529411764705882, "grad_norm": 3.932332825382061, "learning_rate": 7.5e-05, "loss": 1.0703, "step": 500 }, { "epoch": 0.23576470588235293, "grad_norm": 4.393851028779818, "learning_rate": 7.515e-05, "loss": 2.25, "step": 501 }, { "epoch": 0.23623529411764707, "grad_norm": 1.9143997493913207, "learning_rate": 7.529999999999999e-05, "loss": 1.1016, "step": 502 }, { "epoch": 0.23670588235294118, "grad_norm": 27.453675288199356, "learning_rate": 7.544999999999999e-05, "loss": 4.6875, "step": 503 }, { "epoch": 0.2371764705882353, "grad_norm": 0.755930185959564, "learning_rate": 7.56e-05, "loss": 0.4746, "step": 504 }, { "epoch": 0.2376470588235294, "grad_norm": 6.48819560399067, "learning_rate": 7.575e-05, "loss": 2.75, "step": 505 }, { "epoch": 0.23811764705882352, "grad_norm": 2.6083994870496148, "learning_rate": 7.589999999999999e-05, "loss": 1.6719, "step": 506 }, { "epoch": 0.23858823529411766, "grad_norm": 0.9001859600422626, "learning_rate": 7.604999999999999e-05, "loss": 0.377, "step": 507 }, { "epoch": 0.23905882352941177, "grad_norm": 2.954426067205615, "learning_rate": 7.62e-05, "loss": 1.6094, "step": 508 }, { "epoch": 0.23952941176470588, "grad_norm": 1.8274933219673835, "learning_rate": 7.635e-05, "loss": 0.5781, "step": 509 }, { "epoch": 0.24, "grad_norm": 3.5535959615111783, "learning_rate": 7.649999999999999e-05, "loss": 1.125, "step": 510 }, { "epoch": 0.24047058823529413, "grad_norm": 0.6253637258949111, "learning_rate": 7.664999999999999e-05, "loss": 0.1953, "step": 511 }, { "epoch": 0.24094117647058824, "grad_norm": 3.813459794041436, "learning_rate": 7.68e-05, "loss": 2.1562, "step": 512 }, { "epoch": 0.24141176470588235, "grad_norm": 2.822073623273197, "learning_rate": 7.695e-05, "loss": 2.4375, "step": 513 }, { "epoch": 0.24188235294117647, "grad_norm": 1.0501812908326404, "learning_rate": 7.709999999999999e-05, "loss": 0.291, "step": 514 }, { "epoch": 0.24235294117647058, "grad_norm": 181.67421375603786, "learning_rate": 7.725e-05, "loss": 2.2656, "step": 515 }, { "epoch": 0.24282352941176472, "grad_norm": 1.4273624389665847, "learning_rate": 7.74e-05, "loss": 0.8555, "step": 516 }, { "epoch": 0.24329411764705883, "grad_norm": 3.2254434151135327, "learning_rate": 7.755e-05, "loss": 2.875, "step": 517 }, { "epoch": 0.24376470588235294, "grad_norm": 1.2420209531536741, "learning_rate": 7.769999999999999e-05, "loss": 0.5156, "step": 518 }, { "epoch": 0.24423529411764705, "grad_norm": 2.162979501412172, "learning_rate": 7.785e-05, "loss": 1.375, "step": 519 }, { "epoch": 0.2447058823529412, "grad_norm": 1.907025236896015, "learning_rate": 7.8e-05, "loss": 1.2891, "step": 520 }, { "epoch": 0.2451764705882353, "grad_norm": 3.3143930293463266, "learning_rate": 7.815e-05, "loss": 2.0469, "step": 521 }, { "epoch": 0.2456470588235294, "grad_norm": 2.2672139746612334, "learning_rate": 7.829999999999999e-05, "loss": 1.0469, "step": 522 }, { "epoch": 0.24611764705882352, "grad_norm": 2.2488005916953773, "learning_rate": 7.845e-05, "loss": 0.8164, "step": 523 }, { "epoch": 0.24658823529411764, "grad_norm": 3.1919009200307618, "learning_rate": 7.86e-05, "loss": 1.7344, "step": 524 }, { "epoch": 0.24705882352941178, "grad_norm": 3.5323252582334983, "learning_rate": 7.874999999999999e-05, "loss": 2.0156, "step": 525 }, { "epoch": 0.2475294117647059, "grad_norm": 2.1751308079555463, "learning_rate": 7.89e-05, "loss": 1.2734, "step": 526 }, { "epoch": 0.248, "grad_norm": 2.4028408992577908, "learning_rate": 7.905e-05, "loss": 1.8359, "step": 527 }, { "epoch": 0.2484705882352941, "grad_norm": 2.0230987932163855, "learning_rate": 7.92e-05, "loss": 1.5391, "step": 528 }, { "epoch": 0.24894117647058825, "grad_norm": 1.9937565050688877, "learning_rate": 7.934999999999999e-05, "loss": 0.9492, "step": 529 }, { "epoch": 0.24941176470588236, "grad_norm": 1.4876540089512416, "learning_rate": 7.95e-05, "loss": 0.7344, "step": 530 }, { "epoch": 0.24988235294117647, "grad_norm": 3.6726611095786486, "learning_rate": 7.965e-05, "loss": 2.125, "step": 531 }, { "epoch": 0.2503529411764706, "grad_norm": 1.601599634289717, "learning_rate": 7.98e-05, "loss": 1.0078, "step": 532 }, { "epoch": 0.2508235294117647, "grad_norm": 2.261018826960819, "learning_rate": 7.994999999999999e-05, "loss": 1.6641, "step": 533 }, { "epoch": 0.25129411764705883, "grad_norm": 1.9991774027011044, "learning_rate": 8.01e-05, "loss": 0.9883, "step": 534 }, { "epoch": 0.25176470588235295, "grad_norm": 2.5010341487987997, "learning_rate": 8.025e-05, "loss": 2.0156, "step": 535 }, { "epoch": 0.25223529411764706, "grad_norm": 1.7672314631520272, "learning_rate": 8.04e-05, "loss": 1.4297, "step": 536 }, { "epoch": 0.25270588235294117, "grad_norm": 2.0013221059362074, "learning_rate": 8.054999999999999e-05, "loss": 1.6484, "step": 537 }, { "epoch": 0.2531764705882353, "grad_norm": 3.1488636862997237, "learning_rate": 8.07e-05, "loss": 3.0938, "step": 538 }, { "epoch": 0.2536470588235294, "grad_norm": 1.5581578603054778, "learning_rate": 8.085e-05, "loss": 1.0703, "step": 539 }, { "epoch": 0.2541176470588235, "grad_norm": 26.522290245699402, "learning_rate": 8.1e-05, "loss": 1.3047, "step": 540 }, { "epoch": 0.25458823529411767, "grad_norm": 1.794852381581537, "learning_rate": 8.115e-05, "loss": 0.9648, "step": 541 }, { "epoch": 0.2550588235294118, "grad_norm": 2.362466776132806, "learning_rate": 8.13e-05, "loss": 2.375, "step": 542 }, { "epoch": 0.2555294117647059, "grad_norm": 2.1758705560663105, "learning_rate": 8.145e-05, "loss": 1.7578, "step": 543 }, { "epoch": 0.256, "grad_norm": 0.9620594101183405, "learning_rate": 8.16e-05, "loss": 0.5039, "step": 544 }, { "epoch": 0.2564705882352941, "grad_norm": 103.68949465946207, "learning_rate": 8.175e-05, "loss": 3.1406, "step": 545 }, { "epoch": 0.25694117647058823, "grad_norm": 3.3587971071764913, "learning_rate": 8.19e-05, "loss": 2.5781, "step": 546 }, { "epoch": 0.25741176470588234, "grad_norm": 1.3593451640209793, "learning_rate": 8.205e-05, "loss": 0.6367, "step": 547 }, { "epoch": 0.25788235294117645, "grad_norm": 92.17932877302165, "learning_rate": 8.22e-05, "loss": 3.5625, "step": 548 }, { "epoch": 0.25835294117647056, "grad_norm": 2.458204091067131, "learning_rate": 8.235e-05, "loss": 1.7422, "step": 549 }, { "epoch": 0.25882352941176473, "grad_norm": 1.9323226011154897, "learning_rate": 8.25e-05, "loss": 1.2344, "step": 550 }, { "epoch": 0.25929411764705884, "grad_norm": 3.2025804387100094, "learning_rate": 8.265e-05, "loss": 2.5625, "step": 551 }, { "epoch": 0.25976470588235295, "grad_norm": 2.618120503477108, "learning_rate": 8.28e-05, "loss": 1.8438, "step": 552 }, { "epoch": 0.26023529411764706, "grad_norm": 16.243679019861844, "learning_rate": 8.295e-05, "loss": 6.3438, "step": 553 }, { "epoch": 0.2607058823529412, "grad_norm": 17.1299087919568, "learning_rate": 8.31e-05, "loss": 2.8594, "step": 554 }, { "epoch": 0.2611764705882353, "grad_norm": 45.01891686900135, "learning_rate": 8.325e-05, "loss": 6.4062, "step": 555 }, { "epoch": 0.2616470588235294, "grad_norm": 29.35214717553288, "learning_rate": 8.34e-05, "loss": 6.3438, "step": 556 }, { "epoch": 0.2621176470588235, "grad_norm": 4.54055244606022, "learning_rate": 8.355e-05, "loss": 1.3438, "step": 557 }, { "epoch": 0.2625882352941176, "grad_norm": 2.543850586714649, "learning_rate": 8.37e-05, "loss": 1.2031, "step": 558 }, { "epoch": 0.2630588235294118, "grad_norm": 2.083429636503197, "learning_rate": 8.385e-05, "loss": 1.125, "step": 559 }, { "epoch": 0.2635294117647059, "grad_norm": 8.60766771025749, "learning_rate": 8.4e-05, "loss": 2.875, "step": 560 }, { "epoch": 0.264, "grad_norm": 2.3752431914454433, "learning_rate": 8.415e-05, "loss": 1.9844, "step": 561 }, { "epoch": 0.2644705882352941, "grad_norm": 2.6479026000906747, "learning_rate": 8.43e-05, "loss": 1.6016, "step": 562 }, { "epoch": 0.26494117647058824, "grad_norm": 3.03879228726528, "learning_rate": 8.444999999999998e-05, "loss": 2.0781, "step": 563 }, { "epoch": 0.26541176470588235, "grad_norm": 2.2605736324453356, "learning_rate": 8.459999999999998e-05, "loss": 1.2734, "step": 564 }, { "epoch": 0.26588235294117646, "grad_norm": 3.272088940723119, "learning_rate": 8.474999999999999e-05, "loss": 1.875, "step": 565 }, { "epoch": 0.26635294117647057, "grad_norm": 2.3700806027162096, "learning_rate": 8.489999999999999e-05, "loss": 1.7969, "step": 566 }, { "epoch": 0.2668235294117647, "grad_norm": 0.9162963764165041, "learning_rate": 8.504999999999998e-05, "loss": 0.5664, "step": 567 }, { "epoch": 0.26729411764705885, "grad_norm": 2.301503881243715, "learning_rate": 8.519999999999998e-05, "loss": 1.2344, "step": 568 }, { "epoch": 0.26776470588235296, "grad_norm": 1.2354877485025753, "learning_rate": 8.534999999999999e-05, "loss": 0.7539, "step": 569 }, { "epoch": 0.26823529411764707, "grad_norm": 1.696443232635525, "learning_rate": 8.549999999999999e-05, "loss": 0.9141, "step": 570 }, { "epoch": 0.2687058823529412, "grad_norm": 2.405682663748704, "learning_rate": 8.564999999999998e-05, "loss": 1.5391, "step": 571 }, { "epoch": 0.2691764705882353, "grad_norm": 3.333111237059073, "learning_rate": 8.579999999999998e-05, "loss": 1.7266, "step": 572 }, { "epoch": 0.2696470588235294, "grad_norm": 1.690834145638598, "learning_rate": 8.594999999999999e-05, "loss": 1.3828, "step": 573 }, { "epoch": 0.2701176470588235, "grad_norm": 6.484441389780814, "learning_rate": 8.609999999999999e-05, "loss": 1.0625, "step": 574 }, { "epoch": 0.27058823529411763, "grad_norm": 3.3676915378265346, "learning_rate": 8.624999999999998e-05, "loss": 0.9961, "step": 575 }, { "epoch": 0.27105882352941174, "grad_norm": 2.3434606180890762, "learning_rate": 8.639999999999999e-05, "loss": 1.3047, "step": 576 }, { "epoch": 0.2715294117647059, "grad_norm": 2.909547825610969, "learning_rate": 8.654999999999999e-05, "loss": 2.5938, "step": 577 }, { "epoch": 0.272, "grad_norm": 1.5890226426838863, "learning_rate": 8.669999999999998e-05, "loss": 1.5234, "step": 578 }, { "epoch": 0.27247058823529413, "grad_norm": 6.2894118959938075, "learning_rate": 8.684999999999998e-05, "loss": 1.6328, "step": 579 }, { "epoch": 0.27294117647058824, "grad_norm": 3.9661637416206874, "learning_rate": 8.699999999999999e-05, "loss": 3.1875, "step": 580 }, { "epoch": 0.27341176470588235, "grad_norm": 1.430961854777645, "learning_rate": 8.714999999999999e-05, "loss": 1.3906, "step": 581 }, { "epoch": 0.27388235294117647, "grad_norm": 3.3531848006972576, "learning_rate": 8.729999999999998e-05, "loss": 2.6562, "step": 582 }, { "epoch": 0.2743529411764706, "grad_norm": 2.5310513816764946, "learning_rate": 8.744999999999998e-05, "loss": 2.2969, "step": 583 }, { "epoch": 0.2748235294117647, "grad_norm": 2.5983076904907163, "learning_rate": 8.759999999999999e-05, "loss": 2.1562, "step": 584 }, { "epoch": 0.2752941176470588, "grad_norm": 1.8870400725344858, "learning_rate": 8.774999999999999e-05, "loss": 1.4609, "step": 585 }, { "epoch": 0.27576470588235297, "grad_norm": 1.5358327173629889, "learning_rate": 8.789999999999998e-05, "loss": 0.9609, "step": 586 }, { "epoch": 0.2762352941176471, "grad_norm": 2.733976183568517, "learning_rate": 8.804999999999999e-05, "loss": 1.0859, "step": 587 }, { "epoch": 0.2767058823529412, "grad_norm": 1.18705320525025, "learning_rate": 8.819999999999999e-05, "loss": 0.7812, "step": 588 }, { "epoch": 0.2771764705882353, "grad_norm": 1.106893697701064, "learning_rate": 8.834999999999999e-05, "loss": 0.6562, "step": 589 }, { "epoch": 0.2776470588235294, "grad_norm": 1.330711694963248, "learning_rate": 8.849999999999998e-05, "loss": 0.9453, "step": 590 }, { "epoch": 0.2781176470588235, "grad_norm": 1.2547913765148508, "learning_rate": 8.864999999999999e-05, "loss": 0.4609, "step": 591 }, { "epoch": 0.27858823529411764, "grad_norm": 2.1963489207108595, "learning_rate": 8.879999999999999e-05, "loss": 2.0156, "step": 592 }, { "epoch": 0.27905882352941175, "grad_norm": 0.8336311301014371, "learning_rate": 8.895e-05, "loss": 0.6758, "step": 593 }, { "epoch": 0.27952941176470586, "grad_norm": 1.9072375278353655, "learning_rate": 8.909999999999998e-05, "loss": 1.6875, "step": 594 }, { "epoch": 0.28, "grad_norm": 3.3201108102614176, "learning_rate": 8.924999999999999e-05, "loss": 3.3594, "step": 595 }, { "epoch": 0.28047058823529414, "grad_norm": 2.040487691462828, "learning_rate": 8.939999999999999e-05, "loss": 2.25, "step": 596 }, { "epoch": 0.28094117647058825, "grad_norm": 2.3038952564765878, "learning_rate": 8.955e-05, "loss": 1.875, "step": 597 }, { "epoch": 0.28141176470588236, "grad_norm": 1.3033060606979125, "learning_rate": 8.969999999999998e-05, "loss": 0.7461, "step": 598 }, { "epoch": 0.2818823529411765, "grad_norm": 1.7085280074754914, "learning_rate": 8.984999999999999e-05, "loss": 1.3984, "step": 599 }, { "epoch": 0.2823529411764706, "grad_norm": 2.0163022249921956, "learning_rate": 8.999999999999999e-05, "loss": 1.0703, "step": 600 }, { "epoch": 0.2828235294117647, "grad_norm": 1.408284538756869, "learning_rate": 9.014999999999998e-05, "loss": 0.9805, "step": 601 }, { "epoch": 0.2832941176470588, "grad_norm": 1.518048478664724, "learning_rate": 9.029999999999999e-05, "loss": 1.125, "step": 602 }, { "epoch": 0.2837647058823529, "grad_norm": 1.7793868156683836, "learning_rate": 9.044999999999999e-05, "loss": 1.0703, "step": 603 }, { "epoch": 0.2842352941176471, "grad_norm": 1.9180575820394137, "learning_rate": 9.059999999999999e-05, "loss": 0.875, "step": 604 }, { "epoch": 0.2847058823529412, "grad_norm": 2.6570161211020435, "learning_rate": 9.074999999999998e-05, "loss": 2.7969, "step": 605 }, { "epoch": 0.2851764705882353, "grad_norm": 2.9518663071048583, "learning_rate": 9.089999999999999e-05, "loss": 3.0156, "step": 606 }, { "epoch": 0.2856470588235294, "grad_norm": 0.9290282051979258, "learning_rate": 9.104999999999999e-05, "loss": 0.4258, "step": 607 }, { "epoch": 0.28611764705882353, "grad_norm": 9.68893673455589, "learning_rate": 9.12e-05, "loss": 1.4609, "step": 608 }, { "epoch": 0.28658823529411764, "grad_norm": 1.6905896570117958, "learning_rate": 9.134999999999998e-05, "loss": 1.2422, "step": 609 }, { "epoch": 0.28705882352941176, "grad_norm": 2.2158184969838413, "learning_rate": 9.149999999999999e-05, "loss": 2.5938, "step": 610 }, { "epoch": 0.28752941176470587, "grad_norm": 4.270232360809073, "learning_rate": 9.164999999999999e-05, "loss": 1.1484, "step": 611 }, { "epoch": 0.288, "grad_norm": 1.5633571817708447, "learning_rate": 9.18e-05, "loss": 1.3438, "step": 612 }, { "epoch": 0.28847058823529415, "grad_norm": 1.8087105614328087, "learning_rate": 9.194999999999999e-05, "loss": 1.7734, "step": 613 }, { "epoch": 0.28894117647058826, "grad_norm": 1.9532982184025598, "learning_rate": 9.209999999999999e-05, "loss": 2.8906, "step": 614 }, { "epoch": 0.28941176470588237, "grad_norm": 1.2218141020023534, "learning_rate": 9.224999999999999e-05, "loss": 0.5469, "step": 615 }, { "epoch": 0.2898823529411765, "grad_norm": 1.9010206455607002, "learning_rate": 9.24e-05, "loss": 1.5938, "step": 616 }, { "epoch": 0.2903529411764706, "grad_norm": 1.0885086573149796, "learning_rate": 9.254999999999999e-05, "loss": 0.543, "step": 617 }, { "epoch": 0.2908235294117647, "grad_norm": 1.9333910091244086, "learning_rate": 9.269999999999999e-05, "loss": 1.8281, "step": 618 }, { "epoch": 0.2912941176470588, "grad_norm": 1.6687377732315916, "learning_rate": 9.285e-05, "loss": 1.3281, "step": 619 }, { "epoch": 0.2917647058823529, "grad_norm": 1.2340564243637673, "learning_rate": 9.3e-05, "loss": 0.9531, "step": 620 }, { "epoch": 0.29223529411764704, "grad_norm": 2.3973145353027085, "learning_rate": 9.314999999999999e-05, "loss": 2.1406, "step": 621 }, { "epoch": 0.29270588235294115, "grad_norm": 1.7949536739447634, "learning_rate": 9.329999999999999e-05, "loss": 1.5625, "step": 622 }, { "epoch": 0.2931764705882353, "grad_norm": 1.9088257680070686, "learning_rate": 9.345e-05, "loss": 1.6641, "step": 623 }, { "epoch": 0.2936470588235294, "grad_norm": 1.0310297606908116, "learning_rate": 9.36e-05, "loss": 0.6445, "step": 624 }, { "epoch": 0.29411764705882354, "grad_norm": 2.230662129267793, "learning_rate": 9.374999999999999e-05, "loss": 2.1406, "step": 625 }, { "epoch": 0.29458823529411765, "grad_norm": 1.8308397017548128, "learning_rate": 9.389999999999999e-05, "loss": 1.7812, "step": 626 }, { "epoch": 0.29505882352941176, "grad_norm": 1.651239736635819, "learning_rate": 9.405e-05, "loss": 1.8359, "step": 627 }, { "epoch": 0.2955294117647059, "grad_norm": 1.782736438639837, "learning_rate": 9.419999999999999e-05, "loss": 2.0, "step": 628 }, { "epoch": 0.296, "grad_norm": 1.0711446808020266, "learning_rate": 9.434999999999999e-05, "loss": 0.6055, "step": 629 }, { "epoch": 0.2964705882352941, "grad_norm": 1.1920238793059108, "learning_rate": 9.449999999999999e-05, "loss": 0.8203, "step": 630 }, { "epoch": 0.2969411764705882, "grad_norm": 1.7626572651890473, "learning_rate": 9.465e-05, "loss": 1.6328, "step": 631 }, { "epoch": 0.2974117647058824, "grad_norm": 0.5950751620669109, "learning_rate": 9.479999999999999e-05, "loss": 0.1582, "step": 632 }, { "epoch": 0.2978823529411765, "grad_norm": 1.2056396909266904, "learning_rate": 9.494999999999999e-05, "loss": 0.6641, "step": 633 }, { "epoch": 0.2983529411764706, "grad_norm": 3.010528630948196, "learning_rate": 9.51e-05, "loss": 1.5469, "step": 634 }, { "epoch": 0.2988235294117647, "grad_norm": 1.7972856720915877, "learning_rate": 9.525e-05, "loss": 2.0, "step": 635 }, { "epoch": 0.2992941176470588, "grad_norm": 2.0593576388571346, "learning_rate": 9.539999999999999e-05, "loss": 1.9844, "step": 636 }, { "epoch": 0.29976470588235293, "grad_norm": 2.5787770902683227, "learning_rate": 9.554999999999999e-05, "loss": 1.7188, "step": 637 }, { "epoch": 0.30023529411764704, "grad_norm": 1.5471963017820547, "learning_rate": 9.57e-05, "loss": 0.9961, "step": 638 }, { "epoch": 0.30070588235294116, "grad_norm": 1.7864972030252086, "learning_rate": 9.585e-05, "loss": 1.3984, "step": 639 }, { "epoch": 0.30117647058823527, "grad_norm": 1.8664342271895902, "learning_rate": 9.599999999999999e-05, "loss": 1.3203, "step": 640 }, { "epoch": 0.30164705882352943, "grad_norm": 0.7721506747500649, "learning_rate": 9.614999999999999e-05, "loss": 0.3242, "step": 641 }, { "epoch": 0.30211764705882355, "grad_norm": 1.749641552751789, "learning_rate": 9.63e-05, "loss": 1.5625, "step": 642 }, { "epoch": 0.30258823529411766, "grad_norm": 1.9306597472639764, "learning_rate": 9.645e-05, "loss": 1.6328, "step": 643 }, { "epoch": 0.30305882352941177, "grad_norm": 1.0413396281215395, "learning_rate": 9.659999999999999e-05, "loss": 0.6484, "step": 644 }, { "epoch": 0.3035294117647059, "grad_norm": 1.8369582382524496, "learning_rate": 9.675e-05, "loss": 1.75, "step": 645 }, { "epoch": 0.304, "grad_norm": 1.6388030495341146, "learning_rate": 9.69e-05, "loss": 0.8516, "step": 646 }, { "epoch": 0.3044705882352941, "grad_norm": 1.5762249409737936, "learning_rate": 9.705e-05, "loss": 2.3281, "step": 647 }, { "epoch": 0.3049411764705882, "grad_norm": 2.0239119924913758, "learning_rate": 9.719999999999999e-05, "loss": 2.3594, "step": 648 }, { "epoch": 0.3054117647058823, "grad_norm": 1.0237050486026555, "learning_rate": 9.735e-05, "loss": 0.4434, "step": 649 }, { "epoch": 0.3058823529411765, "grad_norm": 1.5679051970167484, "learning_rate": 9.75e-05, "loss": 1.2969, "step": 650 }, { "epoch": 0.3063529411764706, "grad_norm": 0.6952814812963671, "learning_rate": 9.764999999999999e-05, "loss": 0.3203, "step": 651 }, { "epoch": 0.3068235294117647, "grad_norm": 2.2155770304030855, "learning_rate": 9.779999999999999e-05, "loss": 2.0938, "step": 652 }, { "epoch": 0.30729411764705883, "grad_norm": 0.9713755766118942, "learning_rate": 9.795e-05, "loss": 0.4102, "step": 653 }, { "epoch": 0.30776470588235294, "grad_norm": 1.4109634563112994, "learning_rate": 9.81e-05, "loss": 1.3984, "step": 654 }, { "epoch": 0.30823529411764705, "grad_norm": 1.3402644784936832, "learning_rate": 9.824999999999999e-05, "loss": 0.8359, "step": 655 }, { "epoch": 0.30870588235294116, "grad_norm": 1.8957197576982, "learning_rate": 9.839999999999999e-05, "loss": 1.5547, "step": 656 }, { "epoch": 0.3091764705882353, "grad_norm": 2.1100893383852664, "learning_rate": 9.855e-05, "loss": 2.125, "step": 657 }, { "epoch": 0.3096470588235294, "grad_norm": 0.7400181344845435, "learning_rate": 9.87e-05, "loss": 0.2949, "step": 658 }, { "epoch": 0.31011764705882355, "grad_norm": 2.666462742965505, "learning_rate": 9.884999999999999e-05, "loss": 2.1562, "step": 659 }, { "epoch": 0.31058823529411766, "grad_norm": 2.3448374848275635, "learning_rate": 9.9e-05, "loss": 2.0312, "step": 660 }, { "epoch": 0.3110588235294118, "grad_norm": 83.52593034250613, "learning_rate": 9.915e-05, "loss": 1.5234, "step": 661 }, { "epoch": 0.3115294117647059, "grad_norm": 27.554070209748655, "learning_rate": 9.93e-05, "loss": 2.0156, "step": 662 }, { "epoch": 0.312, "grad_norm": 1.5020976154516583, "learning_rate": 9.944999999999999e-05, "loss": 1.25, "step": 663 }, { "epoch": 0.3124705882352941, "grad_norm": 1.8969877384931964, "learning_rate": 9.96e-05, "loss": 1.7578, "step": 664 }, { "epoch": 0.3129411764705882, "grad_norm": 2.020062933411951, "learning_rate": 9.975e-05, "loss": 1.8438, "step": 665 }, { "epoch": 0.31341176470588233, "grad_norm": 23.60957063431457, "learning_rate": 9.99e-05, "loss": 2.1094, "step": 666 }, { "epoch": 0.31388235294117645, "grad_norm": 4.779585905428112, "learning_rate": 0.00010004999999999999, "loss": 1.6641, "step": 667 }, { "epoch": 0.3143529411764706, "grad_norm": 6.400117013595347, "learning_rate": 0.0001002, "loss": 1.6719, "step": 668 }, { "epoch": 0.3148235294117647, "grad_norm": 3.1065389398248047, "learning_rate": 0.00010035, "loss": 1.1875, "step": 669 }, { "epoch": 0.31529411764705884, "grad_norm": 4.748728095083961, "learning_rate": 0.0001005, "loss": 2.9062, "step": 670 }, { "epoch": 0.31576470588235295, "grad_norm": 8.129867504534374, "learning_rate": 0.00010065, "loss": 2.2188, "step": 671 }, { "epoch": 0.31623529411764706, "grad_norm": 4.85004391161815, "learning_rate": 0.0001008, "loss": 1.0625, "step": 672 }, { "epoch": 0.31670588235294117, "grad_norm": 2.5221127626295146, "learning_rate": 0.00010095, "loss": 1.6875, "step": 673 }, { "epoch": 0.3171764705882353, "grad_norm": 3.981005808019661, "learning_rate": 0.0001011, "loss": 2.2188, "step": 674 }, { "epoch": 0.3176470588235294, "grad_norm": 5.415354261532626, "learning_rate": 0.00010125, "loss": 3.0469, "step": 675 }, { "epoch": 0.3181176470588235, "grad_norm": 1.7983349383636076, "learning_rate": 0.0001014, "loss": 0.9531, "step": 676 }, { "epoch": 0.31858823529411767, "grad_norm": 382.0838755020059, "learning_rate": 0.00010155, "loss": 1.7344, "step": 677 }, { "epoch": 0.3190588235294118, "grad_norm": 6.650710040629383, "learning_rate": 0.00010169999999999999, "loss": 2.5781, "step": 678 }, { "epoch": 0.3195294117647059, "grad_norm": 2.962934933980561, "learning_rate": 0.00010185, "loss": 2.7031, "step": 679 }, { "epoch": 0.32, "grad_norm": 22.70614745723712, "learning_rate": 0.000102, "loss": 21.125, "step": 680 }, { "epoch": 0.3204705882352941, "grad_norm": 3.5490334790879503, "learning_rate": 0.00010215, "loss": 1.7969, "step": 681 }, { "epoch": 0.32094117647058823, "grad_norm": 17.287259951122284, "learning_rate": 0.00010229999999999999, "loss": 3.125, "step": 682 }, { "epoch": 0.32141176470588234, "grad_norm": 25.52773975124387, "learning_rate": 0.00010245, "loss": 2.5, "step": 683 }, { "epoch": 0.32188235294117645, "grad_norm": 24.794134281399703, "learning_rate": 0.0001026, "loss": 8.3125, "step": 684 }, { "epoch": 0.32235294117647056, "grad_norm": 2.846871881007648, "learning_rate": 0.00010275, "loss": 0.8203, "step": 685 }, { "epoch": 0.32282352941176473, "grad_norm": 8.00260958904032, "learning_rate": 0.0001029, "loss": 2.8594, "step": 686 }, { "epoch": 0.32329411764705884, "grad_norm": 25.110534402437857, "learning_rate": 0.00010305, "loss": 3.9688, "step": 687 }, { "epoch": 0.32376470588235295, "grad_norm": 7.0289532022299355, "learning_rate": 0.00010319999999999999, "loss": 2.25, "step": 688 }, { "epoch": 0.32423529411764707, "grad_norm": 12.925449339080554, "learning_rate": 0.00010334999999999998, "loss": 3.4844, "step": 689 }, { "epoch": 0.3247058823529412, "grad_norm": 5.407693492336986, "learning_rate": 0.00010349999999999998, "loss": 1.3047, "step": 690 }, { "epoch": 0.3251764705882353, "grad_norm": 3.9192093709874145, "learning_rate": 0.00010364999999999999, "loss": 1.9062, "step": 691 }, { "epoch": 0.3256470588235294, "grad_norm": 10.885077270776879, "learning_rate": 0.00010379999999999999, "loss": 2.4375, "step": 692 }, { "epoch": 0.3261176470588235, "grad_norm": 4.869729748558815, "learning_rate": 0.00010394999999999998, "loss": 4.3438, "step": 693 }, { "epoch": 0.3265882352941176, "grad_norm": 3.677771438919913, "learning_rate": 0.00010409999999999998, "loss": 2.3906, "step": 694 }, { "epoch": 0.3270588235294118, "grad_norm": 6.967503896005864, "learning_rate": 0.00010424999999999999, "loss": 1.8984, "step": 695 }, { "epoch": 0.3275294117647059, "grad_norm": 4.60215160612489, "learning_rate": 0.00010439999999999999, "loss": 3.3125, "step": 696 }, { "epoch": 0.328, "grad_norm": 6.393591706257614, "learning_rate": 0.00010454999999999998, "loss": 1.3359, "step": 697 }, { "epoch": 0.3284705882352941, "grad_norm": 3.2789987796087696, "learning_rate": 0.00010469999999999998, "loss": 2.7344, "step": 698 }, { "epoch": 0.32894117647058824, "grad_norm": 3.724362303917412, "learning_rate": 0.00010484999999999999, "loss": 2.8906, "step": 699 }, { "epoch": 0.32941176470588235, "grad_norm": 3.785399461786648, "learning_rate": 0.00010499999999999999, "loss": 2.0156, "step": 700 }, { "epoch": 0.32988235294117646, "grad_norm": 4.220984283660478, "learning_rate": 0.00010514999999999998, "loss": 2.9219, "step": 701 }, { "epoch": 0.33035294117647057, "grad_norm": 5.6645246671320875, "learning_rate": 0.00010529999999999998, "loss": 2.25, "step": 702 }, { "epoch": 0.3308235294117647, "grad_norm": 14.378941960734775, "learning_rate": 0.00010544999999999999, "loss": 2.6875, "step": 703 }, { "epoch": 0.33129411764705885, "grad_norm": 3.286212448895768, "learning_rate": 0.00010559999999999998, "loss": 2.7969, "step": 704 }, { "epoch": 0.33176470588235296, "grad_norm": 9.226359293877424, "learning_rate": 0.00010574999999999998, "loss": 2.1562, "step": 705 }, { "epoch": 0.3322352941176471, "grad_norm": 5.426822268167521, "learning_rate": 0.00010589999999999999, "loss": 1.6016, "step": 706 }, { "epoch": 0.3327058823529412, "grad_norm": 3.5739161821574172, "learning_rate": 0.00010604999999999999, "loss": 1.8359, "step": 707 }, { "epoch": 0.3331764705882353, "grad_norm": 2.9317023910099365, "learning_rate": 0.00010619999999999998, "loss": 2.3281, "step": 708 }, { "epoch": 0.3336470588235294, "grad_norm": 7.52732908418385, "learning_rate": 0.00010634999999999998, "loss": 2.3906, "step": 709 }, { "epoch": 0.3341176470588235, "grad_norm": 2.5387616562926456, "learning_rate": 0.00010649999999999999, "loss": 2.7969, "step": 710 }, { "epoch": 0.33458823529411763, "grad_norm": 3.3281567152069695, "learning_rate": 0.00010664999999999999, "loss": 1.4141, "step": 711 }, { "epoch": 0.33505882352941174, "grad_norm": 8.319934634256212, "learning_rate": 0.00010679999999999998, "loss": 2.125, "step": 712 }, { "epoch": 0.3355294117647059, "grad_norm": 8.375175241002621, "learning_rate": 0.00010694999999999998, "loss": 2.3281, "step": 713 }, { "epoch": 0.336, "grad_norm": 4.105161643755107, "learning_rate": 0.00010709999999999999, "loss": 2.25, "step": 714 }, { "epoch": 0.33647058823529413, "grad_norm": 2.3377832595888477, "learning_rate": 0.00010724999999999999, "loss": 2.2656, "step": 715 }, { "epoch": 0.33694117647058824, "grad_norm": 6.86639886735363, "learning_rate": 0.00010739999999999998, "loss": 3.4375, "step": 716 }, { "epoch": 0.33741176470588236, "grad_norm": 7.13645746488047, "learning_rate": 0.00010754999999999999, "loss": 1.2109, "step": 717 }, { "epoch": 0.33788235294117647, "grad_norm": 5.666506887086987, "learning_rate": 0.00010769999999999999, "loss": 1.0234, "step": 718 }, { "epoch": 0.3383529411764706, "grad_norm": 5.135557491051689, "learning_rate": 0.00010784999999999999, "loss": 2.5156, "step": 719 }, { "epoch": 0.3388235294117647, "grad_norm": 1.4559625635244633, "learning_rate": 0.00010799999999999998, "loss": 1.5156, "step": 720 }, { "epoch": 0.3392941176470588, "grad_norm": 3.333094031770154, "learning_rate": 0.00010814999999999999, "loss": 2.8906, "step": 721 }, { "epoch": 0.33976470588235297, "grad_norm": 5.361745665430523, "learning_rate": 0.00010829999999999999, "loss": 2.0156, "step": 722 }, { "epoch": 0.3402352941176471, "grad_norm": 4.510507270890005, "learning_rate": 0.00010845, "loss": 1.3203, "step": 723 }, { "epoch": 0.3407058823529412, "grad_norm": 3.617283056933018, "learning_rate": 0.00010859999999999998, "loss": 2.5469, "step": 724 }, { "epoch": 0.3411764705882353, "grad_norm": 5.177574172936291, "learning_rate": 0.00010874999999999999, "loss": 1.2734, "step": 725 }, { "epoch": 0.3416470588235294, "grad_norm": 1.0621575649315953, "learning_rate": 0.00010889999999999999, "loss": 0.6445, "step": 726 }, { "epoch": 0.3421176470588235, "grad_norm": 2.9676708434425914, "learning_rate": 0.00010904999999999998, "loss": 2.8906, "step": 727 }, { "epoch": 0.34258823529411764, "grad_norm": 1.745595124304583, "learning_rate": 0.00010919999999999998, "loss": 0.832, "step": 728 }, { "epoch": 0.34305882352941175, "grad_norm": 1.74302687704269, "learning_rate": 0.00010934999999999999, "loss": 0.7656, "step": 729 }, { "epoch": 0.34352941176470586, "grad_norm": 6.355215244175679, "learning_rate": 0.00010949999999999999, "loss": 1.2891, "step": 730 }, { "epoch": 0.344, "grad_norm": 5.882217248450229, "learning_rate": 0.00010964999999999998, "loss": 1.9062, "step": 731 }, { "epoch": 0.34447058823529414, "grad_norm": 1.3752895972221713, "learning_rate": 0.00010979999999999999, "loss": 1.3281, "step": 732 }, { "epoch": 0.34494117647058825, "grad_norm": 10.507061889640521, "learning_rate": 0.00010994999999999999, "loss": 2.6719, "step": 733 }, { "epoch": 0.34541176470588236, "grad_norm": 4.717629256825174, "learning_rate": 0.00011009999999999999, "loss": 0.9609, "step": 734 }, { "epoch": 0.3458823529411765, "grad_norm": 7.923176685621176, "learning_rate": 0.00011024999999999998, "loss": 3.1719, "step": 735 }, { "epoch": 0.3463529411764706, "grad_norm": 12.349399777092989, "learning_rate": 0.00011039999999999999, "loss": 3.2344, "step": 736 }, { "epoch": 0.3468235294117647, "grad_norm": 9.712389876392386, "learning_rate": 0.00011054999999999999, "loss": 1.5547, "step": 737 }, { "epoch": 0.3472941176470588, "grad_norm": 3.237973343336742, "learning_rate": 0.0001107, "loss": 1.3047, "step": 738 }, { "epoch": 0.3477647058823529, "grad_norm": 7.1225984234498245, "learning_rate": 0.00011084999999999998, "loss": 2.2656, "step": 739 }, { "epoch": 0.34823529411764703, "grad_norm": 5.661624298914657, "learning_rate": 0.00011099999999999999, "loss": 1.7344, "step": 740 }, { "epoch": 0.3487058823529412, "grad_norm": 5.083656879080953, "learning_rate": 0.00011114999999999999, "loss": 2.1719, "step": 741 }, { "epoch": 0.3491764705882353, "grad_norm": 4.43668913746894, "learning_rate": 0.0001113, "loss": 3.7812, "step": 742 }, { "epoch": 0.3496470588235294, "grad_norm": 8.629463415616101, "learning_rate": 0.00011144999999999998, "loss": 0.6719, "step": 743 }, { "epoch": 0.35011764705882353, "grad_norm": 2.253843555577398, "learning_rate": 0.00011159999999999999, "loss": 1.5, "step": 744 }, { "epoch": 0.35058823529411764, "grad_norm": 4.500671635880171, "learning_rate": 0.00011174999999999999, "loss": 1.7266, "step": 745 }, { "epoch": 0.35105882352941176, "grad_norm": 2.992209849311298, "learning_rate": 0.0001119, "loss": 2.2656, "step": 746 }, { "epoch": 0.35152941176470587, "grad_norm": 4.548662884234274, "learning_rate": 0.00011204999999999999, "loss": 1.8906, "step": 747 }, { "epoch": 0.352, "grad_norm": 3.302970103887827, "learning_rate": 0.00011219999999999999, "loss": 2.5938, "step": 748 }, { "epoch": 0.3524705882352941, "grad_norm": 3.197820510595025, "learning_rate": 0.00011235, "loss": 0.9805, "step": 749 }, { "epoch": 0.35294117647058826, "grad_norm": 4.366271111526693, "learning_rate": 0.0001125, "loss": 1.7578, "step": 750 }, { "epoch": 0.35341176470588237, "grad_norm": 1.4425183590417416, "learning_rate": 0.00011264999999999999, "loss": 0.8711, "step": 751 }, { "epoch": 0.3538823529411765, "grad_norm": 3.0378630483134206, "learning_rate": 0.00011279999999999999, "loss": 2.8125, "step": 752 }, { "epoch": 0.3543529411764706, "grad_norm": 5.720584740158784, "learning_rate": 0.00011295, "loss": 1.6484, "step": 753 }, { "epoch": 0.3548235294117647, "grad_norm": 6.5151975485046085, "learning_rate": 0.00011309999999999998, "loss": 3.25, "step": 754 }, { "epoch": 0.3552941176470588, "grad_norm": 2.494128577984884, "learning_rate": 0.00011324999999999999, "loss": 2.1562, "step": 755 }, { "epoch": 0.3557647058823529, "grad_norm": 1.8511184552887412, "learning_rate": 0.00011339999999999999, "loss": 0.8438, "step": 756 }, { "epoch": 0.35623529411764704, "grad_norm": 5.196381347893739, "learning_rate": 0.00011355, "loss": 2.4531, "step": 757 }, { "epoch": 0.35670588235294115, "grad_norm": 1.7489133402141663, "learning_rate": 0.00011369999999999999, "loss": 1.6562, "step": 758 }, { "epoch": 0.3571764705882353, "grad_norm": 5.094968959978825, "learning_rate": 0.00011384999999999999, "loss": 1.0625, "step": 759 }, { "epoch": 0.35764705882352943, "grad_norm": 2.163215030610562, "learning_rate": 0.00011399999999999999, "loss": 2.0156, "step": 760 }, { "epoch": 0.35811764705882354, "grad_norm": 2.1937512421087253, "learning_rate": 0.00011415, "loss": 0.5117, "step": 761 }, { "epoch": 0.35858823529411765, "grad_norm": 16.090350786367278, "learning_rate": 0.00011429999999999999, "loss": 3.5938, "step": 762 }, { "epoch": 0.35905882352941176, "grad_norm": 6.2393581046007, "learning_rate": 0.00011444999999999999, "loss": 1.7812, "step": 763 }, { "epoch": 0.3595294117647059, "grad_norm": 3.6755937951932522, "learning_rate": 0.0001146, "loss": 2.5312, "step": 764 }, { "epoch": 0.36, "grad_norm": 7.202621815418557, "learning_rate": 0.00011475, "loss": 1.9766, "step": 765 }, { "epoch": 0.3604705882352941, "grad_norm": 7.434769691231414, "learning_rate": 0.00011489999999999999, "loss": 1.5312, "step": 766 }, { "epoch": 0.3609411764705882, "grad_norm": 2.084552654376382, "learning_rate": 0.00011504999999999999, "loss": 1.2188, "step": 767 }, { "epoch": 0.3614117647058824, "grad_norm": 3.102630075667738, "learning_rate": 0.0001152, "loss": 1.9766, "step": 768 }, { "epoch": 0.3618823529411765, "grad_norm": 3.861668052812877, "learning_rate": 0.00011535, "loss": 1.1953, "step": 769 }, { "epoch": 0.3623529411764706, "grad_norm": 4.457926426248789, "learning_rate": 0.00011549999999999999, "loss": 2.8438, "step": 770 }, { "epoch": 0.3628235294117647, "grad_norm": 2.6071472473670507, "learning_rate": 0.00011564999999999999, "loss": 3.0312, "step": 771 }, { "epoch": 0.3632941176470588, "grad_norm": 1.1115997304724583, "learning_rate": 0.0001158, "loss": 0.5781, "step": 772 }, { "epoch": 0.36376470588235293, "grad_norm": 4.969848007720913, "learning_rate": 0.00011595, "loss": 1.6641, "step": 773 }, { "epoch": 0.36423529411764705, "grad_norm": 2.084831740670106, "learning_rate": 0.00011609999999999999, "loss": 1.9297, "step": 774 }, { "epoch": 0.36470588235294116, "grad_norm": 4.853669937116771, "learning_rate": 0.00011624999999999999, "loss": 1.0547, "step": 775 }, { "epoch": 0.36517647058823527, "grad_norm": 2.6949345149706474, "learning_rate": 0.0001164, "loss": 1.8828, "step": 776 }, { "epoch": 0.36564705882352944, "grad_norm": 6.600094364117095, "learning_rate": 0.00011654999999999999, "loss": 2.7188, "step": 777 }, { "epoch": 0.36611764705882355, "grad_norm": 8.709985710040721, "learning_rate": 0.00011669999999999999, "loss": 2.125, "step": 778 }, { "epoch": 0.36658823529411766, "grad_norm": 6.127523147924978, "learning_rate": 0.00011685, "loss": 2.0781, "step": 779 }, { "epoch": 0.36705882352941177, "grad_norm": 6.98893413648487, "learning_rate": 0.000117, "loss": 1.8438, "step": 780 }, { "epoch": 0.3675294117647059, "grad_norm": 2.5967773720248837, "learning_rate": 0.00011714999999999999, "loss": 0.3496, "step": 781 }, { "epoch": 0.368, "grad_norm": 3.8040172068370928, "learning_rate": 0.00011729999999999999, "loss": 2.4219, "step": 782 }, { "epoch": 0.3684705882352941, "grad_norm": 2.161061134988291, "learning_rate": 0.00011745, "loss": 0.8281, "step": 783 }, { "epoch": 0.3689411764705882, "grad_norm": 8.90448410203564, "learning_rate": 0.0001176, "loss": 1.2734, "step": 784 }, { "epoch": 0.36941176470588233, "grad_norm": 1.5520378164948825, "learning_rate": 0.00011774999999999999, "loss": 0.4219, "step": 785 }, { "epoch": 0.3698823529411765, "grad_norm": 10.003645473327442, "learning_rate": 0.00011789999999999999, "loss": 2.1875, "step": 786 }, { "epoch": 0.3703529411764706, "grad_norm": 8.027778252818866, "learning_rate": 0.00011805, "loss": 1.9453, "step": 787 }, { "epoch": 0.3708235294117647, "grad_norm": 5.332494706262508, "learning_rate": 0.0001182, "loss": 0.6094, "step": 788 }, { "epoch": 0.37129411764705883, "grad_norm": 6.408029745011118, "learning_rate": 0.00011834999999999999, "loss": 2.1562, "step": 789 }, { "epoch": 0.37176470588235294, "grad_norm": 2.9242067961999836, "learning_rate": 0.0001185, "loss": 1.2656, "step": 790 }, { "epoch": 0.37223529411764705, "grad_norm": 4.277207184213033, "learning_rate": 0.00011865, "loss": 1.2812, "step": 791 }, { "epoch": 0.37270588235294116, "grad_norm": 3.7362298996756094, "learning_rate": 0.0001188, "loss": 0.8711, "step": 792 }, { "epoch": 0.3731764705882353, "grad_norm": 6.609215508016978, "learning_rate": 0.00011894999999999999, "loss": 1.1641, "step": 793 }, { "epoch": 0.3736470588235294, "grad_norm": 8.841047334554569, "learning_rate": 0.0001191, "loss": 3.4844, "step": 794 }, { "epoch": 0.37411764705882355, "grad_norm": 5.761559973019724, "learning_rate": 0.00011925, "loss": 2.0469, "step": 795 }, { "epoch": 0.37458823529411767, "grad_norm": 4.132409996550319, "learning_rate": 0.0001194, "loss": 0.582, "step": 796 }, { "epoch": 0.3750588235294118, "grad_norm": 10.32734030199969, "learning_rate": 0.00011954999999999999, "loss": 0.6758, "step": 797 }, { "epoch": 0.3755294117647059, "grad_norm": 6.1662605265724855, "learning_rate": 0.0001197, "loss": 3.9688, "step": 798 }, { "epoch": 0.376, "grad_norm": 5.624462669737977, "learning_rate": 0.00011985, "loss": 3.5938, "step": 799 }, { "epoch": 0.3764705882352941, "grad_norm": 2.0596606264885344, "learning_rate": 0.00011999999999999999, "loss": 1.2422, "step": 800 }, { "epoch": 0.3769411764705882, "grad_norm": 3.060805490336039, "learning_rate": 0.00012014999999999999, "loss": 0.6211, "step": 801 }, { "epoch": 0.37741176470588234, "grad_norm": 3.4622799630266607, "learning_rate": 0.0001203, "loss": 2.1406, "step": 802 }, { "epoch": 0.37788235294117645, "grad_norm": 2.838058658152987, "learning_rate": 0.00012045, "loss": 1.125, "step": 803 }, { "epoch": 0.3783529411764706, "grad_norm": 1.26672099466583, "learning_rate": 0.00012059999999999999, "loss": 0.5391, "step": 804 }, { "epoch": 0.3788235294117647, "grad_norm": 3.9230477206964642, "learning_rate": 0.00012075, "loss": 1.2734, "step": 805 }, { "epoch": 0.37929411764705884, "grad_norm": 2.4538792284080246, "learning_rate": 0.0001209, "loss": 1.8516, "step": 806 }, { "epoch": 0.37976470588235295, "grad_norm": 2.7854447989411852, "learning_rate": 0.00012105, "loss": 2.9219, "step": 807 }, { "epoch": 0.38023529411764706, "grad_norm": 1.7550970555756524, "learning_rate": 0.00012119999999999999, "loss": 2.0938, "step": 808 }, { "epoch": 0.38070588235294117, "grad_norm": 2.536405276874539, "learning_rate": 0.00012135, "loss": 1.4844, "step": 809 }, { "epoch": 0.3811764705882353, "grad_norm": 3.172478085688269, "learning_rate": 0.0001215, "loss": 2.1719, "step": 810 }, { "epoch": 0.3816470588235294, "grad_norm": 4.59311306671457, "learning_rate": 0.00012165, "loss": 2.2344, "step": 811 }, { "epoch": 0.3821176470588235, "grad_norm": 1.2544385788052603, "learning_rate": 0.00012179999999999999, "loss": 0.9141, "step": 812 }, { "epoch": 0.3825882352941177, "grad_norm": 1.9137439352705372, "learning_rate": 0.00012194999999999998, "loss": 1.3906, "step": 813 }, { "epoch": 0.3830588235294118, "grad_norm": 2.3304340149924996, "learning_rate": 0.00012209999999999999, "loss": 1.6641, "step": 814 }, { "epoch": 0.3835294117647059, "grad_norm": 1.1996899816623003, "learning_rate": 0.00012225, "loss": 0.75, "step": 815 }, { "epoch": 0.384, "grad_norm": 7.240714778219302, "learning_rate": 0.0001224, "loss": 2.9375, "step": 816 }, { "epoch": 0.3844705882352941, "grad_norm": 14.662774165350617, "learning_rate": 0.00012254999999999997, "loss": 0.4004, "step": 817 }, { "epoch": 0.38494117647058823, "grad_norm": 6.021333469540227, "learning_rate": 0.00012269999999999997, "loss": 1.7969, "step": 818 }, { "epoch": 0.38541176470588234, "grad_norm": 1.9896590945128019, "learning_rate": 0.00012284999999999998, "loss": 1.3906, "step": 819 }, { "epoch": 0.38588235294117645, "grad_norm": 1.635381699325542, "learning_rate": 0.00012299999999999998, "loss": 0.9844, "step": 820 }, { "epoch": 0.38635294117647057, "grad_norm": 8.215482137524841, "learning_rate": 0.00012314999999999998, "loss": 2.1875, "step": 821 }, { "epoch": 0.38682352941176473, "grad_norm": 2.975202425956973, "learning_rate": 0.0001233, "loss": 3.4219, "step": 822 }, { "epoch": 0.38729411764705884, "grad_norm": 7.297068597138973, "learning_rate": 0.00012345, "loss": 1.4062, "step": 823 }, { "epoch": 0.38776470588235296, "grad_norm": 2.111704078133022, "learning_rate": 0.0001236, "loss": 1.6641, "step": 824 }, { "epoch": 0.38823529411764707, "grad_norm": 2.179412766533086, "learning_rate": 0.00012374999999999997, "loss": 1.9766, "step": 825 }, { "epoch": 0.3887058823529412, "grad_norm": 1.365709046048984, "learning_rate": 0.00012389999999999998, "loss": 0.8047, "step": 826 }, { "epoch": 0.3891764705882353, "grad_norm": 1.270963011181373, "learning_rate": 0.00012404999999999998, "loss": 0.8906, "step": 827 }, { "epoch": 0.3896470588235294, "grad_norm": 2.15978711927697, "learning_rate": 0.00012419999999999998, "loss": 1.375, "step": 828 }, { "epoch": 0.3901176470588235, "grad_norm": 2.3034598812593146, "learning_rate": 0.00012435, "loss": 1.9922, "step": 829 }, { "epoch": 0.3905882352941176, "grad_norm": 1.6464801553088104, "learning_rate": 0.0001245, "loss": 1.3281, "step": 830 }, { "epoch": 0.3910588235294118, "grad_norm": 2.506633501703921, "learning_rate": 0.00012465, "loss": 1.1016, "step": 831 }, { "epoch": 0.3915294117647059, "grad_norm": 1.5138172104616023, "learning_rate": 0.00012479999999999997, "loss": 0.8438, "step": 832 }, { "epoch": 0.392, "grad_norm": 3.961759736738533, "learning_rate": 0.00012494999999999997, "loss": 2.2656, "step": 833 }, { "epoch": 0.3924705882352941, "grad_norm": 1.2289415980188916, "learning_rate": 0.00012509999999999998, "loss": 0.6758, "step": 834 }, { "epoch": 0.39294117647058824, "grad_norm": 2.3028566416092224, "learning_rate": 0.00012524999999999998, "loss": 1.4297, "step": 835 }, { "epoch": 0.39341176470588235, "grad_norm": 2.343482737963072, "learning_rate": 0.00012539999999999999, "loss": 1.4922, "step": 836 }, { "epoch": 0.39388235294117646, "grad_norm": 1.8366470447751242, "learning_rate": 0.00012555, "loss": 1.7656, "step": 837 }, { "epoch": 0.3943529411764706, "grad_norm": 1.65800683893794, "learning_rate": 0.0001257, "loss": 1.5547, "step": 838 }, { "epoch": 0.3948235294117647, "grad_norm": 1.9566005374576492, "learning_rate": 0.00012585, "loss": 2.0625, "step": 839 }, { "epoch": 0.3952941176470588, "grad_norm": 9.099593167388884, "learning_rate": 0.00012599999999999997, "loss": 0.6094, "step": 840 }, { "epoch": 0.39576470588235296, "grad_norm": 2.3077951912869987, "learning_rate": 0.00012614999999999998, "loss": 2.4219, "step": 841 }, { "epoch": 0.3962352941176471, "grad_norm": 0.8672318219698784, "learning_rate": 0.00012629999999999998, "loss": 0.6172, "step": 842 }, { "epoch": 0.3967058823529412, "grad_norm": 4.973321747355839, "learning_rate": 0.00012644999999999998, "loss": 1.1016, "step": 843 }, { "epoch": 0.3971764705882353, "grad_norm": 1.2325413667314142, "learning_rate": 0.0001266, "loss": 0.5195, "step": 844 }, { "epoch": 0.3976470588235294, "grad_norm": 3.8630315672791795, "learning_rate": 0.00012675, "loss": 2.6562, "step": 845 }, { "epoch": 0.3981176470588235, "grad_norm": 1.4704114008128015, "learning_rate": 0.0001269, "loss": 0.9609, "step": 846 }, { "epoch": 0.39858823529411763, "grad_norm": 0.9130925469803512, "learning_rate": 0.00012705, "loss": 0.3809, "step": 847 }, { "epoch": 0.39905882352941174, "grad_norm": 2.55208532581264, "learning_rate": 0.00012719999999999997, "loss": 1.8438, "step": 848 }, { "epoch": 0.39952941176470586, "grad_norm": 1.9198105560541416, "learning_rate": 0.00012734999999999998, "loss": 0.9766, "step": 849 }, { "epoch": 0.4, "grad_norm": 0.7800114740164429, "learning_rate": 0.00012749999999999998, "loss": 0.2852, "step": 850 }, { "epoch": 0.40047058823529413, "grad_norm": 1.689801302328832, "learning_rate": 0.00012764999999999999, "loss": 0.9062, "step": 851 }, { "epoch": 0.40094117647058825, "grad_norm": 1.5714711624066593, "learning_rate": 0.0001278, "loss": 0.8516, "step": 852 }, { "epoch": 0.40141176470588236, "grad_norm": 1.382423208889802, "learning_rate": 0.00012795, "loss": 0.918, "step": 853 }, { "epoch": 0.40188235294117647, "grad_norm": 1.96369872292485, "learning_rate": 0.0001281, "loss": 1.5703, "step": 854 }, { "epoch": 0.4023529411764706, "grad_norm": 1.5316022843227979, "learning_rate": 0.00012824999999999997, "loss": 1.5547, "step": 855 }, { "epoch": 0.4028235294117647, "grad_norm": 1.9049425534342532, "learning_rate": 0.00012839999999999998, "loss": 1.8984, "step": 856 }, { "epoch": 0.4032941176470588, "grad_norm": 1.7204486759512865, "learning_rate": 0.00012854999999999998, "loss": 1.5938, "step": 857 }, { "epoch": 0.4037647058823529, "grad_norm": 10.127190258718109, "learning_rate": 0.00012869999999999998, "loss": 2.9844, "step": 858 }, { "epoch": 0.4042352941176471, "grad_norm": 2.402617512014234, "learning_rate": 0.00012885, "loss": 2.5469, "step": 859 }, { "epoch": 0.4047058823529412, "grad_norm": 1.0223864069015516, "learning_rate": 0.000129, "loss": 0.2754, "step": 860 }, { "epoch": 0.4051764705882353, "grad_norm": 42.36938820215977, "learning_rate": 0.00012915, "loss": 5.5938, "step": 861 }, { "epoch": 0.4056470588235294, "grad_norm": 2.124062243047209, "learning_rate": 0.0001293, "loss": 2.0781, "step": 862 }, { "epoch": 0.4061176470588235, "grad_norm": 3.1133835861506047, "learning_rate": 0.00012944999999999998, "loss": 0.6875, "step": 863 }, { "epoch": 0.40658823529411764, "grad_norm": 17.755276752659697, "learning_rate": 0.00012959999999999998, "loss": 2.0938, "step": 864 }, { "epoch": 0.40705882352941175, "grad_norm": 21.996240387245223, "learning_rate": 0.00012974999999999998, "loss": 2.5469, "step": 865 }, { "epoch": 0.40752941176470586, "grad_norm": 121.76861455909415, "learning_rate": 0.00012989999999999999, "loss": 4.5, "step": 866 }, { "epoch": 0.408, "grad_norm": 6.749233146336641, "learning_rate": 0.00013005, "loss": 1.0547, "step": 867 }, { "epoch": 0.40847058823529414, "grad_norm": 2.0950225234801665, "learning_rate": 0.0001302, "loss": 0.5703, "step": 868 }, { "epoch": 0.40894117647058825, "grad_norm": 2.969115120261145, "learning_rate": 0.00013035, "loss": 0.5625, "step": 869 }, { "epoch": 0.40941176470588236, "grad_norm": 1.6561235101586955, "learning_rate": 0.0001305, "loss": 1.1797, "step": 870 }, { "epoch": 0.4098823529411765, "grad_norm": 1.944665970885084, "learning_rate": 0.00013064999999999998, "loss": 2.2344, "step": 871 }, { "epoch": 0.4103529411764706, "grad_norm": 1.1297803066122032, "learning_rate": 0.00013079999999999998, "loss": 0.7109, "step": 872 }, { "epoch": 0.4108235294117647, "grad_norm": 1.1856413230381884, "learning_rate": 0.00013094999999999998, "loss": 0.8711, "step": 873 }, { "epoch": 0.4112941176470588, "grad_norm": 12.933676570377264, "learning_rate": 0.0001311, "loss": 2.0625, "step": 874 }, { "epoch": 0.4117647058823529, "grad_norm": 4.77407961291982, "learning_rate": 0.00013125, "loss": 3.2344, "step": 875 }, { "epoch": 0.41223529411764703, "grad_norm": 3.8721689655427816, "learning_rate": 0.0001314, "loss": 1.3594, "step": 876 }, { "epoch": 0.4127058823529412, "grad_norm": 1.2617337696096316, "learning_rate": 0.00013155, "loss": 0.5625, "step": 877 }, { "epoch": 0.4131764705882353, "grad_norm": 1.8752829695723372, "learning_rate": 0.00013169999999999998, "loss": 1.3281, "step": 878 }, { "epoch": 0.4136470588235294, "grad_norm": 1.4384928126459322, "learning_rate": 0.00013184999999999998, "loss": 1.1875, "step": 879 }, { "epoch": 0.41411764705882353, "grad_norm": 0.9875962094710508, "learning_rate": 0.00013199999999999998, "loss": 0.459, "step": 880 }, { "epoch": 0.41458823529411765, "grad_norm": 1.5716906230159233, "learning_rate": 0.00013215, "loss": 0.8203, "step": 881 }, { "epoch": 0.41505882352941176, "grad_norm": 2.259967695801349, "learning_rate": 0.0001323, "loss": 2.1562, "step": 882 }, { "epoch": 0.41552941176470587, "grad_norm": 1.2533585173261976, "learning_rate": 0.00013245, "loss": 1.1875, "step": 883 }, { "epoch": 0.416, "grad_norm": 2.1337648946192505, "learning_rate": 0.0001326, "loss": 1.6641, "step": 884 }, { "epoch": 0.4164705882352941, "grad_norm": 1.2128653114164156, "learning_rate": 0.00013275, "loss": 0.9922, "step": 885 }, { "epoch": 0.41694117647058826, "grad_norm": 1.1088200859251203, "learning_rate": 0.00013289999999999998, "loss": 1.0625, "step": 886 }, { "epoch": 0.41741176470588237, "grad_norm": 1.68505390469374, "learning_rate": 0.00013304999999999998, "loss": 1.0859, "step": 887 }, { "epoch": 0.4178823529411765, "grad_norm": 2.9316678974056734, "learning_rate": 0.00013319999999999999, "loss": 2.0781, "step": 888 }, { "epoch": 0.4183529411764706, "grad_norm": 2.041799050504816, "learning_rate": 0.00013335, "loss": 2.1094, "step": 889 }, { "epoch": 0.4188235294117647, "grad_norm": 1.705046104705279, "learning_rate": 0.0001335, "loss": 1.2578, "step": 890 }, { "epoch": 0.4192941176470588, "grad_norm": 1.261736809970102, "learning_rate": 0.00013365, "loss": 0.582, "step": 891 }, { "epoch": 0.41976470588235293, "grad_norm": 2.257799121192345, "learning_rate": 0.0001338, "loss": 1.4141, "step": 892 }, { "epoch": 0.42023529411764704, "grad_norm": 1.6458292454231804, "learning_rate": 0.00013395, "loss": 1.1875, "step": 893 }, { "epoch": 0.42070588235294115, "grad_norm": 2.544476180371121, "learning_rate": 0.00013409999999999998, "loss": 2.4844, "step": 894 }, { "epoch": 0.4211764705882353, "grad_norm": 1.564704264138278, "learning_rate": 0.00013424999999999998, "loss": 1.2891, "step": 895 }, { "epoch": 0.42164705882352943, "grad_norm": 3.062190701128591, "learning_rate": 0.0001344, "loss": 2.0781, "step": 896 }, { "epoch": 0.42211764705882354, "grad_norm": 1.5370335393850765, "learning_rate": 0.00013455, "loss": 0.7422, "step": 897 }, { "epoch": 0.42258823529411765, "grad_norm": 1.3197147847345967, "learning_rate": 0.0001347, "loss": 0.6992, "step": 898 }, { "epoch": 0.42305882352941176, "grad_norm": 3.4543139063985717, "learning_rate": 0.00013485, "loss": 3.125, "step": 899 }, { "epoch": 0.4235294117647059, "grad_norm": 0.7582694388793428, "learning_rate": 0.000135, "loss": 0.2617, "step": 900 }, { "epoch": 0.424, "grad_norm": 1.2573601196491573, "learning_rate": 0.00013514999999999998, "loss": 0.8711, "step": 901 }, { "epoch": 0.4244705882352941, "grad_norm": 2.277054611852735, "learning_rate": 0.00013529999999999998, "loss": 1.6562, "step": 902 }, { "epoch": 0.4249411764705882, "grad_norm": 1.418937691951044, "learning_rate": 0.00013544999999999999, "loss": 1.0859, "step": 903 }, { "epoch": 0.4254117647058824, "grad_norm": 1.169563370112269, "learning_rate": 0.0001356, "loss": 0.8047, "step": 904 }, { "epoch": 0.4258823529411765, "grad_norm": 2.1175120870373196, "learning_rate": 0.00013575, "loss": 1.7734, "step": 905 }, { "epoch": 0.4263529411764706, "grad_norm": 1.2732660820982804, "learning_rate": 0.0001359, "loss": 0.5625, "step": 906 }, { "epoch": 0.4268235294117647, "grad_norm": 1.9757285391385437, "learning_rate": 0.00013605, "loss": 1.9922, "step": 907 }, { "epoch": 0.4272941176470588, "grad_norm": 2.3672543959877683, "learning_rate": 0.0001362, "loss": 2.625, "step": 908 }, { "epoch": 0.42776470588235294, "grad_norm": 1.3219210589526866, "learning_rate": 0.00013634999999999998, "loss": 0.9727, "step": 909 }, { "epoch": 0.42823529411764705, "grad_norm": 1.2741133687569393, "learning_rate": 0.00013649999999999998, "loss": 1.0625, "step": 910 }, { "epoch": 0.42870588235294116, "grad_norm": 1.3342013322146011, "learning_rate": 0.00013665, "loss": 0.8984, "step": 911 }, { "epoch": 0.42917647058823527, "grad_norm": 1.8608810765599302, "learning_rate": 0.0001368, "loss": 1.5703, "step": 912 }, { "epoch": 0.42964705882352944, "grad_norm": 0.9392793417863108, "learning_rate": 0.00013695, "loss": 0.6289, "step": 913 }, { "epoch": 0.43011764705882355, "grad_norm": 1.8438893620284886, "learning_rate": 0.0001371, "loss": 1.5234, "step": 914 }, { "epoch": 0.43058823529411766, "grad_norm": 1.5472876469897232, "learning_rate": 0.00013725, "loss": 1.2891, "step": 915 }, { "epoch": 0.43105882352941177, "grad_norm": 8.350149566799656, "learning_rate": 0.0001374, "loss": 1.7344, "step": 916 }, { "epoch": 0.4315294117647059, "grad_norm": 3.316116708448448, "learning_rate": 0.00013754999999999998, "loss": 3.375, "step": 917 }, { "epoch": 0.432, "grad_norm": 1.0003499248899905, "learning_rate": 0.00013769999999999999, "loss": 0.4492, "step": 918 }, { "epoch": 0.4324705882352941, "grad_norm": 4.879682058192092, "learning_rate": 0.00013785, "loss": 0.8398, "step": 919 }, { "epoch": 0.4329411764705882, "grad_norm": 3.1648058894640223, "learning_rate": 0.000138, "loss": 1.7578, "step": 920 }, { "epoch": 0.43341176470588233, "grad_norm": 1.4951099140899906, "learning_rate": 0.00013815, "loss": 1.4141, "step": 921 }, { "epoch": 0.4338823529411765, "grad_norm": 2.5560500968806883, "learning_rate": 0.0001383, "loss": 2.2812, "step": 922 }, { "epoch": 0.4343529411764706, "grad_norm": 2.039669494768399, "learning_rate": 0.00013845, "loss": 2.2969, "step": 923 }, { "epoch": 0.4348235294117647, "grad_norm": 1.3390863370570343, "learning_rate": 0.0001386, "loss": 1.2266, "step": 924 }, { "epoch": 0.43529411764705883, "grad_norm": 1.7595579624253426, "learning_rate": 0.00013874999999999998, "loss": 1.6797, "step": 925 }, { "epoch": 0.43576470588235294, "grad_norm": 2.626828791064907, "learning_rate": 0.0001389, "loss": 3.375, "step": 926 }, { "epoch": 0.43623529411764705, "grad_norm": 0.7213686715179265, "learning_rate": 0.00013905, "loss": 0.2148, "step": 927 }, { "epoch": 0.43670588235294117, "grad_norm": 1.0793027206489765, "learning_rate": 0.0001392, "loss": 0.7539, "step": 928 }, { "epoch": 0.4371764705882353, "grad_norm": 4.3138620505650325, "learning_rate": 0.00013935, "loss": 2.3125, "step": 929 }, { "epoch": 0.4376470588235294, "grad_norm": 1.1038467306153277, "learning_rate": 0.0001395, "loss": 0.9219, "step": 930 }, { "epoch": 0.43811764705882356, "grad_norm": 1.9295362734926758, "learning_rate": 0.00013965, "loss": 1.6797, "step": 931 }, { "epoch": 0.43858823529411767, "grad_norm": 2.783087851501884, "learning_rate": 0.00013979999999999998, "loss": 2.9375, "step": 932 }, { "epoch": 0.4390588235294118, "grad_norm": 1.2950966335798468, "learning_rate": 0.00013995, "loss": 0.75, "step": 933 }, { "epoch": 0.4395294117647059, "grad_norm": 1.817190586298299, "learning_rate": 0.0001401, "loss": 1.7109, "step": 934 }, { "epoch": 0.44, "grad_norm": 1.5871017822938613, "learning_rate": 0.00014025, "loss": 1.4844, "step": 935 }, { "epoch": 0.4404705882352941, "grad_norm": 1.050441948703312, "learning_rate": 0.0001404, "loss": 0.4785, "step": 936 }, { "epoch": 0.4409411764705882, "grad_norm": 2.0022973721393797, "learning_rate": 0.00014055, "loss": 1.5156, "step": 937 }, { "epoch": 0.44141176470588234, "grad_norm": 1.9853164161979906, "learning_rate": 0.00014069999999999998, "loss": 1.9141, "step": 938 }, { "epoch": 0.44188235294117645, "grad_norm": 1.112884760084247, "learning_rate": 0.00014084999999999998, "loss": 0.7305, "step": 939 }, { "epoch": 0.4423529411764706, "grad_norm": 1.3731623982281542, "learning_rate": 0.00014099999999999998, "loss": 0.7305, "step": 940 }, { "epoch": 0.4428235294117647, "grad_norm": 1.4284753428173056, "learning_rate": 0.00014115, "loss": 0.875, "step": 941 }, { "epoch": 0.44329411764705884, "grad_norm": 1.470583708812905, "learning_rate": 0.0001413, "loss": 1.0234, "step": 942 }, { "epoch": 0.44376470588235295, "grad_norm": 70.47300367726889, "learning_rate": 0.00014144999999999997, "loss": 2.9531, "step": 943 }, { "epoch": 0.44423529411764706, "grad_norm": 0.9444714193771558, "learning_rate": 0.00014159999999999997, "loss": 0.6367, "step": 944 }, { "epoch": 0.4447058823529412, "grad_norm": 1.0161350479240252, "learning_rate": 0.00014174999999999998, "loss": 0.6953, "step": 945 }, { "epoch": 0.4451764705882353, "grad_norm": 2.74669983969902, "learning_rate": 0.00014189999999999998, "loss": 1.0078, "step": 946 }, { "epoch": 0.4456470588235294, "grad_norm": 27.037512447657384, "learning_rate": 0.00014204999999999998, "loss": 2.7188, "step": 947 }, { "epoch": 0.4461176470588235, "grad_norm": 1.190643891792503, "learning_rate": 0.0001422, "loss": 0.0693, "step": 948 }, { "epoch": 0.4465882352941177, "grad_norm": 0.997790018172666, "learning_rate": 0.00014235, "loss": 0.8242, "step": 949 }, { "epoch": 0.4470588235294118, "grad_norm": 3.69165763583222, "learning_rate": 0.0001425, "loss": 2.8594, "step": 950 }, { "epoch": 0.4475294117647059, "grad_norm": 0.9523429200706877, "learning_rate": 0.00014264999999999997, "loss": 0.6602, "step": 951 }, { "epoch": 0.448, "grad_norm": 1.231942638208678, "learning_rate": 0.00014279999999999997, "loss": 0.7969, "step": 952 }, { "epoch": 0.4484705882352941, "grad_norm": 1.3159356115690628, "learning_rate": 0.00014294999999999998, "loss": 0.7422, "step": 953 }, { "epoch": 0.44894117647058823, "grad_norm": 3.0758847881001095, "learning_rate": 0.00014309999999999998, "loss": 0.8789, "step": 954 }, { "epoch": 0.44941176470588234, "grad_norm": 2.7444035038500894, "learning_rate": 0.00014324999999999999, "loss": 1.9453, "step": 955 }, { "epoch": 0.44988235294117646, "grad_norm": 1.3310461191987455, "learning_rate": 0.0001434, "loss": 1.1875, "step": 956 }, { "epoch": 0.45035294117647057, "grad_norm": 1.2006250629393644, "learning_rate": 0.00014355, "loss": 0.7891, "step": 957 }, { "epoch": 0.4508235294117647, "grad_norm": 0.8373471392293964, "learning_rate": 0.00014369999999999997, "loss": 0.3945, "step": 958 }, { "epoch": 0.45129411764705885, "grad_norm": 2.1401951027062274, "learning_rate": 0.00014384999999999997, "loss": 1.4844, "step": 959 }, { "epoch": 0.45176470588235296, "grad_norm": 1.672883025205646, "learning_rate": 0.00014399999999999998, "loss": 1.6406, "step": 960 }, { "epoch": 0.45223529411764707, "grad_norm": 1.4124012103870827, "learning_rate": 0.00014414999999999998, "loss": 1.1797, "step": 961 }, { "epoch": 0.4527058823529412, "grad_norm": 0.6510010515240789, "learning_rate": 0.00014429999999999998, "loss": 0.1689, "step": 962 }, { "epoch": 0.4531764705882353, "grad_norm": 1.5812517755335285, "learning_rate": 0.00014445, "loss": 1.375, "step": 963 }, { "epoch": 0.4536470588235294, "grad_norm": 2.5908344599773354, "learning_rate": 0.0001446, "loss": 1.5391, "step": 964 }, { "epoch": 0.4541176470588235, "grad_norm": 1.229016313812347, "learning_rate": 0.00014475, "loss": 1.1406, "step": 965 }, { "epoch": 0.4545882352941176, "grad_norm": 1.9252450284721385, "learning_rate": 0.00014489999999999997, "loss": 2.2656, "step": 966 }, { "epoch": 0.45505882352941174, "grad_norm": 2.5218183871655353, "learning_rate": 0.00014504999999999997, "loss": 2.7656, "step": 967 }, { "epoch": 0.4555294117647059, "grad_norm": 1.094765964187049, "learning_rate": 0.00014519999999999998, "loss": 0.668, "step": 968 }, { "epoch": 0.456, "grad_norm": 1.0224936770258446, "learning_rate": 0.00014534999999999998, "loss": 0.8984, "step": 969 }, { "epoch": 0.45647058823529413, "grad_norm": 2.496164580545364, "learning_rate": 0.00014549999999999999, "loss": 2.5625, "step": 970 }, { "epoch": 0.45694117647058824, "grad_norm": 0.9342131036060345, "learning_rate": 0.00014565, "loss": 0.4316, "step": 971 }, { "epoch": 0.45741176470588235, "grad_norm": 1.507134942044389, "learning_rate": 0.0001458, "loss": 1.9844, "step": 972 }, { "epoch": 0.45788235294117646, "grad_norm": 0.8463785319269537, "learning_rate": 0.00014595, "loss": 0.2734, "step": 973 }, { "epoch": 0.4583529411764706, "grad_norm": 2.0231128964037124, "learning_rate": 0.00014609999999999997, "loss": 2.4062, "step": 974 }, { "epoch": 0.4588235294117647, "grad_norm": 1.327505077402603, "learning_rate": 0.00014624999999999998, "loss": 0.5781, "step": 975 }, { "epoch": 0.4592941176470588, "grad_norm": 2.0666055559336907, "learning_rate": 0.00014639999999999998, "loss": 1.6875, "step": 976 }, { "epoch": 0.45976470588235296, "grad_norm": 1.5123948413348822, "learning_rate": 0.00014654999999999998, "loss": 1.5781, "step": 977 }, { "epoch": 0.4602352941176471, "grad_norm": 1.224139722826558, "learning_rate": 0.0001467, "loss": 0.9297, "step": 978 }, { "epoch": 0.4607058823529412, "grad_norm": 2.2793421360394475, "learning_rate": 0.00014685, "loss": 1.4141, "step": 979 }, { "epoch": 0.4611764705882353, "grad_norm": 1.7257136329057259, "learning_rate": 0.000147, "loss": 1.4141, "step": 980 }, { "epoch": 0.4616470588235294, "grad_norm": 1.7508063486716625, "learning_rate": 0.00014714999999999997, "loss": 1.3359, "step": 981 }, { "epoch": 0.4621176470588235, "grad_norm": 1.3446609570345247, "learning_rate": 0.00014729999999999998, "loss": 1.5156, "step": 982 }, { "epoch": 0.46258823529411763, "grad_norm": 1.5015833852788285, "learning_rate": 0.00014744999999999998, "loss": 1.8281, "step": 983 }, { "epoch": 0.46305882352941174, "grad_norm": 2.2257853130307486, "learning_rate": 0.00014759999999999998, "loss": 1.5938, "step": 984 }, { "epoch": 0.46352941176470586, "grad_norm": 0.6757992266640765, "learning_rate": 0.00014774999999999999, "loss": 0.3867, "step": 985 }, { "epoch": 0.464, "grad_norm": 1.1550613290060905, "learning_rate": 0.0001479, "loss": 1.0547, "step": 986 }, { "epoch": 0.46447058823529414, "grad_norm": 1.6191126322067835, "learning_rate": 0.00014805, "loss": 1.7188, "step": 987 }, { "epoch": 0.46494117647058825, "grad_norm": 0.7524878240223318, "learning_rate": 0.0001482, "loss": 0.373, "step": 988 }, { "epoch": 0.46541176470588236, "grad_norm": 1.5431120431502336, "learning_rate": 0.00014834999999999997, "loss": 1.1406, "step": 989 }, { "epoch": 0.46588235294117647, "grad_norm": 2.046386267123101, "learning_rate": 0.00014849999999999998, "loss": 1.8672, "step": 990 }, { "epoch": 0.4663529411764706, "grad_norm": 0.63923583950829, "learning_rate": 0.00014864999999999998, "loss": 0.2832, "step": 991 }, { "epoch": 0.4668235294117647, "grad_norm": 1.0071358764958573, "learning_rate": 0.00014879999999999998, "loss": 0.7109, "step": 992 }, { "epoch": 0.4672941176470588, "grad_norm": 1.7378329604882732, "learning_rate": 0.00014895, "loss": 1.8203, "step": 993 }, { "epoch": 0.4677647058823529, "grad_norm": 1.1952256358907138, "learning_rate": 0.0001491, "loss": 0.7734, "step": 994 }, { "epoch": 0.4682352941176471, "grad_norm": 1.0213325182873427, "learning_rate": 0.00014925, "loss": 0.8984, "step": 995 }, { "epoch": 0.4687058823529412, "grad_norm": 0.7781932679635575, "learning_rate": 0.0001494, "loss": 0.4062, "step": 996 }, { "epoch": 0.4691764705882353, "grad_norm": 1.5700130788687474, "learning_rate": 0.00014954999999999998, "loss": 1.3906, "step": 997 }, { "epoch": 0.4696470588235294, "grad_norm": 1.4806008530330494, "learning_rate": 0.00014969999999999998, "loss": 1.2188, "step": 998 }, { "epoch": 0.47011764705882353, "grad_norm": 1.4453703874515549, "learning_rate": 0.00014984999999999998, "loss": 1.4766, "step": 999 }, { "epoch": 0.47058823529411764, "grad_norm": 0.9080006579296541, "learning_rate": 0.00015, "loss": 0.7227, "step": 1000 }, { "epoch": 0.47105882352941175, "grad_norm": 0.323449325081325, "learning_rate": 0.00015014999999999996, "loss": 0.0713, "step": 1001 }, { "epoch": 0.47152941176470586, "grad_norm": 2.1364644340943073, "learning_rate": 0.0001503, "loss": 2.2656, "step": 1002 }, { "epoch": 0.472, "grad_norm": 1.77810145161306, "learning_rate": 0.00015044999999999997, "loss": 2.125, "step": 1003 }, { "epoch": 0.47247058823529414, "grad_norm": 1.0373099837909079, "learning_rate": 0.00015059999999999997, "loss": 0.5, "step": 1004 }, { "epoch": 0.47294117647058825, "grad_norm": 1.066211003346138, "learning_rate": 0.00015074999999999998, "loss": 0.6328, "step": 1005 }, { "epoch": 0.47341176470588237, "grad_norm": 0.7354076727164033, "learning_rate": 0.00015089999999999998, "loss": 0.3164, "step": 1006 }, { "epoch": 0.4738823529411765, "grad_norm": 2.9901322033730895, "learning_rate": 0.00015104999999999996, "loss": 1.7656, "step": 1007 }, { "epoch": 0.4743529411764706, "grad_norm": 1.0141624483548644, "learning_rate": 0.0001512, "loss": 0.4766, "step": 1008 }, { "epoch": 0.4748235294117647, "grad_norm": 1.6610984062906413, "learning_rate": 0.00015134999999999997, "loss": 1.8281, "step": 1009 }, { "epoch": 0.4752941176470588, "grad_norm": 1.6010160378681542, "learning_rate": 0.0001515, "loss": 1.1719, "step": 1010 }, { "epoch": 0.4757647058823529, "grad_norm": 2.185240997668636, "learning_rate": 0.00015164999999999997, "loss": 2.9688, "step": 1011 }, { "epoch": 0.47623529411764703, "grad_norm": 2.075245074724073, "learning_rate": 0.00015179999999999998, "loss": 2.4375, "step": 1012 }, { "epoch": 0.4767058823529412, "grad_norm": 1.554953085900097, "learning_rate": 0.00015194999999999998, "loss": 1.7578, "step": 1013 }, { "epoch": 0.4771764705882353, "grad_norm": 1.4272507291975893, "learning_rate": 0.00015209999999999998, "loss": 0.9648, "step": 1014 }, { "epoch": 0.4776470588235294, "grad_norm": 0.8171349475508963, "learning_rate": 0.00015224999999999996, "loss": 0.4961, "step": 1015 }, { "epoch": 0.47811764705882354, "grad_norm": 0.7383308402302279, "learning_rate": 0.0001524, "loss": 0.2256, "step": 1016 }, { "epoch": 0.47858823529411765, "grad_norm": 1.0139510073056615, "learning_rate": 0.00015254999999999997, "loss": 0.8984, "step": 1017 }, { "epoch": 0.47905882352941176, "grad_norm": 1.050880120420397, "learning_rate": 0.0001527, "loss": 0.6992, "step": 1018 }, { "epoch": 0.47952941176470587, "grad_norm": 1.0150611387786774, "learning_rate": 0.00015284999999999997, "loss": 0.5391, "step": 1019 }, { "epoch": 0.48, "grad_norm": 1.6052440603928484, "learning_rate": 0.00015299999999999998, "loss": 1.5234, "step": 1020 }, { "epoch": 0.4804705882352941, "grad_norm": 0.7984189671178196, "learning_rate": 0.00015314999999999998, "loss": 0.6719, "step": 1021 }, { "epoch": 0.48094117647058826, "grad_norm": 1.299768637649481, "learning_rate": 0.00015329999999999999, "loss": 1.1562, "step": 1022 }, { "epoch": 0.4814117647058824, "grad_norm": 1.3084092917997754, "learning_rate": 0.00015344999999999996, "loss": 1.1953, "step": 1023 }, { "epoch": 0.4818823529411765, "grad_norm": 0.46404466296682967, "learning_rate": 0.0001536, "loss": 0.1572, "step": 1024 }, { "epoch": 0.4823529411764706, "grad_norm": 1.473452840773952, "learning_rate": 0.00015374999999999997, "loss": 1.5781, "step": 1025 }, { "epoch": 0.4828235294117647, "grad_norm": 1.4627991614293914, "learning_rate": 0.0001539, "loss": 1.1172, "step": 1026 }, { "epoch": 0.4832941176470588, "grad_norm": 1.292101954666966, "learning_rate": 0.00015404999999999998, "loss": 1.7891, "step": 1027 }, { "epoch": 0.48376470588235293, "grad_norm": 1.070207287040961, "learning_rate": 0.00015419999999999998, "loss": 0.9609, "step": 1028 }, { "epoch": 0.48423529411764704, "grad_norm": 1.7977798360970743, "learning_rate": 0.00015434999999999998, "loss": 1.8672, "step": 1029 }, { "epoch": 0.48470588235294115, "grad_norm": 0.42033159170429074, "learning_rate": 0.0001545, "loss": 0.1011, "step": 1030 }, { "epoch": 0.4851764705882353, "grad_norm": 2.774529766342851, "learning_rate": 0.00015464999999999996, "loss": 2.7188, "step": 1031 }, { "epoch": 0.48564705882352943, "grad_norm": 1.2873334967070664, "learning_rate": 0.0001548, "loss": 1.2969, "step": 1032 }, { "epoch": 0.48611764705882354, "grad_norm": 1.3493859434766124, "learning_rate": 0.00015494999999999997, "loss": 1.3984, "step": 1033 }, { "epoch": 0.48658823529411765, "grad_norm": 1.2683167364789143, "learning_rate": 0.0001551, "loss": 1.3359, "step": 1034 }, { "epoch": 0.48705882352941177, "grad_norm": 0.48931737213097604, "learning_rate": 0.00015524999999999998, "loss": 0.1855, "step": 1035 }, { "epoch": 0.4875294117647059, "grad_norm": 1.6147594607918783, "learning_rate": 0.00015539999999999998, "loss": 1.6562, "step": 1036 }, { "epoch": 0.488, "grad_norm": 0.5287060152398435, "learning_rate": 0.00015554999999999999, "loss": 0.2715, "step": 1037 }, { "epoch": 0.4884705882352941, "grad_norm": 1.1067939793563948, "learning_rate": 0.0001557, "loss": 1.0156, "step": 1038 }, { "epoch": 0.4889411764705882, "grad_norm": 1.1693264434100747, "learning_rate": 0.00015584999999999997, "loss": 1.1328, "step": 1039 }, { "epoch": 0.4894117647058824, "grad_norm": 1.4955838605851073, "learning_rate": 0.000156, "loss": 1.8281, "step": 1040 }, { "epoch": 0.4898823529411765, "grad_norm": 1.4711579567832722, "learning_rate": 0.00015614999999999997, "loss": 1.6875, "step": 1041 }, { "epoch": 0.4903529411764706, "grad_norm": 1.4883486359304607, "learning_rate": 0.0001563, "loss": 1.9453, "step": 1042 }, { "epoch": 0.4908235294117647, "grad_norm": 0.46179041959598466, "learning_rate": 0.00015644999999999998, "loss": 0.1299, "step": 1043 }, { "epoch": 0.4912941176470588, "grad_norm": 1.1953992546657997, "learning_rate": 0.00015659999999999998, "loss": 1.0, "step": 1044 }, { "epoch": 0.49176470588235294, "grad_norm": 1.5636592639543303, "learning_rate": 0.00015675, "loss": 1.4297, "step": 1045 }, { "epoch": 0.49223529411764705, "grad_norm": 1.4884671779877476, "learning_rate": 0.0001569, "loss": 1.4141, "step": 1046 }, { "epoch": 0.49270588235294116, "grad_norm": 0.6803900831776399, "learning_rate": 0.00015704999999999997, "loss": 0.4648, "step": 1047 }, { "epoch": 0.49317647058823527, "grad_norm": 1.6367185388719834, "learning_rate": 0.0001572, "loss": 1.4844, "step": 1048 }, { "epoch": 0.49364705882352944, "grad_norm": 1.2380368278394813, "learning_rate": 0.00015734999999999998, "loss": 1.0156, "step": 1049 }, { "epoch": 0.49411764705882355, "grad_norm": 0.9648610972824493, "learning_rate": 0.00015749999999999998, "loss": 0.4414, "step": 1050 }, { "epoch": 0.49458823529411766, "grad_norm": 1.5947774369787013, "learning_rate": 0.00015764999999999998, "loss": 1.3594, "step": 1051 }, { "epoch": 0.4950588235294118, "grad_norm": 1.2372262590508007, "learning_rate": 0.0001578, "loss": 1.2031, "step": 1052 }, { "epoch": 0.4955294117647059, "grad_norm": 0.8112768768691062, "learning_rate": 0.00015794999999999996, "loss": 0.3535, "step": 1053 }, { "epoch": 0.496, "grad_norm": 0.8632572264575874, "learning_rate": 0.0001581, "loss": 0.6055, "step": 1054 }, { "epoch": 0.4964705882352941, "grad_norm": 2.313625300299155, "learning_rate": 0.00015824999999999997, "loss": 1.8516, "step": 1055 }, { "epoch": 0.4969411764705882, "grad_norm": 1.42507116751158, "learning_rate": 0.0001584, "loss": 1.7969, "step": 1056 }, { "epoch": 0.49741176470588233, "grad_norm": 1.8732094070940568, "learning_rate": 0.00015854999999999998, "loss": 1.7812, "step": 1057 }, { "epoch": 0.4978823529411765, "grad_norm": 1.2240323170003504, "learning_rate": 0.00015869999999999998, "loss": 1.0547, "step": 1058 }, { "epoch": 0.4983529411764706, "grad_norm": 1.8752452090864205, "learning_rate": 0.00015884999999999999, "loss": 2.0156, "step": 1059 }, { "epoch": 0.4988235294117647, "grad_norm": 1.5221506197637784, "learning_rate": 0.000159, "loss": 1.4688, "step": 1060 }, { "epoch": 0.49929411764705883, "grad_norm": 2.5005176474980138, "learning_rate": 0.00015914999999999997, "loss": 1.8672, "step": 1061 }, { "epoch": 0.49976470588235294, "grad_norm": 0.7723313506411676, "learning_rate": 0.0001593, "loss": 0.8398, "step": 1062 }, { "epoch": 0.5002352941176471, "grad_norm": 1.1751181714536076, "learning_rate": 0.00015944999999999997, "loss": 0.8672, "step": 1063 }, { "epoch": 0.5007058823529412, "grad_norm": 1.3801359313200248, "learning_rate": 0.0001596, "loss": 0.9141, "step": 1064 }, { "epoch": 0.5011764705882353, "grad_norm": 2.2860826400213905, "learning_rate": 0.00015974999999999998, "loss": 1.9297, "step": 1065 }, { "epoch": 0.5016470588235294, "grad_norm": 1.4681318513141164, "learning_rate": 0.00015989999999999998, "loss": 1.6094, "step": 1066 }, { "epoch": 0.5021176470588236, "grad_norm": 0.9846228866942419, "learning_rate": 0.00016005, "loss": 0.4355, "step": 1067 }, { "epoch": 0.5025882352941177, "grad_norm": 1.1726623969259107, "learning_rate": 0.0001602, "loss": 1.0078, "step": 1068 }, { "epoch": 0.5030588235294118, "grad_norm": 1.8354069795276235, "learning_rate": 0.00016034999999999997, "loss": 1.6641, "step": 1069 }, { "epoch": 0.5035294117647059, "grad_norm": 1.3218904599450725, "learning_rate": 0.0001605, "loss": 1.0312, "step": 1070 }, { "epoch": 0.504, "grad_norm": 0.9982198464049802, "learning_rate": 0.00016064999999999997, "loss": 0.5234, "step": 1071 }, { "epoch": 0.5044705882352941, "grad_norm": 2.023950581155754, "learning_rate": 0.0001608, "loss": 2.0781, "step": 1072 }, { "epoch": 0.5049411764705882, "grad_norm": 1.0630312305938914, "learning_rate": 0.00016094999999999998, "loss": 1.1875, "step": 1073 }, { "epoch": 0.5054117647058823, "grad_norm": 2.105830950342115, "learning_rate": 0.00016109999999999999, "loss": 1.2891, "step": 1074 }, { "epoch": 0.5058823529411764, "grad_norm": 1.3805789513596325, "learning_rate": 0.00016125, "loss": 1.75, "step": 1075 }, { "epoch": 0.5063529411764706, "grad_norm": 1.0267527302606183, "learning_rate": 0.0001614, "loss": 0.6953, "step": 1076 }, { "epoch": 0.5068235294117647, "grad_norm": 0.9804155318197224, "learning_rate": 0.00016154999999999997, "loss": 0.4297, "step": 1077 }, { "epoch": 0.5072941176470588, "grad_norm": 2.2441842251948088, "learning_rate": 0.0001617, "loss": 2.9531, "step": 1078 }, { "epoch": 0.5077647058823529, "grad_norm": 3.670868989699834, "learning_rate": 0.00016184999999999998, "loss": 2.8438, "step": 1079 }, { "epoch": 0.508235294117647, "grad_norm": 1.2307423750662108, "learning_rate": 0.000162, "loss": 0.8164, "step": 1080 }, { "epoch": 0.5087058823529412, "grad_norm": 0.34323018922501175, "learning_rate": 0.00016214999999999998, "loss": 0.032, "step": 1081 }, { "epoch": 0.5091764705882353, "grad_norm": 1.8686921972794683, "learning_rate": 0.0001623, "loss": 1.8672, "step": 1082 }, { "epoch": 0.5096470588235295, "grad_norm": 1.965428526856117, "learning_rate": 0.00016245, "loss": 2.6875, "step": 1083 }, { "epoch": 0.5101176470588236, "grad_norm": 1.3312260633885618, "learning_rate": 0.0001626, "loss": 0.9727, "step": 1084 }, { "epoch": 0.5105882352941177, "grad_norm": 1.9767243028155774, "learning_rate": 0.00016274999999999997, "loss": 2.3594, "step": 1085 }, { "epoch": 0.5110588235294118, "grad_norm": 1.8232318854752305, "learning_rate": 0.0001629, "loss": 1.9297, "step": 1086 }, { "epoch": 0.5115294117647059, "grad_norm": 1.3756105161032666, "learning_rate": 0.00016304999999999998, "loss": 1.7812, "step": 1087 }, { "epoch": 0.512, "grad_norm": 1.6492105099454484, "learning_rate": 0.0001632, "loss": 1.7266, "step": 1088 }, { "epoch": 0.5124705882352941, "grad_norm": 0.7709322328636233, "learning_rate": 0.00016334999999999999, "loss": 0.5117, "step": 1089 }, { "epoch": 0.5129411764705882, "grad_norm": 1.4327100131176729, "learning_rate": 0.0001635, "loss": 1.1484, "step": 1090 }, { "epoch": 0.5134117647058823, "grad_norm": 1.0970014044498182, "learning_rate": 0.00016365, "loss": 0.6953, "step": 1091 }, { "epoch": 0.5138823529411765, "grad_norm": 1.1380282038802778, "learning_rate": 0.0001638, "loss": 0.6562, "step": 1092 }, { "epoch": 0.5143529411764706, "grad_norm": 0.9283054017892142, "learning_rate": 0.00016394999999999997, "loss": 0.8047, "step": 1093 }, { "epoch": 0.5148235294117647, "grad_norm": 1.7768929001598548, "learning_rate": 0.0001641, "loss": 1.6797, "step": 1094 }, { "epoch": 0.5152941176470588, "grad_norm": 0.8192948546011964, "learning_rate": 0.00016424999999999998, "loss": 0.5391, "step": 1095 }, { "epoch": 0.5157647058823529, "grad_norm": 1.4208029108435616, "learning_rate": 0.0001644, "loss": 1.7891, "step": 1096 }, { "epoch": 0.516235294117647, "grad_norm": 1.4312632603449844, "learning_rate": 0.00016455, "loss": 1.9688, "step": 1097 }, { "epoch": 0.5167058823529411, "grad_norm": 1.654617849939787, "learning_rate": 0.0001647, "loss": 1.9453, "step": 1098 }, { "epoch": 0.5171764705882353, "grad_norm": 1.41545425894382, "learning_rate": 0.00016485, "loss": 1.4219, "step": 1099 }, { "epoch": 0.5176470588235295, "grad_norm": 1.1699084930976398, "learning_rate": 0.000165, "loss": 1.2656, "step": 1100 }, { "epoch": 0.5181176470588236, "grad_norm": 0.9568901594997031, "learning_rate": 0.00016514999999999998, "loss": 1.5625, "step": 1101 }, { "epoch": 0.5185882352941177, "grad_norm": 0.5959605719669413, "learning_rate": 0.0001653, "loss": 0.2637, "step": 1102 }, { "epoch": 0.5190588235294118, "grad_norm": 1.013515758190378, "learning_rate": 0.00016544999999999998, "loss": 0.7461, "step": 1103 }, { "epoch": 0.5195294117647059, "grad_norm": 0.9086339199819344, "learning_rate": 0.0001656, "loss": 0.4824, "step": 1104 }, { "epoch": 0.52, "grad_norm": 2.242874039891644, "learning_rate": 0.00016575, "loss": 3.2344, "step": 1105 }, { "epoch": 0.5204705882352941, "grad_norm": 1.2794452046009548, "learning_rate": 0.0001659, "loss": 1.3438, "step": 1106 }, { "epoch": 0.5209411764705882, "grad_norm": 1.3412533477651463, "learning_rate": 0.00016604999999999997, "loss": 1.5859, "step": 1107 }, { "epoch": 0.5214117647058824, "grad_norm": 0.7038529906073343, "learning_rate": 0.0001662, "loss": 0.3379, "step": 1108 }, { "epoch": 0.5218823529411765, "grad_norm": 0.9430068550553208, "learning_rate": 0.00016634999999999998, "loss": 0.8984, "step": 1109 }, { "epoch": 0.5223529411764706, "grad_norm": 1.3042368932730508, "learning_rate": 0.0001665, "loss": 1.1562, "step": 1110 }, { "epoch": 0.5228235294117647, "grad_norm": 0.9270493306199515, "learning_rate": 0.00016664999999999998, "loss": 0.793, "step": 1111 }, { "epoch": 0.5232941176470588, "grad_norm": 1.104842209331734, "learning_rate": 0.0001668, "loss": 1.7812, "step": 1112 }, { "epoch": 0.5237647058823529, "grad_norm": 0.9638536073771986, "learning_rate": 0.00016695, "loss": 0.5391, "step": 1113 }, { "epoch": 0.524235294117647, "grad_norm": 4.900570950807382, "learning_rate": 0.0001671, "loss": 0.7734, "step": 1114 }, { "epoch": 0.5247058823529411, "grad_norm": 0.8425689594145025, "learning_rate": 0.00016724999999999997, "loss": 0.6172, "step": 1115 }, { "epoch": 0.5251764705882352, "grad_norm": 1.4514579544719817, "learning_rate": 0.0001674, "loss": 2.0469, "step": 1116 }, { "epoch": 0.5256470588235295, "grad_norm": 0.7525945108487916, "learning_rate": 0.00016754999999999998, "loss": 0.4473, "step": 1117 }, { "epoch": 0.5261176470588236, "grad_norm": 1.5440592569717662, "learning_rate": 0.0001677, "loss": 1.25, "step": 1118 }, { "epoch": 0.5265882352941177, "grad_norm": 0.7789646182077192, "learning_rate": 0.00016785, "loss": 0.7031, "step": 1119 }, { "epoch": 0.5270588235294118, "grad_norm": 1.1502402272397807, "learning_rate": 0.000168, "loss": 1.1094, "step": 1120 }, { "epoch": 0.5275294117647059, "grad_norm": 1.5008902904483543, "learning_rate": 0.00016815, "loss": 2.2188, "step": 1121 }, { "epoch": 0.528, "grad_norm": 0.7642270232111805, "learning_rate": 0.0001683, "loss": 0.3164, "step": 1122 }, { "epoch": 0.5284705882352941, "grad_norm": 0.044492638088858255, "learning_rate": 0.00016844999999999997, "loss": 0.0115, "step": 1123 }, { "epoch": 0.5289411764705882, "grad_norm": 0.8335212880501053, "learning_rate": 0.0001686, "loss": 0.4707, "step": 1124 }, { "epoch": 0.5294117647058824, "grad_norm": 1.193995012083794, "learning_rate": 0.00016874999999999998, "loss": 0.9688, "step": 1125 }, { "epoch": 0.5298823529411765, "grad_norm": 0.8896656681338116, "learning_rate": 0.00016889999999999996, "loss": 0.5508, "step": 1126 }, { "epoch": 0.5303529411764706, "grad_norm": 0.5912623839969333, "learning_rate": 0.00016905, "loss": 0.2891, "step": 1127 }, { "epoch": 0.5308235294117647, "grad_norm": 2.4710668996727376, "learning_rate": 0.00016919999999999997, "loss": 2.3281, "step": 1128 }, { "epoch": 0.5312941176470588, "grad_norm": 0.519353835775605, "learning_rate": 0.00016935, "loss": 0.1953, "step": 1129 }, { "epoch": 0.5317647058823529, "grad_norm": 1.9418900338792426, "learning_rate": 0.00016949999999999997, "loss": 1.9453, "step": 1130 }, { "epoch": 0.532235294117647, "grad_norm": 1.1335178487178137, "learning_rate": 0.00016964999999999998, "loss": 0.8906, "step": 1131 }, { "epoch": 0.5327058823529411, "grad_norm": 1.3722989070656004, "learning_rate": 0.00016979999999999998, "loss": 1.7031, "step": 1132 }, { "epoch": 0.5331764705882353, "grad_norm": 1.0572967216634566, "learning_rate": 0.00016994999999999998, "loss": 0.7891, "step": 1133 }, { "epoch": 0.5336470588235294, "grad_norm": 0.42395532992031915, "learning_rate": 0.00017009999999999996, "loss": 0.1226, "step": 1134 }, { "epoch": 0.5341176470588235, "grad_norm": 1.2320820088546593, "learning_rate": 0.00017025, "loss": 0.9297, "step": 1135 }, { "epoch": 0.5345882352941177, "grad_norm": 1.5492585655765176, "learning_rate": 0.00017039999999999997, "loss": 1.3828, "step": 1136 }, { "epoch": 0.5350588235294118, "grad_norm": 1.2873192203776944, "learning_rate": 0.00017055, "loss": 1.0312, "step": 1137 }, { "epoch": 0.5355294117647059, "grad_norm": 1.0922311938130511, "learning_rate": 0.00017069999999999998, "loss": 1.1953, "step": 1138 }, { "epoch": 0.536, "grad_norm": 1.6984180756642924, "learning_rate": 0.00017084999999999998, "loss": 1.6172, "step": 1139 }, { "epoch": 0.5364705882352941, "grad_norm": 0.8969765142693796, "learning_rate": 0.00017099999999999998, "loss": 0.4961, "step": 1140 }, { "epoch": 0.5369411764705883, "grad_norm": 0.8954073423971042, "learning_rate": 0.00017114999999999999, "loss": 0.7695, "step": 1141 }, { "epoch": 0.5374117647058824, "grad_norm": 1.090493949674412, "learning_rate": 0.00017129999999999996, "loss": 0.793, "step": 1142 }, { "epoch": 0.5378823529411765, "grad_norm": 1.4162608849040252, "learning_rate": 0.00017145, "loss": 1.3438, "step": 1143 }, { "epoch": 0.5383529411764706, "grad_norm": 1.524770247312696, "learning_rate": 0.00017159999999999997, "loss": 1.2891, "step": 1144 }, { "epoch": 0.5388235294117647, "grad_norm": 1.8697460209526906, "learning_rate": 0.00017175, "loss": 1.7656, "step": 1145 }, { "epoch": 0.5392941176470588, "grad_norm": 0.8927958135068038, "learning_rate": 0.00017189999999999998, "loss": 0.582, "step": 1146 }, { "epoch": 0.5397647058823529, "grad_norm": 1.7854027211487034, "learning_rate": 0.00017204999999999998, "loss": 1.9531, "step": 1147 }, { "epoch": 0.540235294117647, "grad_norm": 1.5273733262237552, "learning_rate": 0.00017219999999999998, "loss": 1.9219, "step": 1148 }, { "epoch": 0.5407058823529411, "grad_norm": 0.7970738383577045, "learning_rate": 0.00017235, "loss": 0.582, "step": 1149 }, { "epoch": 0.5411764705882353, "grad_norm": 1.3839194427167374, "learning_rate": 0.00017249999999999996, "loss": 1.6484, "step": 1150 }, { "epoch": 0.5416470588235294, "grad_norm": 1.0010472347707566, "learning_rate": 0.00017265, "loss": 0.7148, "step": 1151 }, { "epoch": 0.5421176470588235, "grad_norm": 1.697075648750002, "learning_rate": 0.00017279999999999997, "loss": 1.4062, "step": 1152 }, { "epoch": 0.5425882352941176, "grad_norm": 1.8016238254962385, "learning_rate": 0.00017294999999999998, "loss": 2.7656, "step": 1153 }, { "epoch": 0.5430588235294118, "grad_norm": 1.0358681328021186, "learning_rate": 0.00017309999999999998, "loss": 1.5859, "step": 1154 }, { "epoch": 0.5435294117647059, "grad_norm": 2.076372805230922, "learning_rate": 0.00017324999999999998, "loss": 2.3281, "step": 1155 }, { "epoch": 0.544, "grad_norm": 0.6317039654701695, "learning_rate": 0.00017339999999999996, "loss": 0.4961, "step": 1156 }, { "epoch": 0.5444705882352942, "grad_norm": 1.570205688109004, "learning_rate": 0.00017355, "loss": 1.7188, "step": 1157 }, { "epoch": 0.5449411764705883, "grad_norm": 1.2487239770051297, "learning_rate": 0.00017369999999999997, "loss": 0.875, "step": 1158 }, { "epoch": 0.5454117647058824, "grad_norm": 1.1143930745051518, "learning_rate": 0.00017385, "loss": 0.7383, "step": 1159 }, { "epoch": 0.5458823529411765, "grad_norm": 0.634124958930558, "learning_rate": 0.00017399999999999997, "loss": 0.2969, "step": 1160 }, { "epoch": 0.5463529411764706, "grad_norm": 0.9836209102722216, "learning_rate": 0.00017414999999999998, "loss": 0.6016, "step": 1161 }, { "epoch": 0.5468235294117647, "grad_norm": 1.3591150546505413, "learning_rate": 0.00017429999999999998, "loss": 1.9453, "step": 1162 }, { "epoch": 0.5472941176470588, "grad_norm": 1.2924709098629323, "learning_rate": 0.00017444999999999998, "loss": 1.2422, "step": 1163 }, { "epoch": 0.5477647058823529, "grad_norm": 3.502130748753388, "learning_rate": 0.00017459999999999996, "loss": 0.7383, "step": 1164 }, { "epoch": 0.548235294117647, "grad_norm": 1.6496783431565911, "learning_rate": 0.00017475, "loss": 1.7891, "step": 1165 }, { "epoch": 0.5487058823529412, "grad_norm": 0.6706417337432858, "learning_rate": 0.00017489999999999997, "loss": 0.3711, "step": 1166 }, { "epoch": 0.5491764705882353, "grad_norm": 0.9431507830861099, "learning_rate": 0.00017505, "loss": 0.9922, "step": 1167 }, { "epoch": 0.5496470588235294, "grad_norm": 0.8994159306017913, "learning_rate": 0.00017519999999999998, "loss": 0.3867, "step": 1168 }, { "epoch": 0.5501176470588235, "grad_norm": 1.2617496637779706, "learning_rate": 0.00017534999999999998, "loss": 1.3516, "step": 1169 }, { "epoch": 0.5505882352941176, "grad_norm": 0.7489008065718201, "learning_rate": 0.00017549999999999998, "loss": 0.4238, "step": 1170 }, { "epoch": 0.5510588235294117, "grad_norm": 2.354523001090102, "learning_rate": 0.00017565, "loss": 1.8359, "step": 1171 }, { "epoch": 0.5515294117647059, "grad_norm": 1.0479355407788526, "learning_rate": 0.00017579999999999996, "loss": 0.6562, "step": 1172 }, { "epoch": 0.552, "grad_norm": 0.964098681401635, "learning_rate": 0.00017595, "loss": 0.7344, "step": 1173 }, { "epoch": 0.5524705882352942, "grad_norm": 2.01604337312908, "learning_rate": 0.00017609999999999997, "loss": 1.2891, "step": 1174 }, { "epoch": 0.5529411764705883, "grad_norm": 1.3077078182400885, "learning_rate": 0.00017625, "loss": 1.4375, "step": 1175 }, { "epoch": 0.5534117647058824, "grad_norm": 1.687466574305239, "learning_rate": 0.00017639999999999998, "loss": 2.375, "step": 1176 }, { "epoch": 0.5538823529411765, "grad_norm": 2.3725936693140035, "learning_rate": 0.00017654999999999998, "loss": 2.5625, "step": 1177 }, { "epoch": 0.5543529411764706, "grad_norm": 4.82777189345173, "learning_rate": 0.00017669999999999999, "loss": 2.4375, "step": 1178 }, { "epoch": 0.5548235294117647, "grad_norm": 5.200483817550286, "learning_rate": 0.00017685, "loss": 3.6094, "step": 1179 }, { "epoch": 0.5552941176470588, "grad_norm": 10.343058713485819, "learning_rate": 0.00017699999999999997, "loss": 4.4062, "step": 1180 }, { "epoch": 0.5557647058823529, "grad_norm": 13.641041076981715, "learning_rate": 0.00017715, "loss": 12.5, "step": 1181 }, { "epoch": 0.556235294117647, "grad_norm": 16.473630278115316, "learning_rate": 0.00017729999999999997, "loss": 4.7812, "step": 1182 }, { "epoch": 0.5567058823529412, "grad_norm": 4.763846026662805, "learning_rate": 0.00017745, "loss": 4.9375, "step": 1183 }, { "epoch": 0.5571764705882353, "grad_norm": 33.600253829350116, "learning_rate": 0.00017759999999999998, "loss": 4.3125, "step": 1184 }, { "epoch": 0.5576470588235294, "grad_norm": 10.748507713491994, "learning_rate": 0.00017774999999999998, "loss": 4.5312, "step": 1185 }, { "epoch": 0.5581176470588235, "grad_norm": 10.745823876702502, "learning_rate": 0.0001779, "loss": 1.0781, "step": 1186 }, { "epoch": 0.5585882352941176, "grad_norm": 2.9390622621992737, "learning_rate": 0.00017805, "loss": 1.3359, "step": 1187 }, { "epoch": 0.5590588235294117, "grad_norm": 8.152557854213937, "learning_rate": 0.00017819999999999997, "loss": 2.2812, "step": 1188 }, { "epoch": 0.5595294117647058, "grad_norm": 2.8657377934567427, "learning_rate": 0.00017835, "loss": 2.125, "step": 1189 }, { "epoch": 0.56, "grad_norm": 2.0267755506005645, "learning_rate": 0.00017849999999999997, "loss": 1.4297, "step": 1190 }, { "epoch": 0.5604705882352942, "grad_norm": 21.43942142517278, "learning_rate": 0.00017865, "loss": 1.0469, "step": 1191 }, { "epoch": 0.5609411764705883, "grad_norm": 11.534208252919681, "learning_rate": 0.00017879999999999998, "loss": 1.6406, "step": 1192 }, { "epoch": 0.5614117647058824, "grad_norm": 31.12429909908912, "learning_rate": 0.00017894999999999999, "loss": 1.7969, "step": 1193 }, { "epoch": 0.5618823529411765, "grad_norm": 3.9258590558429978, "learning_rate": 0.0001791, "loss": 2.2656, "step": 1194 }, { "epoch": 0.5623529411764706, "grad_norm": 6.977278479821107, "learning_rate": 0.00017925, "loss": 1.3359, "step": 1195 }, { "epoch": 0.5628235294117647, "grad_norm": 5.499666923237638, "learning_rate": 0.00017939999999999997, "loss": 0.6836, "step": 1196 }, { "epoch": 0.5632941176470588, "grad_norm": 5.799878546151647, "learning_rate": 0.00017955, "loss": 2.1094, "step": 1197 }, { "epoch": 0.563764705882353, "grad_norm": 2.1247605904842612, "learning_rate": 0.00017969999999999998, "loss": 2.1406, "step": 1198 }, { "epoch": 0.5642352941176471, "grad_norm": 1.2080602082156109, "learning_rate": 0.00017984999999999998, "loss": 0.6328, "step": 1199 }, { "epoch": 0.5647058823529412, "grad_norm": 2.8722623055724306, "learning_rate": 0.00017999999999999998, "loss": 1.5781, "step": 1200 }, { "epoch": 0.5651764705882353, "grad_norm": 8.1962738338975, "learning_rate": 0.00018015, "loss": 1.4844, "step": 1201 }, { "epoch": 0.5656470588235294, "grad_norm": 33.015348889867276, "learning_rate": 0.00018029999999999996, "loss": 4.5625, "step": 1202 }, { "epoch": 0.5661176470588235, "grad_norm": 342.46057103363404, "learning_rate": 0.00018045, "loss": 8.4375, "step": 1203 }, { "epoch": 0.5665882352941176, "grad_norm": 22.14741095659656, "learning_rate": 0.00018059999999999997, "loss": 1.0312, "step": 1204 }, { "epoch": 0.5670588235294117, "grad_norm": 1.4323933688870232, "learning_rate": 0.00018075, "loss": 0.7773, "step": 1205 }, { "epoch": 0.5675294117647058, "grad_norm": 11.052005852227506, "learning_rate": 0.00018089999999999998, "loss": 3.75, "step": 1206 }, { "epoch": 0.568, "grad_norm": 25.4746441640324, "learning_rate": 0.00018104999999999998, "loss": 8.75, "step": 1207 }, { "epoch": 0.5684705882352942, "grad_norm": 27.302713816795254, "learning_rate": 0.00018119999999999999, "loss": 12.6875, "step": 1208 }, { "epoch": 0.5689411764705883, "grad_norm": 17.22016878332364, "learning_rate": 0.00018135, "loss": 1.7109, "step": 1209 }, { "epoch": 0.5694117647058824, "grad_norm": 8.976185416798469, "learning_rate": 0.00018149999999999997, "loss": 5.7188, "step": 1210 }, { "epoch": 0.5698823529411765, "grad_norm": 20.600003420376666, "learning_rate": 0.00018165, "loss": 18.625, "step": 1211 }, { "epoch": 0.5703529411764706, "grad_norm": 19.172564819626295, "learning_rate": 0.00018179999999999997, "loss": 11.1875, "step": 1212 }, { "epoch": 0.5708235294117647, "grad_norm": 1.3738651517588694, "learning_rate": 0.00018195, "loss": 0.5938, "step": 1213 }, { "epoch": 0.5712941176470588, "grad_norm": 22.057965751835376, "learning_rate": 0.00018209999999999998, "loss": 8.5, "step": 1214 }, { "epoch": 0.571764705882353, "grad_norm": 38.86637758545123, "learning_rate": 0.00018224999999999998, "loss": 8.125, "step": 1215 }, { "epoch": 0.5722352941176471, "grad_norm": 119.99717044399296, "learning_rate": 0.0001824, "loss": 6.5625, "step": 1216 }, { "epoch": 0.5727058823529412, "grad_norm": 14.404007444987686, "learning_rate": 0.00018255, "loss": 6.0625, "step": 1217 }, { "epoch": 0.5731764705882353, "grad_norm": 7.446477398849134, "learning_rate": 0.00018269999999999997, "loss": 3.7031, "step": 1218 }, { "epoch": 0.5736470588235294, "grad_norm": 9.170528876616093, "learning_rate": 0.00018285, "loss": 4.9062, "step": 1219 }, { "epoch": 0.5741176470588235, "grad_norm": 10.996720600950539, "learning_rate": 0.00018299999999999998, "loss": 4.9375, "step": 1220 }, { "epoch": 0.5745882352941176, "grad_norm": 10.990318705070772, "learning_rate": 0.00018315, "loss": 3.7344, "step": 1221 }, { "epoch": 0.5750588235294117, "grad_norm": 6.186583162661532, "learning_rate": 0.00018329999999999998, "loss": 3.7656, "step": 1222 }, { "epoch": 0.5755294117647058, "grad_norm": 13.362324312068754, "learning_rate": 0.00018345, "loss": 4.9375, "step": 1223 }, { "epoch": 0.576, "grad_norm": 9.379384534781844, "learning_rate": 0.0001836, "loss": 4.4688, "step": 1224 }, { "epoch": 0.5764705882352941, "grad_norm": 12.148472993050863, "learning_rate": 0.00018375, "loss": 1.6562, "step": 1225 }, { "epoch": 0.5769411764705883, "grad_norm": 15.834945165687392, "learning_rate": 0.00018389999999999997, "loss": 1.9453, "step": 1226 }, { "epoch": 0.5774117647058824, "grad_norm": 1.3932395168121297, "learning_rate": 0.00018405, "loss": 1.7344, "step": 1227 }, { "epoch": 0.5778823529411765, "grad_norm": 14.659869875708992, "learning_rate": 0.00018419999999999998, "loss": 3.6719, "step": 1228 }, { "epoch": 0.5783529411764706, "grad_norm": 5.208741880854496, "learning_rate": 0.00018435, "loss": 1.8594, "step": 1229 }, { "epoch": 0.5788235294117647, "grad_norm": 0.9771546003201599, "learning_rate": 0.00018449999999999999, "loss": 1.1484, "step": 1230 }, { "epoch": 0.5792941176470588, "grad_norm": 4.606801915546807, "learning_rate": 0.00018465, "loss": 2.5938, "step": 1231 }, { "epoch": 0.579764705882353, "grad_norm": 7.4399082417998565, "learning_rate": 0.0001848, "loss": 1.6016, "step": 1232 }, { "epoch": 0.5802352941176471, "grad_norm": 9.737638470658078, "learning_rate": 0.00018495, "loss": 1.4609, "step": 1233 }, { "epoch": 0.5807058823529412, "grad_norm": 12.24575325619218, "learning_rate": 0.00018509999999999997, "loss": 4.0312, "step": 1234 }, { "epoch": 0.5811764705882353, "grad_norm": 2.2275459002635034, "learning_rate": 0.00018525, "loss": 1.3203, "step": 1235 }, { "epoch": 0.5816470588235294, "grad_norm": 6.3117318635893245, "learning_rate": 0.00018539999999999998, "loss": 2.4375, "step": 1236 }, { "epoch": 0.5821176470588235, "grad_norm": 11.766709178290247, "learning_rate": 0.00018555, "loss": 4.75, "step": 1237 }, { "epoch": 0.5825882352941176, "grad_norm": 16.554477384733953, "learning_rate": 0.0001857, "loss": 1.5312, "step": 1238 }, { "epoch": 0.5830588235294117, "grad_norm": 15.5675716726039, "learning_rate": 0.00018585, "loss": 1.7266, "step": 1239 }, { "epoch": 0.5835294117647059, "grad_norm": 2.9427106221749635, "learning_rate": 0.000186, "loss": 2.8125, "step": 1240 }, { "epoch": 0.584, "grad_norm": 8.091596202742442, "learning_rate": 0.00018615, "loss": 2.4375, "step": 1241 }, { "epoch": 0.5844705882352941, "grad_norm": 7.616393510726176, "learning_rate": 0.00018629999999999997, "loss": 2.375, "step": 1242 }, { "epoch": 0.5849411764705882, "grad_norm": 6.0570018034373465, "learning_rate": 0.00018645, "loss": 1.8828, "step": 1243 }, { "epoch": 0.5854117647058823, "grad_norm": 2.0957379490444303, "learning_rate": 0.00018659999999999998, "loss": 1.2344, "step": 1244 }, { "epoch": 0.5858823529411765, "grad_norm": 2.0983766728981603, "learning_rate": 0.00018675, "loss": 1.3984, "step": 1245 }, { "epoch": 0.5863529411764706, "grad_norm": 3.0012360671107468, "learning_rate": 0.0001869, "loss": 1.6172, "step": 1246 }, { "epoch": 0.5868235294117647, "grad_norm": 4.588478925410474, "learning_rate": 0.00018705, "loss": 2.7969, "step": 1247 }, { "epoch": 0.5872941176470589, "grad_norm": 3.782591498241262, "learning_rate": 0.0001872, "loss": 2.4062, "step": 1248 }, { "epoch": 0.587764705882353, "grad_norm": 11.442137690716976, "learning_rate": 0.00018735, "loss": 1.5234, "step": 1249 }, { "epoch": 0.5882352941176471, "grad_norm": 2.5252863364913276, "learning_rate": 0.00018749999999999998, "loss": 2.3906, "step": 1250 }, { "epoch": 0.5887058823529412, "grad_norm": 1.3896453299372746, "learning_rate": 0.00018764999999999998, "loss": 1.375, "step": 1251 }, { "epoch": 0.5891764705882353, "grad_norm": 6.465188383800039, "learning_rate": 0.00018779999999999998, "loss": 2.0938, "step": 1252 }, { "epoch": 0.5896470588235294, "grad_norm": 6.833299319411409, "learning_rate": 0.00018794999999999996, "loss": 1.8672, "step": 1253 }, { "epoch": 0.5901176470588235, "grad_norm": 11.34128675466404, "learning_rate": 0.0001881, "loss": 4.5938, "step": 1254 }, { "epoch": 0.5905882352941176, "grad_norm": 2.357604349682435, "learning_rate": 0.00018824999999999997, "loss": 2.5, "step": 1255 }, { "epoch": 0.5910588235294117, "grad_norm": 8.957211114055811, "learning_rate": 0.00018839999999999997, "loss": 1.8906, "step": 1256 }, { "epoch": 0.5915294117647059, "grad_norm": 9.716938927138445, "learning_rate": 0.00018854999999999998, "loss": 1.8281, "step": 1257 }, { "epoch": 0.592, "grad_norm": 3.675441357512808, "learning_rate": 0.00018869999999999998, "loss": 1.6953, "step": 1258 }, { "epoch": 0.5924705882352941, "grad_norm": 5.848814818232086, "learning_rate": 0.00018884999999999996, "loss": 2.7188, "step": 1259 }, { "epoch": 0.5929411764705882, "grad_norm": 1.086751859551453, "learning_rate": 0.00018899999999999999, "loss": 0.8477, "step": 1260 }, { "epoch": 0.5934117647058823, "grad_norm": 4.808929015965673, "learning_rate": 0.00018914999999999996, "loss": 2.0625, "step": 1261 }, { "epoch": 0.5938823529411764, "grad_norm": 1.7598769176614857, "learning_rate": 0.0001893, "loss": 0.8594, "step": 1262 }, { "epoch": 0.5943529411764706, "grad_norm": 1.5000377482587537, "learning_rate": 0.00018944999999999997, "loss": 1.5625, "step": 1263 }, { "epoch": 0.5948235294117648, "grad_norm": 1.4743320029284075, "learning_rate": 0.00018959999999999997, "loss": 1.2812, "step": 1264 }, { "epoch": 0.5952941176470589, "grad_norm": 4.934552090318183, "learning_rate": 0.00018974999999999998, "loss": 2.2031, "step": 1265 }, { "epoch": 0.595764705882353, "grad_norm": 2.185615652441494, "learning_rate": 0.00018989999999999998, "loss": 1.7344, "step": 1266 }, { "epoch": 0.5962352941176471, "grad_norm": 1.803496398122275, "learning_rate": 0.00019004999999999996, "loss": 2.6875, "step": 1267 }, { "epoch": 0.5967058823529412, "grad_norm": 5.270661273549223, "learning_rate": 0.0001902, "loss": 1.2891, "step": 1268 }, { "epoch": 0.5971764705882353, "grad_norm": 2.390586177187981, "learning_rate": 0.00019034999999999996, "loss": 2.7969, "step": 1269 }, { "epoch": 0.5976470588235294, "grad_norm": 3.3443493740090116, "learning_rate": 0.0001905, "loss": 3.625, "step": 1270 }, { "epoch": 0.5981176470588235, "grad_norm": 2.1378448887286985, "learning_rate": 0.00019064999999999997, "loss": 0.6953, "step": 1271 }, { "epoch": 0.5985882352941176, "grad_norm": 1.808912031818017, "learning_rate": 0.00019079999999999998, "loss": 2.75, "step": 1272 }, { "epoch": 0.5990588235294118, "grad_norm": 5.150535493798967, "learning_rate": 0.00019094999999999998, "loss": 3.6719, "step": 1273 }, { "epoch": 0.5995294117647059, "grad_norm": 4.970059015798898, "learning_rate": 0.00019109999999999998, "loss": 2.2656, "step": 1274 }, { "epoch": 0.6, "grad_norm": 1.6234424774864735, "learning_rate": 0.00019124999999999996, "loss": 1.8281, "step": 1275 }, { "epoch": 0.6004705882352941, "grad_norm": 1.054491900890652, "learning_rate": 0.0001914, "loss": 0.9414, "step": 1276 }, { "epoch": 0.6009411764705882, "grad_norm": 2.2888061508535587, "learning_rate": 0.00019154999999999997, "loss": 1.2344, "step": 1277 }, { "epoch": 0.6014117647058823, "grad_norm": 6.217997023191655, "learning_rate": 0.0001917, "loss": 1.3047, "step": 1278 }, { "epoch": 0.6018823529411764, "grad_norm": 1.5397187918304491, "learning_rate": 0.00019184999999999997, "loss": 1.0, "step": 1279 }, { "epoch": 0.6023529411764705, "grad_norm": 6.583844699354481, "learning_rate": 0.00019199999999999998, "loss": 2.1094, "step": 1280 }, { "epoch": 0.6028235294117648, "grad_norm": 4.0424796397381195, "learning_rate": 0.00019214999999999998, "loss": 1.3672, "step": 1281 }, { "epoch": 0.6032941176470589, "grad_norm": 4.756396484534194, "learning_rate": 0.00019229999999999999, "loss": 3.4219, "step": 1282 }, { "epoch": 0.603764705882353, "grad_norm": 3.693340406279324, "learning_rate": 0.00019244999999999996, "loss": 1.6562, "step": 1283 }, { "epoch": 0.6042352941176471, "grad_norm": 1.9143331992778927, "learning_rate": 0.0001926, "loss": 1.3516, "step": 1284 }, { "epoch": 0.6047058823529412, "grad_norm": 3.51205089009988, "learning_rate": 0.00019274999999999997, "loss": 1.1641, "step": 1285 }, { "epoch": 0.6051764705882353, "grad_norm": 3.1435089130907423, "learning_rate": 0.0001929, "loss": 0.8398, "step": 1286 }, { "epoch": 0.6056470588235294, "grad_norm": 3.131417435122467, "learning_rate": 0.00019304999999999998, "loss": 1.2188, "step": 1287 }, { "epoch": 0.6061176470588235, "grad_norm": 1.2132598837537072, "learning_rate": 0.00019319999999999998, "loss": 1.4688, "step": 1288 }, { "epoch": 0.6065882352941177, "grad_norm": 1.2349576676952456, "learning_rate": 0.00019334999999999998, "loss": 1.3828, "step": 1289 }, { "epoch": 0.6070588235294118, "grad_norm": 5.438588645459011, "learning_rate": 0.0001935, "loss": 2.9219, "step": 1290 }, { "epoch": 0.6075294117647059, "grad_norm": 1.7915222637818715, "learning_rate": 0.00019364999999999996, "loss": 2.3594, "step": 1291 }, { "epoch": 0.608, "grad_norm": 1.0668545083890926, "learning_rate": 0.0001938, "loss": 1.7188, "step": 1292 }, { "epoch": 0.6084705882352941, "grad_norm": 0.8272993019262135, "learning_rate": 0.00019394999999999997, "loss": 0.8398, "step": 1293 }, { "epoch": 0.6089411764705882, "grad_norm": 0.9684561815410431, "learning_rate": 0.0001941, "loss": 0.5391, "step": 1294 }, { "epoch": 0.6094117647058823, "grad_norm": 0.8200501398932668, "learning_rate": 0.00019424999999999998, "loss": 0.3867, "step": 1295 }, { "epoch": 0.6098823529411764, "grad_norm": 1.8280559121214301, "learning_rate": 0.00019439999999999998, "loss": 1.9609, "step": 1296 }, { "epoch": 0.6103529411764705, "grad_norm": 1.4312776666329403, "learning_rate": 0.00019454999999999999, "loss": 2.2812, "step": 1297 }, { "epoch": 0.6108235294117647, "grad_norm": 1.3139150959281043, "learning_rate": 0.0001947, "loss": 1.6406, "step": 1298 }, { "epoch": 0.6112941176470589, "grad_norm": 1.6117912476186074, "learning_rate": 0.00019484999999999997, "loss": 2.5625, "step": 1299 }, { "epoch": 0.611764705882353, "grad_norm": 1.2550360454870473, "learning_rate": 0.000195, "loss": 1.8516, "step": 1300 }, { "epoch": 0.6122352941176471, "grad_norm": 1.5319044861407247, "learning_rate": 0.00019514999999999997, "loss": 2.1719, "step": 1301 }, { "epoch": 0.6127058823529412, "grad_norm": 1.5832522874317743, "learning_rate": 0.00019529999999999998, "loss": 1.5469, "step": 1302 }, { "epoch": 0.6131764705882353, "grad_norm": 1.5196076672247496, "learning_rate": 0.00019544999999999998, "loss": 3.0312, "step": 1303 }, { "epoch": 0.6136470588235294, "grad_norm": 1.9500706503862841, "learning_rate": 0.00019559999999999998, "loss": 1.9375, "step": 1304 }, { "epoch": 0.6141176470588235, "grad_norm": 0.6328137702914746, "learning_rate": 0.00019574999999999996, "loss": 0.2441, "step": 1305 }, { "epoch": 0.6145882352941177, "grad_norm": 1.5024872025741904, "learning_rate": 0.0001959, "loss": 1.9062, "step": 1306 }, { "epoch": 0.6150588235294118, "grad_norm": 0.8616626389431238, "learning_rate": 0.00019604999999999997, "loss": 0.6016, "step": 1307 }, { "epoch": 0.6155294117647059, "grad_norm": 1.6596824278987552, "learning_rate": 0.0001962, "loss": 3.5312, "step": 1308 }, { "epoch": 0.616, "grad_norm": 1.5475689701721878, "learning_rate": 0.00019634999999999998, "loss": 2.2031, "step": 1309 }, { "epoch": 0.6164705882352941, "grad_norm": 1.0221803076641156, "learning_rate": 0.00019649999999999998, "loss": 0.5664, "step": 1310 }, { "epoch": 0.6169411764705882, "grad_norm": 2.140668778724408, "learning_rate": 0.00019664999999999998, "loss": 2.5312, "step": 1311 }, { "epoch": 0.6174117647058823, "grad_norm": 1.1498419844862697, "learning_rate": 0.00019679999999999999, "loss": 1.5625, "step": 1312 }, { "epoch": 0.6178823529411764, "grad_norm": 0.4834209872124955, "learning_rate": 0.00019694999999999996, "loss": 0.1348, "step": 1313 }, { "epoch": 0.6183529411764705, "grad_norm": 1.6026024276591762, "learning_rate": 0.0001971, "loss": 2.4062, "step": 1314 }, { "epoch": 0.6188235294117647, "grad_norm": 1.1795190169805945, "learning_rate": 0.00019724999999999997, "loss": 1.7969, "step": 1315 }, { "epoch": 0.6192941176470588, "grad_norm": 1.2616539946908136, "learning_rate": 0.0001974, "loss": 1.5938, "step": 1316 }, { "epoch": 0.619764705882353, "grad_norm": 1.731706779751924, "learning_rate": 0.00019754999999999998, "loss": 2.1094, "step": 1317 }, { "epoch": 0.6202352941176471, "grad_norm": 0.698544749984617, "learning_rate": 0.00019769999999999998, "loss": 0.4355, "step": 1318 }, { "epoch": 0.6207058823529412, "grad_norm": 0.5431863456776242, "learning_rate": 0.00019784999999999998, "loss": 0.4785, "step": 1319 }, { "epoch": 0.6211764705882353, "grad_norm": 1.010237278639061, "learning_rate": 0.000198, "loss": 0.8711, "step": 1320 }, { "epoch": 0.6216470588235294, "grad_norm": 0.7816925718629162, "learning_rate": 0.00019814999999999996, "loss": 0.6328, "step": 1321 }, { "epoch": 0.6221176470588236, "grad_norm": 1.0162842983506146, "learning_rate": 0.0001983, "loss": 0.5664, "step": 1322 }, { "epoch": 0.6225882352941177, "grad_norm": 1.007477183880227, "learning_rate": 0.00019844999999999997, "loss": 1.1875, "step": 1323 }, { "epoch": 0.6230588235294118, "grad_norm": 1.1317246872613282, "learning_rate": 0.0001986, "loss": 1.125, "step": 1324 }, { "epoch": 0.6235294117647059, "grad_norm": 1.2275221381605346, "learning_rate": 0.00019874999999999998, "loss": 1.0234, "step": 1325 }, { "epoch": 0.624, "grad_norm": 1.224502205767086, "learning_rate": 0.00019889999999999998, "loss": 1.2266, "step": 1326 }, { "epoch": 0.6244705882352941, "grad_norm": 0.990137568484362, "learning_rate": 0.00019905, "loss": 0.5625, "step": 1327 }, { "epoch": 0.6249411764705882, "grad_norm": 1.4117422011393284, "learning_rate": 0.0001992, "loss": 1.1875, "step": 1328 }, { "epoch": 0.6254117647058823, "grad_norm": 1.2998318802070297, "learning_rate": 0.00019934999999999997, "loss": 1.4062, "step": 1329 }, { "epoch": 0.6258823529411764, "grad_norm": 0.6982305356354722, "learning_rate": 0.0001995, "loss": 0.4941, "step": 1330 }, { "epoch": 0.6263529411764706, "grad_norm": 2.441844338186576, "learning_rate": 0.00019964999999999997, "loss": 2.5156, "step": 1331 }, { "epoch": 0.6268235294117647, "grad_norm": 1.0175657713371917, "learning_rate": 0.0001998, "loss": 0.8047, "step": 1332 }, { "epoch": 0.6272941176470588, "grad_norm": 0.8604852797888968, "learning_rate": 0.00019994999999999998, "loss": 0.3906, "step": 1333 }, { "epoch": 0.6277647058823529, "grad_norm": 2.3673150699068954, "learning_rate": 0.00020009999999999998, "loss": 2.3438, "step": 1334 }, { "epoch": 0.6282352941176471, "grad_norm": 1.277978584663187, "learning_rate": 0.00020025, "loss": 1.1406, "step": 1335 }, { "epoch": 0.6287058823529412, "grad_norm": 0.8357595816771352, "learning_rate": 0.0002004, "loss": 0.582, "step": 1336 }, { "epoch": 0.6291764705882353, "grad_norm": 1.3083324508459822, "learning_rate": 0.00020054999999999997, "loss": 0.9453, "step": 1337 }, { "epoch": 0.6296470588235294, "grad_norm": 1.5738530457292614, "learning_rate": 0.0002007, "loss": 2.0, "step": 1338 }, { "epoch": 0.6301176470588236, "grad_norm": 1.4878470577277392, "learning_rate": 0.00020084999999999998, "loss": 1.0078, "step": 1339 }, { "epoch": 0.6305882352941177, "grad_norm": 1.2153946751304117, "learning_rate": 0.000201, "loss": 1.4375, "step": 1340 }, { "epoch": 0.6310588235294118, "grad_norm": 1.433754770541183, "learning_rate": 0.00020114999999999998, "loss": 1.3203, "step": 1341 }, { "epoch": 0.6315294117647059, "grad_norm": 1.0647448441909535, "learning_rate": 0.0002013, "loss": 0.5586, "step": 1342 }, { "epoch": 0.632, "grad_norm": 0.9568047840896265, "learning_rate": 0.00020145, "loss": 0.7422, "step": 1343 }, { "epoch": 0.6324705882352941, "grad_norm": 2.0548119988797584, "learning_rate": 0.0002016, "loss": 3.125, "step": 1344 }, { "epoch": 0.6329411764705882, "grad_norm": 1.6847579252860003, "learning_rate": 0.00020174999999999997, "loss": 2.8438, "step": 1345 }, { "epoch": 0.6334117647058823, "grad_norm": 1.7964238140050173, "learning_rate": 0.0002019, "loss": 2.4688, "step": 1346 }, { "epoch": 0.6338823529411765, "grad_norm": 2.5905970336938577, "learning_rate": 0.00020204999999999998, "loss": 2.8125, "step": 1347 }, { "epoch": 0.6343529411764706, "grad_norm": 1.3430968896516717, "learning_rate": 0.0002022, "loss": 1.2344, "step": 1348 }, { "epoch": 0.6348235294117647, "grad_norm": 2.505211224382069, "learning_rate": 0.00020234999999999999, "loss": 2.1719, "step": 1349 }, { "epoch": 0.6352941176470588, "grad_norm": 2.250722882956878, "learning_rate": 0.0002025, "loss": 2.3438, "step": 1350 }, { "epoch": 0.6357647058823529, "grad_norm": 1.1436369684979377, "learning_rate": 0.00020264999999999997, "loss": 0.8438, "step": 1351 }, { "epoch": 0.636235294117647, "grad_norm": 2.223067004009781, "learning_rate": 0.0002028, "loss": 3.2031, "step": 1352 }, { "epoch": 0.6367058823529411, "grad_norm": 1.5275499741988559, "learning_rate": 0.00020294999999999997, "loss": 1.7344, "step": 1353 }, { "epoch": 0.6371764705882353, "grad_norm": 0.8936741869328302, "learning_rate": 0.0002031, "loss": 0.4121, "step": 1354 }, { "epoch": 0.6376470588235295, "grad_norm": 0.8335866295538953, "learning_rate": 0.00020324999999999998, "loss": 0.75, "step": 1355 }, { "epoch": 0.6381176470588236, "grad_norm": 1.3345523798554002, "learning_rate": 0.00020339999999999998, "loss": 1.8672, "step": 1356 }, { "epoch": 0.6385882352941177, "grad_norm": 0.775726868708015, "learning_rate": 0.00020355, "loss": 0.334, "step": 1357 }, { "epoch": 0.6390588235294118, "grad_norm": 2.164973248465016, "learning_rate": 0.0002037, "loss": 2.4688, "step": 1358 }, { "epoch": 0.6395294117647059, "grad_norm": 2.0226062868081645, "learning_rate": 0.00020384999999999997, "loss": 2.4844, "step": 1359 }, { "epoch": 0.64, "grad_norm": 1.3994216178001753, "learning_rate": 0.000204, "loss": 2.1406, "step": 1360 }, { "epoch": 0.6404705882352941, "grad_norm": 1.8592756651980105, "learning_rate": 0.00020414999999999997, "loss": 1.6328, "step": 1361 }, { "epoch": 0.6409411764705882, "grad_norm": 0.8929778869869025, "learning_rate": 0.0002043, "loss": 0.6406, "step": 1362 }, { "epoch": 0.6414117647058823, "grad_norm": 1.098687499862879, "learning_rate": 0.00020444999999999998, "loss": 1.1953, "step": 1363 }, { "epoch": 0.6418823529411765, "grad_norm": 1.6409314764619458, "learning_rate": 0.00020459999999999999, "loss": 2.3906, "step": 1364 }, { "epoch": 0.6423529411764706, "grad_norm": 1.3216807867533067, "learning_rate": 0.00020475, "loss": 1.2422, "step": 1365 }, { "epoch": 0.6428235294117647, "grad_norm": 1.632596682921491, "learning_rate": 0.0002049, "loss": 1.8125, "step": 1366 }, { "epoch": 0.6432941176470588, "grad_norm": 1.7472956882421469, "learning_rate": 0.00020504999999999997, "loss": 1.4453, "step": 1367 }, { "epoch": 0.6437647058823529, "grad_norm": 1.4354991914002648, "learning_rate": 0.0002052, "loss": 2.0469, "step": 1368 }, { "epoch": 0.644235294117647, "grad_norm": 1.4376270823191664, "learning_rate": 0.00020534999999999998, "loss": 2.2344, "step": 1369 }, { "epoch": 0.6447058823529411, "grad_norm": 1.380630804333649, "learning_rate": 0.0002055, "loss": 0.9805, "step": 1370 }, { "epoch": 0.6451764705882352, "grad_norm": 1.254365287286634, "learning_rate": 0.00020564999999999998, "loss": 1.7422, "step": 1371 }, { "epoch": 0.6456470588235295, "grad_norm": 1.039968194765694, "learning_rate": 0.0002058, "loss": 0.9453, "step": 1372 }, { "epoch": 0.6461176470588236, "grad_norm": 0.9799732783280451, "learning_rate": 0.00020595, "loss": 1.0547, "step": 1373 }, { "epoch": 0.6465882352941177, "grad_norm": 1.5274847010483548, "learning_rate": 0.0002061, "loss": 1.6797, "step": 1374 }, { "epoch": 0.6470588235294118, "grad_norm": 0.9161685426940543, "learning_rate": 0.00020624999999999997, "loss": 0.4824, "step": 1375 }, { "epoch": 0.6475294117647059, "grad_norm": 1.7584097928259101, "learning_rate": 0.00020639999999999998, "loss": 2.8125, "step": 1376 }, { "epoch": 0.648, "grad_norm": 0.9757555091794015, "learning_rate": 0.00020654999999999998, "loss": 0.8555, "step": 1377 }, { "epoch": 0.6484705882352941, "grad_norm": 0.8323742692911722, "learning_rate": 0.00020669999999999996, "loss": 0.1709, "step": 1378 }, { "epoch": 0.6489411764705882, "grad_norm": 0.645649697471499, "learning_rate": 0.00020684999999999999, "loss": 0.5781, "step": 1379 }, { "epoch": 0.6494117647058824, "grad_norm": 1.285562754542017, "learning_rate": 0.00020699999999999996, "loss": 1.2422, "step": 1380 }, { "epoch": 0.6498823529411765, "grad_norm": 2.1461370746901385, "learning_rate": 0.00020715, "loss": 2.4688, "step": 1381 }, { "epoch": 0.6503529411764706, "grad_norm": 0.7738592485882102, "learning_rate": 0.00020729999999999997, "loss": 0.1875, "step": 1382 }, { "epoch": 0.6508235294117647, "grad_norm": 1.9097135309951643, "learning_rate": 0.00020744999999999997, "loss": 1.4453, "step": 1383 }, { "epoch": 0.6512941176470588, "grad_norm": 2.7723063822095417, "learning_rate": 0.00020759999999999998, "loss": 1.625, "step": 1384 }, { "epoch": 0.6517647058823529, "grad_norm": 1.085251338103916, "learning_rate": 0.00020774999999999998, "loss": 0.8984, "step": 1385 }, { "epoch": 0.652235294117647, "grad_norm": 1.9605019623556126, "learning_rate": 0.00020789999999999996, "loss": 2.0938, "step": 1386 }, { "epoch": 0.6527058823529411, "grad_norm": 0.7465797638066002, "learning_rate": 0.00020805, "loss": 0.459, "step": 1387 }, { "epoch": 0.6531764705882352, "grad_norm": 1.492177591281165, "learning_rate": 0.00020819999999999996, "loss": 1.8438, "step": 1388 }, { "epoch": 0.6536470588235294, "grad_norm": 3.4559992892587204, "learning_rate": 0.00020835, "loss": 3.7188, "step": 1389 }, { "epoch": 0.6541176470588236, "grad_norm": 0.8438106892975688, "learning_rate": 0.00020849999999999997, "loss": 0.5117, "step": 1390 }, { "epoch": 0.6545882352941177, "grad_norm": 1.3086455066672436, "learning_rate": 0.00020864999999999998, "loss": 1.9453, "step": 1391 }, { "epoch": 0.6550588235294118, "grad_norm": 1.864801583121659, "learning_rate": 0.00020879999999999998, "loss": 1.6406, "step": 1392 }, { "epoch": 0.6555294117647059, "grad_norm": 1.1837237126630222, "learning_rate": 0.00020894999999999998, "loss": 0.832, "step": 1393 }, { "epoch": 0.656, "grad_norm": 2.0410235404956882, "learning_rate": 0.00020909999999999996, "loss": 1.9453, "step": 1394 }, { "epoch": 0.6564705882352941, "grad_norm": 1.4619691562228858, "learning_rate": 0.00020925, "loss": 1.9609, "step": 1395 }, { "epoch": 0.6569411764705883, "grad_norm": 0.7928238982739924, "learning_rate": 0.00020939999999999997, "loss": 0.4473, "step": 1396 }, { "epoch": 0.6574117647058824, "grad_norm": 1.981179251518406, "learning_rate": 0.00020955, "loss": 1.9766, "step": 1397 }, { "epoch": 0.6578823529411765, "grad_norm": 1.5475172877897718, "learning_rate": 0.00020969999999999997, "loss": 1.4531, "step": 1398 }, { "epoch": 0.6583529411764706, "grad_norm": 1.689851507920826, "learning_rate": 0.00020984999999999998, "loss": 2.3125, "step": 1399 }, { "epoch": 0.6588235294117647, "grad_norm": 1.0680115710810625, "learning_rate": 0.00020999999999999998, "loss": 0.7305, "step": 1400 }, { "epoch": 0.6592941176470588, "grad_norm": 1.9885464859936162, "learning_rate": 0.00021014999999999999, "loss": 2.1719, "step": 1401 }, { "epoch": 0.6597647058823529, "grad_norm": 1.8192377313766697, "learning_rate": 0.00021029999999999996, "loss": 2.875, "step": 1402 }, { "epoch": 0.660235294117647, "grad_norm": 0.8025720528923774, "learning_rate": 0.00021045, "loss": 0.2891, "step": 1403 }, { "epoch": 0.6607058823529411, "grad_norm": 2.175245686333478, "learning_rate": 0.00021059999999999997, "loss": 1.9688, "step": 1404 }, { "epoch": 0.6611764705882353, "grad_norm": 1.601257001672794, "learning_rate": 0.00021074999999999997, "loss": 2.3125, "step": 1405 }, { "epoch": 0.6616470588235294, "grad_norm": 1.5398556740673683, "learning_rate": 0.00021089999999999998, "loss": 1.1953, "step": 1406 }, { "epoch": 0.6621176470588235, "grad_norm": 1.4326913155971892, "learning_rate": 0.00021104999999999998, "loss": 1.2734, "step": 1407 }, { "epoch": 0.6625882352941177, "grad_norm": 1.9276681363527115, "learning_rate": 0.00021119999999999996, "loss": 2.5781, "step": 1408 }, { "epoch": 0.6630588235294118, "grad_norm": 1.4034104064679815, "learning_rate": 0.00021135, "loss": 1.6719, "step": 1409 }, { "epoch": 0.6635294117647059, "grad_norm": 1.100357033764561, "learning_rate": 0.00021149999999999996, "loss": 0.7227, "step": 1410 }, { "epoch": 0.664, "grad_norm": 1.6109938645319417, "learning_rate": 0.00021165, "loss": 2.75, "step": 1411 }, { "epoch": 0.6644705882352941, "grad_norm": 1.9032219261958827, "learning_rate": 0.00021179999999999997, "loss": 1.625, "step": 1412 }, { "epoch": 0.6649411764705883, "grad_norm": 2.1076054103398683, "learning_rate": 0.00021194999999999997, "loss": 3.0156, "step": 1413 }, { "epoch": 0.6654117647058824, "grad_norm": 1.5347283113917058, "learning_rate": 0.00021209999999999998, "loss": 0.8164, "step": 1414 }, { "epoch": 0.6658823529411765, "grad_norm": 1.7521084564738143, "learning_rate": 0.00021224999999999998, "loss": 1.6641, "step": 1415 }, { "epoch": 0.6663529411764706, "grad_norm": 1.391516002459208, "learning_rate": 0.00021239999999999996, "loss": 1.8984, "step": 1416 }, { "epoch": 0.6668235294117647, "grad_norm": 1.3172273801727206, "learning_rate": 0.00021255, "loss": 1.3906, "step": 1417 }, { "epoch": 0.6672941176470588, "grad_norm": 1.4111830889753982, "learning_rate": 0.00021269999999999997, "loss": 1.5625, "step": 1418 }, { "epoch": 0.6677647058823529, "grad_norm": 1.0755289493922937, "learning_rate": 0.00021285, "loss": 0.832, "step": 1419 }, { "epoch": 0.668235294117647, "grad_norm": 1.2911488908442674, "learning_rate": 0.00021299999999999997, "loss": 1.1562, "step": 1420 }, { "epoch": 0.6687058823529411, "grad_norm": 1.9778469549232571, "learning_rate": 0.00021314999999999998, "loss": 1.7734, "step": 1421 }, { "epoch": 0.6691764705882353, "grad_norm": 1.0620941273726903, "learning_rate": 0.00021329999999999998, "loss": 0.8008, "step": 1422 }, { "epoch": 0.6696470588235294, "grad_norm": 1.5151621423489374, "learning_rate": 0.00021344999999999998, "loss": 1.625, "step": 1423 }, { "epoch": 0.6701176470588235, "grad_norm": 1.8316192761462509, "learning_rate": 0.00021359999999999996, "loss": 1.5547, "step": 1424 }, { "epoch": 0.6705882352941176, "grad_norm": 1.6593480266694924, "learning_rate": 0.00021375, "loss": 1.6406, "step": 1425 }, { "epoch": 0.6710588235294118, "grad_norm": 1.046288198924309, "learning_rate": 0.00021389999999999997, "loss": 0.7148, "step": 1426 }, { "epoch": 0.6715294117647059, "grad_norm": 1.0697469275962361, "learning_rate": 0.00021405, "loss": 0.7109, "step": 1427 }, { "epoch": 0.672, "grad_norm": 2.4414624507047638, "learning_rate": 0.00021419999999999998, "loss": 1.9922, "step": 1428 }, { "epoch": 0.6724705882352942, "grad_norm": 1.3684355935246844, "learning_rate": 0.00021434999999999998, "loss": 1.3984, "step": 1429 }, { "epoch": 0.6729411764705883, "grad_norm": 2.734127924284009, "learning_rate": 0.00021449999999999998, "loss": 2.8281, "step": 1430 }, { "epoch": 0.6734117647058824, "grad_norm": 0.9247346117831885, "learning_rate": 0.00021464999999999999, "loss": 0.7656, "step": 1431 }, { "epoch": 0.6738823529411765, "grad_norm": 1.6051875435023237, "learning_rate": 0.00021479999999999996, "loss": 1.7891, "step": 1432 }, { "epoch": 0.6743529411764706, "grad_norm": 1.337731170578227, "learning_rate": 0.00021495, "loss": 1.1016, "step": 1433 }, { "epoch": 0.6748235294117647, "grad_norm": 2.2465509034072735, "learning_rate": 0.00021509999999999997, "loss": 2.3594, "step": 1434 }, { "epoch": 0.6752941176470588, "grad_norm": 0.9931589044440652, "learning_rate": 0.00021525, "loss": 0.5039, "step": 1435 }, { "epoch": 0.6757647058823529, "grad_norm": 0.9556824989613544, "learning_rate": 0.00021539999999999998, "loss": 0.5391, "step": 1436 }, { "epoch": 0.676235294117647, "grad_norm": 3.205618588736833, "learning_rate": 0.00021554999999999998, "loss": 2.75, "step": 1437 }, { "epoch": 0.6767058823529412, "grad_norm": 1.056469353406231, "learning_rate": 0.00021569999999999998, "loss": 0.8242, "step": 1438 }, { "epoch": 0.6771764705882353, "grad_norm": 0.99854708125576, "learning_rate": 0.00021585, "loss": 1.2109, "step": 1439 }, { "epoch": 0.6776470588235294, "grad_norm": 2.5797210711505327, "learning_rate": 0.00021599999999999996, "loss": 3.3125, "step": 1440 }, { "epoch": 0.6781176470588235, "grad_norm": 0.8340078231078543, "learning_rate": 0.00021615, "loss": 0.4707, "step": 1441 }, { "epoch": 0.6785882352941176, "grad_norm": 2.329825335900077, "learning_rate": 0.00021629999999999997, "loss": 2.3281, "step": 1442 }, { "epoch": 0.6790588235294117, "grad_norm": 0.8791687484269752, "learning_rate": 0.00021645, "loss": 0.4004, "step": 1443 }, { "epoch": 0.6795294117647059, "grad_norm": 2.03392980123954, "learning_rate": 0.00021659999999999998, "loss": 2.5, "step": 1444 }, { "epoch": 0.68, "grad_norm": 2.327693617126854, "learning_rate": 0.00021674999999999998, "loss": 2.2969, "step": 1445 }, { "epoch": 0.6804705882352942, "grad_norm": 1.6133847532745378, "learning_rate": 0.0002169, "loss": 1.5234, "step": 1446 }, { "epoch": 0.6809411764705883, "grad_norm": 0.8996203711358411, "learning_rate": 0.00021705, "loss": 0.5508, "step": 1447 }, { "epoch": 0.6814117647058824, "grad_norm": 1.3348279037532134, "learning_rate": 0.00021719999999999997, "loss": 1.9688, "step": 1448 }, { "epoch": 0.6818823529411765, "grad_norm": 1.9847779605973566, "learning_rate": 0.00021735, "loss": 2.5, "step": 1449 }, { "epoch": 0.6823529411764706, "grad_norm": 0.6239024267511352, "learning_rate": 0.00021749999999999997, "loss": 0.1816, "step": 1450 }, { "epoch": 0.6828235294117647, "grad_norm": 2.2450493877144058, "learning_rate": 0.00021764999999999998, "loss": 2.25, "step": 1451 }, { "epoch": 0.6832941176470588, "grad_norm": 1.173937319064627, "learning_rate": 0.00021779999999999998, "loss": 0.7656, "step": 1452 }, { "epoch": 0.6837647058823529, "grad_norm": 0.7801546382524509, "learning_rate": 0.00021794999999999999, "loss": 0.4219, "step": 1453 }, { "epoch": 0.684235294117647, "grad_norm": 1.7277608593503526, "learning_rate": 0.00021809999999999996, "loss": 2.1406, "step": 1454 }, { "epoch": 0.6847058823529412, "grad_norm": 1.4493207873984995, "learning_rate": 0.00021825, "loss": 1.1797, "step": 1455 }, { "epoch": 0.6851764705882353, "grad_norm": 0.5794200174897507, "learning_rate": 0.00021839999999999997, "loss": 0.1777, "step": 1456 }, { "epoch": 0.6856470588235294, "grad_norm": 1.7366136712821805, "learning_rate": 0.00021855, "loss": 1.8906, "step": 1457 }, { "epoch": 0.6861176470588235, "grad_norm": 1.0605127324078836, "learning_rate": 0.00021869999999999998, "loss": 0.6758, "step": 1458 }, { "epoch": 0.6865882352941176, "grad_norm": 1.0676861578729298, "learning_rate": 0.00021884999999999998, "loss": 0.8398, "step": 1459 }, { "epoch": 0.6870588235294117, "grad_norm": 1.1224324808354391, "learning_rate": 0.00021899999999999998, "loss": 1.2734, "step": 1460 }, { "epoch": 0.6875294117647058, "grad_norm": 1.7440888356093442, "learning_rate": 0.00021915, "loss": 1.8984, "step": 1461 }, { "epoch": 0.688, "grad_norm": 1.8141986565912422, "learning_rate": 0.00021929999999999996, "loss": 2.1094, "step": 1462 }, { "epoch": 0.6884705882352942, "grad_norm": 1.4001369105653045, "learning_rate": 0.00021945, "loss": 1.4062, "step": 1463 }, { "epoch": 0.6889411764705883, "grad_norm": 2.944994687723538, "learning_rate": 0.00021959999999999997, "loss": 3.0312, "step": 1464 }, { "epoch": 0.6894117647058824, "grad_norm": 1.1420134326560472, "learning_rate": 0.00021975, "loss": 0.8867, "step": 1465 }, { "epoch": 0.6898823529411765, "grad_norm": 2.396662543038314, "learning_rate": 0.00021989999999999998, "loss": 1.2734, "step": 1466 }, { "epoch": 0.6903529411764706, "grad_norm": 1.8014948888834355, "learning_rate": 0.00022004999999999998, "loss": 0.8984, "step": 1467 }, { "epoch": 0.6908235294117647, "grad_norm": 2.289817028019277, "learning_rate": 0.00022019999999999999, "loss": 0.9297, "step": 1468 }, { "epoch": 0.6912941176470588, "grad_norm": 1.0638647976024704, "learning_rate": 0.00022035, "loss": 0.6562, "step": 1469 }, { "epoch": 0.691764705882353, "grad_norm": 1.566935426956839, "learning_rate": 0.00022049999999999997, "loss": 1.4219, "step": 1470 }, { "epoch": 0.6922352941176471, "grad_norm": 2.2214487955635582, "learning_rate": 0.00022065, "loss": 2.25, "step": 1471 }, { "epoch": 0.6927058823529412, "grad_norm": 1.3889550281586485, "learning_rate": 0.00022079999999999997, "loss": 1.2578, "step": 1472 }, { "epoch": 0.6931764705882353, "grad_norm": 2.2985357924177867, "learning_rate": 0.00022095, "loss": 1.8047, "step": 1473 }, { "epoch": 0.6936470588235294, "grad_norm": 1.509769092079812, "learning_rate": 0.00022109999999999998, "loss": 1.5781, "step": 1474 }, { "epoch": 0.6941176470588235, "grad_norm": 2.294771478254613, "learning_rate": 0.00022124999999999998, "loss": 2.9531, "step": 1475 }, { "epoch": 0.6945882352941176, "grad_norm": 0.9564120299648162, "learning_rate": 0.0002214, "loss": 0.6953, "step": 1476 }, { "epoch": 0.6950588235294117, "grad_norm": 2.153150373825726, "learning_rate": 0.00022155, "loss": 1.9531, "step": 1477 }, { "epoch": 0.6955294117647058, "grad_norm": 1.0740740111888911, "learning_rate": 0.00022169999999999997, "loss": 0.9375, "step": 1478 }, { "epoch": 0.696, "grad_norm": 1.0564228109770686, "learning_rate": 0.00022185, "loss": 0.5547, "step": 1479 }, { "epoch": 0.6964705882352941, "grad_norm": 1.881888123037215, "learning_rate": 0.00022199999999999998, "loss": 1.9062, "step": 1480 }, { "epoch": 0.6969411764705883, "grad_norm": 1.8815787788103868, "learning_rate": 0.00022215, "loss": 2.25, "step": 1481 }, { "epoch": 0.6974117647058824, "grad_norm": 2.2920622208999486, "learning_rate": 0.00022229999999999998, "loss": 2.3438, "step": 1482 }, { "epoch": 0.6978823529411765, "grad_norm": 1.2711898786667801, "learning_rate": 0.00022244999999999999, "loss": 0.7461, "step": 1483 }, { "epoch": 0.6983529411764706, "grad_norm": 1.6146301189465475, "learning_rate": 0.0002226, "loss": 1.8516, "step": 1484 }, { "epoch": 0.6988235294117647, "grad_norm": 1.0477737483432903, "learning_rate": 0.00022275, "loss": 0.7773, "step": 1485 }, { "epoch": 0.6992941176470588, "grad_norm": 1.4043179347237658, "learning_rate": 0.00022289999999999997, "loss": 1.2344, "step": 1486 }, { "epoch": 0.699764705882353, "grad_norm": 2.2529803889402356, "learning_rate": 0.00022305, "loss": 2.0938, "step": 1487 }, { "epoch": 0.7002352941176471, "grad_norm": 2.0599666402458126, "learning_rate": 0.00022319999999999998, "loss": 1.9766, "step": 1488 }, { "epoch": 0.7007058823529412, "grad_norm": 1.6457212421728413, "learning_rate": 0.00022335, "loss": 1.6406, "step": 1489 }, { "epoch": 0.7011764705882353, "grad_norm": 0.8558369186056171, "learning_rate": 0.00022349999999999998, "loss": 0.5508, "step": 1490 }, { "epoch": 0.7016470588235294, "grad_norm": 1.7233109953208154, "learning_rate": 0.00022365, "loss": 1.1094, "step": 1491 }, { "epoch": 0.7021176470588235, "grad_norm": 1.90200712739352, "learning_rate": 0.0002238, "loss": 2.5, "step": 1492 }, { "epoch": 0.7025882352941176, "grad_norm": 0.771203116434876, "learning_rate": 0.00022395, "loss": 0.3164, "step": 1493 }, { "epoch": 0.7030588235294117, "grad_norm": 1.741315770980422, "learning_rate": 0.00022409999999999997, "loss": 1.5469, "step": 1494 }, { "epoch": 0.7035294117647058, "grad_norm": 2.0506622935062944, "learning_rate": 0.00022425, "loss": 1.5469, "step": 1495 }, { "epoch": 0.704, "grad_norm": 0.48150215094086457, "learning_rate": 0.00022439999999999998, "loss": 0.0986, "step": 1496 }, { "epoch": 0.7044705882352941, "grad_norm": 1.8010577309912614, "learning_rate": 0.00022455, "loss": 1.5234, "step": 1497 }, { "epoch": 0.7049411764705882, "grad_norm": 1.115001136577085, "learning_rate": 0.0002247, "loss": 0.6445, "step": 1498 }, { "epoch": 0.7054117647058824, "grad_norm": 1.1448550536068178, "learning_rate": 0.00022485, "loss": 0.3516, "step": 1499 }, { "epoch": 0.7058823529411765, "grad_norm": 2.002466878118243, "learning_rate": 0.000225, "loss": 1.875, "step": 1500 }, { "epoch": 0.7063529411764706, "grad_norm": 2.243390877644146, "learning_rate": 0.00022514999999999997, "loss": 2.1719, "step": 1501 }, { "epoch": 0.7068235294117647, "grad_norm": 0.7933052857721004, "learning_rate": 0.00022529999999999997, "loss": 0.9141, "step": 1502 }, { "epoch": 0.7072941176470589, "grad_norm": 0.8210389385309065, "learning_rate": 0.00022544999999999995, "loss": 0.3691, "step": 1503 }, { "epoch": 0.707764705882353, "grad_norm": 1.2436449108085548, "learning_rate": 0.00022559999999999998, "loss": 0.9336, "step": 1504 }, { "epoch": 0.7082352941176471, "grad_norm": 1.0843690237572787, "learning_rate": 0.00022574999999999996, "loss": 0.5703, "step": 1505 }, { "epoch": 0.7087058823529412, "grad_norm": 2.0005745467191764, "learning_rate": 0.0002259, "loss": 1.2031, "step": 1506 }, { "epoch": 0.7091764705882353, "grad_norm": 1.6576221653086185, "learning_rate": 0.00022604999999999997, "loss": 1.8438, "step": 1507 }, { "epoch": 0.7096470588235294, "grad_norm": 0.7914855770707301, "learning_rate": 0.00022619999999999997, "loss": 0.3867, "step": 1508 }, { "epoch": 0.7101176470588235, "grad_norm": 2.3299303212195497, "learning_rate": 0.00022634999999999997, "loss": 1.7969, "step": 1509 }, { "epoch": 0.7105882352941176, "grad_norm": 0.6732455877600194, "learning_rate": 0.00022649999999999998, "loss": 0.2539, "step": 1510 }, { "epoch": 0.7110588235294117, "grad_norm": 1.7551164347084025, "learning_rate": 0.00022664999999999995, "loss": 1.2031, "step": 1511 }, { "epoch": 0.7115294117647059, "grad_norm": 5.057364880797265, "learning_rate": 0.00022679999999999998, "loss": 1.7734, "step": 1512 }, { "epoch": 0.712, "grad_norm": 1.6127455012739549, "learning_rate": 0.00022694999999999996, "loss": 1.1172, "step": 1513 }, { "epoch": 0.7124705882352941, "grad_norm": 1.4770123396963786, "learning_rate": 0.0002271, "loss": 1.0625, "step": 1514 }, { "epoch": 0.7129411764705882, "grad_norm": 1.8314775616499797, "learning_rate": 0.00022724999999999997, "loss": 2.1719, "step": 1515 }, { "epoch": 0.7134117647058823, "grad_norm": 1.22415625340187, "learning_rate": 0.00022739999999999997, "loss": 0.9688, "step": 1516 }, { "epoch": 0.7138823529411765, "grad_norm": 0.8816228615847451, "learning_rate": 0.00022754999999999997, "loss": 0.4199, "step": 1517 }, { "epoch": 0.7143529411764706, "grad_norm": 1.3667127865359627, "learning_rate": 0.00022769999999999998, "loss": 1.3672, "step": 1518 }, { "epoch": 0.7148235294117647, "grad_norm": 1.7736022765405606, "learning_rate": 0.00022784999999999995, "loss": 1.8281, "step": 1519 }, { "epoch": 0.7152941176470589, "grad_norm": 1.7186572886588192, "learning_rate": 0.00022799999999999999, "loss": 2.2812, "step": 1520 }, { "epoch": 0.715764705882353, "grad_norm": 1.3258637415091505, "learning_rate": 0.00022814999999999996, "loss": 1.6328, "step": 1521 }, { "epoch": 0.7162352941176471, "grad_norm": 1.0543161864957695, "learning_rate": 0.0002283, "loss": 0.9805, "step": 1522 }, { "epoch": 0.7167058823529412, "grad_norm": 1.5464315562485307, "learning_rate": 0.00022844999999999997, "loss": 0.7109, "step": 1523 }, { "epoch": 0.7171764705882353, "grad_norm": 1.5684669204484483, "learning_rate": 0.00022859999999999997, "loss": 0.9883, "step": 1524 }, { "epoch": 0.7176470588235294, "grad_norm": 0.6469908491895732, "learning_rate": 0.00022874999999999998, "loss": 0.207, "step": 1525 }, { "epoch": 0.7181176470588235, "grad_norm": 2.004026336265847, "learning_rate": 0.00022889999999999998, "loss": 1.3906, "step": 1526 }, { "epoch": 0.7185882352941176, "grad_norm": 1.8170979960405542, "learning_rate": 0.00022904999999999996, "loss": 1.9141, "step": 1527 }, { "epoch": 0.7190588235294118, "grad_norm": 1.4372800730974022, "learning_rate": 0.0002292, "loss": 1.5469, "step": 1528 }, { "epoch": 0.7195294117647059, "grad_norm": 0.8942940080145666, "learning_rate": 0.00022934999999999996, "loss": 0.8594, "step": 1529 }, { "epoch": 0.72, "grad_norm": 1.512262933673524, "learning_rate": 0.0002295, "loss": 1.1484, "step": 1530 }, { "epoch": 0.7204705882352941, "grad_norm": 1.457618068476937, "learning_rate": 0.00022964999999999997, "loss": 1.7109, "step": 1531 }, { "epoch": 0.7209411764705882, "grad_norm": 1.0983612541187378, "learning_rate": 0.00022979999999999997, "loss": 1.375, "step": 1532 }, { "epoch": 0.7214117647058823, "grad_norm": 1.4516118679980714, "learning_rate": 0.00022994999999999998, "loss": 1.375, "step": 1533 }, { "epoch": 0.7218823529411764, "grad_norm": 1.1128723438519699, "learning_rate": 0.00023009999999999998, "loss": 0.3789, "step": 1534 }, { "epoch": 0.7223529411764706, "grad_norm": 1.497934739548582, "learning_rate": 0.00023024999999999996, "loss": 1.7188, "step": 1535 }, { "epoch": 0.7228235294117648, "grad_norm": 1.372715174660939, "learning_rate": 0.0002304, "loss": 1.3828, "step": 1536 }, { "epoch": 0.7232941176470589, "grad_norm": 1.3723885549815154, "learning_rate": 0.00023054999999999997, "loss": 1.4297, "step": 1537 }, { "epoch": 0.723764705882353, "grad_norm": 1.0235887706309914, "learning_rate": 0.0002307, "loss": 0.5938, "step": 1538 }, { "epoch": 0.7242352941176471, "grad_norm": 1.8630104414550583, "learning_rate": 0.00023084999999999997, "loss": 2.0469, "step": 1539 }, { "epoch": 0.7247058823529412, "grad_norm": 0.9915032563993458, "learning_rate": 0.00023099999999999998, "loss": 1.1641, "step": 1540 }, { "epoch": 0.7251764705882353, "grad_norm": 0.717858298863884, "learning_rate": 0.00023114999999999998, "loss": 0.2227, "step": 1541 }, { "epoch": 0.7256470588235294, "grad_norm": 2.0450152314309706, "learning_rate": 0.00023129999999999998, "loss": 1.8438, "step": 1542 }, { "epoch": 0.7261176470588235, "grad_norm": 2.0721736412634613, "learning_rate": 0.00023144999999999996, "loss": 2.6094, "step": 1543 }, { "epoch": 0.7265882352941176, "grad_norm": 0.6397941839698859, "learning_rate": 0.0002316, "loss": 0.2344, "step": 1544 }, { "epoch": 0.7270588235294118, "grad_norm": 0.9132668553410773, "learning_rate": 0.00023174999999999997, "loss": 1.0859, "step": 1545 }, { "epoch": 0.7275294117647059, "grad_norm": 1.8548162292064398, "learning_rate": 0.0002319, "loss": 2.4062, "step": 1546 }, { "epoch": 0.728, "grad_norm": 1.410596845458364, "learning_rate": 0.00023204999999999998, "loss": 1.4453, "step": 1547 }, { "epoch": 0.7284705882352941, "grad_norm": 1.7237562931107906, "learning_rate": 0.00023219999999999998, "loss": 1.4844, "step": 1548 }, { "epoch": 0.7289411764705882, "grad_norm": 1.3819354139709086, "learning_rate": 0.00023234999999999998, "loss": 1.2812, "step": 1549 }, { "epoch": 0.7294117647058823, "grad_norm": 1.1726805462869492, "learning_rate": 0.00023249999999999999, "loss": 0.7773, "step": 1550 }, { "epoch": 0.7298823529411764, "grad_norm": 1.485430388540664, "learning_rate": 0.00023264999999999996, "loss": 1.9062, "step": 1551 }, { "epoch": 0.7303529411764705, "grad_norm": 1.376234509381933, "learning_rate": 0.0002328, "loss": 0.9648, "step": 1552 }, { "epoch": 0.7308235294117648, "grad_norm": 1.345762944200761, "learning_rate": 0.00023294999999999997, "loss": 1.4609, "step": 1553 }, { "epoch": 0.7312941176470589, "grad_norm": 1.5286219587648229, "learning_rate": 0.00023309999999999997, "loss": 1.5703, "step": 1554 }, { "epoch": 0.731764705882353, "grad_norm": 1.8647334336544776, "learning_rate": 0.00023324999999999998, "loss": 2.4375, "step": 1555 }, { "epoch": 0.7322352941176471, "grad_norm": 1.3375179053463675, "learning_rate": 0.00023339999999999998, "loss": 1.8516, "step": 1556 }, { "epoch": 0.7327058823529412, "grad_norm": 1.1603943383671513, "learning_rate": 0.00023354999999999996, "loss": 0.7734, "step": 1557 }, { "epoch": 0.7331764705882353, "grad_norm": 1.410258768465599, "learning_rate": 0.0002337, "loss": 1.1328, "step": 1558 }, { "epoch": 0.7336470588235294, "grad_norm": 1.4318068971208664, "learning_rate": 0.00023384999999999997, "loss": 1.2812, "step": 1559 }, { "epoch": 0.7341176470588235, "grad_norm": 1.5221337626758675, "learning_rate": 0.000234, "loss": 1.3828, "step": 1560 }, { "epoch": 0.7345882352941177, "grad_norm": 1.0973267960636235, "learning_rate": 0.00023414999999999997, "loss": 1.0391, "step": 1561 }, { "epoch": 0.7350588235294118, "grad_norm": 2.9161969643041767, "learning_rate": 0.00023429999999999998, "loss": 1.6172, "step": 1562 }, { "epoch": 0.7355294117647059, "grad_norm": 2.823473824012162, "learning_rate": 0.00023444999999999998, "loss": 2.0781, "step": 1563 }, { "epoch": 0.736, "grad_norm": 6.127048125116006, "learning_rate": 0.00023459999999999998, "loss": 1.3594, "step": 1564 }, { "epoch": 0.7364705882352941, "grad_norm": 0.9647008685907865, "learning_rate": 0.00023474999999999996, "loss": 0.4375, "step": 1565 }, { "epoch": 0.7369411764705882, "grad_norm": 1.827622066511383, "learning_rate": 0.0002349, "loss": 1.1562, "step": 1566 }, { "epoch": 0.7374117647058823, "grad_norm": 2.718654564874134, "learning_rate": 0.00023504999999999997, "loss": 2.9062, "step": 1567 }, { "epoch": 0.7378823529411764, "grad_norm": 1.5867506545060592, "learning_rate": 0.0002352, "loss": 1.3203, "step": 1568 }, { "epoch": 0.7383529411764705, "grad_norm": 1.7708237816711876, "learning_rate": 0.00023534999999999997, "loss": 1.3594, "step": 1569 }, { "epoch": 0.7388235294117647, "grad_norm": 1.950407755690421, "learning_rate": 0.00023549999999999998, "loss": 1.4219, "step": 1570 }, { "epoch": 0.7392941176470588, "grad_norm": 1.9657908305653, "learning_rate": 0.00023564999999999998, "loss": 1.5703, "step": 1571 }, { "epoch": 0.739764705882353, "grad_norm": 3.612883362887462, "learning_rate": 0.00023579999999999999, "loss": 1.7344, "step": 1572 }, { "epoch": 0.7402352941176471, "grad_norm": 1.7275762885737662, "learning_rate": 0.00023594999999999996, "loss": 1.5, "step": 1573 }, { "epoch": 0.7407058823529412, "grad_norm": 1.7317822390588622, "learning_rate": 0.0002361, "loss": 1.3203, "step": 1574 }, { "epoch": 0.7411764705882353, "grad_norm": 2.5418209216459418, "learning_rate": 0.00023624999999999997, "loss": 2.9219, "step": 1575 }, { "epoch": 0.7416470588235294, "grad_norm": 1.227362967827865, "learning_rate": 0.0002364, "loss": 0.7422, "step": 1576 }, { "epoch": 0.7421176470588235, "grad_norm": 1.2207874201157425, "learning_rate": 0.00023654999999999998, "loss": 0.75, "step": 1577 }, { "epoch": 0.7425882352941177, "grad_norm": 1.132546316850959, "learning_rate": 0.00023669999999999998, "loss": 0.6484, "step": 1578 }, { "epoch": 0.7430588235294118, "grad_norm": 1.5546424082267905, "learning_rate": 0.00023684999999999998, "loss": 1.3828, "step": 1579 }, { "epoch": 0.7435294117647059, "grad_norm": 1.6559778806318168, "learning_rate": 0.000237, "loss": 1.4609, "step": 1580 }, { "epoch": 0.744, "grad_norm": 0.7856082941580724, "learning_rate": 0.00023714999999999996, "loss": 0.3633, "step": 1581 }, { "epoch": 0.7444705882352941, "grad_norm": 3.2138769219959804, "learning_rate": 0.0002373, "loss": 1.6875, "step": 1582 }, { "epoch": 0.7449411764705882, "grad_norm": 1.1291936150456767, "learning_rate": 0.00023744999999999997, "loss": 0.582, "step": 1583 }, { "epoch": 0.7454117647058823, "grad_norm": 1.1285013912490307, "learning_rate": 0.0002376, "loss": 0.6445, "step": 1584 }, { "epoch": 0.7458823529411764, "grad_norm": 0.8408855396505677, "learning_rate": 0.00023774999999999998, "loss": 0.5625, "step": 1585 }, { "epoch": 0.7463529411764706, "grad_norm": 2.056151207753172, "learning_rate": 0.00023789999999999998, "loss": 1.8594, "step": 1586 }, { "epoch": 0.7468235294117647, "grad_norm": 1.9822558981691043, "learning_rate": 0.00023804999999999999, "loss": 2.1094, "step": 1587 }, { "epoch": 0.7472941176470588, "grad_norm": 2.2919018598412935, "learning_rate": 0.0002382, "loss": 1.7188, "step": 1588 }, { "epoch": 0.7477647058823529, "grad_norm": 0.9559798473573774, "learning_rate": 0.00023834999999999997, "loss": 0.5, "step": 1589 }, { "epoch": 0.7482352941176471, "grad_norm": 1.0086704682274432, "learning_rate": 0.0002385, "loss": 0.4219, "step": 1590 }, { "epoch": 0.7487058823529412, "grad_norm": 1.4875727269532775, "learning_rate": 0.00023864999999999997, "loss": 1.3359, "step": 1591 }, { "epoch": 0.7491764705882353, "grad_norm": 0.7712649599816807, "learning_rate": 0.0002388, "loss": 0.3613, "step": 1592 }, { "epoch": 0.7496470588235294, "grad_norm": 1.6479173214539227, "learning_rate": 0.00023894999999999998, "loss": 1.3359, "step": 1593 }, { "epoch": 0.7501176470588236, "grad_norm": 2.544171273577595, "learning_rate": 0.00023909999999999998, "loss": 3.625, "step": 1594 }, { "epoch": 0.7505882352941177, "grad_norm": 1.9113624450591145, "learning_rate": 0.00023925, "loss": 1.1953, "step": 1595 }, { "epoch": 0.7510588235294118, "grad_norm": 0.5953350856292811, "learning_rate": 0.0002394, "loss": 0.2949, "step": 1596 }, { "epoch": 0.7515294117647059, "grad_norm": 1.833491802772106, "learning_rate": 0.00023954999999999997, "loss": 1.7812, "step": 1597 }, { "epoch": 0.752, "grad_norm": 1.2885414722241315, "learning_rate": 0.0002397, "loss": 0.7227, "step": 1598 }, { "epoch": 0.7524705882352941, "grad_norm": 1.5304114595398233, "learning_rate": 0.00023984999999999998, "loss": 1.75, "step": 1599 }, { "epoch": 0.7529411764705882, "grad_norm": 2.1056275550470667, "learning_rate": 0.00023999999999999998, "loss": 2.1094, "step": 1600 }, { "epoch": 0.7534117647058823, "grad_norm": 1.4333120011289902, "learning_rate": 0.00024014999999999998, "loss": 0.6094, "step": 1601 }, { "epoch": 0.7538823529411764, "grad_norm": 0.6683847361405261, "learning_rate": 0.00024029999999999999, "loss": 0.1602, "step": 1602 }, { "epoch": 0.7543529411764706, "grad_norm": 1.702927388811074, "learning_rate": 0.00024044999999999996, "loss": 2.0625, "step": 1603 }, { "epoch": 0.7548235294117647, "grad_norm": 1.3066847024680368, "learning_rate": 0.0002406, "loss": 1.2109, "step": 1604 }, { "epoch": 0.7552941176470588, "grad_norm": 0.866031860810396, "learning_rate": 0.00024074999999999997, "loss": 0.2734, "step": 1605 }, { "epoch": 0.7557647058823529, "grad_norm": 1.1029141761789774, "learning_rate": 0.0002409, "loss": 0.6367, "step": 1606 }, { "epoch": 0.756235294117647, "grad_norm": 1.198148358896885, "learning_rate": 0.00024104999999999998, "loss": 1.6406, "step": 1607 }, { "epoch": 0.7567058823529412, "grad_norm": 2.007341021163878, "learning_rate": 0.00024119999999999998, "loss": 1.6328, "step": 1608 }, { "epoch": 0.7571764705882353, "grad_norm": 0.9118883446436854, "learning_rate": 0.00024134999999999998, "loss": 0.3887, "step": 1609 }, { "epoch": 0.7576470588235295, "grad_norm": 0.48394426528658346, "learning_rate": 0.0002415, "loss": 0.1553, "step": 1610 }, { "epoch": 0.7581176470588236, "grad_norm": 1.6902988358906939, "learning_rate": 0.00024164999999999996, "loss": 1.6484, "step": 1611 }, { "epoch": 0.7585882352941177, "grad_norm": 1.9779363065188846, "learning_rate": 0.0002418, "loss": 2.0156, "step": 1612 }, { "epoch": 0.7590588235294118, "grad_norm": 1.4494602804207777, "learning_rate": 0.00024194999999999997, "loss": 1.1406, "step": 1613 }, { "epoch": 0.7595294117647059, "grad_norm": 0.9568049336920649, "learning_rate": 0.0002421, "loss": 0.6211, "step": 1614 }, { "epoch": 0.76, "grad_norm": 1.9393886515034995, "learning_rate": 0.00024224999999999998, "loss": 1.2344, "step": 1615 }, { "epoch": 0.7604705882352941, "grad_norm": 0.8172766835717122, "learning_rate": 0.00024239999999999998, "loss": 0.3047, "step": 1616 }, { "epoch": 0.7609411764705882, "grad_norm": 1.2893320942487936, "learning_rate": 0.00024255, "loss": 1.0312, "step": 1617 }, { "epoch": 0.7614117647058823, "grad_norm": 1.107617047218808, "learning_rate": 0.0002427, "loss": 0.543, "step": 1618 }, { "epoch": 0.7618823529411765, "grad_norm": 2.6594585161192303, "learning_rate": 0.00024284999999999997, "loss": 2.2969, "step": 1619 }, { "epoch": 0.7623529411764706, "grad_norm": 1.644519792343315, "learning_rate": 0.000243, "loss": 1.4688, "step": 1620 }, { "epoch": 0.7628235294117647, "grad_norm": 2.385499383662702, "learning_rate": 0.00024314999999999997, "loss": 1.7422, "step": 1621 }, { "epoch": 0.7632941176470588, "grad_norm": 5.181595798760868, "learning_rate": 0.0002433, "loss": 1.4219, "step": 1622 }, { "epoch": 0.7637647058823529, "grad_norm": 1.3199011853752238, "learning_rate": 0.00024344999999999998, "loss": 1.1797, "step": 1623 }, { "epoch": 0.764235294117647, "grad_norm": 1.9746806432617297, "learning_rate": 0.00024359999999999999, "loss": 1.2734, "step": 1624 }, { "epoch": 0.7647058823529411, "grad_norm": 1.3517038635234677, "learning_rate": 0.00024375, "loss": 0.9531, "step": 1625 }, { "epoch": 0.7651764705882353, "grad_norm": 3.1740987089031836, "learning_rate": 0.00024389999999999997, "loss": 2.25, "step": 1626 }, { "epoch": 0.7656470588235295, "grad_norm": 2.1223662399327754, "learning_rate": 0.00024404999999999997, "loss": 1.4453, "step": 1627 }, { "epoch": 0.7661176470588236, "grad_norm": 1.75359090512374, "learning_rate": 0.00024419999999999997, "loss": 0.9453, "step": 1628 }, { "epoch": 0.7665882352941177, "grad_norm": 2.2630424013336667, "learning_rate": 0.00024435, "loss": 1.5547, "step": 1629 }, { "epoch": 0.7670588235294118, "grad_norm": 1.6399694084326204, "learning_rate": 0.0002445, "loss": 1.3672, "step": 1630 }, { "epoch": 0.7675294117647059, "grad_norm": 1.1277491632590553, "learning_rate": 0.00024464999999999996, "loss": 0.5312, "step": 1631 }, { "epoch": 0.768, "grad_norm": 1.7890187056810953, "learning_rate": 0.0002448, "loss": 1.6016, "step": 1632 }, { "epoch": 0.7684705882352941, "grad_norm": 1.4312890055630076, "learning_rate": 0.00024494999999999996, "loss": 1.4688, "step": 1633 }, { "epoch": 0.7689411764705882, "grad_norm": 1.172001193296477, "learning_rate": 0.00024509999999999994, "loss": 0.9805, "step": 1634 }, { "epoch": 0.7694117647058824, "grad_norm": 1.7294262342919235, "learning_rate": 0.00024524999999999997, "loss": 1.7734, "step": 1635 }, { "epoch": 0.7698823529411765, "grad_norm": 1.2802482312692713, "learning_rate": 0.00024539999999999995, "loss": 1.5625, "step": 1636 }, { "epoch": 0.7703529411764706, "grad_norm": 1.2270008654419409, "learning_rate": 0.00024555, "loss": 0.416, "step": 1637 }, { "epoch": 0.7708235294117647, "grad_norm": 1.1541002398397084, "learning_rate": 0.00024569999999999995, "loss": 0.7617, "step": 1638 }, { "epoch": 0.7712941176470588, "grad_norm": 2.5742140191612375, "learning_rate": 0.00024585, "loss": 2.375, "step": 1639 }, { "epoch": 0.7717647058823529, "grad_norm": 1.047245690588938, "learning_rate": 0.00024599999999999996, "loss": 0.6133, "step": 1640 }, { "epoch": 0.772235294117647, "grad_norm": 0.9031580393677878, "learning_rate": 0.00024615, "loss": 0.5078, "step": 1641 }, { "epoch": 0.7727058823529411, "grad_norm": 1.7078382466155801, "learning_rate": 0.00024629999999999997, "loss": 1.1406, "step": 1642 }, { "epoch": 0.7731764705882352, "grad_norm": 1.8143034615800226, "learning_rate": 0.00024645, "loss": 1.4922, "step": 1643 }, { "epoch": 0.7736470588235295, "grad_norm": 2.033303300961258, "learning_rate": 0.0002466, "loss": 2.3906, "step": 1644 }, { "epoch": 0.7741176470588236, "grad_norm": 122.36543041725905, "learning_rate": 0.00024675, "loss": 3.75, "step": 1645 }, { "epoch": 0.7745882352941177, "grad_norm": 1.4397788532610756, "learning_rate": 0.0002469, "loss": 1.6328, "step": 1646 }, { "epoch": 0.7750588235294118, "grad_norm": 0.9580563965781728, "learning_rate": 0.00024704999999999996, "loss": 0.4297, "step": 1647 }, { "epoch": 0.7755294117647059, "grad_norm": 1.5441745057039158, "learning_rate": 0.0002472, "loss": 1.2734, "step": 1648 }, { "epoch": 0.776, "grad_norm": 1.6324834431628839, "learning_rate": 0.00024734999999999997, "loss": 1.0859, "step": 1649 }, { "epoch": 0.7764705882352941, "grad_norm": 3.793041177936619, "learning_rate": 0.00024749999999999994, "loss": 1.9688, "step": 1650 }, { "epoch": 0.7769411764705882, "grad_norm": 1.6661744006160537, "learning_rate": 0.00024765, "loss": 1.2031, "step": 1651 }, { "epoch": 0.7774117647058824, "grad_norm": 1.5402299746043697, "learning_rate": 0.00024779999999999995, "loss": 1.5781, "step": 1652 }, { "epoch": 0.7778823529411765, "grad_norm": 1.3896855357160591, "learning_rate": 0.00024795, "loss": 0.8906, "step": 1653 }, { "epoch": 0.7783529411764706, "grad_norm": 1.026101807086388, "learning_rate": 0.00024809999999999996, "loss": 0.4688, "step": 1654 }, { "epoch": 0.7788235294117647, "grad_norm": 3.813912554776395, "learning_rate": 0.00024825, "loss": 2.3594, "step": 1655 }, { "epoch": 0.7792941176470588, "grad_norm": 2.4872806243488936, "learning_rate": 0.00024839999999999997, "loss": 1.5469, "step": 1656 }, { "epoch": 0.7797647058823529, "grad_norm": 2.636176812131022, "learning_rate": 0.00024855, "loss": 2.5, "step": 1657 }, { "epoch": 0.780235294117647, "grad_norm": 1.4794920484584495, "learning_rate": 0.0002487, "loss": 1.4062, "step": 1658 }, { "epoch": 0.7807058823529411, "grad_norm": 3.464164102464666, "learning_rate": 0.00024885, "loss": 1.5391, "step": 1659 }, { "epoch": 0.7811764705882352, "grad_norm": 2.489042694515551, "learning_rate": 0.000249, "loss": 0.9531, "step": 1660 }, { "epoch": 0.7816470588235294, "grad_norm": 1.5960671374012279, "learning_rate": 0.00024914999999999996, "loss": 0.9219, "step": 1661 }, { "epoch": 0.7821176470588236, "grad_norm": 2.02628891502077, "learning_rate": 0.0002493, "loss": 1.5234, "step": 1662 }, { "epoch": 0.7825882352941177, "grad_norm": 1.0376011783296513, "learning_rate": 0.00024944999999999996, "loss": 0.5312, "step": 1663 }, { "epoch": 0.7830588235294118, "grad_norm": 0.9235661553608794, "learning_rate": 0.00024959999999999994, "loss": 0.5781, "step": 1664 }, { "epoch": 0.7835294117647059, "grad_norm": 1.3511412054401808, "learning_rate": 0.00024974999999999997, "loss": 0.8203, "step": 1665 }, { "epoch": 0.784, "grad_norm": 1.7813783856966914, "learning_rate": 0.00024989999999999995, "loss": 1.7031, "step": 1666 }, { "epoch": 0.7844705882352941, "grad_norm": 1.0188645845560467, "learning_rate": 0.00025005, "loss": 0.5508, "step": 1667 }, { "epoch": 0.7849411764705883, "grad_norm": 2.007300591644676, "learning_rate": 0.00025019999999999996, "loss": 1.9375, "step": 1668 }, { "epoch": 0.7854117647058824, "grad_norm": 1.8393687250078081, "learning_rate": 0.00025035, "loss": 1.8281, "step": 1669 }, { "epoch": 0.7858823529411765, "grad_norm": 0.9786651130820581, "learning_rate": 0.00025049999999999996, "loss": 0.6484, "step": 1670 }, { "epoch": 0.7863529411764706, "grad_norm": 1.0875319238112904, "learning_rate": 0.00025065, "loss": 0.9727, "step": 1671 }, { "epoch": 0.7868235294117647, "grad_norm": 2.226293110793674, "learning_rate": 0.00025079999999999997, "loss": 2.3281, "step": 1672 }, { "epoch": 0.7872941176470588, "grad_norm": 3.329347685743579, "learning_rate": 0.00025095, "loss": 1.7344, "step": 1673 }, { "epoch": 0.7877647058823529, "grad_norm": 1.6220240832144248, "learning_rate": 0.0002511, "loss": 1.5312, "step": 1674 }, { "epoch": 0.788235294117647, "grad_norm": 1.1763280219451535, "learning_rate": 0.00025125, "loss": 1.0234, "step": 1675 }, { "epoch": 0.7887058823529411, "grad_norm": 1.4759938622868143, "learning_rate": 0.0002514, "loss": 0.707, "step": 1676 }, { "epoch": 0.7891764705882353, "grad_norm": 1.320431736698532, "learning_rate": 0.00025154999999999996, "loss": 0.5742, "step": 1677 }, { "epoch": 0.7896470588235294, "grad_norm": 1.939692127661355, "learning_rate": 0.0002517, "loss": 1.5938, "step": 1678 }, { "epoch": 0.7901176470588235, "grad_norm": 1.144832048958532, "learning_rate": 0.00025184999999999997, "loss": 0.4258, "step": 1679 }, { "epoch": 0.7905882352941176, "grad_norm": 1.579794304126324, "learning_rate": 0.00025199999999999995, "loss": 1.0391, "step": 1680 }, { "epoch": 0.7910588235294118, "grad_norm": 1.8042156658797235, "learning_rate": 0.00025215, "loss": 1.5391, "step": 1681 }, { "epoch": 0.7915294117647059, "grad_norm": 1.1909558920700065, "learning_rate": 0.00025229999999999995, "loss": 0.875, "step": 1682 }, { "epoch": 0.792, "grad_norm": 0.8788411554880425, "learning_rate": 0.00025245, "loss": 0.4141, "step": 1683 }, { "epoch": 0.7924705882352941, "grad_norm": 2.1033217407327265, "learning_rate": 0.00025259999999999996, "loss": 2.1094, "step": 1684 }, { "epoch": 0.7929411764705883, "grad_norm": 1.3542822346142636, "learning_rate": 0.00025275, "loss": 1.3047, "step": 1685 }, { "epoch": 0.7934117647058824, "grad_norm": 1.59103894653004, "learning_rate": 0.00025289999999999997, "loss": 1.3203, "step": 1686 }, { "epoch": 0.7938823529411765, "grad_norm": 1.3062294374864978, "learning_rate": 0.00025305, "loss": 0.9648, "step": 1687 }, { "epoch": 0.7943529411764706, "grad_norm": 1.5549166237249772, "learning_rate": 0.0002532, "loss": 1.6406, "step": 1688 }, { "epoch": 0.7948235294117647, "grad_norm": 1.3358385252050642, "learning_rate": 0.00025335, "loss": 1.4922, "step": 1689 }, { "epoch": 0.7952941176470588, "grad_norm": 1.5096025816483338, "learning_rate": 0.0002535, "loss": 1.0234, "step": 1690 }, { "epoch": 0.7957647058823529, "grad_norm": 1.277311134679351, "learning_rate": 0.00025365, "loss": 0.5859, "step": 1691 }, { "epoch": 0.796235294117647, "grad_norm": 23.383803894597744, "learning_rate": 0.0002538, "loss": 1.4141, "step": 1692 }, { "epoch": 0.7967058823529412, "grad_norm": 1.6362063069209536, "learning_rate": 0.00025394999999999997, "loss": 1.5938, "step": 1693 }, { "epoch": 0.7971764705882353, "grad_norm": 1.4170332771030587, "learning_rate": 0.0002541, "loss": 0.7188, "step": 1694 }, { "epoch": 0.7976470588235294, "grad_norm": 1.2302687427051828, "learning_rate": 0.00025425, "loss": 0.9023, "step": 1695 }, { "epoch": 0.7981176470588235, "grad_norm": 1.6480853020623096, "learning_rate": 0.00025439999999999995, "loss": 0.9766, "step": 1696 }, { "epoch": 0.7985882352941176, "grad_norm": 0.9863220047105035, "learning_rate": 0.00025455, "loss": 0.6836, "step": 1697 }, { "epoch": 0.7990588235294117, "grad_norm": 1.4856696476907088, "learning_rate": 0.00025469999999999996, "loss": 1.2656, "step": 1698 }, { "epoch": 0.7995294117647059, "grad_norm": 1.058483286183272, "learning_rate": 0.00025485, "loss": 0.4023, "step": 1699 }, { "epoch": 0.8, "grad_norm": 1.9695340338149352, "learning_rate": 0.00025499999999999996, "loss": 2.2344, "step": 1700 }, { "epoch": 0.8004705882352942, "grad_norm": 2.1465165784436318, "learning_rate": 0.00025515, "loss": 1.9922, "step": 1701 }, { "epoch": 0.8009411764705883, "grad_norm": 1.8101352935971586, "learning_rate": 0.00025529999999999997, "loss": 2.2656, "step": 1702 }, { "epoch": 0.8014117647058824, "grad_norm": 2.117760453976525, "learning_rate": 0.00025545, "loss": 1.9219, "step": 1703 }, { "epoch": 0.8018823529411765, "grad_norm": 0.8533170516800425, "learning_rate": 0.0002556, "loss": 0.5508, "step": 1704 }, { "epoch": 0.8023529411764706, "grad_norm": 1.4637774830717134, "learning_rate": 0.00025575, "loss": 1.1016, "step": 1705 }, { "epoch": 0.8028235294117647, "grad_norm": 1.3555987696315348, "learning_rate": 0.0002559, "loss": 0.7852, "step": 1706 }, { "epoch": 0.8032941176470588, "grad_norm": 2.560771489282394, "learning_rate": 0.00025604999999999996, "loss": 2.3125, "step": 1707 }, { "epoch": 0.8037647058823529, "grad_norm": 1.0306628123876878, "learning_rate": 0.0002562, "loss": 0.5508, "step": 1708 }, { "epoch": 0.804235294117647, "grad_norm": 1.9773685241593426, "learning_rate": 0.00025634999999999997, "loss": 2.0312, "step": 1709 }, { "epoch": 0.8047058823529412, "grad_norm": 1.0619296172202344, "learning_rate": 0.00025649999999999995, "loss": 0.4746, "step": 1710 }, { "epoch": 0.8051764705882353, "grad_norm": 1.0339259284256908, "learning_rate": 0.00025665, "loss": 0.4082, "step": 1711 }, { "epoch": 0.8056470588235294, "grad_norm": 1.3690579744155091, "learning_rate": 0.00025679999999999995, "loss": 0.5898, "step": 1712 }, { "epoch": 0.8061176470588235, "grad_norm": 2.7589802089563333, "learning_rate": 0.00025695, "loss": 1.3906, "step": 1713 }, { "epoch": 0.8065882352941176, "grad_norm": 1.5029409151208086, "learning_rate": 0.00025709999999999996, "loss": 0.9727, "step": 1714 }, { "epoch": 0.8070588235294117, "grad_norm": 1.250860432976604, "learning_rate": 0.00025725, "loss": 0.6992, "step": 1715 }, { "epoch": 0.8075294117647058, "grad_norm": 2.0883991636157977, "learning_rate": 0.00025739999999999997, "loss": 1.5547, "step": 1716 }, { "epoch": 0.808, "grad_norm": 1.3700699386460427, "learning_rate": 0.00025755, "loss": 1.0703, "step": 1717 }, { "epoch": 0.8084705882352942, "grad_norm": 1.6952119726456198, "learning_rate": 0.0002577, "loss": 1.3516, "step": 1718 }, { "epoch": 0.8089411764705883, "grad_norm": 1.7963755095051996, "learning_rate": 0.00025785, "loss": 1.5625, "step": 1719 }, { "epoch": 0.8094117647058824, "grad_norm": 0.2814184314414709, "learning_rate": 0.000258, "loss": 0.042, "step": 1720 }, { "epoch": 0.8098823529411765, "grad_norm": 1.0758585129139648, "learning_rate": 0.00025815, "loss": 0.6367, "step": 1721 }, { "epoch": 0.8103529411764706, "grad_norm": 2.507394793596463, "learning_rate": 0.0002583, "loss": 1.4844, "step": 1722 }, { "epoch": 0.8108235294117647, "grad_norm": 1.723814395219692, "learning_rate": 0.00025844999999999997, "loss": 0.9609, "step": 1723 }, { "epoch": 0.8112941176470588, "grad_norm": 1.1144866187769231, "learning_rate": 0.0002586, "loss": 0.7188, "step": 1724 }, { "epoch": 0.8117647058823529, "grad_norm": 1.3651449995308815, "learning_rate": 0.00025875, "loss": 0.8555, "step": 1725 }, { "epoch": 0.812235294117647, "grad_norm": 1.625059203644096, "learning_rate": 0.00025889999999999995, "loss": 1.6094, "step": 1726 }, { "epoch": 0.8127058823529412, "grad_norm": 2.4355911598783635, "learning_rate": 0.00025905, "loss": 2.8906, "step": 1727 }, { "epoch": 0.8131764705882353, "grad_norm": 1.279477376700869, "learning_rate": 0.00025919999999999996, "loss": 1.1328, "step": 1728 }, { "epoch": 0.8136470588235294, "grad_norm": 1.653836763184889, "learning_rate": 0.00025935, "loss": 1.7969, "step": 1729 }, { "epoch": 0.8141176470588235, "grad_norm": 1.1492424428479786, "learning_rate": 0.00025949999999999997, "loss": 0.6562, "step": 1730 }, { "epoch": 0.8145882352941176, "grad_norm": 1.264155904369615, "learning_rate": 0.00025965, "loss": 1.0625, "step": 1731 }, { "epoch": 0.8150588235294117, "grad_norm": 1.8180620713267293, "learning_rate": 0.00025979999999999997, "loss": 1.3047, "step": 1732 }, { "epoch": 0.8155294117647058, "grad_norm": 1.5157251810135792, "learning_rate": 0.00025995, "loss": 1.5547, "step": 1733 }, { "epoch": 0.816, "grad_norm": 1.4672531328419973, "learning_rate": 0.0002601, "loss": 1.5781, "step": 1734 }, { "epoch": 0.8164705882352942, "grad_norm": 0.8625874754842066, "learning_rate": 0.00026025, "loss": 0.3633, "step": 1735 }, { "epoch": 0.8169411764705883, "grad_norm": 1.4510396951725502, "learning_rate": 0.0002604, "loss": 0.832, "step": 1736 }, { "epoch": 0.8174117647058824, "grad_norm": 0.96606806738535, "learning_rate": 0.00026055, "loss": 0.416, "step": 1737 }, { "epoch": 0.8178823529411765, "grad_norm": 1.2287884183159148, "learning_rate": 0.0002607, "loss": 0.5078, "step": 1738 }, { "epoch": 0.8183529411764706, "grad_norm": 1.8604460331835662, "learning_rate": 0.00026084999999999997, "loss": 1.6094, "step": 1739 }, { "epoch": 0.8188235294117647, "grad_norm": 1.6620069733889482, "learning_rate": 0.000261, "loss": 1.2188, "step": 1740 }, { "epoch": 0.8192941176470588, "grad_norm": 2.1737214899500055, "learning_rate": 0.00026115, "loss": 1.3203, "step": 1741 }, { "epoch": 0.819764705882353, "grad_norm": 1.88268977885947, "learning_rate": 0.00026129999999999995, "loss": 2.1406, "step": 1742 }, { "epoch": 0.8202352941176471, "grad_norm": 1.2787845032059382, "learning_rate": 0.00026145, "loss": 1.0547, "step": 1743 }, { "epoch": 0.8207058823529412, "grad_norm": 2.482232649262694, "learning_rate": 0.00026159999999999996, "loss": 2.5938, "step": 1744 }, { "epoch": 0.8211764705882353, "grad_norm": 2.834076113355291, "learning_rate": 0.00026175, "loss": 2.2188, "step": 1745 }, { "epoch": 0.8216470588235294, "grad_norm": 1.5647822294894491, "learning_rate": 0.00026189999999999997, "loss": 1.4375, "step": 1746 }, { "epoch": 0.8221176470588235, "grad_norm": 1.8418019084734822, "learning_rate": 0.00026205, "loss": 1.6094, "step": 1747 }, { "epoch": 0.8225882352941176, "grad_norm": 1.7837960545962517, "learning_rate": 0.0002622, "loss": 1.7969, "step": 1748 }, { "epoch": 0.8230588235294117, "grad_norm": 0.960657343926191, "learning_rate": 0.00026235, "loss": 0.7812, "step": 1749 }, { "epoch": 0.8235294117647058, "grad_norm": 1.8245928959639706, "learning_rate": 0.0002625, "loss": 1.4922, "step": 1750 }, { "epoch": 0.824, "grad_norm": 1.9444339154053976, "learning_rate": 0.00026264999999999996, "loss": 1.5078, "step": 1751 }, { "epoch": 0.8244705882352941, "grad_norm": 1.0164698317200827, "learning_rate": 0.0002628, "loss": 0.7461, "step": 1752 }, { "epoch": 0.8249411764705883, "grad_norm": 1.8919751204493014, "learning_rate": 0.00026294999999999997, "loss": 1.2812, "step": 1753 }, { "epoch": 0.8254117647058824, "grad_norm": 1.3413478811695063, "learning_rate": 0.0002631, "loss": 0.6602, "step": 1754 }, { "epoch": 0.8258823529411765, "grad_norm": 2.3436247463836346, "learning_rate": 0.00026325, "loss": 1.8359, "step": 1755 }, { "epoch": 0.8263529411764706, "grad_norm": 0.7273343732154273, "learning_rate": 0.00026339999999999995, "loss": 0.3398, "step": 1756 }, { "epoch": 0.8268235294117647, "grad_norm": 1.5970795826870012, "learning_rate": 0.00026355, "loss": 1.5, "step": 1757 }, { "epoch": 0.8272941176470588, "grad_norm": 1.4991928029699326, "learning_rate": 0.00026369999999999996, "loss": 0.9102, "step": 1758 }, { "epoch": 0.827764705882353, "grad_norm": 1.4019505933116991, "learning_rate": 0.00026384999999999994, "loss": 1.0078, "step": 1759 }, { "epoch": 0.8282352941176471, "grad_norm": 1.2746230587316993, "learning_rate": 0.00026399999999999997, "loss": 0.8398, "step": 1760 }, { "epoch": 0.8287058823529412, "grad_norm": 1.7143125611053354, "learning_rate": 0.00026414999999999994, "loss": 1.3828, "step": 1761 }, { "epoch": 0.8291764705882353, "grad_norm": 2.0451135420211424, "learning_rate": 0.0002643, "loss": 1.7578, "step": 1762 }, { "epoch": 0.8296470588235294, "grad_norm": 1.699520621019585, "learning_rate": 0.00026444999999999995, "loss": 1.8359, "step": 1763 }, { "epoch": 0.8301176470588235, "grad_norm": 1.1102976358453671, "learning_rate": 0.0002646, "loss": 0.5664, "step": 1764 }, { "epoch": 0.8305882352941176, "grad_norm": 0.929364136439921, "learning_rate": 0.00026474999999999996, "loss": 0.5156, "step": 1765 }, { "epoch": 0.8310588235294117, "grad_norm": 0.6671766166669685, "learning_rate": 0.0002649, "loss": 0.2021, "step": 1766 }, { "epoch": 0.8315294117647059, "grad_norm": 1.2398524523430796, "learning_rate": 0.00026504999999999996, "loss": 0.5742, "step": 1767 }, { "epoch": 0.832, "grad_norm": 2.321091992697232, "learning_rate": 0.0002652, "loss": 1.2344, "step": 1768 }, { "epoch": 0.8324705882352941, "grad_norm": 3.1680057598230156, "learning_rate": 0.00026534999999999997, "loss": 1.4375, "step": 1769 }, { "epoch": 0.8329411764705882, "grad_norm": 1.6893440626207727, "learning_rate": 0.0002655, "loss": 1.625, "step": 1770 }, { "epoch": 0.8334117647058824, "grad_norm": 1.761814654191755, "learning_rate": 0.00026565, "loss": 2.0312, "step": 1771 }, { "epoch": 0.8338823529411765, "grad_norm": 1.9475911721441697, "learning_rate": 0.00026579999999999996, "loss": 1.5234, "step": 1772 }, { "epoch": 0.8343529411764706, "grad_norm": 1.9638292845350842, "learning_rate": 0.00026595, "loss": 1.0938, "step": 1773 }, { "epoch": 0.8348235294117647, "grad_norm": 1.6467012051910133, "learning_rate": 0.00026609999999999996, "loss": 1.3047, "step": 1774 }, { "epoch": 0.8352941176470589, "grad_norm": 1.1107250791428946, "learning_rate": 0.00026624999999999994, "loss": 0.5859, "step": 1775 }, { "epoch": 0.835764705882353, "grad_norm": 1.397307583499717, "learning_rate": 0.00026639999999999997, "loss": 0.9648, "step": 1776 }, { "epoch": 0.8362352941176471, "grad_norm": 2.1118494774738603, "learning_rate": 0.00026654999999999995, "loss": 2.0938, "step": 1777 }, { "epoch": 0.8367058823529412, "grad_norm": 1.6250118436786058, "learning_rate": 0.0002667, "loss": 1.6016, "step": 1778 }, { "epoch": 0.8371764705882353, "grad_norm": 0.9296864728463674, "learning_rate": 0.00026684999999999995, "loss": 0.7734, "step": 1779 }, { "epoch": 0.8376470588235294, "grad_norm": 0.9872945894226386, "learning_rate": 0.000267, "loss": 0.7344, "step": 1780 }, { "epoch": 0.8381176470588235, "grad_norm": 2.114385722448446, "learning_rate": 0.00026714999999999996, "loss": 1.6328, "step": 1781 }, { "epoch": 0.8385882352941176, "grad_norm": 1.6250024053247314, "learning_rate": 0.0002673, "loss": 1.2969, "step": 1782 }, { "epoch": 0.8390588235294117, "grad_norm": 1.4492690387707339, "learning_rate": 0.00026744999999999997, "loss": 1.0, "step": 1783 }, { "epoch": 0.8395294117647059, "grad_norm": 2.0250356563222702, "learning_rate": 0.0002676, "loss": 1.0547, "step": 1784 }, { "epoch": 0.84, "grad_norm": 1.3053439481411022, "learning_rate": 0.00026775, "loss": 0.9805, "step": 1785 }, { "epoch": 0.8404705882352941, "grad_norm": 0.8325571228127052, "learning_rate": 0.0002679, "loss": 0.4375, "step": 1786 }, { "epoch": 0.8409411764705882, "grad_norm": 1.8702852272908912, "learning_rate": 0.00026805, "loss": 1.5703, "step": 1787 }, { "epoch": 0.8414117647058823, "grad_norm": 1.4786358353038511, "learning_rate": 0.00026819999999999996, "loss": 1.1328, "step": 1788 }, { "epoch": 0.8418823529411765, "grad_norm": 1.7424824611559537, "learning_rate": 0.00026835, "loss": 1.2109, "step": 1789 }, { "epoch": 0.8423529411764706, "grad_norm": 2.179808690066035, "learning_rate": 0.00026849999999999997, "loss": 1.8828, "step": 1790 }, { "epoch": 0.8428235294117647, "grad_norm": 1.1827210116050284, "learning_rate": 0.00026864999999999994, "loss": 1.3594, "step": 1791 }, { "epoch": 0.8432941176470589, "grad_norm": 0.7408096393003057, "learning_rate": 0.0002688, "loss": 0.3066, "step": 1792 }, { "epoch": 0.843764705882353, "grad_norm": 0.9730048850237704, "learning_rate": 0.00026894999999999995, "loss": 0.5391, "step": 1793 }, { "epoch": 0.8442352941176471, "grad_norm": 2.8390320528235566, "learning_rate": 0.0002691, "loss": 2.4688, "step": 1794 }, { "epoch": 0.8447058823529412, "grad_norm": 0.7321845780102332, "learning_rate": 0.00026924999999999996, "loss": 0.2754, "step": 1795 }, { "epoch": 0.8451764705882353, "grad_norm": 2.119075993486297, "learning_rate": 0.0002694, "loss": 2.625, "step": 1796 }, { "epoch": 0.8456470588235294, "grad_norm": 1.4085790370070221, "learning_rate": 0.00026954999999999997, "loss": 1.1328, "step": 1797 }, { "epoch": 0.8461176470588235, "grad_norm": 1.306110641884099, "learning_rate": 0.0002697, "loss": 1.1719, "step": 1798 }, { "epoch": 0.8465882352941176, "grad_norm": 1.4986776952065053, "learning_rate": 0.00026984999999999997, "loss": 1.7891, "step": 1799 }, { "epoch": 0.8470588235294118, "grad_norm": 1.7845780252563157, "learning_rate": 0.00027, "loss": 1.7031, "step": 1800 }, { "epoch": 0.8475294117647059, "grad_norm": 0.8792743715801468, "learning_rate": 0.00027015, "loss": 0.4922, "step": 1801 }, { "epoch": 0.848, "grad_norm": 0.5581629630069683, "learning_rate": 0.00027029999999999996, "loss": 0.2002, "step": 1802 }, { "epoch": 0.8484705882352941, "grad_norm": 2.0572104169981538, "learning_rate": 0.00027045, "loss": 2.125, "step": 1803 }, { "epoch": 0.8489411764705882, "grad_norm": 1.446037353076896, "learning_rate": 0.00027059999999999996, "loss": 1.1484, "step": 1804 }, { "epoch": 0.8494117647058823, "grad_norm": 1.511919037962428, "learning_rate": 0.00027074999999999994, "loss": 1.2188, "step": 1805 }, { "epoch": 0.8498823529411764, "grad_norm": 2.1342443496689762, "learning_rate": 0.00027089999999999997, "loss": 1.6953, "step": 1806 }, { "epoch": 0.8503529411764705, "grad_norm": 2.260505207813265, "learning_rate": 0.00027104999999999995, "loss": 1.7422, "step": 1807 }, { "epoch": 0.8508235294117648, "grad_norm": 2.0870471222127382, "learning_rate": 0.0002712, "loss": 1.5469, "step": 1808 }, { "epoch": 0.8512941176470589, "grad_norm": 0.7317239977518591, "learning_rate": 0.00027134999999999995, "loss": 0.2197, "step": 1809 }, { "epoch": 0.851764705882353, "grad_norm": 1.4488154172313332, "learning_rate": 0.0002715, "loss": 1.1875, "step": 1810 }, { "epoch": 0.8522352941176471, "grad_norm": 1.4745456992810886, "learning_rate": 0.00027164999999999996, "loss": 1.2344, "step": 1811 }, { "epoch": 0.8527058823529412, "grad_norm": 1.7120949946483182, "learning_rate": 0.0002718, "loss": 1.3281, "step": 1812 }, { "epoch": 0.8531764705882353, "grad_norm": 1.6075735055226321, "learning_rate": 0.00027194999999999997, "loss": 1.8438, "step": 1813 }, { "epoch": 0.8536470588235294, "grad_norm": 1.6762866861991366, "learning_rate": 0.0002721, "loss": 1.5391, "step": 1814 }, { "epoch": 0.8541176470588235, "grad_norm": 1.0388238682238977, "learning_rate": 0.00027225, "loss": 0.4961, "step": 1815 }, { "epoch": 0.8545882352941176, "grad_norm": 1.1463310996242728, "learning_rate": 0.0002724, "loss": 0.7188, "step": 1816 }, { "epoch": 0.8550588235294118, "grad_norm": 1.3848938054294877, "learning_rate": 0.00027255, "loss": 1.0547, "step": 1817 }, { "epoch": 0.8555294117647059, "grad_norm": 1.4873215964734678, "learning_rate": 0.00027269999999999996, "loss": 1.375, "step": 1818 }, { "epoch": 0.856, "grad_norm": 0.9800358368086188, "learning_rate": 0.00027285, "loss": 0.9219, "step": 1819 }, { "epoch": 0.8564705882352941, "grad_norm": 0.999630196376986, "learning_rate": 0.00027299999999999997, "loss": 0.9531, "step": 1820 }, { "epoch": 0.8569411764705882, "grad_norm": 2.5505415472021893, "learning_rate": 0.00027314999999999994, "loss": 2.25, "step": 1821 }, { "epoch": 0.8574117647058823, "grad_norm": 1.303558516222661, "learning_rate": 0.0002733, "loss": 0.8594, "step": 1822 }, { "epoch": 0.8578823529411764, "grad_norm": 1.7005778058787901, "learning_rate": 0.00027344999999999995, "loss": 2.3125, "step": 1823 }, { "epoch": 0.8583529411764705, "grad_norm": 1.0514857770200314, "learning_rate": 0.0002736, "loss": 0.6797, "step": 1824 }, { "epoch": 0.8588235294117647, "grad_norm": 0.8547484430516051, "learning_rate": 0.00027374999999999996, "loss": 0.5703, "step": 1825 }, { "epoch": 0.8592941176470589, "grad_norm": 1.855563736353677, "learning_rate": 0.0002739, "loss": 1.8125, "step": 1826 }, { "epoch": 0.859764705882353, "grad_norm": 0.9462397967885745, "learning_rate": 0.00027404999999999997, "loss": 0.4258, "step": 1827 }, { "epoch": 0.8602352941176471, "grad_norm": 1.5305001536094316, "learning_rate": 0.0002742, "loss": 1.4531, "step": 1828 }, { "epoch": 0.8607058823529412, "grad_norm": 2.0960439788342717, "learning_rate": 0.00027435, "loss": 1.75, "step": 1829 }, { "epoch": 0.8611764705882353, "grad_norm": 0.780894428849561, "learning_rate": 0.0002745, "loss": 0.3613, "step": 1830 }, { "epoch": 0.8616470588235294, "grad_norm": 2.1204433090441923, "learning_rate": 0.00027465, "loss": 2.2812, "step": 1831 }, { "epoch": 0.8621176470588235, "grad_norm": 1.5115727022488876, "learning_rate": 0.0002748, "loss": 0.9844, "step": 1832 }, { "epoch": 0.8625882352941177, "grad_norm": 1.308319540936614, "learning_rate": 0.00027495, "loss": 1.5547, "step": 1833 }, { "epoch": 0.8630588235294118, "grad_norm": 1.3475874083964217, "learning_rate": 0.00027509999999999996, "loss": 1.2734, "step": 1834 }, { "epoch": 0.8635294117647059, "grad_norm": 1.7888288202759663, "learning_rate": 0.00027525, "loss": 1.2578, "step": 1835 }, { "epoch": 0.864, "grad_norm": 1.771244451119353, "learning_rate": 0.00027539999999999997, "loss": 1.9609, "step": 1836 }, { "epoch": 0.8644705882352941, "grad_norm": 0.5025089406442049, "learning_rate": 0.00027554999999999995, "loss": 0.0635, "step": 1837 }, { "epoch": 0.8649411764705882, "grad_norm": 1.7913030364678104, "learning_rate": 0.0002757, "loss": 2.0625, "step": 1838 }, { "epoch": 0.8654117647058823, "grad_norm": 1.542036624808241, "learning_rate": 0.00027584999999999996, "loss": 1.7422, "step": 1839 }, { "epoch": 0.8658823529411764, "grad_norm": 0.8709114309892164, "learning_rate": 0.000276, "loss": 0.3711, "step": 1840 }, { "epoch": 0.8663529411764705, "grad_norm": 1.0744447935664836, "learning_rate": 0.00027614999999999996, "loss": 0.5508, "step": 1841 }, { "epoch": 0.8668235294117647, "grad_norm": 1.318671316481504, "learning_rate": 0.0002763, "loss": 0.8945, "step": 1842 }, { "epoch": 0.8672941176470588, "grad_norm": 2.0367957347342394, "learning_rate": 0.00027644999999999997, "loss": 2.4062, "step": 1843 }, { "epoch": 0.867764705882353, "grad_norm": 1.6145452929983926, "learning_rate": 0.0002766, "loss": 1.3906, "step": 1844 }, { "epoch": 0.8682352941176471, "grad_norm": 1.4546468582557657, "learning_rate": 0.00027675, "loss": 1.4531, "step": 1845 }, { "epoch": 0.8687058823529412, "grad_norm": 1.5011215084381482, "learning_rate": 0.0002769, "loss": 1.2031, "step": 1846 }, { "epoch": 0.8691764705882353, "grad_norm": 1.187205242036273, "learning_rate": 0.00027705, "loss": 0.7344, "step": 1847 }, { "epoch": 0.8696470588235294, "grad_norm": 1.1900087877023122, "learning_rate": 0.0002772, "loss": 0.5469, "step": 1848 }, { "epoch": 0.8701176470588236, "grad_norm": 1.4443382012817827, "learning_rate": 0.00027735, "loss": 0.8086, "step": 1849 }, { "epoch": 0.8705882352941177, "grad_norm": 1.8446205528221018, "learning_rate": 0.00027749999999999997, "loss": 1.6016, "step": 1850 }, { "epoch": 0.8710588235294118, "grad_norm": 1.4878512026567572, "learning_rate": 0.00027764999999999995, "loss": 1.4609, "step": 1851 }, { "epoch": 0.8715294117647059, "grad_norm": 1.3665584413290472, "learning_rate": 0.0002778, "loss": 1.2422, "step": 1852 }, { "epoch": 0.872, "grad_norm": 0.9842596511047415, "learning_rate": 0.00027794999999999995, "loss": 0.5273, "step": 1853 }, { "epoch": 0.8724705882352941, "grad_norm": 0.9752734790731974, "learning_rate": 0.0002781, "loss": 0.6016, "step": 1854 }, { "epoch": 0.8729411764705882, "grad_norm": 1.6737167411765417, "learning_rate": 0.00027824999999999996, "loss": 1.2109, "step": 1855 }, { "epoch": 0.8734117647058823, "grad_norm": 1.2591688359543074, "learning_rate": 0.0002784, "loss": 0.9414, "step": 1856 }, { "epoch": 0.8738823529411764, "grad_norm": 1.4855729761531027, "learning_rate": 0.00027854999999999997, "loss": 1.2109, "step": 1857 }, { "epoch": 0.8743529411764706, "grad_norm": 1.0499467038528687, "learning_rate": 0.0002787, "loss": 0.8672, "step": 1858 }, { "epoch": 0.8748235294117647, "grad_norm": 1.0014536655678874, "learning_rate": 0.00027885, "loss": 0.3516, "step": 1859 }, { "epoch": 0.8752941176470588, "grad_norm": 1.869313875881754, "learning_rate": 0.000279, "loss": 2.0938, "step": 1860 }, { "epoch": 0.8757647058823529, "grad_norm": 1.0171717201397186, "learning_rate": 0.00027915, "loss": 0.8164, "step": 1861 }, { "epoch": 0.8762352941176471, "grad_norm": 1.48363077619619, "learning_rate": 0.0002793, "loss": 1.1484, "step": 1862 }, { "epoch": 0.8767058823529412, "grad_norm": 1.2981764815173074, "learning_rate": 0.00027945, "loss": 0.918, "step": 1863 }, { "epoch": 0.8771764705882353, "grad_norm": 1.8940727218334412, "learning_rate": 0.00027959999999999997, "loss": 2.2031, "step": 1864 }, { "epoch": 0.8776470588235294, "grad_norm": 1.3458441601562534, "learning_rate": 0.00027975, "loss": 0.8828, "step": 1865 }, { "epoch": 0.8781176470588236, "grad_norm": 1.6774290163167018, "learning_rate": 0.0002799, "loss": 1.4531, "step": 1866 }, { "epoch": 0.8785882352941177, "grad_norm": 1.7344989960860473, "learning_rate": 0.00028004999999999995, "loss": 1.9922, "step": 1867 }, { "epoch": 0.8790588235294118, "grad_norm": 0.7662515211087357, "learning_rate": 0.0002802, "loss": 0.8477, "step": 1868 }, { "epoch": 0.8795294117647059, "grad_norm": 1.6017214192626026, "learning_rate": 0.00028034999999999996, "loss": 1.6641, "step": 1869 }, { "epoch": 0.88, "grad_norm": 1.9500281515450548, "learning_rate": 0.0002805, "loss": 2.0156, "step": 1870 }, { "epoch": 0.8804705882352941, "grad_norm": 1.064997325404619, "learning_rate": 0.00028064999999999996, "loss": 0.5664, "step": 1871 }, { "epoch": 0.8809411764705882, "grad_norm": 1.0356182467716502, "learning_rate": 0.0002808, "loss": 0.6367, "step": 1872 }, { "epoch": 0.8814117647058823, "grad_norm": 1.4955998548826168, "learning_rate": 0.00028094999999999997, "loss": 1.1016, "step": 1873 }, { "epoch": 0.8818823529411765, "grad_norm": 0.6545893976738184, "learning_rate": 0.0002811, "loss": 0.1582, "step": 1874 }, { "epoch": 0.8823529411764706, "grad_norm": 1.3969529869896722, "learning_rate": 0.00028125, "loss": 1.5703, "step": 1875 }, { "epoch": 0.8828235294117647, "grad_norm": 1.6515751566009065, "learning_rate": 0.00028139999999999996, "loss": 1.2969, "step": 1876 }, { "epoch": 0.8832941176470588, "grad_norm": 1.6795021809543373, "learning_rate": 0.00028155, "loss": 1.1172, "step": 1877 }, { "epoch": 0.8837647058823529, "grad_norm": 0.6894705918503777, "learning_rate": 0.00028169999999999996, "loss": 0.3027, "step": 1878 }, { "epoch": 0.884235294117647, "grad_norm": 1.902410187957115, "learning_rate": 0.00028185, "loss": 1.4141, "step": 1879 }, { "epoch": 0.8847058823529412, "grad_norm": 0.9331832862538378, "learning_rate": 0.00028199999999999997, "loss": 0.3535, "step": 1880 }, { "epoch": 0.8851764705882353, "grad_norm": 2.161896723136463, "learning_rate": 0.00028215, "loss": 2.0, "step": 1881 }, { "epoch": 0.8856470588235295, "grad_norm": 1.5737964870592913, "learning_rate": 0.0002823, "loss": 1.5312, "step": 1882 }, { "epoch": 0.8861176470588236, "grad_norm": 0.8623933683881055, "learning_rate": 0.00028244999999999995, "loss": 0.6133, "step": 1883 }, { "epoch": 0.8865882352941177, "grad_norm": 0.9310853544322704, "learning_rate": 0.0002826, "loss": 0.5859, "step": 1884 }, { "epoch": 0.8870588235294118, "grad_norm": 0.7580839520205144, "learning_rate": 0.00028274999999999996, "loss": 0.2578, "step": 1885 }, { "epoch": 0.8875294117647059, "grad_norm": 1.761282479818138, "learning_rate": 0.00028289999999999994, "loss": 1.3672, "step": 1886 }, { "epoch": 0.888, "grad_norm": 0.5998948565408597, "learning_rate": 0.00028304999999999997, "loss": 0.2285, "step": 1887 }, { "epoch": 0.8884705882352941, "grad_norm": 1.2694302293558757, "learning_rate": 0.00028319999999999994, "loss": 1.0312, "step": 1888 }, { "epoch": 0.8889411764705882, "grad_norm": 1.281969031457098, "learning_rate": 0.00028335, "loss": 1.4297, "step": 1889 }, { "epoch": 0.8894117647058823, "grad_norm": 1.1008096601462596, "learning_rate": 0.00028349999999999995, "loss": 0.6641, "step": 1890 }, { "epoch": 0.8898823529411765, "grad_norm": 1.0048979359087702, "learning_rate": 0.00028365, "loss": 0.6172, "step": 1891 }, { "epoch": 0.8903529411764706, "grad_norm": 1.3259325180374282, "learning_rate": 0.00028379999999999996, "loss": 0.7578, "step": 1892 }, { "epoch": 0.8908235294117647, "grad_norm": 1.3259399981344788, "learning_rate": 0.00028395, "loss": 1.1016, "step": 1893 }, { "epoch": 0.8912941176470588, "grad_norm": 1.7222564799254751, "learning_rate": 0.00028409999999999997, "loss": 1.6641, "step": 1894 }, { "epoch": 0.8917647058823529, "grad_norm": 0.6583452946595615, "learning_rate": 0.00028425, "loss": 0.1748, "step": 1895 }, { "epoch": 0.892235294117647, "grad_norm": 1.135327520910731, "learning_rate": 0.0002844, "loss": 0.6016, "step": 1896 }, { "epoch": 0.8927058823529411, "grad_norm": 1.5825127728790436, "learning_rate": 0.00028455, "loss": 1.5859, "step": 1897 }, { "epoch": 0.8931764705882353, "grad_norm": 1.5837813036562478, "learning_rate": 0.0002847, "loss": 0.918, "step": 1898 }, { "epoch": 0.8936470588235295, "grad_norm": 1.2327556516601574, "learning_rate": 0.00028484999999999996, "loss": 0.8281, "step": 1899 }, { "epoch": 0.8941176470588236, "grad_norm": 1.8521745895922737, "learning_rate": 0.000285, "loss": 1.3516, "step": 1900 }, { "epoch": 0.8945882352941177, "grad_norm": 1.482014758267969, "learning_rate": 0.00028514999999999997, "loss": 1.4844, "step": 1901 }, { "epoch": 0.8950588235294118, "grad_norm": 1.516667528691898, "learning_rate": 0.00028529999999999994, "loss": 1.2266, "step": 1902 }, { "epoch": 0.8955294117647059, "grad_norm": 1.2900224369610698, "learning_rate": 0.00028544999999999997, "loss": 1.1953, "step": 1903 }, { "epoch": 0.896, "grad_norm": 0.7913621729970252, "learning_rate": 0.00028559999999999995, "loss": 0.3066, "step": 1904 }, { "epoch": 0.8964705882352941, "grad_norm": 2.106465112485022, "learning_rate": 0.00028575, "loss": 1.2812, "step": 1905 }, { "epoch": 0.8969411764705882, "grad_norm": 1.7340692431553768, "learning_rate": 0.00028589999999999996, "loss": 1.4922, "step": 1906 }, { "epoch": 0.8974117647058824, "grad_norm": 1.3882224786477977, "learning_rate": 0.00028605, "loss": 1.0312, "step": 1907 }, { "epoch": 0.8978823529411765, "grad_norm": 1.048271240007174, "learning_rate": 0.00028619999999999996, "loss": 0.6641, "step": 1908 }, { "epoch": 0.8983529411764706, "grad_norm": 2.1953069437284567, "learning_rate": 0.00028635, "loss": 1.9453, "step": 1909 }, { "epoch": 0.8988235294117647, "grad_norm": 1.2171330307164887, "learning_rate": 0.00028649999999999997, "loss": 0.9922, "step": 1910 }, { "epoch": 0.8992941176470588, "grad_norm": 2.046737659508775, "learning_rate": 0.00028665, "loss": 2.0938, "step": 1911 }, { "epoch": 0.8997647058823529, "grad_norm": 2.2161750571097816, "learning_rate": 0.0002868, "loss": 1.625, "step": 1912 }, { "epoch": 0.900235294117647, "grad_norm": 2.232197333381478, "learning_rate": 0.00028694999999999995, "loss": 1.5312, "step": 1913 }, { "epoch": 0.9007058823529411, "grad_norm": 2.8744597987253226, "learning_rate": 0.0002871, "loss": 2.0625, "step": 1914 }, { "epoch": 0.9011764705882352, "grad_norm": 1.5242315911993316, "learning_rate": 0.00028724999999999996, "loss": 1.2891, "step": 1915 }, { "epoch": 0.9016470588235294, "grad_norm": 1.511978846430635, "learning_rate": 0.00028739999999999994, "loss": 0.6758, "step": 1916 }, { "epoch": 0.9021176470588236, "grad_norm": 1.6536391237754928, "learning_rate": 0.00028754999999999997, "loss": 0.7383, "step": 1917 }, { "epoch": 0.9025882352941177, "grad_norm": 0.8237025807477955, "learning_rate": 0.00028769999999999995, "loss": 0.4473, "step": 1918 }, { "epoch": 0.9030588235294118, "grad_norm": 3.168969447117972, "learning_rate": 0.00028785, "loss": 2.375, "step": 1919 }, { "epoch": 0.9035294117647059, "grad_norm": 0.8396244598272764, "learning_rate": 0.00028799999999999995, "loss": 0.3418, "step": 1920 }, { "epoch": 0.904, "grad_norm": 2.0928551901990184, "learning_rate": 0.00028815, "loss": 2.0312, "step": 1921 }, { "epoch": 0.9044705882352941, "grad_norm": 1.3862523876709587, "learning_rate": 0.00028829999999999996, "loss": 1.0469, "step": 1922 }, { "epoch": 0.9049411764705882, "grad_norm": 1.8355156768240217, "learning_rate": 0.00028845, "loss": 1.1719, "step": 1923 }, { "epoch": 0.9054117647058824, "grad_norm": 2.0639062431637103, "learning_rate": 0.00028859999999999997, "loss": 1.3828, "step": 1924 }, { "epoch": 0.9058823529411765, "grad_norm": 1.5540174170414134, "learning_rate": 0.00028875, "loss": 1.1953, "step": 1925 }, { "epoch": 0.9063529411764706, "grad_norm": 2.9170079580365393, "learning_rate": 0.0002889, "loss": 3.5156, "step": 1926 }, { "epoch": 0.9068235294117647, "grad_norm": 1.482405890903312, "learning_rate": 0.00028905, "loss": 1.4141, "step": 1927 }, { "epoch": 0.9072941176470588, "grad_norm": 0.6751005793675615, "learning_rate": 0.0002892, "loss": 0.2217, "step": 1928 }, { "epoch": 0.9077647058823529, "grad_norm": 1.3466416398477714, "learning_rate": 0.00028934999999999996, "loss": 1.0312, "step": 1929 }, { "epoch": 0.908235294117647, "grad_norm": 0.5490493599020242, "learning_rate": 0.0002895, "loss": 0.1079, "step": 1930 }, { "epoch": 0.9087058823529411, "grad_norm": 1.0140331262366542, "learning_rate": 0.00028964999999999997, "loss": 0.6523, "step": 1931 }, { "epoch": 0.9091764705882353, "grad_norm": 1.6640706156039278, "learning_rate": 0.00028979999999999994, "loss": 1.0859, "step": 1932 }, { "epoch": 0.9096470588235294, "grad_norm": 1.5269897376536785, "learning_rate": 0.00028995, "loss": 1.5391, "step": 1933 }, { "epoch": 0.9101176470588235, "grad_norm": 0.9934784856962249, "learning_rate": 0.00029009999999999995, "loss": 0.5508, "step": 1934 }, { "epoch": 0.9105882352941177, "grad_norm": 1.0229727466513499, "learning_rate": 0.00029025, "loss": 0.6523, "step": 1935 }, { "epoch": 0.9110588235294118, "grad_norm": 0.8721612284826549, "learning_rate": 0.00029039999999999996, "loss": 0.7109, "step": 1936 }, { "epoch": 0.9115294117647059, "grad_norm": 1.5880845116537083, "learning_rate": 0.00029055, "loss": 1.0, "step": 1937 }, { "epoch": 0.912, "grad_norm": 0.9061254139361724, "learning_rate": 0.00029069999999999996, "loss": 0.5391, "step": 1938 }, { "epoch": 0.9124705882352941, "grad_norm": 0.8754952865802467, "learning_rate": 0.00029085, "loss": 0.4766, "step": 1939 }, { "epoch": 0.9129411764705883, "grad_norm": 1.740112812975539, "learning_rate": 0.00029099999999999997, "loss": 1.9922, "step": 1940 }, { "epoch": 0.9134117647058824, "grad_norm": 1.576601794708666, "learning_rate": 0.00029115, "loss": 1.1875, "step": 1941 }, { "epoch": 0.9138823529411765, "grad_norm": 2.307412020550301, "learning_rate": 0.0002913, "loss": 2.5312, "step": 1942 }, { "epoch": 0.9143529411764706, "grad_norm": 1.0079363159287973, "learning_rate": 0.00029145, "loss": 0.4922, "step": 1943 }, { "epoch": 0.9148235294117647, "grad_norm": 1.5621860706237294, "learning_rate": 0.0002916, "loss": 1.0625, "step": 1944 }, { "epoch": 0.9152941176470588, "grad_norm": 0.7657541988599627, "learning_rate": 0.00029174999999999996, "loss": 0.3008, "step": 1945 }, { "epoch": 0.9157647058823529, "grad_norm": 1.581278410351497, "learning_rate": 0.0002919, "loss": 1.9062, "step": 1946 }, { "epoch": 0.916235294117647, "grad_norm": 1.3669466246576458, "learning_rate": 0.00029204999999999997, "loss": 0.9922, "step": 1947 }, { "epoch": 0.9167058823529411, "grad_norm": 1.7656653701564196, "learning_rate": 0.00029219999999999995, "loss": 1.8594, "step": 1948 }, { "epoch": 0.9171764705882353, "grad_norm": 1.3961921633916086, "learning_rate": 0.00029235, "loss": 1.1406, "step": 1949 }, { "epoch": 0.9176470588235294, "grad_norm": 1.5861270248927268, "learning_rate": 0.00029249999999999995, "loss": 2.0, "step": 1950 }, { "epoch": 0.9181176470588235, "grad_norm": 1.4480952224641226, "learning_rate": 0.00029265, "loss": 0.7695, "step": 1951 }, { "epoch": 0.9185882352941176, "grad_norm": 1.6435017747439689, "learning_rate": 0.00029279999999999996, "loss": 1.3672, "step": 1952 }, { "epoch": 0.9190588235294118, "grad_norm": 0.7625704558260535, "learning_rate": 0.00029295, "loss": 0.332, "step": 1953 }, { "epoch": 0.9195294117647059, "grad_norm": 1.2526890094500327, "learning_rate": 0.00029309999999999997, "loss": 1.3125, "step": 1954 }, { "epoch": 0.92, "grad_norm": 0.7430323195960039, "learning_rate": 0.00029325, "loss": 0.418, "step": 1955 }, { "epoch": 0.9204705882352942, "grad_norm": 1.4004816033321779, "learning_rate": 0.0002934, "loss": 0.8164, "step": 1956 }, { "epoch": 0.9209411764705883, "grad_norm": 1.4628283148458832, "learning_rate": 0.00029355, "loss": 1.1641, "step": 1957 }, { "epoch": 0.9214117647058824, "grad_norm": 1.7743619442541911, "learning_rate": 0.0002937, "loss": 1.0469, "step": 1958 }, { "epoch": 0.9218823529411765, "grad_norm": 1.0471997150236227, "learning_rate": 0.00029384999999999996, "loss": 0.6445, "step": 1959 }, { "epoch": 0.9223529411764706, "grad_norm": 1.1160861816791552, "learning_rate": 0.000294, "loss": 0.9609, "step": 1960 }, { "epoch": 0.9228235294117647, "grad_norm": 2.034805473213131, "learning_rate": 0.00029414999999999997, "loss": 1.375, "step": 1961 }, { "epoch": 0.9232941176470588, "grad_norm": 1.820546653067549, "learning_rate": 0.00029429999999999994, "loss": 1.625, "step": 1962 }, { "epoch": 0.9237647058823529, "grad_norm": 1.3168086709955602, "learning_rate": 0.00029445, "loss": 1.2656, "step": 1963 }, { "epoch": 0.924235294117647, "grad_norm": 1.5126555589192676, "learning_rate": 0.00029459999999999995, "loss": 1.2031, "step": 1964 }, { "epoch": 0.9247058823529412, "grad_norm": 1.59692130251038, "learning_rate": 0.00029475, "loss": 1.4844, "step": 1965 }, { "epoch": 0.9251764705882353, "grad_norm": 1.1886933604458758, "learning_rate": 0.00029489999999999996, "loss": 1.125, "step": 1966 }, { "epoch": 0.9256470588235294, "grad_norm": 2.1497774504799008, "learning_rate": 0.00029505, "loss": 2.25, "step": 1967 }, { "epoch": 0.9261176470588235, "grad_norm": 0.9646434759299843, "learning_rate": 0.00029519999999999997, "loss": 0.6641, "step": 1968 }, { "epoch": 0.9265882352941176, "grad_norm": 1.0885893940482971, "learning_rate": 0.00029535, "loss": 0.6719, "step": 1969 }, { "epoch": 0.9270588235294117, "grad_norm": 0.6702168748775503, "learning_rate": 0.00029549999999999997, "loss": 0.2432, "step": 1970 }, { "epoch": 0.9275294117647059, "grad_norm": 1.7358145811624697, "learning_rate": 0.00029565, "loss": 1.6016, "step": 1971 }, { "epoch": 0.928, "grad_norm": 0.8563413409013754, "learning_rate": 0.0002958, "loss": 0.4707, "step": 1972 }, { "epoch": 0.9284705882352942, "grad_norm": 1.1111126149813824, "learning_rate": 0.00029595, "loss": 0.8633, "step": 1973 }, { "epoch": 0.9289411764705883, "grad_norm": 1.8589114530361657, "learning_rate": 0.0002961, "loss": 2.0, "step": 1974 }, { "epoch": 0.9294117647058824, "grad_norm": 1.0143689254101833, "learning_rate": 0.00029624999999999996, "loss": 0.9219, "step": 1975 }, { "epoch": 0.9298823529411765, "grad_norm": 1.440812137931272, "learning_rate": 0.0002964, "loss": 0.9258, "step": 1976 }, { "epoch": 0.9303529411764706, "grad_norm": 1.1398637185687954, "learning_rate": 0.00029654999999999997, "loss": 0.6289, "step": 1977 }, { "epoch": 0.9308235294117647, "grad_norm": 1.9500278129518032, "learning_rate": 0.00029669999999999995, "loss": 1.8828, "step": 1978 }, { "epoch": 0.9312941176470588, "grad_norm": 0.92407617560594, "learning_rate": 0.00029685, "loss": 0.793, "step": 1979 }, { "epoch": 0.9317647058823529, "grad_norm": 0.6889132710718094, "learning_rate": 0.00029699999999999996, "loss": 0.2773, "step": 1980 }, { "epoch": 0.932235294117647, "grad_norm": 0.8351449518258167, "learning_rate": 0.00029715, "loss": 0.6133, "step": 1981 }, { "epoch": 0.9327058823529412, "grad_norm": 1.6571506994638707, "learning_rate": 0.00029729999999999996, "loss": 1.3359, "step": 1982 }, { "epoch": 0.9331764705882353, "grad_norm": 0.7722029700047522, "learning_rate": 0.00029745, "loss": 0.3594, "step": 1983 }, { "epoch": 0.9336470588235294, "grad_norm": 0.7936226564715714, "learning_rate": 0.00029759999999999997, "loss": 0.293, "step": 1984 }, { "epoch": 0.9341176470588235, "grad_norm": 2.912784511628962, "learning_rate": 0.00029775, "loss": 2.6406, "step": 1985 }, { "epoch": 0.9345882352941176, "grad_norm": 1.3721614250062653, "learning_rate": 0.0002979, "loss": 1.0312, "step": 1986 }, { "epoch": 0.9350588235294117, "grad_norm": 0.8286174012929082, "learning_rate": 0.00029805, "loss": 0.5508, "step": 1987 }, { "epoch": 0.9355294117647058, "grad_norm": 0.9488758577422828, "learning_rate": 0.0002982, "loss": 0.4355, "step": 1988 }, { "epoch": 0.936, "grad_norm": 0.6107073964011922, "learning_rate": 0.00029835, "loss": 0.2402, "step": 1989 }, { "epoch": 0.9364705882352942, "grad_norm": 1.0831295578087647, "learning_rate": 0.0002985, "loss": 0.6016, "step": 1990 }, { "epoch": 0.9369411764705883, "grad_norm": 1.0556833950357094, "learning_rate": 0.00029864999999999997, "loss": 0.9844, "step": 1991 }, { "epoch": 0.9374117647058824, "grad_norm": 1.1520020706859495, "learning_rate": 0.0002988, "loss": 0.7852, "step": 1992 }, { "epoch": 0.9378823529411765, "grad_norm": 1.7204244039289607, "learning_rate": 0.00029895, "loss": 1.4375, "step": 1993 }, { "epoch": 0.9383529411764706, "grad_norm": 1.2587168866252427, "learning_rate": 0.00029909999999999995, "loss": 1.1016, "step": 1994 }, { "epoch": 0.9388235294117647, "grad_norm": 1.7162051182091083, "learning_rate": 0.00029925, "loss": 1.4844, "step": 1995 }, { "epoch": 0.9392941176470588, "grad_norm": 1.477234227000156, "learning_rate": 0.00029939999999999996, "loss": 1.4375, "step": 1996 }, { "epoch": 0.939764705882353, "grad_norm": 0.8924491040861816, "learning_rate": 0.00029955, "loss": 0.3184, "step": 1997 }, { "epoch": 0.9402352941176471, "grad_norm": 2.0565103042474697, "learning_rate": 0.00029969999999999997, "loss": 1.7266, "step": 1998 }, { "epoch": 0.9407058823529412, "grad_norm": 0.7108636344037773, "learning_rate": 0.00029985, "loss": 0.6367, "step": 1999 }, { "epoch": 0.9411764705882353, "grad_norm": 0.6303157263469227, "learning_rate": 0.0003, "loss": 0.4004, "step": 2000 }, { "epoch": 0.9416470588235294, "grad_norm": 2.2450044275910757, "learning_rate": 0.0002999999935946321, "loss": 2.5312, "step": 2001 }, { "epoch": 0.9421176470588235, "grad_norm": 1.523510679250402, "learning_rate": 0.00029999997437852905, "loss": 1.6719, "step": 2002 }, { "epoch": 0.9425882352941176, "grad_norm": 1.8591424204906948, "learning_rate": 0.0002999999423516924, "loss": 1.5078, "step": 2003 }, { "epoch": 0.9430588235294117, "grad_norm": 1.2675269139977, "learning_rate": 0.00029999989751412494, "loss": 0.9727, "step": 2004 }, { "epoch": 0.9435294117647058, "grad_norm": 1.0141579821407924, "learning_rate": 0.00029999983986583054, "loss": 0.6562, "step": 2005 }, { "epoch": 0.944, "grad_norm": 1.2585359327692847, "learning_rate": 0.000299999769406814, "loss": 1.1562, "step": 2006 }, { "epoch": 0.9444705882352942, "grad_norm": 0.8406928164549186, "learning_rate": 0.0002999996861370814, "loss": 0.2891, "step": 2007 }, { "epoch": 0.9449411764705883, "grad_norm": 1.4742068053470612, "learning_rate": 0.0002999995900566399, "loss": 1.5078, "step": 2008 }, { "epoch": 0.9454117647058824, "grad_norm": 1.0420761943260115, "learning_rate": 0.0002999994811654977, "loss": 0.6992, "step": 2009 }, { "epoch": 0.9458823529411765, "grad_norm": 1.6176289659741347, "learning_rate": 0.000299999359463664, "loss": 1.3203, "step": 2010 }, { "epoch": 0.9463529411764706, "grad_norm": 1.6455873117634463, "learning_rate": 0.0002999992249511493, "loss": 1.6094, "step": 2011 }, { "epoch": 0.9468235294117647, "grad_norm": 1.4950220832897256, "learning_rate": 0.00029999907762796505, "loss": 1.6328, "step": 2012 }, { "epoch": 0.9472941176470588, "grad_norm": 1.5123893746783974, "learning_rate": 0.00029999891749412375, "loss": 1.8906, "step": 2013 }, { "epoch": 0.947764705882353, "grad_norm": 0.8824126802911931, "learning_rate": 0.0002999987445496392, "loss": 0.6641, "step": 2014 }, { "epoch": 0.9482352941176471, "grad_norm": 0.9836988509907784, "learning_rate": 0.00029999855879452616, "loss": 0.7227, "step": 2015 }, { "epoch": 0.9487058823529412, "grad_norm": 2.0449356647238317, "learning_rate": 0.00029999836022880045, "loss": 1.25, "step": 2016 }, { "epoch": 0.9491764705882353, "grad_norm": 1.6995473234393084, "learning_rate": 0.00029999814885247906, "loss": 1.8203, "step": 2017 }, { "epoch": 0.9496470588235294, "grad_norm": 1.1850469050789727, "learning_rate": 0.00029999792466558, "loss": 1.0781, "step": 2018 }, { "epoch": 0.9501176470588235, "grad_norm": 1.8918855843374396, "learning_rate": 0.0002999976876681224, "loss": 1.4062, "step": 2019 }, { "epoch": 0.9505882352941176, "grad_norm": 1.4429591126881285, "learning_rate": 0.00029999743786012656, "loss": 1.3125, "step": 2020 }, { "epoch": 0.9510588235294117, "grad_norm": 1.8455956273954257, "learning_rate": 0.0002999971752416138, "loss": 1.5156, "step": 2021 }, { "epoch": 0.9515294117647058, "grad_norm": 1.0528881112101722, "learning_rate": 0.00029999689981260655, "loss": 0.6836, "step": 2022 }, { "epoch": 0.952, "grad_norm": 1.2507143534488083, "learning_rate": 0.0002999966115731283, "loss": 0.6133, "step": 2023 }, { "epoch": 0.9524705882352941, "grad_norm": 0.9364955741531843, "learning_rate": 0.00029999631052320374, "loss": 0.4688, "step": 2024 }, { "epoch": 0.9529411764705882, "grad_norm": 1.6164645180163042, "learning_rate": 0.0002999959966628585, "loss": 1.375, "step": 2025 }, { "epoch": 0.9534117647058824, "grad_norm": 1.6087363090662377, "learning_rate": 0.0002999956699921194, "loss": 1.6875, "step": 2026 }, { "epoch": 0.9538823529411765, "grad_norm": 0.7683646657709712, "learning_rate": 0.0002999953305110144, "loss": 0.7422, "step": 2027 }, { "epoch": 0.9543529411764706, "grad_norm": 0.8812243264845627, "learning_rate": 0.0002999949782195724, "loss": 0.3535, "step": 2028 }, { "epoch": 0.9548235294117647, "grad_norm": 0.8043966656411469, "learning_rate": 0.0002999946131178236, "loss": 0.2461, "step": 2029 }, { "epoch": 0.9552941176470588, "grad_norm": 2.838877538937649, "learning_rate": 0.0002999942352057991, "loss": 2.1719, "step": 2030 }, { "epoch": 0.955764705882353, "grad_norm": 1.7719081679682145, "learning_rate": 0.00029999384448353114, "loss": 1.3359, "step": 2031 }, { "epoch": 0.9562352941176471, "grad_norm": 1.4131560395076506, "learning_rate": 0.00029999344095105326, "loss": 1.1094, "step": 2032 }, { "epoch": 0.9567058823529412, "grad_norm": 1.6569016217967296, "learning_rate": 0.0002999930246083997, "loss": 1.5312, "step": 2033 }, { "epoch": 0.9571764705882353, "grad_norm": 1.376253163878324, "learning_rate": 0.0002999925954556062, "loss": 1.3203, "step": 2034 }, { "epoch": 0.9576470588235294, "grad_norm": 2.202678824689756, "learning_rate": 0.0002999921534927093, "loss": 2.1875, "step": 2035 }, { "epoch": 0.9581176470588235, "grad_norm": 1.8040257653436043, "learning_rate": 0.00029999169871974677, "loss": 1.4062, "step": 2036 }, { "epoch": 0.9585882352941176, "grad_norm": 1.1097255592925594, "learning_rate": 0.0002999912311367575, "loss": 0.7148, "step": 2037 }, { "epoch": 0.9590588235294117, "grad_norm": 1.708807838172086, "learning_rate": 0.00029999075074378136, "loss": 0.9062, "step": 2038 }, { "epoch": 0.9595294117647059, "grad_norm": 1.0075126989160998, "learning_rate": 0.00029999025754085944, "loss": 0.5508, "step": 2039 }, { "epoch": 0.96, "grad_norm": 1.1920301253061458, "learning_rate": 0.00029998975152803383, "loss": 1.2109, "step": 2040 }, { "epoch": 0.9604705882352941, "grad_norm": 0.9369359499901654, "learning_rate": 0.0002999892327053477, "loss": 0.5742, "step": 2041 }, { "epoch": 0.9609411764705882, "grad_norm": 1.3599516678668682, "learning_rate": 0.0002999887010728454, "loss": 1.1953, "step": 2042 }, { "epoch": 0.9614117647058823, "grad_norm": 0.95373778777284, "learning_rate": 0.0002999881566305724, "loss": 0.377, "step": 2043 }, { "epoch": 0.9618823529411765, "grad_norm": 1.5478278087912039, "learning_rate": 0.0002999875993785751, "loss": 1.3516, "step": 2044 }, { "epoch": 0.9623529411764706, "grad_norm": 0.8384493916976392, "learning_rate": 0.0002999870293169011, "loss": 0.4238, "step": 2045 }, { "epoch": 0.9628235294117647, "grad_norm": 0.6665072171688131, "learning_rate": 0.00029998644644559916, "loss": 0.3262, "step": 2046 }, { "epoch": 0.9632941176470589, "grad_norm": 0.8530740254081632, "learning_rate": 0.000299985850764719, "loss": 0.3477, "step": 2047 }, { "epoch": 0.963764705882353, "grad_norm": 0.7999475465137352, "learning_rate": 0.00029998524227431146, "loss": 0.334, "step": 2048 }, { "epoch": 0.9642352941176471, "grad_norm": 1.390259015337499, "learning_rate": 0.00029998462097442866, "loss": 0.625, "step": 2049 }, { "epoch": 0.9647058823529412, "grad_norm": 2.145477810825693, "learning_rate": 0.0002999839868651235, "loss": 2.0938, "step": 2050 }, { "epoch": 0.9651764705882353, "grad_norm": 2.1061064259609776, "learning_rate": 0.0002999833399464502, "loss": 2.1875, "step": 2051 }, { "epoch": 0.9656470588235294, "grad_norm": 2.3148560220139034, "learning_rate": 0.000299982680218464, "loss": 1.7109, "step": 2052 }, { "epoch": 0.9661176470588235, "grad_norm": 0.5321991898883789, "learning_rate": 0.00029998200768122123, "loss": 0.0723, "step": 2053 }, { "epoch": 0.9665882352941176, "grad_norm": 1.3554135951260713, "learning_rate": 0.00029998132233477936, "loss": 0.7656, "step": 2054 }, { "epoch": 0.9670588235294117, "grad_norm": 2.2710946327248576, "learning_rate": 0.0002999806241791969, "loss": 1.9297, "step": 2055 }, { "epoch": 0.9675294117647059, "grad_norm": 1.3662621586723298, "learning_rate": 0.0002999799132145335, "loss": 0.8633, "step": 2056 }, { "epoch": 0.968, "grad_norm": 1.492953543755651, "learning_rate": 0.00029997918944084986, "loss": 1.0156, "step": 2057 }, { "epoch": 0.9684705882352941, "grad_norm": 1.2759060865965164, "learning_rate": 0.0002999784528582078, "loss": 1.2031, "step": 2058 }, { "epoch": 0.9689411764705882, "grad_norm": 1.7961095348437999, "learning_rate": 0.0002999777034666702, "loss": 2.0156, "step": 2059 }, { "epoch": 0.9694117647058823, "grad_norm": 2.501704717149007, "learning_rate": 0.00029997694126630115, "loss": 3.1562, "step": 2060 }, { "epoch": 0.9698823529411764, "grad_norm": 1.5770143712574292, "learning_rate": 0.00029997616625716567, "loss": 1.3438, "step": 2061 }, { "epoch": 0.9703529411764706, "grad_norm": 1.0945747728606219, "learning_rate": 0.0002999753784393299, "loss": 0.7344, "step": 2062 }, { "epoch": 0.9708235294117648, "grad_norm": 1.4790959825502827, "learning_rate": 0.00029997457781286125, "loss": 1.3281, "step": 2063 }, { "epoch": 0.9712941176470589, "grad_norm": 1.7988174040075577, "learning_rate": 0.000299973764377828, "loss": 1.0547, "step": 2064 }, { "epoch": 0.971764705882353, "grad_norm": 0.7810399815677338, "learning_rate": 0.00029997293813429965, "loss": 0.293, "step": 2065 }, { "epoch": 0.9722352941176471, "grad_norm": 1.4706114126659064, "learning_rate": 0.0002999720990823468, "loss": 1.3828, "step": 2066 }, { "epoch": 0.9727058823529412, "grad_norm": 1.1408097717964252, "learning_rate": 0.0002999712472220411, "loss": 0.9102, "step": 2067 }, { "epoch": 0.9731764705882353, "grad_norm": 0.6808501890409953, "learning_rate": 0.0002999703825534552, "loss": 0.5273, "step": 2068 }, { "epoch": 0.9736470588235294, "grad_norm": 1.1546829727764634, "learning_rate": 0.0002999695050766631, "loss": 1.1328, "step": 2069 }, { "epoch": 0.9741176470588235, "grad_norm": 1.3273190814690534, "learning_rate": 0.0002999686147917396, "loss": 1.0781, "step": 2070 }, { "epoch": 0.9745882352941176, "grad_norm": 1.6242261321601925, "learning_rate": 0.0002999677116987609, "loss": 0.9688, "step": 2071 }, { "epoch": 0.9750588235294118, "grad_norm": 1.0872135098933962, "learning_rate": 0.00029996679579780394, "loss": 0.75, "step": 2072 }, { "epoch": 0.9755294117647059, "grad_norm": 1.62475790876946, "learning_rate": 0.0002999658670889471, "loss": 1.4531, "step": 2073 }, { "epoch": 0.976, "grad_norm": 0.9748510537191981, "learning_rate": 0.0002999649255722696, "loss": 0.6914, "step": 2074 }, { "epoch": 0.9764705882352941, "grad_norm": 0.9103662947095018, "learning_rate": 0.0002999639712478519, "loss": 0.4316, "step": 2075 }, { "epoch": 0.9769411764705882, "grad_norm": 0.9531115455709347, "learning_rate": 0.00029996300411577547, "loss": 0.3926, "step": 2076 }, { "epoch": 0.9774117647058823, "grad_norm": 1.619708135221749, "learning_rate": 0.00029996202417612294, "loss": 1.3359, "step": 2077 }, { "epoch": 0.9778823529411764, "grad_norm": 0.8483312516453186, "learning_rate": 0.000299961031428978, "loss": 0.4629, "step": 2078 }, { "epoch": 0.9783529411764705, "grad_norm": 0.9862456341226806, "learning_rate": 0.00029996002587442537, "loss": 0.9531, "step": 2079 }, { "epoch": 0.9788235294117648, "grad_norm": 2.014452048965287, "learning_rate": 0.000299959007512551, "loss": 2.2031, "step": 2080 }, { "epoch": 0.9792941176470589, "grad_norm": 1.3117822963332324, "learning_rate": 0.00029995797634344186, "loss": 1.2734, "step": 2081 }, { "epoch": 0.979764705882353, "grad_norm": 0.9396538971314962, "learning_rate": 0.000299956932367186, "loss": 0.4707, "step": 2082 }, { "epoch": 0.9802352941176471, "grad_norm": 1.0358925186279473, "learning_rate": 0.0002999558755838726, "loss": 0.6758, "step": 2083 }, { "epoch": 0.9807058823529412, "grad_norm": 0.38156137418634767, "learning_rate": 0.00029995480599359186, "loss": 0.0718, "step": 2084 }, { "epoch": 0.9811764705882353, "grad_norm": 1.693177525312365, "learning_rate": 0.00029995372359643515, "loss": 1.3516, "step": 2085 }, { "epoch": 0.9816470588235294, "grad_norm": 1.4598340835549908, "learning_rate": 0.00029995262839249497, "loss": 1.2578, "step": 2086 }, { "epoch": 0.9821176470588235, "grad_norm": 1.7092754093513005, "learning_rate": 0.0002999515203818648, "loss": 1.4922, "step": 2087 }, { "epoch": 0.9825882352941177, "grad_norm": 1.2437703141417076, "learning_rate": 0.0002999503995646393, "loss": 0.9492, "step": 2088 }, { "epoch": 0.9830588235294118, "grad_norm": 0.8415739809573628, "learning_rate": 0.00029994926594091413, "loss": 0.6211, "step": 2089 }, { "epoch": 0.9835294117647059, "grad_norm": 1.6131785335716438, "learning_rate": 0.00029994811951078616, "loss": 0.7266, "step": 2090 }, { "epoch": 0.984, "grad_norm": 2.0331565105485354, "learning_rate": 0.0002999469602743533, "loss": 1.7891, "step": 2091 }, { "epoch": 0.9844705882352941, "grad_norm": 1.339693986383191, "learning_rate": 0.0002999457882317146, "loss": 0.5938, "step": 2092 }, { "epoch": 0.9849411764705882, "grad_norm": 1.3552250552137666, "learning_rate": 0.00029994460338297, "loss": 1.4219, "step": 2093 }, { "epoch": 0.9854117647058823, "grad_norm": 1.83691950255306, "learning_rate": 0.0002999434057282209, "loss": 1.6328, "step": 2094 }, { "epoch": 0.9858823529411764, "grad_norm": 1.6177890991028394, "learning_rate": 0.0002999421952675694, "loss": 1.375, "step": 2095 }, { "epoch": 0.9863529411764705, "grad_norm": 1.9603972254574316, "learning_rate": 0.00029994097200111903, "loss": 1.3828, "step": 2096 }, { "epoch": 0.9868235294117647, "grad_norm": 1.2908771371824266, "learning_rate": 0.0002999397359289741, "loss": 1.2344, "step": 2097 }, { "epoch": 0.9872941176470589, "grad_norm": 1.3269782742550877, "learning_rate": 0.00029993848705124035, "loss": 1.5, "step": 2098 }, { "epoch": 0.987764705882353, "grad_norm": 0.9024658682467955, "learning_rate": 0.0002999372253680244, "loss": 0.3887, "step": 2099 }, { "epoch": 0.9882352941176471, "grad_norm": 1.5834497739722224, "learning_rate": 0.0002999359508794339, "loss": 1.7422, "step": 2100 }, { "epoch": 0.9887058823529412, "grad_norm": 1.0790499859367149, "learning_rate": 0.0002999346635855778, "loss": 0.5664, "step": 2101 }, { "epoch": 0.9891764705882353, "grad_norm": 1.3437994062059644, "learning_rate": 0.00029993336348656594, "loss": 1.1484, "step": 2102 }, { "epoch": 0.9896470588235294, "grad_norm": 1.1712990112056725, "learning_rate": 0.00029993205058250947, "loss": 0.7852, "step": 2103 }, { "epoch": 0.9901176470588235, "grad_norm": 1.0833236561132884, "learning_rate": 0.00029993072487352043, "loss": 0.6367, "step": 2104 }, { "epoch": 0.9905882352941177, "grad_norm": 1.7265432045923061, "learning_rate": 0.0002999293863597121, "loss": 1.6719, "step": 2105 }, { "epoch": 0.9910588235294118, "grad_norm": 1.556613510667465, "learning_rate": 0.00029992803504119876, "loss": 1.1484, "step": 2106 }, { "epoch": 0.9915294117647059, "grad_norm": 1.0794661595034658, "learning_rate": 0.0002999266709180959, "loss": 1.2578, "step": 2107 }, { "epoch": 0.992, "grad_norm": 1.4118213450836048, "learning_rate": 0.0002999252939905199, "loss": 1.0312, "step": 2108 }, { "epoch": 0.9924705882352941, "grad_norm": 1.9594628421190552, "learning_rate": 0.0002999239042585884, "loss": 2.0781, "step": 2109 }, { "epoch": 0.9929411764705882, "grad_norm": 1.4614014117866745, "learning_rate": 0.0002999225017224201, "loss": 1.3594, "step": 2110 }, { "epoch": 0.9934117647058823, "grad_norm": 1.2254166699903823, "learning_rate": 0.0002999210863821349, "loss": 1.2031, "step": 2111 }, { "epoch": 0.9938823529411764, "grad_norm": 1.891120688353206, "learning_rate": 0.0002999196582378535, "loss": 2.6562, "step": 2112 }, { "epoch": 0.9943529411764706, "grad_norm": 1.2438835958059773, "learning_rate": 0.0002999182172896979, "loss": 1.4219, "step": 2113 }, { "epoch": 0.9948235294117647, "grad_norm": 1.2016591197515851, "learning_rate": 0.0002999167635377912, "loss": 1.4375, "step": 2114 }, { "epoch": 0.9952941176470588, "grad_norm": 0.9152039477590899, "learning_rate": 0.00029991529698225757, "loss": 0.4199, "step": 2115 }, { "epoch": 0.995764705882353, "grad_norm": 0.4942688434472492, "learning_rate": 0.0002999138176232223, "loss": 0.2461, "step": 2116 }, { "epoch": 0.9962352941176471, "grad_norm": 0.9688434846567662, "learning_rate": 0.00029991232546081156, "loss": 0.6289, "step": 2117 }, { "epoch": 0.9967058823529412, "grad_norm": 0.5355720855492044, "learning_rate": 0.000299910820495153, "loss": 0.2539, "step": 2118 }, { "epoch": 0.9971764705882353, "grad_norm": 1.5362439750895867, "learning_rate": 0.00029990930272637506, "loss": 0.8594, "step": 2119 }, { "epoch": 0.9976470588235294, "grad_norm": 3.1741941207658835, "learning_rate": 0.00029990777215460737, "loss": 2.7812, "step": 2120 }, { "epoch": 0.9981176470588236, "grad_norm": 1.3056906646155373, "learning_rate": 0.0002999062287799806, "loss": 1.5391, "step": 2121 }, { "epoch": 0.9985882352941177, "grad_norm": 2.0823503624832425, "learning_rate": 0.0002999046726026266, "loss": 2.3438, "step": 2122 }, { "epoch": 0.9990588235294118, "grad_norm": 1.148111168257156, "learning_rate": 0.00029990310362267833, "loss": 1.0469, "step": 2123 }, { "epoch": 0.9995294117647059, "grad_norm": 1.2527519775516205, "learning_rate": 0.0002999015218402697, "loss": 0.8008, "step": 2124 }, { "epoch": 1.0, "grad_norm": 1.209091872102859, "learning_rate": 0.0002998999272555359, "loss": 0.8281, "step": 2125 }, { "epoch": 1.0, "eval_loss": 1.1416575908660889, "eval_runtime": 37.5302, "eval_samples_per_second": 19.984, "eval_steps_per_second": 9.992, "step": 2125 }, { "epoch": 1.0004705882352942, "grad_norm": 0.4614182221036976, "learning_rate": 0.000299898319868613, "loss": 0.1396, "step": 2126 }, { "epoch": 1.0009411764705882, "grad_norm": 1.8944369131651955, "learning_rate": 0.00029989669967963835, "loss": 1.4062, "step": 2127 }, { "epoch": 1.0014117647058824, "grad_norm": 1.8088318950690383, "learning_rate": 0.00029989506668875033, "loss": 1.9062, "step": 2128 }, { "epoch": 1.0018823529411764, "grad_norm": 1.64189020624665, "learning_rate": 0.00029989342089608835, "loss": 1.2656, "step": 2129 }, { "epoch": 1.0023529411764707, "grad_norm": 1.10948941003675, "learning_rate": 0.000299891762301793, "loss": 0.7422, "step": 2130 }, { "epoch": 1.0028235294117647, "grad_norm": 1.2668082955383464, "learning_rate": 0.00029989009090600595, "loss": 0.8984, "step": 2131 }, { "epoch": 1.003294117647059, "grad_norm": 1.5491420936348475, "learning_rate": 0.00029988840670887, "loss": 1.2578, "step": 2132 }, { "epoch": 1.003764705882353, "grad_norm": 1.4678551866526524, "learning_rate": 0.0002998867097105288, "loss": 1.1719, "step": 2133 }, { "epoch": 1.0042352941176471, "grad_norm": 1.099218812923787, "learning_rate": 0.00029988499991112747, "loss": 0.707, "step": 2134 }, { "epoch": 1.0047058823529411, "grad_norm": 1.1004627665199824, "learning_rate": 0.000299883277310812, "loss": 0.793, "step": 2135 }, { "epoch": 1.0051764705882353, "grad_norm": 0.7300071958896653, "learning_rate": 0.0002998815419097294, "loss": 0.2871, "step": 2136 }, { "epoch": 1.0056470588235293, "grad_norm": 1.0091830289924906, "learning_rate": 0.0002998797937080279, "loss": 0.6641, "step": 2137 }, { "epoch": 1.0061176470588236, "grad_norm": 1.4034872117750905, "learning_rate": 0.000299878032705857, "loss": 0.8516, "step": 2138 }, { "epoch": 1.0065882352941176, "grad_norm": 0.5783804320784245, "learning_rate": 0.0002998762589033669, "loss": 0.1943, "step": 2139 }, { "epoch": 1.0070588235294118, "grad_norm": 1.6400263097399677, "learning_rate": 0.00029987447230070917, "loss": 1.5312, "step": 2140 }, { "epoch": 1.0075294117647058, "grad_norm": 0.4545144415246435, "learning_rate": 0.00029987267289803635, "loss": 0.1338, "step": 2141 }, { "epoch": 1.008, "grad_norm": 1.3591570222025633, "learning_rate": 0.00029987086069550216, "loss": 0.9922, "step": 2142 }, { "epoch": 1.008470588235294, "grad_norm": 1.2307388470366134, "learning_rate": 0.0002998690356932613, "loss": 0.7617, "step": 2143 }, { "epoch": 1.0089411764705882, "grad_norm": 1.245220032548855, "learning_rate": 0.0002998671978914698, "loss": 0.8438, "step": 2144 }, { "epoch": 1.0094117647058825, "grad_norm": 0.8761084761795632, "learning_rate": 0.00029986534729028444, "loss": 0.3867, "step": 2145 }, { "epoch": 1.0098823529411765, "grad_norm": 1.4859544759603287, "learning_rate": 0.00029986348388986333, "loss": 1.5625, "step": 2146 }, { "epoch": 1.0103529411764707, "grad_norm": 1.5794307371631207, "learning_rate": 0.00029986160769036565, "loss": 1.3125, "step": 2147 }, { "epoch": 1.0108235294117647, "grad_norm": 1.5988923437887297, "learning_rate": 0.0002998597186919515, "loss": 1.4453, "step": 2148 }, { "epoch": 1.011294117647059, "grad_norm": 1.03180060943181, "learning_rate": 0.0002998578168947824, "loss": 1.0391, "step": 2149 }, { "epoch": 1.011764705882353, "grad_norm": 1.1395037719589105, "learning_rate": 0.00029985590229902073, "loss": 0.8945, "step": 2150 }, { "epoch": 1.0122352941176471, "grad_norm": 1.2118124734640088, "learning_rate": 0.0002998539749048299, "loss": 0.9062, "step": 2151 }, { "epoch": 1.0127058823529411, "grad_norm": 1.068673501813249, "learning_rate": 0.0002998520347123746, "loss": 0.7109, "step": 2152 }, { "epoch": 1.0131764705882353, "grad_norm": 1.1569010772316808, "learning_rate": 0.00029985008172182055, "loss": 0.6211, "step": 2153 }, { "epoch": 1.0136470588235293, "grad_norm": 0.7773199782844809, "learning_rate": 0.00029984811593333446, "loss": 0.3926, "step": 2154 }, { "epoch": 1.0141176470588236, "grad_norm": 1.5340615330783471, "learning_rate": 0.00029984613734708427, "loss": 1.0312, "step": 2155 }, { "epoch": 1.0145882352941176, "grad_norm": 1.9100668125572895, "learning_rate": 0.00029984414596323896, "loss": 2.3438, "step": 2156 }, { "epoch": 1.0150588235294118, "grad_norm": 0.9476730226918606, "learning_rate": 0.00029984214178196863, "loss": 0.5078, "step": 2157 }, { "epoch": 1.0155294117647058, "grad_norm": 1.5020343647940368, "learning_rate": 0.00029984012480344444, "loss": 1.4844, "step": 2158 }, { "epoch": 1.016, "grad_norm": 1.3685324341124352, "learning_rate": 0.0002998380950278386, "loss": 1.0547, "step": 2159 }, { "epoch": 1.016470588235294, "grad_norm": 1.267328568606121, "learning_rate": 0.00029983605245532453, "loss": 0.9492, "step": 2160 }, { "epoch": 1.0169411764705882, "grad_norm": 1.8054746001585984, "learning_rate": 0.0002998339970860766, "loss": 2.2344, "step": 2161 }, { "epoch": 1.0174117647058825, "grad_norm": 1.9658510723283589, "learning_rate": 0.00029983192892027037, "loss": 1.5703, "step": 2162 }, { "epoch": 1.0178823529411765, "grad_norm": 1.4717309419281392, "learning_rate": 0.0002998298479580825, "loss": 1.2891, "step": 2163 }, { "epoch": 1.0183529411764707, "grad_norm": 1.6031248464388375, "learning_rate": 0.0002998277541996907, "loss": 1.4219, "step": 2164 }, { "epoch": 1.0188235294117647, "grad_norm": 2.4574339613260108, "learning_rate": 0.0002998256476452738, "loss": 1.8672, "step": 2165 }, { "epoch": 1.019294117647059, "grad_norm": 1.1812746977255753, "learning_rate": 0.0002998235282950117, "loss": 1.1094, "step": 2166 }, { "epoch": 1.019764705882353, "grad_norm": 0.9453756694841725, "learning_rate": 0.0002998213961490854, "loss": 0.4629, "step": 2167 }, { "epoch": 1.0202352941176471, "grad_norm": 0.9254227352085574, "learning_rate": 0.00029981925120767705, "loss": 0.457, "step": 2168 }, { "epoch": 1.0207058823529411, "grad_norm": 0.8572129013442659, "learning_rate": 0.00029981709347096975, "loss": 0.2559, "step": 2169 }, { "epoch": 1.0211764705882354, "grad_norm": 5.34644949214636, "learning_rate": 0.00029981492293914777, "loss": 1.8906, "step": 2170 }, { "epoch": 1.0216470588235294, "grad_norm": 2.417857616932129, "learning_rate": 0.0002998127396123966, "loss": 2.2812, "step": 2171 }, { "epoch": 1.0221176470588236, "grad_norm": 4.23036692558461, "learning_rate": 0.00029981054349090264, "loss": 3.0156, "step": 2172 }, { "epoch": 1.0225882352941176, "grad_norm": 2.0188134617678104, "learning_rate": 0.0002998083345748534, "loss": 1.5547, "step": 2173 }, { "epoch": 1.0230588235294118, "grad_norm": 1.3840259407302733, "learning_rate": 0.0002998061128644376, "loss": 0.8086, "step": 2174 }, { "epoch": 1.0235294117647058, "grad_norm": 1.3989006842967744, "learning_rate": 0.00029980387835984494, "loss": 1.1875, "step": 2175 }, { "epoch": 1.024, "grad_norm": 1.0181914020109342, "learning_rate": 0.0002998016310612663, "loss": 0.5273, "step": 2176 }, { "epoch": 1.024470588235294, "grad_norm": 1.3101722054172307, "learning_rate": 0.0002997993709688937, "loss": 1.2188, "step": 2177 }, { "epoch": 1.0249411764705882, "grad_norm": 1.7148380207701992, "learning_rate": 0.0002997970980829199, "loss": 1.3359, "step": 2178 }, { "epoch": 1.0254117647058822, "grad_norm": 1.546886938080219, "learning_rate": 0.0002997948124035393, "loss": 1.4609, "step": 2179 }, { "epoch": 1.0258823529411765, "grad_norm": 1.6630269994772526, "learning_rate": 0.00029979251393094685, "loss": 1.25, "step": 2180 }, { "epoch": 1.0263529411764707, "grad_norm": 1.080143409332578, "learning_rate": 0.00029979020266533913, "loss": 0.6836, "step": 2181 }, { "epoch": 1.0268235294117647, "grad_norm": 1.6892536539055951, "learning_rate": 0.0002997878786069133, "loss": 1.6719, "step": 2182 }, { "epoch": 1.027294117647059, "grad_norm": 1.7749424048456424, "learning_rate": 0.0002997855417558679, "loss": 1.6641, "step": 2183 }, { "epoch": 1.027764705882353, "grad_norm": 1.5901291612140231, "learning_rate": 0.0002997831921124026, "loss": 1.5781, "step": 2184 }, { "epoch": 1.0282352941176471, "grad_norm": 0.9891965624269258, "learning_rate": 0.000299780829676718, "loss": 0.8438, "step": 2185 }, { "epoch": 1.0287058823529411, "grad_norm": 1.1111876277802593, "learning_rate": 0.00029977845444901586, "loss": 0.8516, "step": 2186 }, { "epoch": 1.0291764705882354, "grad_norm": 0.6764261080482176, "learning_rate": 0.0002997760664294991, "loss": 0.2598, "step": 2187 }, { "epoch": 1.0296470588235294, "grad_norm": 1.2701347747613665, "learning_rate": 0.00029977366561837154, "loss": 1.3438, "step": 2188 }, { "epoch": 1.0301176470588236, "grad_norm": 0.8929823604310491, "learning_rate": 0.0002997712520158383, "loss": 0.3457, "step": 2189 }, { "epoch": 1.0305882352941176, "grad_norm": 1.8865276758436964, "learning_rate": 0.00029976882562210563, "loss": 1.625, "step": 2190 }, { "epoch": 1.0310588235294118, "grad_norm": 1.462304225142293, "learning_rate": 0.00029976638643738055, "loss": 1.7422, "step": 2191 }, { "epoch": 1.0315294117647058, "grad_norm": 1.1779144823739758, "learning_rate": 0.0002997639344618714, "loss": 0.9688, "step": 2192 }, { "epoch": 1.032, "grad_norm": 1.9671921951994733, "learning_rate": 0.0002997614696957878, "loss": 2.2656, "step": 2193 }, { "epoch": 1.032470588235294, "grad_norm": 1.3008078686902553, "learning_rate": 0.00029975899213934004, "loss": 1.1016, "step": 2194 }, { "epoch": 1.0329411764705883, "grad_norm": 1.6948694941918336, "learning_rate": 0.0002997565017927398, "loss": 2.7188, "step": 2195 }, { "epoch": 1.0334117647058823, "grad_norm": 1.3272761180167334, "learning_rate": 0.00029975399865619977, "loss": 0.8047, "step": 2196 }, { "epoch": 1.0338823529411765, "grad_norm": 0.6629490250725791, "learning_rate": 0.00029975148272993373, "loss": 0.7188, "step": 2197 }, { "epoch": 1.0343529411764707, "grad_norm": 1.4178464257590842, "learning_rate": 0.0002997489540141565, "loss": 1.6172, "step": 2198 }, { "epoch": 1.0348235294117647, "grad_norm": 0.875933102677497, "learning_rate": 0.00029974641250908406, "loss": 0.4961, "step": 2199 }, { "epoch": 1.035294117647059, "grad_norm": 1.641066598697134, "learning_rate": 0.0002997438582149335, "loss": 1.7969, "step": 2200 }, { "epoch": 1.035764705882353, "grad_norm": 1.59231618860599, "learning_rate": 0.000299741291131923, "loss": 1.4375, "step": 2201 }, { "epoch": 1.0362352941176471, "grad_norm": 1.3976366476362883, "learning_rate": 0.00029973871126027173, "loss": 1.5234, "step": 2202 }, { "epoch": 1.0367058823529411, "grad_norm": 0.6279671640629052, "learning_rate": 0.00029973611860020013, "loss": 0.2256, "step": 2203 }, { "epoch": 1.0371764705882354, "grad_norm": 1.1283666576058295, "learning_rate": 0.0002997335131519294, "loss": 1.0859, "step": 2204 }, { "epoch": 1.0376470588235294, "grad_norm": 1.5670683207326177, "learning_rate": 0.0002997308949156823, "loss": 1.7891, "step": 2205 }, { "epoch": 1.0381176470588236, "grad_norm": 1.0051777650809868, "learning_rate": 0.00029972826389168235, "loss": 0.5586, "step": 2206 }, { "epoch": 1.0385882352941176, "grad_norm": 1.2257907523741858, "learning_rate": 0.00029972562008015425, "loss": 0.9336, "step": 2207 }, { "epoch": 1.0390588235294118, "grad_norm": 1.6436720036982253, "learning_rate": 0.0002997229634813238, "loss": 1.8125, "step": 2208 }, { "epoch": 1.0395294117647058, "grad_norm": 1.58911910785769, "learning_rate": 0.0002997202940954179, "loss": 1.3125, "step": 2209 }, { "epoch": 1.04, "grad_norm": 1.4062617345537651, "learning_rate": 0.00029971761192266447, "loss": 1.2031, "step": 2210 }, { "epoch": 1.040470588235294, "grad_norm": 1.955543298612465, "learning_rate": 0.0002997149169632926, "loss": 1.8516, "step": 2211 }, { "epoch": 1.0409411764705883, "grad_norm": 2.124696716014365, "learning_rate": 0.0002997122092175325, "loss": 1.625, "step": 2212 }, { "epoch": 1.0414117647058823, "grad_norm": 1.0021804167912918, "learning_rate": 0.00029970948868561547, "loss": 0.5195, "step": 2213 }, { "epoch": 1.0418823529411765, "grad_norm": 1.1629434037661663, "learning_rate": 0.00029970675536777365, "loss": 0.8633, "step": 2214 }, { "epoch": 1.0423529411764705, "grad_norm": 1.4296538366321752, "learning_rate": 0.0002997040092642407, "loss": 0.8359, "step": 2215 }, { "epoch": 1.0428235294117647, "grad_norm": 1.6409757183264684, "learning_rate": 0.00029970125037525106, "loss": 1.6953, "step": 2216 }, { "epoch": 1.043294117647059, "grad_norm": 0.9868634509172743, "learning_rate": 0.00029969847870104033, "loss": 1.4453, "step": 2217 }, { "epoch": 1.043764705882353, "grad_norm": 0.6340242906806879, "learning_rate": 0.00029969569424184527, "loss": 0.3008, "step": 2218 }, { "epoch": 1.0442352941176472, "grad_norm": 2.1609298438388733, "learning_rate": 0.0002996928969979037, "loss": 1.8906, "step": 2219 }, { "epoch": 1.0447058823529412, "grad_norm": 1.8508260605307252, "learning_rate": 0.00029969008696945445, "loss": 0.9805, "step": 2220 }, { "epoch": 1.0451764705882354, "grad_norm": 1.2741133578981976, "learning_rate": 0.00029968726415673753, "loss": 1.2656, "step": 2221 }, { "epoch": 1.0456470588235294, "grad_norm": 1.472323310301332, "learning_rate": 0.00029968442855999406, "loss": 1.3359, "step": 2222 }, { "epoch": 1.0461176470588236, "grad_norm": 1.2704143817057714, "learning_rate": 0.0002996815801794662, "loss": 1.3281, "step": 2223 }, { "epoch": 1.0465882352941176, "grad_norm": 0.7787653280956991, "learning_rate": 0.0002996787190153972, "loss": 0.2734, "step": 2224 }, { "epoch": 1.0470588235294118, "grad_norm": 0.9832881017775783, "learning_rate": 0.0002996758450680314, "loss": 0.7148, "step": 2225 }, { "epoch": 1.0475294117647058, "grad_norm": 1.6172028256447226, "learning_rate": 0.0002996729583376143, "loss": 1.75, "step": 2226 }, { "epoch": 1.048, "grad_norm": 1.5584216091802092, "learning_rate": 0.00029967005882439244, "loss": 1.4609, "step": 2227 }, { "epoch": 1.048470588235294, "grad_norm": 1.493929793943094, "learning_rate": 0.0002996671465286134, "loss": 1.3125, "step": 2228 }, { "epoch": 1.0489411764705883, "grad_norm": 1.1888532304941126, "learning_rate": 0.00029966422145052594, "loss": 0.7422, "step": 2229 }, { "epoch": 1.0494117647058823, "grad_norm": 2.5311397457885882, "learning_rate": 0.00029966128359037986, "loss": 1.3594, "step": 2230 }, { "epoch": 1.0498823529411765, "grad_norm": 1.476760098090724, "learning_rate": 0.0002996583329484261, "loss": 1.0156, "step": 2231 }, { "epoch": 1.0503529411764705, "grad_norm": 1.5872789834194, "learning_rate": 0.00029965536952491665, "loss": 1.5234, "step": 2232 }, { "epoch": 1.0508235294117647, "grad_norm": 1.0611523092790296, "learning_rate": 0.00029965239332010454, "loss": 0.6797, "step": 2233 }, { "epoch": 1.0512941176470587, "grad_norm": 2.111885419976075, "learning_rate": 0.000299649404334244, "loss": 2.4219, "step": 2234 }, { "epoch": 1.051764705882353, "grad_norm": 1.3180177720325268, "learning_rate": 0.0002996464025675903, "loss": 1.3281, "step": 2235 }, { "epoch": 1.0522352941176472, "grad_norm": 1.8950194234474929, "learning_rate": 0.00029964338802039987, "loss": 1.2734, "step": 2236 }, { "epoch": 1.0527058823529412, "grad_norm": 1.1684597326196828, "learning_rate": 0.00029964036069293006, "loss": 1.1094, "step": 2237 }, { "epoch": 1.0531764705882354, "grad_norm": 1.2923516973457414, "learning_rate": 0.0002996373205854395, "loss": 1.1016, "step": 2238 }, { "epoch": 1.0536470588235294, "grad_norm": 1.7103067583530875, "learning_rate": 0.00029963426769818774, "loss": 2.2344, "step": 2239 }, { "epoch": 1.0541176470588236, "grad_norm": 0.6776464801312039, "learning_rate": 0.0002996312020314356, "loss": 0.2617, "step": 2240 }, { "epoch": 1.0545882352941176, "grad_norm": 0.8526708210726798, "learning_rate": 0.00029962812358544485, "loss": 0.4082, "step": 2241 }, { "epoch": 1.0550588235294118, "grad_norm": 1.4102962735810312, "learning_rate": 0.0002996250323604785, "loss": 1.3594, "step": 2242 }, { "epoch": 1.0555294117647058, "grad_norm": 1.9784222099589859, "learning_rate": 0.00029962192835680037, "loss": 1.7422, "step": 2243 }, { "epoch": 1.056, "grad_norm": 0.9121895899992642, "learning_rate": 0.0002996188115746757, "loss": 0.3105, "step": 2244 }, { "epoch": 1.056470588235294, "grad_norm": 1.3476468145835816, "learning_rate": 0.0002996156820143707, "loss": 1.3672, "step": 2245 }, { "epoch": 1.0569411764705883, "grad_norm": 1.2454927970808432, "learning_rate": 0.00029961253967615255, "loss": 1.0156, "step": 2246 }, { "epoch": 1.0574117647058823, "grad_norm": 1.3260166276870489, "learning_rate": 0.0002996093845602897, "loss": 1.5391, "step": 2247 }, { "epoch": 1.0578823529411765, "grad_norm": 1.2888888920861372, "learning_rate": 0.00029960621666705155, "loss": 1.2656, "step": 2248 }, { "epoch": 1.0583529411764705, "grad_norm": 1.1829373598303703, "learning_rate": 0.00029960303599670866, "loss": 0.8281, "step": 2249 }, { "epoch": 1.0588235294117647, "grad_norm": 1.6363149743641923, "learning_rate": 0.0002995998425495327, "loss": 1.25, "step": 2250 }, { "epoch": 1.0592941176470587, "grad_norm": 1.558859137043801, "learning_rate": 0.0002995966363257964, "loss": 1.1953, "step": 2251 }, { "epoch": 1.059764705882353, "grad_norm": 1.6905643120510385, "learning_rate": 0.0002995934173257736, "loss": 1.0391, "step": 2252 }, { "epoch": 1.0602352941176472, "grad_norm": 1.2508464151108807, "learning_rate": 0.0002995901855497393, "loss": 1.0938, "step": 2253 }, { "epoch": 1.0607058823529412, "grad_norm": 1.0398952583328844, "learning_rate": 0.00029958694099796934, "loss": 0.7344, "step": 2254 }, { "epoch": 1.0611764705882354, "grad_norm": 0.9422305210441172, "learning_rate": 0.0002995836836707409, "loss": 0.7383, "step": 2255 }, { "epoch": 1.0616470588235294, "grad_norm": 1.433187122949616, "learning_rate": 0.00029958041356833217, "loss": 1.1641, "step": 2256 }, { "epoch": 1.0621176470588236, "grad_norm": 1.05484374094721, "learning_rate": 0.00029957713069102244, "loss": 0.3242, "step": 2257 }, { "epoch": 1.0625882352941176, "grad_norm": 1.2772361008991981, "learning_rate": 0.00029957383503909204, "loss": 1.0391, "step": 2258 }, { "epoch": 1.0630588235294118, "grad_norm": 1.2938290285970593, "learning_rate": 0.00029957052661282256, "loss": 1.1719, "step": 2259 }, { "epoch": 1.0635294117647058, "grad_norm": 0.6766956130755808, "learning_rate": 0.0002995672054124964, "loss": 0.3379, "step": 2260 }, { "epoch": 1.064, "grad_norm": 0.8151268892703185, "learning_rate": 0.0002995638714383973, "loss": 0.7773, "step": 2261 }, { "epoch": 1.064470588235294, "grad_norm": 1.3295975284716839, "learning_rate": 0.00029956052469080994, "loss": 1.2891, "step": 2262 }, { "epoch": 1.0649411764705883, "grad_norm": 1.2357662600016184, "learning_rate": 0.0002995571651700202, "loss": 1.0781, "step": 2263 }, { "epoch": 1.0654117647058823, "grad_norm": 1.19809140006016, "learning_rate": 0.000299553792876315, "loss": 1.0391, "step": 2264 }, { "epoch": 1.0658823529411765, "grad_norm": 0.7113987826282702, "learning_rate": 0.0002995504078099823, "loss": 0.1943, "step": 2265 }, { "epoch": 1.0663529411764705, "grad_norm": 1.0043821815160754, "learning_rate": 0.00029954700997131126, "loss": 0.582, "step": 2266 }, { "epoch": 1.0668235294117647, "grad_norm": 1.2377711084676584, "learning_rate": 0.000299543599360592, "loss": 0.9258, "step": 2267 }, { "epoch": 1.0672941176470587, "grad_norm": 1.1834013821522382, "learning_rate": 0.0002995401759781159, "loss": 0.8203, "step": 2268 }, { "epoch": 1.067764705882353, "grad_norm": 2.2366512007705595, "learning_rate": 0.00029953673982417527, "loss": 2.0469, "step": 2269 }, { "epoch": 1.0682352941176472, "grad_norm": 1.3508836808933329, "learning_rate": 0.00029953329089906357, "loss": 1.7031, "step": 2270 }, { "epoch": 1.0687058823529412, "grad_norm": 0.8747863084764077, "learning_rate": 0.0002995298292030754, "loss": 0.6641, "step": 2271 }, { "epoch": 1.0691764705882354, "grad_norm": 1.3710060223739169, "learning_rate": 0.00029952635473650633, "loss": 1.25, "step": 2272 }, { "epoch": 1.0696470588235294, "grad_norm": 1.3642227593218286, "learning_rate": 0.00029952286749965315, "loss": 1.6094, "step": 2273 }, { "epoch": 1.0701176470588236, "grad_norm": 1.2376515829758734, "learning_rate": 0.0002995193674928137, "loss": 1.125, "step": 2274 }, { "epoch": 1.0705882352941176, "grad_norm": 1.4667552998644124, "learning_rate": 0.0002995158547162869, "loss": 1.4375, "step": 2275 }, { "epoch": 1.0710588235294118, "grad_norm": 0.6989055903574612, "learning_rate": 0.00029951232917037264, "loss": 0.4883, "step": 2276 }, { "epoch": 1.0715294117647058, "grad_norm": 1.0398012270027448, "learning_rate": 0.0002995087908553722, "loss": 1.0859, "step": 2277 }, { "epoch": 1.072, "grad_norm": 1.4947365448386358, "learning_rate": 0.00029950523977158766, "loss": 1.3906, "step": 2278 }, { "epoch": 1.072470588235294, "grad_norm": 1.5061423507193825, "learning_rate": 0.00029950167591932233, "loss": 1.2031, "step": 2279 }, { "epoch": 1.0729411764705883, "grad_norm": 1.856480174003646, "learning_rate": 0.0002994980992988806, "loss": 2.0781, "step": 2280 }, { "epoch": 1.0734117647058823, "grad_norm": 1.066916438728288, "learning_rate": 0.0002994945099105679, "loss": 0.5352, "step": 2281 }, { "epoch": 1.0738823529411765, "grad_norm": 1.0379627280561792, "learning_rate": 0.00029949090775469077, "loss": 0.6055, "step": 2282 }, { "epoch": 1.0743529411764705, "grad_norm": 1.4125478885895093, "learning_rate": 0.00029948729283155683, "loss": 1.0312, "step": 2283 }, { "epoch": 1.0748235294117647, "grad_norm": 1.1299516918688663, "learning_rate": 0.0002994836651414749, "loss": 0.4961, "step": 2284 }, { "epoch": 1.0752941176470587, "grad_norm": 1.3431945752960788, "learning_rate": 0.00029948002468475475, "loss": 1.1797, "step": 2285 }, { "epoch": 1.075764705882353, "grad_norm": 1.012021426773734, "learning_rate": 0.0002994763714617073, "loss": 0.6484, "step": 2286 }, { "epoch": 1.076235294117647, "grad_norm": 1.3180989431933934, "learning_rate": 0.00029947270547264456, "loss": 0.4551, "step": 2287 }, { "epoch": 1.0767058823529412, "grad_norm": 1.2840765856528904, "learning_rate": 0.00029946902671787955, "loss": 1.4453, "step": 2288 }, { "epoch": 1.0771764705882352, "grad_norm": 1.380521608813661, "learning_rate": 0.00029946533519772657, "loss": 1.3672, "step": 2289 }, { "epoch": 1.0776470588235294, "grad_norm": 1.785275159310709, "learning_rate": 0.0002994616309125008, "loss": 2.0625, "step": 2290 }, { "epoch": 1.0781176470588236, "grad_norm": 1.7093894031173562, "learning_rate": 0.0002994579138625187, "loss": 0.6328, "step": 2291 }, { "epoch": 1.0785882352941176, "grad_norm": 1.3639528921126174, "learning_rate": 0.00029945418404809767, "loss": 1.0469, "step": 2292 }, { "epoch": 1.0790588235294118, "grad_norm": 1.4723871127014356, "learning_rate": 0.0002994504414695562, "loss": 1.9453, "step": 2293 }, { "epoch": 1.0795294117647058, "grad_norm": 1.4022173999294811, "learning_rate": 0.000299446686127214, "loss": 0.7969, "step": 2294 }, { "epoch": 1.08, "grad_norm": 0.8147154127475561, "learning_rate": 0.00029944291802139184, "loss": 0.4043, "step": 2295 }, { "epoch": 1.080470588235294, "grad_norm": 1.4526880335785506, "learning_rate": 0.00029943913715241137, "loss": 1.125, "step": 2296 }, { "epoch": 1.0809411764705883, "grad_norm": 2.0253778752074, "learning_rate": 0.0002994353435205956, "loss": 2.0, "step": 2297 }, { "epoch": 1.0814117647058823, "grad_norm": 1.3349013207591538, "learning_rate": 0.00029943153712626855, "loss": 2.1719, "step": 2298 }, { "epoch": 1.0818823529411765, "grad_norm": 1.0195188367490433, "learning_rate": 0.0002994277179697553, "loss": 1.0625, "step": 2299 }, { "epoch": 1.0823529411764705, "grad_norm": 0.7111158385058227, "learning_rate": 0.000299423886051382, "loss": 0.5703, "step": 2300 }, { "epoch": 1.0828235294117647, "grad_norm": 0.5077588397297659, "learning_rate": 0.00029942004137147586, "loss": 0.1523, "step": 2301 }, { "epoch": 1.0832941176470587, "grad_norm": 1.6087604972880882, "learning_rate": 0.0002994161839303653, "loss": 1.3672, "step": 2302 }, { "epoch": 1.083764705882353, "grad_norm": 0.9469165227080185, "learning_rate": 0.00029941231372837974, "loss": 0.9102, "step": 2303 }, { "epoch": 1.084235294117647, "grad_norm": 4.477307427475758, "learning_rate": 0.0002994084307658497, "loss": 0.4961, "step": 2304 }, { "epoch": 1.0847058823529412, "grad_norm": 1.914723005889693, "learning_rate": 0.0002994045350431069, "loss": 1.4922, "step": 2305 }, { "epoch": 1.0851764705882352, "grad_norm": 1.4472947819308917, "learning_rate": 0.00029940062656048396, "loss": 1.9453, "step": 2306 }, { "epoch": 1.0856470588235294, "grad_norm": 1.1515997075784248, "learning_rate": 0.0002993967053183147, "loss": 1.0625, "step": 2307 }, { "epoch": 1.0861176470588236, "grad_norm": 0.7047870858949558, "learning_rate": 0.000299392771316934, "loss": 0.4082, "step": 2308 }, { "epoch": 1.0865882352941176, "grad_norm": 0.9257258054948674, "learning_rate": 0.00029938882455667786, "loss": 0.957, "step": 2309 }, { "epoch": 1.0870588235294119, "grad_norm": 0.8944510270200621, "learning_rate": 0.00029938486503788337, "loss": 0.3262, "step": 2310 }, { "epoch": 1.0875294117647059, "grad_norm": 1.2330447971332918, "learning_rate": 0.0002993808927608887, "loss": 0.6914, "step": 2311 }, { "epoch": 1.088, "grad_norm": 1.0943934921887675, "learning_rate": 0.000299376907726033, "loss": 0.8281, "step": 2312 }, { "epoch": 1.088470588235294, "grad_norm": 2.0658925839687394, "learning_rate": 0.0002993729099336568, "loss": 2.2812, "step": 2313 }, { "epoch": 1.0889411764705883, "grad_norm": 4.346736481919329, "learning_rate": 0.00029936889938410137, "loss": 0.8984, "step": 2314 }, { "epoch": 1.0894117647058823, "grad_norm": 1.7960875387298574, "learning_rate": 0.0002993648760777092, "loss": 1.7578, "step": 2315 }, { "epoch": 1.0898823529411765, "grad_norm": 1.1655505774783894, "learning_rate": 0.0002993608400148241, "loss": 0.9023, "step": 2316 }, { "epoch": 1.0903529411764705, "grad_norm": 1.185991500516807, "learning_rate": 0.00029935679119579066, "loss": 0.9883, "step": 2317 }, { "epoch": 1.0908235294117647, "grad_norm": 1.7285682489915313, "learning_rate": 0.0002993527296209546, "loss": 1.9375, "step": 2318 }, { "epoch": 1.0912941176470587, "grad_norm": 1.6359492143271737, "learning_rate": 0.0002993486552906629, "loss": 1.5859, "step": 2319 }, { "epoch": 1.091764705882353, "grad_norm": 0.7781948190570495, "learning_rate": 0.00029934456820526344, "loss": 0.4453, "step": 2320 }, { "epoch": 1.092235294117647, "grad_norm": 1.3830104924080906, "learning_rate": 0.00029934046836510537, "loss": 1.1719, "step": 2321 }, { "epoch": 1.0927058823529412, "grad_norm": 1.9461332891924332, "learning_rate": 0.00029933635577053877, "loss": 1.8672, "step": 2322 }, { "epoch": 1.0931764705882352, "grad_norm": 1.8909133857393257, "learning_rate": 0.00029933223042191487, "loss": 1.3281, "step": 2323 }, { "epoch": 1.0936470588235294, "grad_norm": 1.6918058601077988, "learning_rate": 0.000299328092319586, "loss": 1.6953, "step": 2324 }, { "epoch": 1.0941176470588236, "grad_norm": 1.4789935478765843, "learning_rate": 0.0002993239414639057, "loss": 1.4219, "step": 2325 }, { "epoch": 1.0945882352941176, "grad_norm": 0.9080442443784335, "learning_rate": 0.0002993197778552283, "loss": 0.3828, "step": 2326 }, { "epoch": 1.0950588235294119, "grad_norm": 1.2322669148304182, "learning_rate": 0.0002993156014939095, "loss": 0.8906, "step": 2327 }, { "epoch": 1.0955294117647059, "grad_norm": 1.2927367833384855, "learning_rate": 0.00029931141238030587, "loss": 0.9141, "step": 2328 }, { "epoch": 1.096, "grad_norm": 1.4390436229597077, "learning_rate": 0.00029930721051477533, "loss": 1.5781, "step": 2329 }, { "epoch": 1.096470588235294, "grad_norm": 1.3178802586621878, "learning_rate": 0.0002993029958976766, "loss": 1.0391, "step": 2330 }, { "epoch": 1.0969411764705883, "grad_norm": 0.5506899845033063, "learning_rate": 0.00029929876852936973, "loss": 0.1787, "step": 2331 }, { "epoch": 1.0974117647058823, "grad_norm": 0.9928419779950589, "learning_rate": 0.00029929452841021573, "loss": 0.832, "step": 2332 }, { "epoch": 1.0978823529411765, "grad_norm": 0.8787882785974855, "learning_rate": 0.00029929027554057675, "loss": 0.4434, "step": 2333 }, { "epoch": 1.0983529411764705, "grad_norm": 0.6445680726959527, "learning_rate": 0.0002992860099208159, "loss": 0.3438, "step": 2334 }, { "epoch": 1.0988235294117648, "grad_norm": 0.6434041295362097, "learning_rate": 0.0002992817315512976, "loss": 0.2148, "step": 2335 }, { "epoch": 1.0992941176470588, "grad_norm": 0.8427056012328112, "learning_rate": 0.0002992774404323872, "loss": 0.3086, "step": 2336 }, { "epoch": 1.099764705882353, "grad_norm": 0.891947008436888, "learning_rate": 0.0002992731365644512, "loss": 0.5078, "step": 2337 }, { "epoch": 1.100235294117647, "grad_norm": 0.8621567112447966, "learning_rate": 0.0002992688199478571, "loss": 0.2656, "step": 2338 }, { "epoch": 1.1007058823529412, "grad_norm": 1.02858883588492, "learning_rate": 0.00029926449058297364, "loss": 0.6484, "step": 2339 }, { "epoch": 1.1011764705882352, "grad_norm": 1.2997713242853701, "learning_rate": 0.0002992601484701706, "loss": 0.9453, "step": 2340 }, { "epoch": 1.1016470588235294, "grad_norm": 1.570344037301121, "learning_rate": 0.0002992557936098187, "loss": 1.3125, "step": 2341 }, { "epoch": 1.1021176470588236, "grad_norm": 1.701202844173099, "learning_rate": 0.00029925142600229, "loss": 1.2578, "step": 2342 }, { "epoch": 1.1025882352941176, "grad_norm": 1.1648205611876916, "learning_rate": 0.00029924704564795735, "loss": 0.4629, "step": 2343 }, { "epoch": 1.1030588235294116, "grad_norm": 3.195350739399087, "learning_rate": 0.000299242652547195, "loss": 1.375, "step": 2344 }, { "epoch": 1.1035294117647059, "grad_norm": 2.221133556684996, "learning_rate": 0.0002992382467003781, "loss": 1.7891, "step": 2345 }, { "epoch": 1.104, "grad_norm": 1.825007250098955, "learning_rate": 0.00029923382810788287, "loss": 2.0, "step": 2346 }, { "epoch": 1.104470588235294, "grad_norm": 1.1156712778893134, "learning_rate": 0.0002992293967700868, "loss": 0.8438, "step": 2347 }, { "epoch": 1.1049411764705883, "grad_norm": 1.5474935671384413, "learning_rate": 0.00029922495268736824, "loss": 0.9141, "step": 2348 }, { "epoch": 1.1054117647058823, "grad_norm": 1.232268343046556, "learning_rate": 0.0002992204958601068, "loss": 1.0781, "step": 2349 }, { "epoch": 1.1058823529411765, "grad_norm": 1.50857422110454, "learning_rate": 0.0002992160262886831, "loss": 1.1406, "step": 2350 }, { "epoch": 1.1063529411764705, "grad_norm": 0.8379545354285325, "learning_rate": 0.0002992115439734788, "loss": 0.6641, "step": 2351 }, { "epoch": 1.1068235294117648, "grad_norm": 0.8968653226765135, "learning_rate": 0.0002992070489148768, "loss": 0.5234, "step": 2352 }, { "epoch": 1.1072941176470588, "grad_norm": 1.179436003204649, "learning_rate": 0.000299202541113261, "loss": 0.7188, "step": 2353 }, { "epoch": 1.107764705882353, "grad_norm": 1.5557677142129596, "learning_rate": 0.0002991980205690163, "loss": 1.1797, "step": 2354 }, { "epoch": 1.108235294117647, "grad_norm": 1.9363247099857335, "learning_rate": 0.0002991934872825289, "loss": 1.7188, "step": 2355 }, { "epoch": 1.1087058823529412, "grad_norm": 1.5721021028394708, "learning_rate": 0.0002991889412541858, "loss": 1.1016, "step": 2356 }, { "epoch": 1.1091764705882352, "grad_norm": 0.88113382988262, "learning_rate": 0.0002991843824843754, "loss": 0.5, "step": 2357 }, { "epoch": 1.1096470588235294, "grad_norm": 1.317944830594401, "learning_rate": 0.000299179810973487, "loss": 1.0625, "step": 2358 }, { "epoch": 1.1101176470588234, "grad_norm": 1.2880957906553365, "learning_rate": 0.000299175226721911, "loss": 1.3828, "step": 2359 }, { "epoch": 1.1105882352941177, "grad_norm": 1.2187660416793578, "learning_rate": 0.00029917062973003897, "loss": 0.5977, "step": 2360 }, { "epoch": 1.1110588235294117, "grad_norm": 1.023996739454288, "learning_rate": 0.0002991660199982634, "loss": 0.5703, "step": 2361 }, { "epoch": 1.1115294117647059, "grad_norm": 1.2567869420329876, "learning_rate": 0.0002991613975269782, "loss": 1.4062, "step": 2362 }, { "epoch": 1.112, "grad_norm": 1.5186533448844657, "learning_rate": 0.0002991567623165779, "loss": 1.9688, "step": 2363 }, { "epoch": 1.112470588235294, "grad_norm": 1.3742873165834337, "learning_rate": 0.00029915211436745855, "loss": 1.2266, "step": 2364 }, { "epoch": 1.1129411764705883, "grad_norm": 1.4559059367357978, "learning_rate": 0.00029914745368001704, "loss": 1.4453, "step": 2365 }, { "epoch": 1.1134117647058823, "grad_norm": 0.9428672110785346, "learning_rate": 0.0002991427802546514, "loss": 0.5195, "step": 2366 }, { "epoch": 1.1138823529411765, "grad_norm": 1.2151860956930407, "learning_rate": 0.0002991380940917608, "loss": 0.9492, "step": 2367 }, { "epoch": 1.1143529411764705, "grad_norm": 1.4237853370308908, "learning_rate": 0.0002991333951917454, "loss": 1.4688, "step": 2368 }, { "epoch": 1.1148235294117648, "grad_norm": 0.7395518605433458, "learning_rate": 0.00029912868355500666, "loss": 0.375, "step": 2369 }, { "epoch": 1.1152941176470588, "grad_norm": 1.1483029361364654, "learning_rate": 0.0002991239591819468, "loss": 0.6211, "step": 2370 }, { "epoch": 1.115764705882353, "grad_norm": 1.705239855547257, "learning_rate": 0.00029911922207296937, "loss": 0.5625, "step": 2371 }, { "epoch": 1.116235294117647, "grad_norm": 1.3588127994458947, "learning_rate": 0.00029911447222847893, "loss": 1.125, "step": 2372 }, { "epoch": 1.1167058823529412, "grad_norm": 0.9045074166880184, "learning_rate": 0.00029910970964888126, "loss": 0.375, "step": 2373 }, { "epoch": 1.1171764705882352, "grad_norm": 1.1171752410937514, "learning_rate": 0.0002991049343345829, "loss": 0.7656, "step": 2374 }, { "epoch": 1.1176470588235294, "grad_norm": 1.48895870157191, "learning_rate": 0.00029910014628599184, "loss": 1.5312, "step": 2375 }, { "epoch": 1.1181176470588234, "grad_norm": 0.962626235801651, "learning_rate": 0.00029909534550351695, "loss": 0.6172, "step": 2376 }, { "epoch": 1.1185882352941177, "grad_norm": 0.6350886562608192, "learning_rate": 0.00029909053198756827, "loss": 0.2402, "step": 2377 }, { "epoch": 1.1190588235294117, "grad_norm": 1.2395273516977792, "learning_rate": 0.00029908570573855683, "loss": 0.8359, "step": 2378 }, { "epoch": 1.1195294117647059, "grad_norm": 0.36032967392063614, "learning_rate": 0.0002990808667568949, "loss": 0.0981, "step": 2379 }, { "epoch": 1.12, "grad_norm": 1.418551506690328, "learning_rate": 0.0002990760150429957, "loss": 1.2344, "step": 2380 }, { "epoch": 1.120470588235294, "grad_norm": 0.9422637281572752, "learning_rate": 0.0002990711505972736, "loss": 0.4668, "step": 2381 }, { "epoch": 1.1209411764705883, "grad_norm": 1.4934921330153386, "learning_rate": 0.00029906627342014407, "loss": 1.4609, "step": 2382 }, { "epoch": 1.1214117647058823, "grad_norm": 1.402271609878635, "learning_rate": 0.00029906138351202356, "loss": 0.9883, "step": 2383 }, { "epoch": 1.1218823529411766, "grad_norm": 1.290771488187728, "learning_rate": 0.0002990564808733298, "loss": 1.1016, "step": 2384 }, { "epoch": 1.1223529411764706, "grad_norm": 0.7681773097182607, "learning_rate": 0.00029905156550448145, "loss": 0.2617, "step": 2385 }, { "epoch": 1.1228235294117648, "grad_norm": 1.0971684178434082, "learning_rate": 0.00029904663740589827, "loss": 0.5117, "step": 2386 }, { "epoch": 1.1232941176470588, "grad_norm": 1.885254398993916, "learning_rate": 0.0002990416965780012, "loss": 1.3906, "step": 2387 }, { "epoch": 1.123764705882353, "grad_norm": 0.9337000560734618, "learning_rate": 0.0002990367430212123, "loss": 0.457, "step": 2388 }, { "epoch": 1.124235294117647, "grad_norm": 1.149059542847512, "learning_rate": 0.0002990317767359544, "loss": 1.1094, "step": 2389 }, { "epoch": 1.1247058823529412, "grad_norm": 1.1917701285786373, "learning_rate": 0.0002990267977226518, "loss": 1.1406, "step": 2390 }, { "epoch": 1.1251764705882352, "grad_norm": 1.2468513355039805, "learning_rate": 0.00029902180598172977, "loss": 1.4453, "step": 2391 }, { "epoch": 1.1256470588235294, "grad_norm": 0.8854236276046997, "learning_rate": 0.00029901680151361447, "loss": 0.5977, "step": 2392 }, { "epoch": 1.1261176470588234, "grad_norm": 1.682876073492398, "learning_rate": 0.00029901178431873346, "loss": 0.9961, "step": 2393 }, { "epoch": 1.1265882352941177, "grad_norm": 0.7930755200548361, "learning_rate": 0.0002990067543975151, "loss": 0.3262, "step": 2394 }, { "epoch": 1.1270588235294117, "grad_norm": 1.0222767871983112, "learning_rate": 0.0002990017117503891, "loss": 0.7578, "step": 2395 }, { "epoch": 1.127529411764706, "grad_norm": 1.7730192977251238, "learning_rate": 0.0002989966563777861, "loss": 0.5469, "step": 2396 }, { "epoch": 1.1280000000000001, "grad_norm": 1.407217691357579, "learning_rate": 0.00029899158828013774, "loss": 0.8281, "step": 2397 }, { "epoch": 1.1284705882352941, "grad_norm": 0.9822459610105848, "learning_rate": 0.000298986507457877, "loss": 0.5195, "step": 2398 }, { "epoch": 1.1289411764705881, "grad_norm": 1.4185467269477792, "learning_rate": 0.0002989814139114377, "loss": 0.7852, "step": 2399 }, { "epoch": 1.1294117647058823, "grad_norm": 1.3407953109981514, "learning_rate": 0.0002989763076412549, "loss": 1.2422, "step": 2400 }, { "epoch": 1.1298823529411766, "grad_norm": 1.677533511734479, "learning_rate": 0.00029897118864776476, "loss": 1.2188, "step": 2401 }, { "epoch": 1.1303529411764706, "grad_norm": 1.2282063207964928, "learning_rate": 0.00029896605693140437, "loss": 0.5703, "step": 2402 }, { "epoch": 1.1308235294117648, "grad_norm": 2.667267393208664, "learning_rate": 0.0002989609124926121, "loss": 2.0625, "step": 2403 }, { "epoch": 1.1312941176470588, "grad_norm": 1.6395496450129021, "learning_rate": 0.0002989557553318272, "loss": 1.5938, "step": 2404 }, { "epoch": 1.131764705882353, "grad_norm": 1.47491420213017, "learning_rate": 0.0002989505854494902, "loss": 1.3203, "step": 2405 }, { "epoch": 1.132235294117647, "grad_norm": 0.6858296288686337, "learning_rate": 0.00029894540284604255, "loss": 0.2676, "step": 2406 }, { "epoch": 1.1327058823529412, "grad_norm": 2.1825157931326133, "learning_rate": 0.000298940207521927, "loss": 2.2344, "step": 2407 }, { "epoch": 1.1331764705882352, "grad_norm": 1.6227682639600558, "learning_rate": 0.00029893499947758706, "loss": 1.7031, "step": 2408 }, { "epoch": 1.1336470588235295, "grad_norm": 1.5403059270957395, "learning_rate": 0.00029892977871346776, "loss": 1.5703, "step": 2409 }, { "epoch": 1.1341176470588235, "grad_norm": 0.8841862363337213, "learning_rate": 0.00029892454523001484, "loss": 0.4746, "step": 2410 }, { "epoch": 1.1345882352941177, "grad_norm": 1.9894527319731643, "learning_rate": 0.00029891929902767526, "loss": 1.0391, "step": 2411 }, { "epoch": 1.1350588235294117, "grad_norm": 1.4859866544341394, "learning_rate": 0.0002989140401068971, "loss": 0.9648, "step": 2412 }, { "epoch": 1.135529411764706, "grad_norm": 1.2212500339913357, "learning_rate": 0.0002989087684681295, "loss": 1.0, "step": 2413 }, { "epoch": 1.1360000000000001, "grad_norm": 1.4643922693430431, "learning_rate": 0.00029890348411182267, "loss": 1.4922, "step": 2414 }, { "epoch": 1.1364705882352941, "grad_norm": 0.7986560155684201, "learning_rate": 0.00029889818703842796, "loss": 0.3184, "step": 2415 }, { "epoch": 1.1369411764705881, "grad_norm": 1.5459053028461656, "learning_rate": 0.0002988928772483977, "loss": 1.7344, "step": 2416 }, { "epoch": 1.1374117647058823, "grad_norm": 1.0644330009954286, "learning_rate": 0.0002988875547421854, "loss": 0.5859, "step": 2417 }, { "epoch": 1.1378823529411766, "grad_norm": 1.393272914078635, "learning_rate": 0.00029888221952024567, "loss": 1.3594, "step": 2418 }, { "epoch": 1.1383529411764706, "grad_norm": 1.6833006822141139, "learning_rate": 0.00029887687158303416, "loss": 1.6641, "step": 2419 }, { "epoch": 1.1388235294117648, "grad_norm": 0.9731264950493067, "learning_rate": 0.00029887151093100755, "loss": 0.5664, "step": 2420 }, { "epoch": 1.1392941176470588, "grad_norm": 1.784011727214658, "learning_rate": 0.00029886613756462365, "loss": 2.1406, "step": 2421 }, { "epoch": 1.139764705882353, "grad_norm": 1.2736381052925343, "learning_rate": 0.00029886075148434146, "loss": 1.7656, "step": 2422 }, { "epoch": 1.140235294117647, "grad_norm": 1.3517143798327984, "learning_rate": 0.0002988553526906209, "loss": 1.3203, "step": 2423 }, { "epoch": 1.1407058823529412, "grad_norm": 2.2436081193910296, "learning_rate": 0.0002988499411839232, "loss": 2.2031, "step": 2424 }, { "epoch": 1.1411764705882352, "grad_norm": 1.6086617561790155, "learning_rate": 0.0002988445169647103, "loss": 1.7109, "step": 2425 }, { "epoch": 1.1416470588235295, "grad_norm": 1.5353505177487414, "learning_rate": 0.00029883908003344554, "loss": 1.5, "step": 2426 }, { "epoch": 1.1421176470588235, "grad_norm": 1.239627038746306, "learning_rate": 0.00029883363039059343, "loss": 1.3203, "step": 2427 }, { "epoch": 1.1425882352941177, "grad_norm": 1.077072818893185, "learning_rate": 0.0002988281680366192, "loss": 1.0234, "step": 2428 }, { "epoch": 1.1430588235294117, "grad_norm": 0.5053621204542565, "learning_rate": 0.0002988226929719894, "loss": 0.6758, "step": 2429 }, { "epoch": 1.143529411764706, "grad_norm": 1.1664336808355025, "learning_rate": 0.0002988172051971717, "loss": 0.9727, "step": 2430 }, { "epoch": 1.144, "grad_norm": 0.8623489048332903, "learning_rate": 0.0002988117047126346, "loss": 0.3066, "step": 2431 }, { "epoch": 1.1444705882352941, "grad_norm": 1.9347642267081266, "learning_rate": 0.00029880619151884813, "loss": 2.1719, "step": 2432 }, { "epoch": 1.1449411764705881, "grad_norm": 1.0909635740015462, "learning_rate": 0.000298800665616283, "loss": 0.6875, "step": 2433 }, { "epoch": 1.1454117647058824, "grad_norm": 1.5303239389175054, "learning_rate": 0.00029879512700541115, "loss": 1.2969, "step": 2434 }, { "epoch": 1.1458823529411766, "grad_norm": 1.581385179793375, "learning_rate": 0.0002987895756867056, "loss": 1.6172, "step": 2435 }, { "epoch": 1.1463529411764706, "grad_norm": 1.593865594039512, "learning_rate": 0.00029878401166064046, "loss": 1.7578, "step": 2436 }, { "epoch": 1.1468235294117648, "grad_norm": 1.7491696327410093, "learning_rate": 0.000298778434927691, "loss": 0.9414, "step": 2437 }, { "epoch": 1.1472941176470588, "grad_norm": 1.0425638956727943, "learning_rate": 0.0002987728454883334, "loss": 0.4336, "step": 2438 }, { "epoch": 1.147764705882353, "grad_norm": 1.094329707814346, "learning_rate": 0.00029876724334304506, "loss": 1.4844, "step": 2439 }, { "epoch": 1.148235294117647, "grad_norm": 1.5484780113725383, "learning_rate": 0.00029876162849230446, "loss": 2.0156, "step": 2440 }, { "epoch": 1.1487058823529412, "grad_norm": 0.9535115472766602, "learning_rate": 0.00029875600093659115, "loss": 1.1172, "step": 2441 }, { "epoch": 1.1491764705882352, "grad_norm": 1.1449823755151025, "learning_rate": 0.00029875036067638565, "loss": 0.7891, "step": 2442 }, { "epoch": 1.1496470588235295, "grad_norm": 1.2962989602635475, "learning_rate": 0.00029874470771216983, "loss": 1.7734, "step": 2443 }, { "epoch": 1.1501176470588235, "grad_norm": 1.2062587316113007, "learning_rate": 0.0002987390420444262, "loss": 1.125, "step": 2444 }, { "epoch": 1.1505882352941177, "grad_norm": 0.9935376619689825, "learning_rate": 0.00029873336367363903, "loss": 1.0156, "step": 2445 }, { "epoch": 1.1510588235294117, "grad_norm": 0.4806555880285856, "learning_rate": 0.00029872767260029294, "loss": 0.1172, "step": 2446 }, { "epoch": 1.151529411764706, "grad_norm": 1.4372559560112086, "learning_rate": 0.00029872196882487417, "loss": 1.0938, "step": 2447 }, { "epoch": 1.152, "grad_norm": 0.8306980847257475, "learning_rate": 0.00029871625234786974, "loss": 0.6055, "step": 2448 }, { "epoch": 1.1524705882352941, "grad_norm": 1.4750752018722084, "learning_rate": 0.0002987105231697679, "loss": 1.1172, "step": 2449 }, { "epoch": 1.1529411764705881, "grad_norm": 1.838144285435725, "learning_rate": 0.000298704781291058, "loss": 2.7812, "step": 2450 }, { "epoch": 1.1534117647058824, "grad_norm": 0.7263220991323303, "learning_rate": 0.0002986990267122304, "loss": 0.2158, "step": 2451 }, { "epoch": 1.1538823529411766, "grad_norm": 0.8126427263956963, "learning_rate": 0.0002986932594337765, "loss": 0.3652, "step": 2452 }, { "epoch": 1.1543529411764706, "grad_norm": 2.2588568998007297, "learning_rate": 0.0002986874794561889, "loss": 2.3125, "step": 2453 }, { "epoch": 1.1548235294117646, "grad_norm": 0.652418915228988, "learning_rate": 0.0002986816867799613, "loss": 0.5547, "step": 2454 }, { "epoch": 1.1552941176470588, "grad_norm": 0.8754232707479211, "learning_rate": 0.0002986758814055883, "loss": 0.7734, "step": 2455 }, { "epoch": 1.155764705882353, "grad_norm": 1.137668649465842, "learning_rate": 0.0002986700633335659, "loss": 0.8477, "step": 2456 }, { "epoch": 1.156235294117647, "grad_norm": 1.9953762584807335, "learning_rate": 0.0002986642325643908, "loss": 2.0312, "step": 2457 }, { "epoch": 1.1567058823529413, "grad_norm": 1.067983713480871, "learning_rate": 0.00029865838909856103, "loss": 0.7812, "step": 2458 }, { "epoch": 1.1571764705882353, "grad_norm": 1.0537844702181212, "learning_rate": 0.0002986525329365757, "loss": 0.6406, "step": 2459 }, { "epoch": 1.1576470588235295, "grad_norm": 2.080555337329314, "learning_rate": 0.00029864666407893487, "loss": 2.5312, "step": 2460 }, { "epoch": 1.1581176470588235, "grad_norm": 1.7334304410247956, "learning_rate": 0.00029864078252613983, "loss": 2.4688, "step": 2461 }, { "epoch": 1.1585882352941177, "grad_norm": 2.176090058720462, "learning_rate": 0.0002986348882786929, "loss": 1.6094, "step": 2462 }, { "epoch": 1.1590588235294117, "grad_norm": 1.0731782559299343, "learning_rate": 0.00029862898133709747, "loss": 0.9688, "step": 2463 }, { "epoch": 1.159529411764706, "grad_norm": 1.5295490110260543, "learning_rate": 0.000298623061701858, "loss": 1.3125, "step": 2464 }, { "epoch": 1.16, "grad_norm": 0.8989419271487356, "learning_rate": 0.00029861712937348007, "loss": 0.6523, "step": 2465 }, { "epoch": 1.1604705882352941, "grad_norm": 2.25409311450963, "learning_rate": 0.0002986111843524703, "loss": 1.8594, "step": 2466 }, { "epoch": 1.1609411764705881, "grad_norm": 0.9166715716763159, "learning_rate": 0.00029860522663933655, "loss": 0.5312, "step": 2467 }, { "epoch": 1.1614117647058824, "grad_norm": 0.9293352261843446, "learning_rate": 0.00029859925623458745, "loss": 0.4316, "step": 2468 }, { "epoch": 1.1618823529411766, "grad_norm": 1.0688832442787, "learning_rate": 0.000298593273138733, "loss": 0.8516, "step": 2469 }, { "epoch": 1.1623529411764706, "grad_norm": 1.456909433677333, "learning_rate": 0.00029858727735228416, "loss": 1.4766, "step": 2470 }, { "epoch": 1.1628235294117646, "grad_norm": 1.3439744227118497, "learning_rate": 0.00029858126887575306, "loss": 1.0859, "step": 2471 }, { "epoch": 1.1632941176470588, "grad_norm": 0.7187837724835605, "learning_rate": 0.00029857524770965285, "loss": 0.3281, "step": 2472 }, { "epoch": 1.163764705882353, "grad_norm": 1.3930188916704764, "learning_rate": 0.00029856921385449763, "loss": 1.1719, "step": 2473 }, { "epoch": 1.164235294117647, "grad_norm": 1.1293216149705128, "learning_rate": 0.0002985631673108029, "loss": 0.7305, "step": 2474 }, { "epoch": 1.1647058823529413, "grad_norm": 1.2584950265671515, "learning_rate": 0.00029855710807908493, "loss": 0.8516, "step": 2475 }, { "epoch": 1.1651764705882353, "grad_norm": 1.0405066670347973, "learning_rate": 0.00029855103615986133, "loss": 0.5312, "step": 2476 }, { "epoch": 1.1656470588235295, "grad_norm": 1.0722218850123317, "learning_rate": 0.0002985449515536505, "loss": 0.7305, "step": 2477 }, { "epoch": 1.1661176470588235, "grad_norm": 1.6738529149819392, "learning_rate": 0.0002985388542609723, "loss": 1.3672, "step": 2478 }, { "epoch": 1.1665882352941177, "grad_norm": 2.0745816915169812, "learning_rate": 0.0002985327442823473, "loss": 2.2344, "step": 2479 }, { "epoch": 1.1670588235294117, "grad_norm": 1.1843065546026053, "learning_rate": 0.00029852662161829744, "loss": 1.2109, "step": 2480 }, { "epoch": 1.167529411764706, "grad_norm": 1.2420914429664924, "learning_rate": 0.0002985204862693455, "loss": 0.9766, "step": 2481 }, { "epoch": 1.168, "grad_norm": 0.9549351392802912, "learning_rate": 0.00029851433823601563, "loss": 0.4648, "step": 2482 }, { "epoch": 1.1684705882352941, "grad_norm": 0.5525191277331448, "learning_rate": 0.00029850817751883276, "loss": 0.2129, "step": 2483 }, { "epoch": 1.1689411764705882, "grad_norm": 1.2353544620609824, "learning_rate": 0.0002985020041183231, "loss": 0.8203, "step": 2484 }, { "epoch": 1.1694117647058824, "grad_norm": 0.9548331925393687, "learning_rate": 0.00029849581803501393, "loss": 0.4355, "step": 2485 }, { "epoch": 1.1698823529411766, "grad_norm": 1.4183499427631445, "learning_rate": 0.0002984896192694335, "loss": 1.2969, "step": 2486 }, { "epoch": 1.1703529411764706, "grad_norm": 1.550809437317743, "learning_rate": 0.00029848340782211125, "loss": 1.6406, "step": 2487 }, { "epoch": 1.1708235294117646, "grad_norm": 0.8004566656645246, "learning_rate": 0.0002984771836935777, "loss": 0.3555, "step": 2488 }, { "epoch": 1.1712941176470588, "grad_norm": 1.2402923301817819, "learning_rate": 0.0002984709468843644, "loss": 1.0938, "step": 2489 }, { "epoch": 1.171764705882353, "grad_norm": 1.4678207791854594, "learning_rate": 0.0002984646973950039, "loss": 1.5156, "step": 2490 }, { "epoch": 1.172235294117647, "grad_norm": 1.3027750478752886, "learning_rate": 0.00029845843522603006, "loss": 1.3047, "step": 2491 }, { "epoch": 1.1727058823529413, "grad_norm": 0.8685144373404778, "learning_rate": 0.0002984521603779777, "loss": 0.6523, "step": 2492 }, { "epoch": 1.1731764705882353, "grad_norm": 0.8940343578291544, "learning_rate": 0.0002984458728513827, "loss": 0.3105, "step": 2493 }, { "epoch": 1.1736470588235295, "grad_norm": 1.5491122938804895, "learning_rate": 0.000298439572646782, "loss": 1.4531, "step": 2494 }, { "epoch": 1.1741176470588235, "grad_norm": 1.4663209041598284, "learning_rate": 0.0002984332597647137, "loss": 1.2578, "step": 2495 }, { "epoch": 1.1745882352941177, "grad_norm": 1.1230575942196261, "learning_rate": 0.000298426934205717, "loss": 0.6406, "step": 2496 }, { "epoch": 1.1750588235294117, "grad_norm": 1.1214772479520658, "learning_rate": 0.00029842059597033204, "loss": 1.0859, "step": 2497 }, { "epoch": 1.175529411764706, "grad_norm": 2.09589880209169, "learning_rate": 0.0002984142450591002, "loss": 1.0312, "step": 2498 }, { "epoch": 1.176, "grad_norm": 0.966955767586349, "learning_rate": 0.0002984078814725639, "loss": 0.582, "step": 2499 }, { "epoch": 1.1764705882352942, "grad_norm": 0.9444834085539022, "learning_rate": 0.0002984015052112665, "loss": 0.4707, "step": 2500 }, { "epoch": 1.1769411764705882, "grad_norm": 0.9644889284728475, "learning_rate": 0.0002983951162757527, "loss": 0.4648, "step": 2501 }, { "epoch": 1.1774117647058824, "grad_norm": 1.562646957526445, "learning_rate": 0.0002983887146665681, "loss": 1.8516, "step": 2502 }, { "epoch": 1.1778823529411764, "grad_norm": 0.9121564997724693, "learning_rate": 0.00029838230038425944, "loss": 0.6094, "step": 2503 }, { "epoch": 1.1783529411764706, "grad_norm": 0.9790997529227353, "learning_rate": 0.0002983758734293745, "loss": 0.8359, "step": 2504 }, { "epoch": 1.1788235294117646, "grad_norm": 1.7882063070794845, "learning_rate": 0.00029836943380246215, "loss": 1.875, "step": 2505 }, { "epoch": 1.1792941176470588, "grad_norm": 1.1458837358724339, "learning_rate": 0.00029836298150407245, "loss": 0.5664, "step": 2506 }, { "epoch": 1.179764705882353, "grad_norm": 1.2103986318812137, "learning_rate": 0.00029835651653475643, "loss": 1.1328, "step": 2507 }, { "epoch": 1.180235294117647, "grad_norm": 1.4475663106413534, "learning_rate": 0.00029835003889506616, "loss": 1.4297, "step": 2508 }, { "epoch": 1.1807058823529413, "grad_norm": 1.3777472112770408, "learning_rate": 0.000298343548585555, "loss": 1.0469, "step": 2509 }, { "epoch": 1.1811764705882353, "grad_norm": 1.8101223151192702, "learning_rate": 0.0002983370456067771, "loss": 1.9609, "step": 2510 }, { "epoch": 1.1816470588235295, "grad_norm": 1.0951414790009126, "learning_rate": 0.0002983305299592879, "loss": 0.7109, "step": 2511 }, { "epoch": 1.1821176470588235, "grad_norm": 1.100932153733525, "learning_rate": 0.0002983240016436439, "loss": 1.125, "step": 2512 }, { "epoch": 1.1825882352941177, "grad_norm": 1.4296202652659185, "learning_rate": 0.0002983174606604027, "loss": 1.4766, "step": 2513 }, { "epoch": 1.1830588235294117, "grad_norm": 1.1188702860884032, "learning_rate": 0.0002983109070101228, "loss": 0.7461, "step": 2514 }, { "epoch": 1.183529411764706, "grad_norm": 1.7464400038278032, "learning_rate": 0.00029830434069336403, "loss": 1.7656, "step": 2515 }, { "epoch": 1.184, "grad_norm": 1.0138219039697567, "learning_rate": 0.0002982977617106871, "loss": 0.5117, "step": 2516 }, { "epoch": 1.1844705882352942, "grad_norm": 1.129151061449467, "learning_rate": 0.0002982911700626539, "loss": 0.8164, "step": 2517 }, { "epoch": 1.1849411764705882, "grad_norm": 0.7078388592681856, "learning_rate": 0.0002982845657498274, "loss": 0.3906, "step": 2518 }, { "epoch": 1.1854117647058824, "grad_norm": 1.4433237789401656, "learning_rate": 0.0002982779487727718, "loss": 1.2656, "step": 2519 }, { "epoch": 1.1858823529411764, "grad_norm": 1.1899633892388002, "learning_rate": 0.0002982713191320519, "loss": 1.1172, "step": 2520 }, { "epoch": 1.1863529411764706, "grad_norm": 1.1429535183292598, "learning_rate": 0.00029826467682823415, "loss": 0.9844, "step": 2521 }, { "epoch": 1.1868235294117646, "grad_norm": 1.2841702426653108, "learning_rate": 0.0002982580218618858, "loss": 0.75, "step": 2522 }, { "epoch": 1.1872941176470588, "grad_norm": 1.9197761105501248, "learning_rate": 0.0002982513542335751, "loss": 1.2422, "step": 2523 }, { "epoch": 1.187764705882353, "grad_norm": 1.2234815409705946, "learning_rate": 0.00029824467394387166, "loss": 0.9531, "step": 2524 }, { "epoch": 1.188235294117647, "grad_norm": 1.2750134504797905, "learning_rate": 0.0002982379809933459, "loss": 1.3203, "step": 2525 }, { "epoch": 1.188705882352941, "grad_norm": 1.2488716585385986, "learning_rate": 0.00029823127538256943, "loss": 1.4141, "step": 2526 }, { "epoch": 1.1891764705882353, "grad_norm": 0.6940196627653201, "learning_rate": 0.000298224557112115, "loss": 0.2334, "step": 2527 }, { "epoch": 1.1896470588235295, "grad_norm": 1.0902218447465655, "learning_rate": 0.0002982178261825563, "loss": 0.3164, "step": 2528 }, { "epoch": 1.1901176470588235, "grad_norm": 1.7648310023545222, "learning_rate": 0.0002982110825944683, "loss": 1.9453, "step": 2529 }, { "epoch": 1.1905882352941177, "grad_norm": 1.1236624566474998, "learning_rate": 0.0002982043263484269, "loss": 0.457, "step": 2530 }, { "epoch": 1.1910588235294117, "grad_norm": 1.4487154490039884, "learning_rate": 0.00029819755744500904, "loss": 1.4922, "step": 2531 }, { "epoch": 1.191529411764706, "grad_norm": 1.4020762270831133, "learning_rate": 0.0002981907758847929, "loss": 1.1641, "step": 2532 }, { "epoch": 1.192, "grad_norm": 0.36771038351912766, "learning_rate": 0.0002981839816683576, "loss": 0.0962, "step": 2533 }, { "epoch": 1.1924705882352942, "grad_norm": 1.4437071955612728, "learning_rate": 0.00029817717479628346, "loss": 1.2734, "step": 2534 }, { "epoch": 1.1929411764705882, "grad_norm": 1.0479837088934953, "learning_rate": 0.00029817035526915177, "loss": 0.5781, "step": 2535 }, { "epoch": 1.1934117647058824, "grad_norm": 1.7826836658515222, "learning_rate": 0.00029816352308754496, "loss": 2.1406, "step": 2536 }, { "epoch": 1.1938823529411764, "grad_norm": 0.6307454837791118, "learning_rate": 0.00029815667825204654, "loss": 0.2539, "step": 2537 }, { "epoch": 1.1943529411764706, "grad_norm": 0.7604137251911648, "learning_rate": 0.0002981498207632411, "loss": 0.3457, "step": 2538 }, { "epoch": 1.1948235294117646, "grad_norm": 1.3257226110459144, "learning_rate": 0.0002981429506217143, "loss": 1.6016, "step": 2539 }, { "epoch": 1.1952941176470588, "grad_norm": 1.1394260076122238, "learning_rate": 0.0002981360678280529, "loss": 0.5547, "step": 2540 }, { "epoch": 1.195764705882353, "grad_norm": 0.4614372002732367, "learning_rate": 0.00029812917238284466, "loss": 0.1001, "step": 2541 }, { "epoch": 1.196235294117647, "grad_norm": 1.0075753370614275, "learning_rate": 0.00029812226428667854, "loss": 0.7109, "step": 2542 }, { "epoch": 1.196705882352941, "grad_norm": 1.2883929753706074, "learning_rate": 0.0002981153435401445, "loss": 0.8594, "step": 2543 }, { "epoch": 1.1971764705882353, "grad_norm": 1.0178803961393534, "learning_rate": 0.00029810841014383366, "loss": 0.3613, "step": 2544 }, { "epoch": 1.1976470588235295, "grad_norm": 1.6132565086772321, "learning_rate": 0.00029810146409833815, "loss": 2.4688, "step": 2545 }, { "epoch": 1.1981176470588235, "grad_norm": 1.327916000643524, "learning_rate": 0.0002980945054042511, "loss": 1.4531, "step": 2546 }, { "epoch": 1.1985882352941177, "grad_norm": 0.44424626995091776, "learning_rate": 0.00029808753406216693, "loss": 0.0972, "step": 2547 }, { "epoch": 1.1990588235294117, "grad_norm": 2.0117046451355143, "learning_rate": 0.000298080550072681, "loss": 2.1719, "step": 2548 }, { "epoch": 1.199529411764706, "grad_norm": 0.8097371031862151, "learning_rate": 0.0002980735534363897, "loss": 0.5742, "step": 2549 }, { "epoch": 1.2, "grad_norm": 1.308428144411917, "learning_rate": 0.0002980665441538907, "loss": 0.6367, "step": 2550 }, { "epoch": 1.2004705882352942, "grad_norm": 1.717882995821359, "learning_rate": 0.00029805952222578257, "loss": 1.4844, "step": 2551 }, { "epoch": 1.2009411764705882, "grad_norm": 1.4038754287303263, "learning_rate": 0.000298052487652665, "loss": 1.2578, "step": 2552 }, { "epoch": 1.2014117647058824, "grad_norm": 1.3041156556152893, "learning_rate": 0.0002980454404351388, "loss": 0.7383, "step": 2553 }, { "epoch": 1.2018823529411764, "grad_norm": 1.56157132533302, "learning_rate": 0.0002980383805738058, "loss": 0.1245, "step": 2554 }, { "epoch": 1.2023529411764706, "grad_norm": 1.0964037178501538, "learning_rate": 0.00029803130806926897, "loss": 0.5625, "step": 2555 }, { "epoch": 1.2028235294117646, "grad_norm": 2.0882869788117, "learning_rate": 0.0002980242229221324, "loss": 2.3594, "step": 2556 }, { "epoch": 1.2032941176470588, "grad_norm": 1.1505931192137722, "learning_rate": 0.00029801712513300105, "loss": 0.9258, "step": 2557 }, { "epoch": 1.2037647058823528, "grad_norm": 1.1618587909004359, "learning_rate": 0.00029801001470248123, "loss": 1.0547, "step": 2558 }, { "epoch": 1.204235294117647, "grad_norm": 1.3156804406463778, "learning_rate": 0.00029800289163118014, "loss": 1.2188, "step": 2559 }, { "epoch": 1.204705882352941, "grad_norm": 0.858435477235104, "learning_rate": 0.00029799575591970623, "loss": 0.3223, "step": 2560 }, { "epoch": 1.2051764705882353, "grad_norm": 0.5346274970416437, "learning_rate": 0.00029798860756866883, "loss": 0.1357, "step": 2561 }, { "epoch": 1.2056470588235295, "grad_norm": 1.647866214733779, "learning_rate": 0.00029798144657867845, "loss": 1.3203, "step": 2562 }, { "epoch": 1.2061176470588235, "grad_norm": 0.5125804767087988, "learning_rate": 0.0002979742729503467, "loss": 0.2617, "step": 2563 }, { "epoch": 1.2065882352941177, "grad_norm": 1.7751830307512853, "learning_rate": 0.0002979670866842862, "loss": 1.2578, "step": 2564 }, { "epoch": 1.2070588235294117, "grad_norm": 0.715331259358151, "learning_rate": 0.0002979598877811107, "loss": 0.7305, "step": 2565 }, { "epoch": 1.207529411764706, "grad_norm": 1.438947219347154, "learning_rate": 0.0002979526762414351, "loss": 1.4688, "step": 2566 }, { "epoch": 1.208, "grad_norm": 1.3004462628764677, "learning_rate": 0.0002979454520658752, "loss": 1.4531, "step": 2567 }, { "epoch": 1.2084705882352942, "grad_norm": 0.40850201105160827, "learning_rate": 0.0002979382152550481, "loss": 0.0615, "step": 2568 }, { "epoch": 1.2089411764705882, "grad_norm": 0.7614004494675692, "learning_rate": 0.0002979309658095717, "loss": 0.4824, "step": 2569 }, { "epoch": 1.2094117647058824, "grad_norm": 1.1507133869223827, "learning_rate": 0.0002979237037300653, "loss": 0.8984, "step": 2570 }, { "epoch": 1.2098823529411764, "grad_norm": 1.230041200446324, "learning_rate": 0.00029791642901714897, "loss": 1.2969, "step": 2571 }, { "epoch": 1.2103529411764706, "grad_norm": 1.4054537853696956, "learning_rate": 0.0002979091416714441, "loss": 1.5234, "step": 2572 }, { "epoch": 1.2108235294117646, "grad_norm": 1.3654142909847815, "learning_rate": 0.00029790184169357307, "loss": 1.3984, "step": 2573 }, { "epoch": 1.2112941176470589, "grad_norm": 1.282219731389197, "learning_rate": 0.00029789452908415923, "loss": 1.3438, "step": 2574 }, { "epoch": 1.2117647058823529, "grad_norm": 1.0150120511559872, "learning_rate": 0.0002978872038438273, "loss": 1.0469, "step": 2575 }, { "epoch": 1.212235294117647, "grad_norm": 0.7208337466449064, "learning_rate": 0.0002978798659732027, "loss": 0.4277, "step": 2576 }, { "epoch": 1.212705882352941, "grad_norm": 1.5750464914294129, "learning_rate": 0.00029787251547291215, "loss": 1.625, "step": 2577 }, { "epoch": 1.2131764705882353, "grad_norm": 1.0858677541948556, "learning_rate": 0.00029786515234358357, "loss": 0.4902, "step": 2578 }, { "epoch": 1.2136470588235295, "grad_norm": 1.453235978947074, "learning_rate": 0.0002978577765858457, "loss": 1.3516, "step": 2579 }, { "epoch": 1.2141176470588235, "grad_norm": 0.8943596224995587, "learning_rate": 0.00029785038820032844, "loss": 0.5625, "step": 2580 }, { "epoch": 1.2145882352941175, "grad_norm": 1.0679229859881527, "learning_rate": 0.0002978429871876628, "loss": 0.7773, "step": 2581 }, { "epoch": 1.2150588235294117, "grad_norm": 1.0888061470915134, "learning_rate": 0.00029783557354848094, "loss": 0.8359, "step": 2582 }, { "epoch": 1.215529411764706, "grad_norm": 1.730788884840852, "learning_rate": 0.0002978281472834159, "loss": 1.4453, "step": 2583 }, { "epoch": 1.216, "grad_norm": 2.9443766454061047, "learning_rate": 0.00029782070839310206, "loss": 3.0156, "step": 2584 }, { "epoch": 1.2164705882352942, "grad_norm": 1.449309959511531, "learning_rate": 0.0002978132568781746, "loss": 1.2344, "step": 2585 }, { "epoch": 1.2169411764705882, "grad_norm": 1.1133807684645096, "learning_rate": 0.00029780579273927006, "loss": 1.0156, "step": 2586 }, { "epoch": 1.2174117647058824, "grad_norm": 1.5097201987719526, "learning_rate": 0.0002977983159770258, "loss": 1.2891, "step": 2587 }, { "epoch": 1.2178823529411764, "grad_norm": 0.5490666398790794, "learning_rate": 0.0002977908265920804, "loss": 0.3203, "step": 2588 }, { "epoch": 1.2183529411764706, "grad_norm": 1.1949720358837432, "learning_rate": 0.00029778332458507353, "loss": 1.4297, "step": 2589 }, { "epoch": 1.2188235294117646, "grad_norm": 1.3145170150657426, "learning_rate": 0.0002977758099566458, "loss": 1.4219, "step": 2590 }, { "epoch": 1.2192941176470589, "grad_norm": 1.0629085059709584, "learning_rate": 0.0002977682827074391, "loss": 0.5664, "step": 2591 }, { "epoch": 1.2197647058823529, "grad_norm": 1.1137252201559642, "learning_rate": 0.0002977607428380963, "loss": 0.7266, "step": 2592 }, { "epoch": 1.220235294117647, "grad_norm": 1.3782409953193013, "learning_rate": 0.00029775319034926125, "loss": 1.2812, "step": 2593 }, { "epoch": 1.220705882352941, "grad_norm": 1.4358646133162676, "learning_rate": 0.000297745625241579, "loss": 1.8359, "step": 2594 }, { "epoch": 1.2211764705882353, "grad_norm": 1.3681123582241221, "learning_rate": 0.0002977380475156958, "loss": 1.2109, "step": 2595 }, { "epoch": 1.2216470588235295, "grad_norm": 0.9144503453190916, "learning_rate": 0.00029773045717225856, "loss": 0.7656, "step": 2596 }, { "epoch": 1.2221176470588235, "grad_norm": 1.2521639382467662, "learning_rate": 0.0002977228542119157, "loss": 1.1406, "step": 2597 }, { "epoch": 1.2225882352941175, "grad_norm": 0.9634437612070478, "learning_rate": 0.0002977152386353165, "loss": 0.4805, "step": 2598 }, { "epoch": 1.2230588235294118, "grad_norm": 1.058462301024472, "learning_rate": 0.0002977076104431114, "loss": 0.8789, "step": 2599 }, { "epoch": 1.223529411764706, "grad_norm": 0.49667106457674787, "learning_rate": 0.00029769996963595184, "loss": 0.1172, "step": 2600 }, { "epoch": 1.224, "grad_norm": 1.4497715227748378, "learning_rate": 0.0002976923162144904, "loss": 1.5312, "step": 2601 }, { "epoch": 1.2244705882352942, "grad_norm": 1.550777187569834, "learning_rate": 0.0002976846501793808, "loss": 1.0234, "step": 2602 }, { "epoch": 1.2249411764705882, "grad_norm": 0.9920853879293023, "learning_rate": 0.00029767697153127766, "loss": 0.5703, "step": 2603 }, { "epoch": 1.2254117647058824, "grad_norm": 1.4962327737055432, "learning_rate": 0.0002976692802708368, "loss": 1.5781, "step": 2604 }, { "epoch": 1.2258823529411764, "grad_norm": 1.3137163637966365, "learning_rate": 0.000297661576398715, "loss": 1.1641, "step": 2605 }, { "epoch": 1.2263529411764706, "grad_norm": 1.5357534331805898, "learning_rate": 0.0002976538599155704, "loss": 1.7969, "step": 2606 }, { "epoch": 1.2268235294117646, "grad_norm": 0.7270067916580502, "learning_rate": 0.00029764613082206183, "loss": 0.3828, "step": 2607 }, { "epoch": 1.2272941176470589, "grad_norm": 0.894941239251672, "learning_rate": 0.0002976383891188496, "loss": 0.5, "step": 2608 }, { "epoch": 1.2277647058823529, "grad_norm": 1.3980904691610367, "learning_rate": 0.0002976306348065947, "loss": 1.625, "step": 2609 }, { "epoch": 1.228235294117647, "grad_norm": 1.2742834356890305, "learning_rate": 0.00029762286788595954, "loss": 1.3984, "step": 2610 }, { "epoch": 1.228705882352941, "grad_norm": 0.5730131720126391, "learning_rate": 0.0002976150883576073, "loss": 0.4023, "step": 2611 }, { "epoch": 1.2291764705882353, "grad_norm": 1.1291735999919497, "learning_rate": 0.0002976072962222025, "loss": 0.7344, "step": 2612 }, { "epoch": 1.2296470588235293, "grad_norm": 1.63812266747462, "learning_rate": 0.0002975994914804106, "loss": 2.125, "step": 2613 }, { "epoch": 1.2301176470588235, "grad_norm": 1.4663872402540974, "learning_rate": 0.00029759167413289815, "loss": 1.2812, "step": 2614 }, { "epoch": 1.2305882352941175, "grad_norm": 1.2320338927853904, "learning_rate": 0.00029758384418033276, "loss": 1.1875, "step": 2615 }, { "epoch": 1.2310588235294118, "grad_norm": 1.2606956255413702, "learning_rate": 0.00029757600162338326, "loss": 0.7461, "step": 2616 }, { "epoch": 1.231529411764706, "grad_norm": 1.2624172124447512, "learning_rate": 0.0002975681464627193, "loss": 1.3281, "step": 2617 }, { "epoch": 1.232, "grad_norm": 1.250138241451433, "learning_rate": 0.00029756027869901183, "loss": 1.1328, "step": 2618 }, { "epoch": 1.2324705882352942, "grad_norm": 1.2323019811564364, "learning_rate": 0.0002975523983329328, "loss": 0.9453, "step": 2619 }, { "epoch": 1.2329411764705882, "grad_norm": 1.0772114450965322, "learning_rate": 0.00029754450536515517, "loss": 0.5781, "step": 2620 }, { "epoch": 1.2334117647058824, "grad_norm": 1.083023305337584, "learning_rate": 0.00029753659979635315, "loss": 0.5117, "step": 2621 }, { "epoch": 1.2338823529411764, "grad_norm": 1.446569666065713, "learning_rate": 0.00029752868162720176, "loss": 1.7812, "step": 2622 }, { "epoch": 1.2343529411764707, "grad_norm": 0.9434927221282221, "learning_rate": 0.0002975207508583774, "loss": 0.7539, "step": 2623 }, { "epoch": 1.2348235294117647, "grad_norm": 0.9899003583337463, "learning_rate": 0.00029751280749055727, "loss": 0.6641, "step": 2624 }, { "epoch": 1.2352941176470589, "grad_norm": 1.1571236867272177, "learning_rate": 0.0002975048515244199, "loss": 1.1875, "step": 2625 }, { "epoch": 1.2357647058823529, "grad_norm": 1.5852717569921497, "learning_rate": 0.0002974968829606447, "loss": 1.6953, "step": 2626 }, { "epoch": 1.236235294117647, "grad_norm": 1.2982911995189463, "learning_rate": 0.0002974889017999122, "loss": 1.5391, "step": 2627 }, { "epoch": 1.236705882352941, "grad_norm": 1.3277150252249235, "learning_rate": 0.00029748090804290403, "loss": 1.7344, "step": 2628 }, { "epoch": 1.2371764705882353, "grad_norm": 1.8500666014490985, "learning_rate": 0.00029747290169030295, "loss": 2.0781, "step": 2629 }, { "epoch": 1.2376470588235293, "grad_norm": 0.617045454147249, "learning_rate": 0.0002974648827427927, "loss": 0.2852, "step": 2630 }, { "epoch": 1.2381176470588235, "grad_norm": 0.8568310481958212, "learning_rate": 0.0002974568512010582, "loss": 0.6094, "step": 2631 }, { "epoch": 1.2385882352941175, "grad_norm": 0.7045660133435908, "learning_rate": 0.00029744880706578526, "loss": 0.3613, "step": 2632 }, { "epoch": 1.2390588235294118, "grad_norm": 1.3666317569597342, "learning_rate": 0.00029744075033766103, "loss": 1.4297, "step": 2633 }, { "epoch": 1.239529411764706, "grad_norm": 0.780021108465248, "learning_rate": 0.0002974326810173735, "loss": 0.3359, "step": 2634 }, { "epoch": 1.24, "grad_norm": 1.3348709888550325, "learning_rate": 0.00029742459910561187, "loss": 1.1328, "step": 2635 }, { "epoch": 1.2404705882352942, "grad_norm": 1.0972761208155315, "learning_rate": 0.00029741650460306636, "loss": 1.0859, "step": 2636 }, { "epoch": 1.2409411764705882, "grad_norm": 1.2691793498974788, "learning_rate": 0.0002974083975104283, "loss": 1.1406, "step": 2637 }, { "epoch": 1.2414117647058824, "grad_norm": 1.252852625919683, "learning_rate": 0.00029740027782839003, "loss": 0.9609, "step": 2638 }, { "epoch": 1.2418823529411764, "grad_norm": 1.661464882507642, "learning_rate": 0.00029739214555764504, "loss": 1.8203, "step": 2639 }, { "epoch": 1.2423529411764707, "grad_norm": 1.25169140371306, "learning_rate": 0.00029738400069888787, "loss": 1.7188, "step": 2640 }, { "epoch": 1.2428235294117647, "grad_norm": 1.0931734282883399, "learning_rate": 0.00029737584325281415, "loss": 0.9414, "step": 2641 }, { "epoch": 1.2432941176470589, "grad_norm": 0.8792389757925849, "learning_rate": 0.0002973676732201205, "loss": 0.3672, "step": 2642 }, { "epoch": 1.2437647058823529, "grad_norm": 0.9790815783298825, "learning_rate": 0.00029735949060150476, "loss": 1.0234, "step": 2643 }, { "epoch": 1.244235294117647, "grad_norm": 1.275949338853641, "learning_rate": 0.0002973512953976657, "loss": 1.5156, "step": 2644 }, { "epoch": 1.244705882352941, "grad_norm": 1.2087292551508428, "learning_rate": 0.0002973430876093033, "loss": 0.5273, "step": 2645 }, { "epoch": 1.2451764705882353, "grad_norm": 0.9693751908062314, "learning_rate": 0.00029733486723711845, "loss": 0.8008, "step": 2646 }, { "epoch": 1.2456470588235293, "grad_norm": 1.0607104368204758, "learning_rate": 0.0002973266342818133, "loss": 0.6367, "step": 2647 }, { "epoch": 1.2461176470588236, "grad_norm": 1.2868842510680456, "learning_rate": 0.00029731838874409093, "loss": 1.4062, "step": 2648 }, { "epoch": 1.2465882352941176, "grad_norm": 1.1495125982719046, "learning_rate": 0.0002973101306246556, "loss": 0.7109, "step": 2649 }, { "epoch": 1.2470588235294118, "grad_norm": 1.1783024959735133, "learning_rate": 0.0002973018599242125, "loss": 0.9375, "step": 2650 }, { "epoch": 1.247529411764706, "grad_norm": 1.2786250807866018, "learning_rate": 0.0002972935766434681, "loss": 1.2969, "step": 2651 }, { "epoch": 1.248, "grad_norm": 1.025890722191823, "learning_rate": 0.00029728528078312976, "loss": 0.9141, "step": 2652 }, { "epoch": 1.248470588235294, "grad_norm": 1.0216853788586813, "learning_rate": 0.00029727697234390604, "loss": 0.8516, "step": 2653 }, { "epoch": 1.2489411764705882, "grad_norm": 1.3081368996285079, "learning_rate": 0.0002972686513265065, "loss": 1.6016, "step": 2654 }, { "epoch": 1.2494117647058824, "grad_norm": 1.118947993713679, "learning_rate": 0.00029726031773164176, "loss": 0.9102, "step": 2655 }, { "epoch": 1.2498823529411764, "grad_norm": 0.8976285004743602, "learning_rate": 0.00029725197156002357, "loss": 0.5664, "step": 2656 }, { "epoch": 1.2503529411764707, "grad_norm": 2.0277758302666, "learning_rate": 0.00029724361281236475, "loss": 1.8984, "step": 2657 }, { "epoch": 1.2508235294117647, "grad_norm": 0.8721446500851994, "learning_rate": 0.0002972352414893792, "loss": 0.4062, "step": 2658 }, { "epoch": 1.251294117647059, "grad_norm": 1.2361797676253938, "learning_rate": 0.00029722685759178187, "loss": 1.1875, "step": 2659 }, { "epoch": 1.251764705882353, "grad_norm": 1.0302257676206117, "learning_rate": 0.0002972184611202886, "loss": 0.9336, "step": 2660 }, { "epoch": 1.2522352941176471, "grad_norm": 1.0693872272264227, "learning_rate": 0.00029721005207561683, "loss": 0.8477, "step": 2661 }, { "epoch": 1.2527058823529411, "grad_norm": 1.7062552013433054, "learning_rate": 0.00029720163045848444, "loss": 1.3672, "step": 2662 }, { "epoch": 1.2531764705882353, "grad_norm": 2.184175435669623, "learning_rate": 0.0002971931962696109, "loss": 2.2969, "step": 2663 }, { "epoch": 1.2536470588235293, "grad_norm": 0.8290558056212421, "learning_rate": 0.00029718474950971626, "loss": 0.7383, "step": 2664 }, { "epoch": 1.2541176470588236, "grad_norm": 0.3729615305517579, "learning_rate": 0.0002971762901795222, "loss": 0.0933, "step": 2665 }, { "epoch": 1.2545882352941176, "grad_norm": 1.535184264078274, "learning_rate": 0.000297167818279751, "loss": 1.25, "step": 2666 }, { "epoch": 1.2550588235294118, "grad_norm": 1.7238580679563529, "learning_rate": 0.0002971593338111263, "loss": 1.2734, "step": 2667 }, { "epoch": 1.255529411764706, "grad_norm": 1.2946232062556764, "learning_rate": 0.0002971508367743727, "loss": 1.4688, "step": 2668 }, { "epoch": 1.256, "grad_norm": 1.3395462012626513, "learning_rate": 0.0002971423271702158, "loss": 1.3594, "step": 2669 }, { "epoch": 1.256470588235294, "grad_norm": 1.0511881549708872, "learning_rate": 0.00029713380499938246, "loss": 0.7109, "step": 2670 }, { "epoch": 1.2569411764705882, "grad_norm": 1.587366510524418, "learning_rate": 0.00029712527026260047, "loss": 2.2031, "step": 2671 }, { "epoch": 1.2574117647058825, "grad_norm": 1.2041976280965772, "learning_rate": 0.0002971167229605988, "loss": 0.957, "step": 2672 }, { "epoch": 1.2578823529411765, "grad_norm": 1.2394534458864261, "learning_rate": 0.00029710816309410737, "loss": 1.2422, "step": 2673 }, { "epoch": 1.2583529411764705, "grad_norm": 1.4494470401381336, "learning_rate": 0.00029709959066385723, "loss": 1.1562, "step": 2674 }, { "epoch": 1.2588235294117647, "grad_norm": 0.9242469581426503, "learning_rate": 0.0002970910056705806, "loss": 0.4609, "step": 2675 }, { "epoch": 1.259294117647059, "grad_norm": 1.8052598876531325, "learning_rate": 0.0002970824081150105, "loss": 2.2656, "step": 2676 }, { "epoch": 1.259764705882353, "grad_norm": 1.2656053137329797, "learning_rate": 0.00029707379799788134, "loss": 1.375, "step": 2677 }, { "epoch": 1.2602352941176471, "grad_norm": 1.524356555707817, "learning_rate": 0.0002970651753199285, "loss": 1.2891, "step": 2678 }, { "epoch": 1.2607058823529411, "grad_norm": 1.465711726906621, "learning_rate": 0.0002970565400818883, "loss": 1.5391, "step": 2679 }, { "epoch": 1.2611764705882353, "grad_norm": 1.113318384477569, "learning_rate": 0.0002970478922844983, "loss": 1.0156, "step": 2680 }, { "epoch": 1.2616470588235293, "grad_norm": 1.429745298607517, "learning_rate": 0.00029703923192849694, "loss": 1.2656, "step": 2681 }, { "epoch": 1.2621176470588236, "grad_norm": 0.6250358741643162, "learning_rate": 0.000297030559014624, "loss": 0.2754, "step": 2682 }, { "epoch": 1.2625882352941176, "grad_norm": 1.5252573034290569, "learning_rate": 0.00029702187354362016, "loss": 1.7656, "step": 2683 }, { "epoch": 1.2630588235294118, "grad_norm": 1.0028864039672312, "learning_rate": 0.0002970131755162272, "loss": 0.6797, "step": 2684 }, { "epoch": 1.263529411764706, "grad_norm": 1.2261566275636964, "learning_rate": 0.0002970044649331879, "loss": 0.9023, "step": 2685 }, { "epoch": 1.264, "grad_norm": 0.9478009208212506, "learning_rate": 0.0002969957417952463, "loss": 0.7578, "step": 2686 }, { "epoch": 1.264470588235294, "grad_norm": 1.2462298184882743, "learning_rate": 0.00029698700610314725, "loss": 1.1641, "step": 2687 }, { "epoch": 1.2649411764705882, "grad_norm": 0.8953475869296476, "learning_rate": 0.000296978257857637, "loss": 0.9258, "step": 2688 }, { "epoch": 1.2654117647058825, "grad_norm": 1.5051810069349163, "learning_rate": 0.00029696949705946255, "loss": 1.2578, "step": 2689 }, { "epoch": 1.2658823529411765, "grad_norm": 0.8498889856924633, "learning_rate": 0.0002969607237093722, "loss": 0.373, "step": 2690 }, { "epoch": 1.2663529411764705, "grad_norm": 1.7670138037441216, "learning_rate": 0.00029695193780811523, "loss": 2.0938, "step": 2691 }, { "epoch": 1.2668235294117647, "grad_norm": 1.0427675428715857, "learning_rate": 0.0002969431393564419, "loss": 0.6211, "step": 2692 }, { "epoch": 1.267294117647059, "grad_norm": 1.6048419101269584, "learning_rate": 0.0002969343283551038, "loss": 1.0625, "step": 2693 }, { "epoch": 1.267764705882353, "grad_norm": 1.393679706775035, "learning_rate": 0.0002969255048048533, "loss": 1.2266, "step": 2694 }, { "epoch": 1.2682352941176471, "grad_norm": 0.7103271138185179, "learning_rate": 0.00029691666870644405, "loss": 0.2451, "step": 2695 }, { "epoch": 1.2687058823529411, "grad_norm": 1.3936173791172568, "learning_rate": 0.0002969078200606307, "loss": 1.6406, "step": 2696 }, { "epoch": 1.2691764705882354, "grad_norm": 2.347219318814232, "learning_rate": 0.0002968989588681688, "loss": 3.3125, "step": 2697 }, { "epoch": 1.2696470588235294, "grad_norm": 0.9674021063761741, "learning_rate": 0.0002968900851298154, "loss": 0.5664, "step": 2698 }, { "epoch": 1.2701176470588236, "grad_norm": 1.6344920176400273, "learning_rate": 0.00029688119884632827, "loss": 1.7109, "step": 2699 }, { "epoch": 1.2705882352941176, "grad_norm": 0.8554171808662792, "learning_rate": 0.0002968723000184662, "loss": 0.4062, "step": 2700 }, { "epoch": 1.2710588235294118, "grad_norm": 0.9026238376950299, "learning_rate": 0.0002968633886469894, "loss": 0.459, "step": 2701 }, { "epoch": 1.271529411764706, "grad_norm": 1.8830025116693105, "learning_rate": 0.00029685446473265876, "loss": 1.9375, "step": 2702 }, { "epoch": 1.272, "grad_norm": 1.3936183582381736, "learning_rate": 0.00029684552827623657, "loss": 1.4766, "step": 2703 }, { "epoch": 1.272470588235294, "grad_norm": 1.569183520904831, "learning_rate": 0.000296836579278486, "loss": 1.3438, "step": 2704 }, { "epoch": 1.2729411764705882, "grad_norm": 1.2239078992268386, "learning_rate": 0.0002968276177401713, "loss": 1.1562, "step": 2705 }, { "epoch": 1.2734117647058825, "grad_norm": 1.2476930129550683, "learning_rate": 0.00029681864366205786, "loss": 0.9297, "step": 2706 }, { "epoch": 1.2738823529411765, "grad_norm": 1.0109367535842069, "learning_rate": 0.0002968096570449121, "loss": 1.0625, "step": 2707 }, { "epoch": 1.2743529411764705, "grad_norm": 1.331154613604552, "learning_rate": 0.0002968006578895016, "loss": 1.6016, "step": 2708 }, { "epoch": 1.2748235294117647, "grad_norm": 1.2334634562467173, "learning_rate": 0.0002967916461965948, "loss": 0.9922, "step": 2709 }, { "epoch": 1.275294117647059, "grad_norm": 1.1399520857700147, "learning_rate": 0.00029678262196696143, "loss": 0.9414, "step": 2710 }, { "epoch": 1.275764705882353, "grad_norm": 1.3290423727057004, "learning_rate": 0.0002967735852013722, "loss": 1.3281, "step": 2711 }, { "epoch": 1.2762352941176471, "grad_norm": 0.9713140969256824, "learning_rate": 0.0002967645359005988, "loss": 0.75, "step": 2712 }, { "epoch": 1.2767058823529411, "grad_norm": 1.3685683049268522, "learning_rate": 0.0002967554740654142, "loss": 1.9141, "step": 2713 }, { "epoch": 1.2771764705882354, "grad_norm": 0.8131239908957085, "learning_rate": 0.0002967463996965923, "loss": 0.6758, "step": 2714 }, { "epoch": 1.2776470588235294, "grad_norm": 1.314878537539117, "learning_rate": 0.0002967373127949081, "loss": 1.7422, "step": 2715 }, { "epoch": 1.2781176470588236, "grad_norm": 0.7382940197054455, "learning_rate": 0.0002967282133611375, "loss": 0.3379, "step": 2716 }, { "epoch": 1.2785882352941176, "grad_norm": 0.9073581649088671, "learning_rate": 0.0002967191013960579, "loss": 0.4121, "step": 2717 }, { "epoch": 1.2790588235294118, "grad_norm": 1.0665858430637303, "learning_rate": 0.00029670997690044745, "loss": 0.8711, "step": 2718 }, { "epoch": 1.2795294117647058, "grad_norm": 1.021924786277104, "learning_rate": 0.0002967008398750853, "loss": 0.7227, "step": 2719 }, { "epoch": 1.28, "grad_norm": 1.3073642355688677, "learning_rate": 0.00029669169032075186, "loss": 1.2344, "step": 2720 }, { "epoch": 1.280470588235294, "grad_norm": 1.3744509320039846, "learning_rate": 0.0002966825282382285, "loss": 1.4375, "step": 2721 }, { "epoch": 1.2809411764705882, "grad_norm": 1.3723327166391992, "learning_rate": 0.0002966733536282978, "loss": 0.9648, "step": 2722 }, { "epoch": 1.2814117647058825, "grad_norm": 1.243315409622818, "learning_rate": 0.0002966641664917433, "loss": 1.0625, "step": 2723 }, { "epoch": 1.2818823529411765, "grad_norm": 1.0911071205181444, "learning_rate": 0.0002966549668293495, "loss": 0.6328, "step": 2724 }, { "epoch": 1.2823529411764705, "grad_norm": 1.197708548205978, "learning_rate": 0.0002966457546419023, "loss": 1.1016, "step": 2725 }, { "epoch": 1.2828235294117647, "grad_norm": 1.3734281316157158, "learning_rate": 0.0002966365299301883, "loss": 1.3984, "step": 2726 }, { "epoch": 1.283294117647059, "grad_norm": 0.4589203426987448, "learning_rate": 0.00029662729269499546, "loss": 0.21, "step": 2727 }, { "epoch": 1.283764705882353, "grad_norm": 1.741306798976007, "learning_rate": 0.00029661804293711254, "loss": 1.8438, "step": 2728 }, { "epoch": 1.2842352941176471, "grad_norm": 1.4707445068060823, "learning_rate": 0.00029660878065732965, "loss": 1.1719, "step": 2729 }, { "epoch": 1.2847058823529411, "grad_norm": 1.3429735318318015, "learning_rate": 0.0002965995058564377, "loss": 1.8359, "step": 2730 }, { "epoch": 1.2851764705882354, "grad_norm": 1.2263891123014472, "learning_rate": 0.000296590218535229, "loss": 1.2109, "step": 2731 }, { "epoch": 1.2856470588235294, "grad_norm": 0.9755719982133633, "learning_rate": 0.00029658091869449657, "loss": 0.6172, "step": 2732 }, { "epoch": 1.2861176470588236, "grad_norm": 1.320069050858779, "learning_rate": 0.0002965716063350347, "loss": 1.2109, "step": 2733 }, { "epoch": 1.2865882352941176, "grad_norm": 1.728288918820385, "learning_rate": 0.0002965622814576387, "loss": 1.4922, "step": 2734 }, { "epoch": 1.2870588235294118, "grad_norm": 1.0880892491760623, "learning_rate": 0.000296552944063105, "loss": 0.6836, "step": 2735 }, { "epoch": 1.2875294117647058, "grad_norm": 1.3599743236363075, "learning_rate": 0.0002965435941522311, "loss": 1.5, "step": 2736 }, { "epoch": 1.288, "grad_norm": 2.1689250249341794, "learning_rate": 0.00029653423172581534, "loss": 2.2656, "step": 2737 }, { "epoch": 1.288470588235294, "grad_norm": 0.5999379326765757, "learning_rate": 0.0002965248567846575, "loss": 0.2432, "step": 2738 }, { "epoch": 1.2889411764705883, "grad_norm": 1.2533366305013571, "learning_rate": 0.0002965154693295582, "loss": 1.0078, "step": 2739 }, { "epoch": 1.2894117647058825, "grad_norm": 0.7362883344487476, "learning_rate": 0.00029650606936131913, "loss": 0.9023, "step": 2740 }, { "epoch": 1.2898823529411765, "grad_norm": 0.5576687824401669, "learning_rate": 0.0002964966568807432, "loss": 0.1328, "step": 2741 }, { "epoch": 1.2903529411764705, "grad_norm": 1.237833694832826, "learning_rate": 0.00029648723188863417, "loss": 0.6914, "step": 2742 }, { "epoch": 1.2908235294117647, "grad_norm": 1.5033133086959474, "learning_rate": 0.000296477794385797, "loss": 1.2656, "step": 2743 }, { "epoch": 1.291294117647059, "grad_norm": 0.9186103312958876, "learning_rate": 0.00029646834437303777, "loss": 0.418, "step": 2744 }, { "epoch": 1.291764705882353, "grad_norm": 0.5841897561244083, "learning_rate": 0.00029645888185116353, "loss": 0.2969, "step": 2745 }, { "epoch": 1.292235294117647, "grad_norm": 0.8333482490188358, "learning_rate": 0.00029644940682098234, "loss": 0.4492, "step": 2746 }, { "epoch": 1.2927058823529411, "grad_norm": 1.5666585813352873, "learning_rate": 0.00029643991928330346, "loss": 1.3594, "step": 2747 }, { "epoch": 1.2931764705882354, "grad_norm": 0.7870580940162619, "learning_rate": 0.0002964304192389372, "loss": 0.8086, "step": 2748 }, { "epoch": 1.2936470588235294, "grad_norm": 1.6239875234737224, "learning_rate": 0.000296420906688695, "loss": 1.7266, "step": 2749 }, { "epoch": 1.2941176470588236, "grad_norm": 1.1483515558141313, "learning_rate": 0.00029641138163338907, "loss": 0.5352, "step": 2750 }, { "epoch": 1.2945882352941176, "grad_norm": 1.1945336009806369, "learning_rate": 0.00029640184407383303, "loss": 1.6172, "step": 2751 }, { "epoch": 1.2950588235294118, "grad_norm": 1.2971924079778465, "learning_rate": 0.00029639229401084137, "loss": 0.9766, "step": 2752 }, { "epoch": 1.2955294117647058, "grad_norm": 0.8268071518455286, "learning_rate": 0.00029638273144522983, "loss": 0.4316, "step": 2753 }, { "epoch": 1.296, "grad_norm": 0.7887502668514007, "learning_rate": 0.00029637315637781494, "loss": 0.4727, "step": 2754 }, { "epoch": 1.296470588235294, "grad_norm": 1.1469689656265893, "learning_rate": 0.0002963635688094146, "loss": 0.5391, "step": 2755 }, { "epoch": 1.2969411764705883, "grad_norm": 1.1336603834763979, "learning_rate": 0.0002963539687408476, "loss": 0.5977, "step": 2756 }, { "epoch": 1.2974117647058825, "grad_norm": 0.9240096326790862, "learning_rate": 0.00029634435617293374, "loss": 0.6406, "step": 2757 }, { "epoch": 1.2978823529411765, "grad_norm": 1.3479757831877062, "learning_rate": 0.00029633473110649407, "loss": 1.0781, "step": 2758 }, { "epoch": 1.2983529411764705, "grad_norm": 1.415524098470123, "learning_rate": 0.00029632509354235063, "loss": 1.2656, "step": 2759 }, { "epoch": 1.2988235294117647, "grad_norm": 1.6423411494758968, "learning_rate": 0.00029631544348132643, "loss": 2.1719, "step": 2760 }, { "epoch": 1.299294117647059, "grad_norm": 1.2831030305011228, "learning_rate": 0.0002963057809242457, "loss": 1.5859, "step": 2761 }, { "epoch": 1.299764705882353, "grad_norm": 1.427500076115843, "learning_rate": 0.00029629610587193365, "loss": 1.2266, "step": 2762 }, { "epoch": 1.300235294117647, "grad_norm": 1.3304159307545076, "learning_rate": 0.00029628641832521663, "loss": 1.0312, "step": 2763 }, { "epoch": 1.3007058823529412, "grad_norm": 1.3494020827851678, "learning_rate": 0.0002962767182849219, "loss": 0.7539, "step": 2764 }, { "epoch": 1.3011764705882354, "grad_norm": 1.0541166116814857, "learning_rate": 0.00029626700575187794, "loss": 0.6133, "step": 2765 }, { "epoch": 1.3016470588235294, "grad_norm": 0.29370300128126625, "learning_rate": 0.0002962572807269143, "loss": 0.0208, "step": 2766 }, { "epoch": 1.3021176470588236, "grad_norm": 1.630920858276207, "learning_rate": 0.00029624754321086146, "loss": 2.1094, "step": 2767 }, { "epoch": 1.3025882352941176, "grad_norm": 0.7596415225558895, "learning_rate": 0.00029623779320455106, "loss": 0.3535, "step": 2768 }, { "epoch": 1.3030588235294118, "grad_norm": 0.7036834956410907, "learning_rate": 0.00029622803070881586, "loss": 0.4238, "step": 2769 }, { "epoch": 1.3035294117647058, "grad_norm": 0.7242036074370336, "learning_rate": 0.00029621825572448965, "loss": 0.6016, "step": 2770 }, { "epoch": 1.304, "grad_norm": 1.0191924646235793, "learning_rate": 0.0002962084682524071, "loss": 0.6211, "step": 2771 }, { "epoch": 1.304470588235294, "grad_norm": 1.4983918600558965, "learning_rate": 0.0002961986682934043, "loss": 1.4922, "step": 2772 }, { "epoch": 1.3049411764705883, "grad_norm": 0.5816936975941951, "learning_rate": 0.0002961888558483181, "loss": 0.1836, "step": 2773 }, { "epoch": 1.3054117647058823, "grad_norm": 1.0397311718152347, "learning_rate": 0.0002961790309179866, "loss": 0.4805, "step": 2774 }, { "epoch": 1.3058823529411765, "grad_norm": 1.286821650862386, "learning_rate": 0.00029616919350324873, "loss": 1.3203, "step": 2775 }, { "epoch": 1.3063529411764705, "grad_norm": 0.8601440162048859, "learning_rate": 0.0002961593436049449, "loss": 0.5, "step": 2776 }, { "epoch": 1.3068235294117647, "grad_norm": 1.1098003576798727, "learning_rate": 0.0002961494812239162, "loss": 0.582, "step": 2777 }, { "epoch": 1.307294117647059, "grad_norm": 1.4151225289915061, "learning_rate": 0.00029613960636100487, "loss": 1.4062, "step": 2778 }, { "epoch": 1.307764705882353, "grad_norm": 0.8554369473919952, "learning_rate": 0.0002961297190170544, "loss": 0.9375, "step": 2779 }, { "epoch": 1.308235294117647, "grad_norm": 1.3556760439461903, "learning_rate": 0.00029611981919290916, "loss": 1.0703, "step": 2780 }, { "epoch": 1.3087058823529412, "grad_norm": 1.3194350667458266, "learning_rate": 0.0002961099068894146, "loss": 1.7266, "step": 2781 }, { "epoch": 1.3091764705882354, "grad_norm": 0.9159808676378232, "learning_rate": 0.00029609998210741744, "loss": 0.4609, "step": 2782 }, { "epoch": 1.3096470588235294, "grad_norm": 1.5988522194304036, "learning_rate": 0.0002960900448477651, "loss": 1.5312, "step": 2783 }, { "epoch": 1.3101176470588236, "grad_norm": 1.4145074543685692, "learning_rate": 0.0002960800951113064, "loss": 1.3516, "step": 2784 }, { "epoch": 1.3105882352941176, "grad_norm": 0.8897072765465237, "learning_rate": 0.00029607013289889106, "loss": 0.3223, "step": 2785 }, { "epoch": 1.3110588235294118, "grad_norm": 1.154079081482838, "learning_rate": 0.0002960601582113699, "loss": 0.9336, "step": 2786 }, { "epoch": 1.3115294117647058, "grad_norm": 1.272337353331597, "learning_rate": 0.00029605017104959475, "loss": 1.0859, "step": 2787 }, { "epoch": 1.312, "grad_norm": 1.0415301817955451, "learning_rate": 0.00029604017141441865, "loss": 1.1797, "step": 2788 }, { "epoch": 1.312470588235294, "grad_norm": 0.5143496778447865, "learning_rate": 0.0002960301593066956, "loss": 0.1777, "step": 2789 }, { "epoch": 1.3129411764705883, "grad_norm": 1.4646559889007684, "learning_rate": 0.0002960201347272807, "loss": 0.8398, "step": 2790 }, { "epoch": 1.3134117647058823, "grad_norm": 0.8349768645295998, "learning_rate": 0.00029601009767703007, "loss": 0.3164, "step": 2791 }, { "epoch": 1.3138823529411765, "grad_norm": 1.7175180445689073, "learning_rate": 0.0002960000481568009, "loss": 1.9375, "step": 2792 }, { "epoch": 1.3143529411764705, "grad_norm": 0.5165727589498609, "learning_rate": 0.0002959899861674515, "loss": 0.2637, "step": 2793 }, { "epoch": 1.3148235294117647, "grad_norm": 0.7763737186411553, "learning_rate": 0.0002959799117098412, "loss": 0.4688, "step": 2794 }, { "epoch": 1.315294117647059, "grad_norm": 1.1881533462297755, "learning_rate": 0.00029596982478483037, "loss": 1.2812, "step": 2795 }, { "epoch": 1.315764705882353, "grad_norm": 1.0512575336799717, "learning_rate": 0.00029595972539328053, "loss": 1.3125, "step": 2796 }, { "epoch": 1.316235294117647, "grad_norm": 1.0964434717208944, "learning_rate": 0.00029594961353605426, "loss": 0.8867, "step": 2797 }, { "epoch": 1.3167058823529412, "grad_norm": 0.7146018928836481, "learning_rate": 0.0002959394892140151, "loss": 0.3008, "step": 2798 }, { "epoch": 1.3171764705882354, "grad_norm": 1.0900044753851759, "learning_rate": 0.00029592935242802774, "loss": 1.5547, "step": 2799 }, { "epoch": 1.3176470588235294, "grad_norm": 1.12903037457856, "learning_rate": 0.0002959192031789579, "loss": 1.0469, "step": 2800 }, { "epoch": 1.3181176470588234, "grad_norm": 0.5248381172775323, "learning_rate": 0.00029590904146767236, "loss": 0.1641, "step": 2801 }, { "epoch": 1.3185882352941176, "grad_norm": 1.2968462012819373, "learning_rate": 0.000295898867295039, "loss": 1.3594, "step": 2802 }, { "epoch": 1.3190588235294118, "grad_norm": 1.1068702921722446, "learning_rate": 0.0002958886806619268, "loss": 1.375, "step": 2803 }, { "epoch": 1.3195294117647058, "grad_norm": 1.361650972415324, "learning_rate": 0.00029587848156920564, "loss": 1.2031, "step": 2804 }, { "epoch": 1.32, "grad_norm": 0.41643461238128215, "learning_rate": 0.0002958682700177466, "loss": 0.0732, "step": 2805 }, { "epoch": 1.320470588235294, "grad_norm": 1.0230651122394776, "learning_rate": 0.00029585804600842187, "loss": 1.0625, "step": 2806 }, { "epoch": 1.3209411764705883, "grad_norm": 1.3354709122355615, "learning_rate": 0.00029584780954210454, "loss": 1.4609, "step": 2807 }, { "epoch": 1.3214117647058823, "grad_norm": 1.572942796705195, "learning_rate": 0.0002958375606196689, "loss": 2.2031, "step": 2808 }, { "epoch": 1.3218823529411765, "grad_norm": 1.7362259481076496, "learning_rate": 0.0002958272992419903, "loss": 1.3672, "step": 2809 }, { "epoch": 1.3223529411764705, "grad_norm": 1.8454331403099162, "learning_rate": 0.000295817025409945, "loss": 1.2578, "step": 2810 }, { "epoch": 1.3228235294117647, "grad_norm": 0.9607738927730646, "learning_rate": 0.00029580673912441057, "loss": 0.5703, "step": 2811 }, { "epoch": 1.323294117647059, "grad_norm": 1.1799991780563932, "learning_rate": 0.00029579644038626546, "loss": 1.0156, "step": 2812 }, { "epoch": 1.323764705882353, "grad_norm": 0.7268840586347983, "learning_rate": 0.00029578612919638913, "loss": 0.3105, "step": 2813 }, { "epoch": 1.324235294117647, "grad_norm": 1.4083777509141313, "learning_rate": 0.0002957758055556623, "loss": 1.625, "step": 2814 }, { "epoch": 1.3247058823529412, "grad_norm": 1.2924933537454695, "learning_rate": 0.0002957654694649667, "loss": 1.2266, "step": 2815 }, { "epoch": 1.3251764705882354, "grad_norm": 0.9988454992108629, "learning_rate": 0.000295755120925185, "loss": 0.832, "step": 2816 }, { "epoch": 1.3256470588235294, "grad_norm": 0.949309421199435, "learning_rate": 0.0002957447599372011, "loss": 0.5508, "step": 2817 }, { "epoch": 1.3261176470588234, "grad_norm": 0.5660556347533222, "learning_rate": 0.00029573438650189974, "loss": 0.2578, "step": 2818 }, { "epoch": 1.3265882352941176, "grad_norm": 1.0020503578289996, "learning_rate": 0.000295724000620167, "loss": 0.7109, "step": 2819 }, { "epoch": 1.3270588235294118, "grad_norm": 0.8810924444307613, "learning_rate": 0.00029571360229288984, "loss": 0.457, "step": 2820 }, { "epoch": 1.3275294117647058, "grad_norm": 1.5975063520506787, "learning_rate": 0.0002957031915209563, "loss": 1.9688, "step": 2821 }, { "epoch": 1.328, "grad_norm": 2.758648467475388, "learning_rate": 0.00029569276830525553, "loss": 1.625, "step": 2822 }, { "epoch": 1.328470588235294, "grad_norm": 1.5561100176279046, "learning_rate": 0.00029568233264667775, "loss": 1.0469, "step": 2823 }, { "epoch": 1.3289411764705883, "grad_norm": 1.3375961763394628, "learning_rate": 0.00029567188454611415, "loss": 1.4766, "step": 2824 }, { "epoch": 1.3294117647058823, "grad_norm": 1.1709106886010587, "learning_rate": 0.00029566142400445716, "loss": 0.918, "step": 2825 }, { "epoch": 1.3298823529411765, "grad_norm": 1.1267138959196923, "learning_rate": 0.0002956509510226001, "loss": 1.0859, "step": 2826 }, { "epoch": 1.3303529411764705, "grad_norm": 1.2359354643997547, "learning_rate": 0.00029564046560143735, "loss": 1.3438, "step": 2827 }, { "epoch": 1.3308235294117647, "grad_norm": 1.1033212927908305, "learning_rate": 0.0002956299677418645, "loss": 0.5703, "step": 2828 }, { "epoch": 1.331294117647059, "grad_norm": 10.369398256936815, "learning_rate": 0.000295619457444778, "loss": 1.0625, "step": 2829 }, { "epoch": 1.331764705882353, "grad_norm": 1.1864520313158116, "learning_rate": 0.00029560893471107573, "loss": 0.75, "step": 2830 }, { "epoch": 1.332235294117647, "grad_norm": 2.1765375586471394, "learning_rate": 0.00029559839954165616, "loss": 2.3594, "step": 2831 }, { "epoch": 1.3327058823529412, "grad_norm": 1.2736707814242787, "learning_rate": 0.00029558785193741914, "loss": 1.6484, "step": 2832 }, { "epoch": 1.3331764705882354, "grad_norm": 1.3897494723655768, "learning_rate": 0.0002955772918992654, "loss": 1.5625, "step": 2833 }, { "epoch": 1.3336470588235294, "grad_norm": 1.282708800068383, "learning_rate": 0.00029556671942809694, "loss": 1.3828, "step": 2834 }, { "epoch": 1.3341176470588234, "grad_norm": 1.1745261247635148, "learning_rate": 0.00029555613452481657, "loss": 1.4922, "step": 2835 }, { "epoch": 1.3345882352941176, "grad_norm": 1.0415652482085125, "learning_rate": 0.00029554553719032844, "loss": 1.2891, "step": 2836 }, { "epoch": 1.3350588235294119, "grad_norm": 0.7937718467126481, "learning_rate": 0.00029553492742553754, "loss": 0.377, "step": 2837 }, { "epoch": 1.3355294117647059, "grad_norm": 1.2104853691322817, "learning_rate": 0.00029552430523134993, "loss": 1.0234, "step": 2838 }, { "epoch": 1.336, "grad_norm": 1.2659319557961528, "learning_rate": 0.0002955136706086729, "loss": 1.2109, "step": 2839 }, { "epoch": 1.336470588235294, "grad_norm": 1.046962272008402, "learning_rate": 0.0002955030235584146, "loss": 0.6836, "step": 2840 }, { "epoch": 1.3369411764705883, "grad_norm": 1.1115427319112787, "learning_rate": 0.00029549236408148453, "loss": 1.0938, "step": 2841 }, { "epoch": 1.3374117647058823, "grad_norm": 1.713157390476687, "learning_rate": 0.00029548169217879285, "loss": 1.4141, "step": 2842 }, { "epoch": 1.3378823529411765, "grad_norm": 0.7804499510737839, "learning_rate": 0.000295471007851251, "loss": 0.2852, "step": 2843 }, { "epoch": 1.3383529411764705, "grad_norm": 1.1681999798460825, "learning_rate": 0.00029546031109977164, "loss": 1.1328, "step": 2844 }, { "epoch": 1.3388235294117647, "grad_norm": 0.6023387872063168, "learning_rate": 0.00029544960192526824, "loss": 0.2246, "step": 2845 }, { "epoch": 1.3392941176470587, "grad_norm": 0.878083423977145, "learning_rate": 0.0002954388803286554, "loss": 0.4531, "step": 2846 }, { "epoch": 1.339764705882353, "grad_norm": 1.2201560238892142, "learning_rate": 0.0002954281463108488, "loss": 1.1484, "step": 2847 }, { "epoch": 1.340235294117647, "grad_norm": 0.7663250311891185, "learning_rate": 0.0002954173998727652, "loss": 0.4863, "step": 2848 }, { "epoch": 1.3407058823529412, "grad_norm": 1.2110941260602157, "learning_rate": 0.00029540664101532227, "loss": 1.4766, "step": 2849 }, { "epoch": 1.3411764705882354, "grad_norm": 2.821837034679463, "learning_rate": 0.0002953958697394391, "loss": 2.4688, "step": 2850 }, { "epoch": 1.3416470588235294, "grad_norm": 0.877516123842763, "learning_rate": 0.0002953850860460355, "loss": 0.375, "step": 2851 }, { "epoch": 1.3421176470588234, "grad_norm": 1.2126993511553124, "learning_rate": 0.00029537428993603237, "loss": 1.0469, "step": 2852 }, { "epoch": 1.3425882352941176, "grad_norm": 1.4582532739144944, "learning_rate": 0.00029536348141035183, "loss": 1.0234, "step": 2853 }, { "epoch": 1.3430588235294119, "grad_norm": 1.3314107767330925, "learning_rate": 0.000295352660469917, "loss": 0.9531, "step": 2854 }, { "epoch": 1.3435294117647059, "grad_norm": 1.3315684359137179, "learning_rate": 0.00029534182711565203, "loss": 1.4297, "step": 2855 }, { "epoch": 1.3439999999999999, "grad_norm": 1.5585961389428524, "learning_rate": 0.0002953309813484821, "loss": 1.1875, "step": 2856 }, { "epoch": 1.344470588235294, "grad_norm": 2.9499711397962596, "learning_rate": 0.0002953201231693336, "loss": 2.5156, "step": 2857 }, { "epoch": 1.3449411764705883, "grad_norm": 0.8050987719734742, "learning_rate": 0.00029530925257913366, "loss": 0.4473, "step": 2858 }, { "epoch": 1.3454117647058823, "grad_norm": 1.5961666904770735, "learning_rate": 0.0002952983695788109, "loss": 0.9805, "step": 2859 }, { "epoch": 1.3458823529411765, "grad_norm": 1.1776848715917776, "learning_rate": 0.00029528747416929463, "loss": 0.9922, "step": 2860 }, { "epoch": 1.3463529411764705, "grad_norm": 1.6236519321203262, "learning_rate": 0.0002952765663515155, "loss": 1.3438, "step": 2861 }, { "epoch": 1.3468235294117648, "grad_norm": 1.4662639121416405, "learning_rate": 0.000295265646126405, "loss": 1.6406, "step": 2862 }, { "epoch": 1.3472941176470588, "grad_norm": 0.6832619484228131, "learning_rate": 0.00029525471349489576, "loss": 0.3027, "step": 2863 }, { "epoch": 1.347764705882353, "grad_norm": 1.2639397842987594, "learning_rate": 0.00029524376845792153, "loss": 1.2109, "step": 2864 }, { "epoch": 1.348235294117647, "grad_norm": 0.9404924704044626, "learning_rate": 0.0002952328110164171, "loss": 0.8477, "step": 2865 }, { "epoch": 1.3487058823529412, "grad_norm": 1.3959411383682372, "learning_rate": 0.0002952218411713182, "loss": 1.2969, "step": 2866 }, { "epoch": 1.3491764705882354, "grad_norm": 1.5717550919789203, "learning_rate": 0.0002952108589235618, "loss": 1.6797, "step": 2867 }, { "epoch": 1.3496470588235294, "grad_norm": 1.0725258958876707, "learning_rate": 0.0002951998642740858, "loss": 1.0781, "step": 2868 }, { "epoch": 1.3501176470588234, "grad_norm": 0.6173707756766968, "learning_rate": 0.00029518885722382913, "loss": 0.375, "step": 2869 }, { "epoch": 1.3505882352941176, "grad_norm": 1.8015347864621523, "learning_rate": 0.00029517783777373194, "loss": 2.5625, "step": 2870 }, { "epoch": 1.3510588235294119, "grad_norm": 1.4182965179528877, "learning_rate": 0.0002951668059247353, "loss": 1.1562, "step": 2871 }, { "epoch": 1.3515294117647059, "grad_norm": 1.340356907965804, "learning_rate": 0.0002951557616777814, "loss": 1.7422, "step": 2872 }, { "epoch": 1.3519999999999999, "grad_norm": 0.44067609771767174, "learning_rate": 0.0002951447050338135, "loss": 0.2051, "step": 2873 }, { "epoch": 1.352470588235294, "grad_norm": 1.5099883755224501, "learning_rate": 0.00029513363599377587, "loss": 0.9609, "step": 2874 }, { "epoch": 1.3529411764705883, "grad_norm": 1.1620548785551859, "learning_rate": 0.00029512255455861375, "loss": 1.0859, "step": 2875 }, { "epoch": 1.3534117647058823, "grad_norm": 1.0881474177791535, "learning_rate": 0.0002951114607292737, "loss": 0.9102, "step": 2876 }, { "epoch": 1.3538823529411765, "grad_norm": 1.3943062261777, "learning_rate": 0.0002951003545067032, "loss": 1.9688, "step": 2877 }, { "epoch": 1.3543529411764705, "grad_norm": 1.7520636681513533, "learning_rate": 0.0002950892358918506, "loss": 1.5859, "step": 2878 }, { "epoch": 1.3548235294117648, "grad_norm": 2.007602400297442, "learning_rate": 0.00029507810488566563, "loss": 2.5938, "step": 2879 }, { "epoch": 1.3552941176470588, "grad_norm": 1.1875616676552496, "learning_rate": 0.000295066961489099, "loss": 0.8086, "step": 2880 }, { "epoch": 1.355764705882353, "grad_norm": 1.1204384436980828, "learning_rate": 0.0002950558057031022, "loss": 1.5703, "step": 2881 }, { "epoch": 1.356235294117647, "grad_norm": 0.35207976565765847, "learning_rate": 0.0002950446375286281, "loss": 0.051, "step": 2882 }, { "epoch": 1.3567058823529412, "grad_norm": 0.9171623483284264, "learning_rate": 0.00029503345696663054, "loss": 0.3848, "step": 2883 }, { "epoch": 1.3571764705882354, "grad_norm": 1.0860346185151633, "learning_rate": 0.0002950222640180644, "loss": 0.7617, "step": 2884 }, { "epoch": 1.3576470588235294, "grad_norm": 0.6741863727938789, "learning_rate": 0.0002950110586838855, "loss": 0.2432, "step": 2885 }, { "epoch": 1.3581176470588234, "grad_norm": 0.7333037600702016, "learning_rate": 0.00029499984096505107, "loss": 0.3555, "step": 2886 }, { "epoch": 1.3585882352941177, "grad_norm": 1.4323412700350484, "learning_rate": 0.00029498861086251887, "loss": 1.9141, "step": 2887 }, { "epoch": 1.3590588235294119, "grad_norm": 1.7635886378178562, "learning_rate": 0.00029497736837724817, "loss": 1.125, "step": 2888 }, { "epoch": 1.3595294117647059, "grad_norm": 1.1428573827346935, "learning_rate": 0.0002949661135101991, "loss": 0.7617, "step": 2889 }, { "epoch": 1.3599999999999999, "grad_norm": 0.8486871728372263, "learning_rate": 0.0002949548462623329, "loss": 0.418, "step": 2890 }, { "epoch": 1.360470588235294, "grad_norm": 0.8776196361892445, "learning_rate": 0.0002949435666346118, "loss": 0.6562, "step": 2891 }, { "epoch": 1.3609411764705883, "grad_norm": 1.2429745244941097, "learning_rate": 0.00029493227462799914, "loss": 0.668, "step": 2892 }, { "epoch": 1.3614117647058823, "grad_norm": 1.5712364040990563, "learning_rate": 0.00029492097024345936, "loss": 1.125, "step": 2893 }, { "epoch": 1.3618823529411765, "grad_norm": 1.262652661703025, "learning_rate": 0.00029490965348195794, "loss": 1.3672, "step": 2894 }, { "epoch": 1.3623529411764705, "grad_norm": 0.7669204952767142, "learning_rate": 0.00029489832434446126, "loss": 0.3457, "step": 2895 }, { "epoch": 1.3628235294117648, "grad_norm": 0.8473698654598484, "learning_rate": 0.000294886982831937, "loss": 1.0469, "step": 2896 }, { "epoch": 1.3632941176470588, "grad_norm": 0.4938891513935479, "learning_rate": 0.00029487562894535375, "loss": 0.1758, "step": 2897 }, { "epoch": 1.363764705882353, "grad_norm": 1.4709657305181534, "learning_rate": 0.0002948642626856812, "loss": 1.7031, "step": 2898 }, { "epoch": 1.364235294117647, "grad_norm": 0.9433979050886642, "learning_rate": 0.00029485288405389, "loss": 0.5859, "step": 2899 }, { "epoch": 1.3647058823529412, "grad_norm": 1.1660301398485573, "learning_rate": 0.000294841493050952, "loss": 1.5781, "step": 2900 }, { "epoch": 1.3651764705882352, "grad_norm": 0.9538580668541065, "learning_rate": 0.00029483008967784007, "loss": 0.5977, "step": 2901 }, { "epoch": 1.3656470588235294, "grad_norm": 0.9116448958211111, "learning_rate": 0.0002948186739355281, "loss": 0.4043, "step": 2902 }, { "epoch": 1.3661176470588234, "grad_norm": 1.2950168960298365, "learning_rate": 0.00029480724582499107, "loss": 0.7617, "step": 2903 }, { "epoch": 1.3665882352941177, "grad_norm": 0.9022043332744849, "learning_rate": 0.0002947958053472049, "loss": 0.3965, "step": 2904 }, { "epoch": 1.3670588235294119, "grad_norm": 1.4902037753048452, "learning_rate": 0.0002947843525031468, "loss": 1.2422, "step": 2905 }, { "epoch": 1.3675294117647059, "grad_norm": 0.9872452838134841, "learning_rate": 0.0002947728872937948, "loss": 0.8125, "step": 2906 }, { "epoch": 1.3679999999999999, "grad_norm": 1.2895406494084998, "learning_rate": 0.00029476140972012815, "loss": 1.2969, "step": 2907 }, { "epoch": 1.368470588235294, "grad_norm": 1.3262764165467258, "learning_rate": 0.000294749919783127, "loss": 1.3438, "step": 2908 }, { "epoch": 1.3689411764705883, "grad_norm": 1.325798966332734, "learning_rate": 0.00029473841748377277, "loss": 1.5859, "step": 2909 }, { "epoch": 1.3694117647058823, "grad_norm": 0.9759981371157899, "learning_rate": 0.0002947269028230477, "loss": 0.8398, "step": 2910 }, { "epoch": 1.3698823529411766, "grad_norm": 1.1012470867396964, "learning_rate": 0.0002947153758019353, "loss": 0.9258, "step": 2911 }, { "epoch": 1.3703529411764706, "grad_norm": 0.8513110717810415, "learning_rate": 0.00029470383642142, "loss": 0.5156, "step": 2912 }, { "epoch": 1.3708235294117648, "grad_norm": 1.2185482261546463, "learning_rate": 0.0002946922846824872, "loss": 0.6836, "step": 2913 }, { "epoch": 1.3712941176470588, "grad_norm": 0.7504641174226755, "learning_rate": 0.00029468072058612367, "loss": 0.7148, "step": 2914 }, { "epoch": 1.371764705882353, "grad_norm": 0.8227028247252777, "learning_rate": 0.0002946691441333169, "loss": 0.5586, "step": 2915 }, { "epoch": 1.372235294117647, "grad_norm": 1.0344090226948561, "learning_rate": 0.0002946575553250556, "loss": 1.0703, "step": 2916 }, { "epoch": 1.3727058823529412, "grad_norm": 1.5582644363222395, "learning_rate": 0.00029464595416232956, "loss": 1.0781, "step": 2917 }, { "epoch": 1.3731764705882352, "grad_norm": 0.637199600544644, "learning_rate": 0.0002946343406461296, "loss": 0.1992, "step": 2918 }, { "epoch": 1.3736470588235294, "grad_norm": 1.016392286054045, "learning_rate": 0.0002946227147774475, "loss": 0.6055, "step": 2919 }, { "epoch": 1.3741176470588234, "grad_norm": 0.822408989687839, "learning_rate": 0.0002946110765572761, "loss": 0.5781, "step": 2920 }, { "epoch": 1.3745882352941177, "grad_norm": 0.63993834455035, "learning_rate": 0.00029459942598660953, "loss": 0.4277, "step": 2921 }, { "epoch": 1.3750588235294119, "grad_norm": 0.8403090651323941, "learning_rate": 0.00029458776306644273, "loss": 0.6289, "step": 2922 }, { "epoch": 1.375529411764706, "grad_norm": 1.6788380257800966, "learning_rate": 0.0002945760877977717, "loss": 2.4062, "step": 2923 }, { "epoch": 1.376, "grad_norm": 0.8292614491502913, "learning_rate": 0.0002945644001815937, "loss": 0.5898, "step": 2924 }, { "epoch": 1.3764705882352941, "grad_norm": 1.730421611576664, "learning_rate": 0.0002945527002189068, "loss": 1.0156, "step": 2925 }, { "epoch": 1.3769411764705883, "grad_norm": 0.7293473536438779, "learning_rate": 0.0002945409879107103, "loss": 0.3496, "step": 2926 }, { "epoch": 1.3774117647058823, "grad_norm": 1.1422315030671313, "learning_rate": 0.00029452926325800445, "loss": 1.2422, "step": 2927 }, { "epoch": 1.3778823529411763, "grad_norm": 1.2380718305156817, "learning_rate": 0.0002945175262617906, "loss": 1.2188, "step": 2928 }, { "epoch": 1.3783529411764706, "grad_norm": 1.2398283252080824, "learning_rate": 0.00029450577692307115, "loss": 1.2812, "step": 2929 }, { "epoch": 1.3788235294117648, "grad_norm": 1.19183753323321, "learning_rate": 0.0002944940152428496, "loss": 1.7656, "step": 2930 }, { "epoch": 1.3792941176470588, "grad_norm": 1.0083313956459494, "learning_rate": 0.0002944822412221303, "loss": 0.6094, "step": 2931 }, { "epoch": 1.379764705882353, "grad_norm": 1.8449648102840444, "learning_rate": 0.00029447045486191904, "loss": 2.1094, "step": 2932 }, { "epoch": 1.380235294117647, "grad_norm": 1.16877660360978, "learning_rate": 0.0002944586561632223, "loss": 1.5859, "step": 2933 }, { "epoch": 1.3807058823529412, "grad_norm": 1.5144573584792445, "learning_rate": 0.0002944468451270477, "loss": 1.7812, "step": 2934 }, { "epoch": 1.3811764705882352, "grad_norm": 0.8809659466109178, "learning_rate": 0.000294435021754404, "loss": 0.4434, "step": 2935 }, { "epoch": 1.3816470588235295, "grad_norm": 1.3217264813095773, "learning_rate": 0.000294423186046301, "loss": 0.9609, "step": 2936 }, { "epoch": 1.3821176470588235, "grad_norm": 1.240648515603188, "learning_rate": 0.00029441133800374957, "loss": 1.3828, "step": 2937 }, { "epoch": 1.3825882352941177, "grad_norm": 1.0933158514708634, "learning_rate": 0.00029439947762776154, "loss": 0.7734, "step": 2938 }, { "epoch": 1.383058823529412, "grad_norm": 0.9343644149744106, "learning_rate": 0.00029438760491934987, "loss": 0.5195, "step": 2939 }, { "epoch": 1.383529411764706, "grad_norm": 1.280165810576224, "learning_rate": 0.0002943757198795284, "loss": 1.6172, "step": 2940 }, { "epoch": 1.384, "grad_norm": 1.2574090946967564, "learning_rate": 0.0002943638225093124, "loss": 1.5703, "step": 2941 }, { "epoch": 1.3844705882352941, "grad_norm": 1.0527095436230025, "learning_rate": 0.00029435191280971783, "loss": 0.6797, "step": 2942 }, { "epoch": 1.3849411764705883, "grad_norm": 1.2918270952988078, "learning_rate": 0.0002943399907817619, "loss": 1.4219, "step": 2943 }, { "epoch": 1.3854117647058823, "grad_norm": 0.6326294012560753, "learning_rate": 0.0002943280564264627, "loss": 0.2832, "step": 2944 }, { "epoch": 1.3858823529411763, "grad_norm": 1.2570869915642575, "learning_rate": 0.0002943161097448396, "loss": 1.4219, "step": 2945 }, { "epoch": 1.3863529411764706, "grad_norm": 0.931630554360452, "learning_rate": 0.0002943041507379128, "loss": 0.5508, "step": 2946 }, { "epoch": 1.3868235294117648, "grad_norm": 1.4093838821917037, "learning_rate": 0.00029429217940670383, "loss": 1.3984, "step": 2947 }, { "epoch": 1.3872941176470588, "grad_norm": 1.1690091971037344, "learning_rate": 0.0002942801957522349, "loss": 1.3281, "step": 2948 }, { "epoch": 1.387764705882353, "grad_norm": 0.9783761984116705, "learning_rate": 0.00029426819977552965, "loss": 0.8633, "step": 2949 }, { "epoch": 1.388235294117647, "grad_norm": 1.1464982324911848, "learning_rate": 0.0002942561914776124, "loss": 1.375, "step": 2950 }, { "epoch": 1.3887058823529412, "grad_norm": 0.6497252220030448, "learning_rate": 0.0002942441708595089, "loss": 0.1992, "step": 2951 }, { "epoch": 1.3891764705882352, "grad_norm": 0.8954419693676565, "learning_rate": 0.0002942321379222457, "loss": 0.4219, "step": 2952 }, { "epoch": 1.3896470588235295, "grad_norm": 1.2101833373511317, "learning_rate": 0.0002942200926668504, "loss": 0.9727, "step": 2953 }, { "epoch": 1.3901176470588235, "grad_norm": 1.3636734782449902, "learning_rate": 0.00029420803509435186, "loss": 1.625, "step": 2954 }, { "epoch": 1.3905882352941177, "grad_norm": 1.7252673330224442, "learning_rate": 0.00029419596520577975, "loss": 1.7578, "step": 2955 }, { "epoch": 1.391058823529412, "grad_norm": 1.4098777174346693, "learning_rate": 0.00029418388300216493, "loss": 1.2109, "step": 2956 }, { "epoch": 1.391529411764706, "grad_norm": 0.37043840114988846, "learning_rate": 0.0002941717884845393, "loss": 0.083, "step": 2957 }, { "epoch": 1.392, "grad_norm": 1.1498316425525197, "learning_rate": 0.00029415968165393577, "loss": 1.6797, "step": 2958 }, { "epoch": 1.3924705882352941, "grad_norm": 1.6576741650719287, "learning_rate": 0.00029414756251138835, "loss": 1.9375, "step": 2959 }, { "epoch": 1.3929411764705883, "grad_norm": 0.7678152820073206, "learning_rate": 0.000294135431057932, "loss": 0.3379, "step": 2960 }, { "epoch": 1.3934117647058823, "grad_norm": 1.1618492864601706, "learning_rate": 0.0002941232872946029, "loss": 1.3438, "step": 2961 }, { "epoch": 1.3938823529411764, "grad_norm": 1.5346268502318912, "learning_rate": 0.00029411113122243814, "loss": 1.1016, "step": 2962 }, { "epoch": 1.3943529411764706, "grad_norm": 0.6364840572941416, "learning_rate": 0.0002940989628424759, "loss": 0.1631, "step": 2963 }, { "epoch": 1.3948235294117648, "grad_norm": 1.2941177332595564, "learning_rate": 0.0002940867821557554, "loss": 0.8984, "step": 2964 }, { "epoch": 1.3952941176470588, "grad_norm": 1.632136447603144, "learning_rate": 0.00029407458916331705, "loss": 1.7969, "step": 2965 }, { "epoch": 1.395764705882353, "grad_norm": 1.199471245565124, "learning_rate": 0.00029406238386620205, "loss": 1.4688, "step": 2966 }, { "epoch": 1.396235294117647, "grad_norm": 0.8590516086351798, "learning_rate": 0.0002940501662654528, "loss": 0.3691, "step": 2967 }, { "epoch": 1.3967058823529412, "grad_norm": 0.985929557678237, "learning_rate": 0.0002940379363621129, "loss": 0.4922, "step": 2968 }, { "epoch": 1.3971764705882352, "grad_norm": 1.354367090287961, "learning_rate": 0.00029402569415722666, "loss": 1.1172, "step": 2969 }, { "epoch": 1.3976470588235295, "grad_norm": 0.9495741257527245, "learning_rate": 0.00029401343965183965, "loss": 1.0, "step": 2970 }, { "epoch": 1.3981176470588235, "grad_norm": 1.587206566584936, "learning_rate": 0.0002940011728469986, "loss": 1.9844, "step": 2971 }, { "epoch": 1.3985882352941177, "grad_norm": 1.416158705688493, "learning_rate": 0.000293988893743751, "loss": 1.3438, "step": 2972 }, { "epoch": 1.3990588235294117, "grad_norm": 1.1799724402168377, "learning_rate": 0.0002939766023431457, "loss": 0.7266, "step": 2973 }, { "epoch": 1.399529411764706, "grad_norm": 1.1275510278752572, "learning_rate": 0.0002939642986462323, "loss": 0.4961, "step": 2974 }, { "epoch": 1.4, "grad_norm": 2.4781761040679364, "learning_rate": 0.0002939519826540616, "loss": 2.4219, "step": 2975 }, { "epoch": 1.4004705882352941, "grad_norm": 1.2021692406694717, "learning_rate": 0.0002939396543676856, "loss": 1.4062, "step": 2976 }, { "epoch": 1.4009411764705884, "grad_norm": 1.1438415691818098, "learning_rate": 0.00029392731378815706, "loss": 0.6484, "step": 2977 }, { "epoch": 1.4014117647058824, "grad_norm": 1.7556889524719055, "learning_rate": 0.00029391496091652993, "loss": 2.0, "step": 2978 }, { "epoch": 1.4018823529411764, "grad_norm": 0.8855891062168082, "learning_rate": 0.0002939025957538593, "loss": 0.75, "step": 2979 }, { "epoch": 1.4023529411764706, "grad_norm": 1.3445767884943252, "learning_rate": 0.0002938902183012011, "loss": 1.2422, "step": 2980 }, { "epoch": 1.4028235294117648, "grad_norm": 1.3684579420879233, "learning_rate": 0.00029387782855961253, "loss": 0.4766, "step": 2981 }, { "epoch": 1.4032941176470588, "grad_norm": 1.1742368213621102, "learning_rate": 0.00029386542653015167, "loss": 0.6992, "step": 2982 }, { "epoch": 1.4037647058823528, "grad_norm": 2.210138415326263, "learning_rate": 0.0002938530122138777, "loss": 1.2109, "step": 2983 }, { "epoch": 1.404235294117647, "grad_norm": 1.5181463484449997, "learning_rate": 0.0002938405856118509, "loss": 1.1719, "step": 2984 }, { "epoch": 1.4047058823529412, "grad_norm": 0.6100512889373992, "learning_rate": 0.0002938281467251326, "loss": 0.3066, "step": 2985 }, { "epoch": 1.4051764705882352, "grad_norm": 0.973900418916263, "learning_rate": 0.00029381569555478505, "loss": 0.8789, "step": 2986 }, { "epoch": 1.4056470588235295, "grad_norm": 1.0188698994333942, "learning_rate": 0.00029380323210187166, "loss": 0.498, "step": 2987 }, { "epoch": 1.4061176470588235, "grad_norm": 0.702762305759476, "learning_rate": 0.00029379075636745694, "loss": 0.4395, "step": 2988 }, { "epoch": 1.4065882352941177, "grad_norm": 1.867384475737045, "learning_rate": 0.0002937782683526063, "loss": 1.7812, "step": 2989 }, { "epoch": 1.4070588235294117, "grad_norm": 0.8689388808477385, "learning_rate": 0.0002937657680583864, "loss": 0.3496, "step": 2990 }, { "epoch": 1.407529411764706, "grad_norm": 0.8914726459737887, "learning_rate": 0.0002937532554858647, "loss": 0.4531, "step": 2991 }, { "epoch": 1.408, "grad_norm": 1.3318582974490778, "learning_rate": 0.0002937407306361099, "loss": 1.6172, "step": 2992 }, { "epoch": 1.4084705882352941, "grad_norm": 1.459407297281059, "learning_rate": 0.0002937281935101916, "loss": 1.2031, "step": 2993 }, { "epoch": 1.4089411764705884, "grad_norm": 1.106929842766656, "learning_rate": 0.0002937156441091806, "loss": 1.1328, "step": 2994 }, { "epoch": 1.4094117647058824, "grad_norm": 0.9199245592516723, "learning_rate": 0.0002937030824341487, "loss": 0.5352, "step": 2995 }, { "epoch": 1.4098823529411764, "grad_norm": 1.1551244943287746, "learning_rate": 0.00029369050848616863, "loss": 0.957, "step": 2996 }, { "epoch": 1.4103529411764706, "grad_norm": 1.428870893266046, "learning_rate": 0.0002936779222663144, "loss": 1.4062, "step": 2997 }, { "epoch": 1.4108235294117648, "grad_norm": 1.7560946127357244, "learning_rate": 0.0002936653237756609, "loss": 1.7969, "step": 2998 }, { "epoch": 1.4112941176470588, "grad_norm": 0.45727452684398884, "learning_rate": 0.00029365271301528406, "loss": 0.1973, "step": 2999 }, { "epoch": 1.4117647058823528, "grad_norm": 0.6767591589432318, "learning_rate": 0.00029364008998626086, "loss": 0.4805, "step": 3000 }, { "epoch": 1.412235294117647, "grad_norm": 1.5737042305598905, "learning_rate": 0.00029362745468966946, "loss": 1.4141, "step": 3001 }, { "epoch": 1.4127058823529413, "grad_norm": 1.1684117712383095, "learning_rate": 0.0002936148071265889, "loss": 0.6953, "step": 3002 }, { "epoch": 1.4131764705882353, "grad_norm": 1.859191940143639, "learning_rate": 0.00029360214729809946, "loss": 1.9219, "step": 3003 }, { "epoch": 1.4136470588235295, "grad_norm": 1.3334452089450164, "learning_rate": 0.00029358947520528227, "loss": 1.2891, "step": 3004 }, { "epoch": 1.4141176470588235, "grad_norm": 1.0140542322892214, "learning_rate": 0.0002935767908492196, "loss": 0.7148, "step": 3005 }, { "epoch": 1.4145882352941177, "grad_norm": 1.2136067650115059, "learning_rate": 0.0002935640942309947, "loss": 0.7188, "step": 3006 }, { "epoch": 1.4150588235294117, "grad_norm": 1.7352850892503497, "learning_rate": 0.000293551385351692, "loss": 1.4297, "step": 3007 }, { "epoch": 1.415529411764706, "grad_norm": 1.999313381844128, "learning_rate": 0.0002935386642123968, "loss": 1.9844, "step": 3008 }, { "epoch": 1.416, "grad_norm": 1.2153639474987674, "learning_rate": 0.00029352593081419573, "loss": 1.0469, "step": 3009 }, { "epoch": 1.4164705882352941, "grad_norm": 1.0059615621435418, "learning_rate": 0.0002935131851581762, "loss": 1.2031, "step": 3010 }, { "epoch": 1.4169411764705884, "grad_norm": 1.525018589117408, "learning_rate": 0.00029350042724542657, "loss": 1.0781, "step": 3011 }, { "epoch": 1.4174117647058824, "grad_norm": 35.53192830970034, "learning_rate": 0.0002934876570770367, "loss": 2.0312, "step": 3012 }, { "epoch": 1.4178823529411764, "grad_norm": 0.9856144615303691, "learning_rate": 0.00029347487465409714, "loss": 0.7617, "step": 3013 }, { "epoch": 1.4183529411764706, "grad_norm": 1.0619558881900841, "learning_rate": 0.0002934620799776995, "loss": 0.543, "step": 3014 }, { "epoch": 1.4188235294117648, "grad_norm": 1.377172294423313, "learning_rate": 0.00029344927304893653, "loss": 0.7148, "step": 3015 }, { "epoch": 1.4192941176470588, "grad_norm": 22.845064382452534, "learning_rate": 0.00029343645386890204, "loss": 1.0078, "step": 3016 }, { "epoch": 1.4197647058823528, "grad_norm": 15.412197465235069, "learning_rate": 0.00029342362243869093, "loss": 0.6875, "step": 3017 }, { "epoch": 1.420235294117647, "grad_norm": 19.970912373359862, "learning_rate": 0.00029341077875939884, "loss": 1.2344, "step": 3018 }, { "epoch": 1.4207058823529413, "grad_norm": 1.3850517997979215, "learning_rate": 0.0002933979228321229, "loss": 1.1719, "step": 3019 }, { "epoch": 1.4211764705882353, "grad_norm": 2.5660226628798477, "learning_rate": 0.00029338505465796095, "loss": 1.5469, "step": 3020 }, { "epoch": 1.4216470588235295, "grad_norm": 5.235428867621642, "learning_rate": 0.00029337217423801206, "loss": 1.1719, "step": 3021 }, { "epoch": 1.4221176470588235, "grad_norm": 3.0152671284981443, "learning_rate": 0.00029335928157337625, "loss": 0.7891, "step": 3022 }, { "epoch": 1.4225882352941177, "grad_norm": 1.4384242078890936, "learning_rate": 0.00029334637666515457, "loss": 1.0859, "step": 3023 }, { "epoch": 1.4230588235294117, "grad_norm": 1.0383296033350389, "learning_rate": 0.00029333345951444923, "loss": 0.5625, "step": 3024 }, { "epoch": 1.423529411764706, "grad_norm": 1.445996433002375, "learning_rate": 0.00029332053012236345, "loss": 1.5703, "step": 3025 }, { "epoch": 1.424, "grad_norm": 1.5321536010371086, "learning_rate": 0.0002933075884900014, "loss": 1.3984, "step": 3026 }, { "epoch": 1.4244705882352942, "grad_norm": 1.6336293815645124, "learning_rate": 0.00029329463461846833, "loss": 1.5156, "step": 3027 }, { "epoch": 1.4249411764705882, "grad_norm": 1.3240859135745013, "learning_rate": 0.00029328166850887066, "loss": 1.3359, "step": 3028 }, { "epoch": 1.4254117647058824, "grad_norm": 0.7289922986676448, "learning_rate": 0.00029326869016231566, "loss": 0.2812, "step": 3029 }, { "epoch": 1.4258823529411764, "grad_norm": 1.4132481016299678, "learning_rate": 0.0002932556995799118, "loss": 1.1797, "step": 3030 }, { "epoch": 1.4263529411764706, "grad_norm": 1.3128436199149227, "learning_rate": 0.00029324269676276856, "loss": 1.1328, "step": 3031 }, { "epoch": 1.4268235294117648, "grad_norm": 0.8908548875535282, "learning_rate": 0.0002932296817119964, "loss": 0.5195, "step": 3032 }, { "epoch": 1.4272941176470588, "grad_norm": 1.7156064866416028, "learning_rate": 0.0002932166544287069, "loss": 1.875, "step": 3033 }, { "epoch": 1.4277647058823528, "grad_norm": 1.383748257928257, "learning_rate": 0.00029320361491401263, "loss": 1.6016, "step": 3034 }, { "epoch": 1.428235294117647, "grad_norm": 1.3234121386971225, "learning_rate": 0.00029319056316902723, "loss": 1.3203, "step": 3035 }, { "epoch": 1.4287058823529413, "grad_norm": 1.0549915366697369, "learning_rate": 0.0002931774991948654, "loss": 1.375, "step": 3036 }, { "epoch": 1.4291764705882353, "grad_norm": 0.8653558868038028, "learning_rate": 0.0002931644229926428, "loss": 0.7578, "step": 3037 }, { "epoch": 1.4296470588235295, "grad_norm": 1.2490335680149054, "learning_rate": 0.00029315133456347633, "loss": 1.4766, "step": 3038 }, { "epoch": 1.4301176470588235, "grad_norm": 1.2612767588649243, "learning_rate": 0.0002931382339084837, "loss": 1.5859, "step": 3039 }, { "epoch": 1.4305882352941177, "grad_norm": 0.7648344495197736, "learning_rate": 0.0002931251210287838, "loss": 1.0078, "step": 3040 }, { "epoch": 1.4310588235294117, "grad_norm": 1.3886402071088555, "learning_rate": 0.00029311199592549656, "loss": 1.3359, "step": 3041 }, { "epoch": 1.431529411764706, "grad_norm": 1.2381812676636033, "learning_rate": 0.00029309885859974294, "loss": 1.2109, "step": 3042 }, { "epoch": 1.432, "grad_norm": 0.7076861815360216, "learning_rate": 0.00029308570905264484, "loss": 0.249, "step": 3043 }, { "epoch": 1.4324705882352942, "grad_norm": 1.5560681033066095, "learning_rate": 0.00029307254728532536, "loss": 1.0625, "step": 3044 }, { "epoch": 1.4329411764705882, "grad_norm": 1.1946420718323703, "learning_rate": 0.00029305937329890863, "loss": 0.8711, "step": 3045 }, { "epoch": 1.4334117647058824, "grad_norm": 1.122768897982592, "learning_rate": 0.00029304618709451966, "loss": 0.8125, "step": 3046 }, { "epoch": 1.4338823529411764, "grad_norm": 1.1318644058756668, "learning_rate": 0.0002930329886732847, "loss": 0.7891, "step": 3047 }, { "epoch": 1.4343529411764706, "grad_norm": 1.5518242224687349, "learning_rate": 0.00029301977803633094, "loss": 0.3379, "step": 3048 }, { "epoch": 1.4348235294117648, "grad_norm": 0.9231090523721306, "learning_rate": 0.0002930065551847866, "loss": 0.6055, "step": 3049 }, { "epoch": 1.4352941176470588, "grad_norm": 1.2735521769593097, "learning_rate": 0.00029299332011978107, "loss": 1.2969, "step": 3050 }, { "epoch": 1.4357647058823528, "grad_norm": 1.3588617876418776, "learning_rate": 0.00029298007284244456, "loss": 1.1328, "step": 3051 }, { "epoch": 1.436235294117647, "grad_norm": 0.9845987354332884, "learning_rate": 0.0002929668133539085, "loss": 0.9453, "step": 3052 }, { "epoch": 1.4367058823529413, "grad_norm": 0.9888805104998835, "learning_rate": 0.0002929535416553054, "loss": 0.9531, "step": 3053 }, { "epoch": 1.4371764705882353, "grad_norm": 1.4335780871039996, "learning_rate": 0.0002929402577477686, "loss": 1.4297, "step": 3054 }, { "epoch": 1.4376470588235293, "grad_norm": 0.8762401747890041, "learning_rate": 0.0002929269616324327, "loss": 0.875, "step": 3055 }, { "epoch": 1.4381176470588235, "grad_norm": 1.0461922940861492, "learning_rate": 0.0002929136533104332, "loss": 0.8281, "step": 3056 }, { "epoch": 1.4385882352941177, "grad_norm": 1.2543505754011655, "learning_rate": 0.0002929003327829067, "loss": 1.5078, "step": 3057 }, { "epoch": 1.4390588235294117, "grad_norm": 1.4122300279550886, "learning_rate": 0.0002928870000509909, "loss": 0.918, "step": 3058 }, { "epoch": 1.439529411764706, "grad_norm": 0.44265401163197166, "learning_rate": 0.0002928736551158244, "loss": 0.3965, "step": 3059 }, { "epoch": 1.44, "grad_norm": 1.401501593231668, "learning_rate": 0.00029286029797854704, "loss": 1.4766, "step": 3060 }, { "epoch": 1.4404705882352942, "grad_norm": 0.8474149702158882, "learning_rate": 0.00029284692864029946, "loss": 0.3438, "step": 3061 }, { "epoch": 1.4409411764705882, "grad_norm": 0.9458534174012263, "learning_rate": 0.0002928335471022235, "loss": 0.6211, "step": 3062 }, { "epoch": 1.4414117647058824, "grad_norm": 1.0063818014619383, "learning_rate": 0.00029282015336546205, "loss": 1.2344, "step": 3063 }, { "epoch": 1.4418823529411764, "grad_norm": 1.1205069791079396, "learning_rate": 0.00029280674743115896, "loss": 0.7109, "step": 3064 }, { "epoch": 1.4423529411764706, "grad_norm": 0.9140496489396549, "learning_rate": 0.0002927933293004592, "loss": 0.4492, "step": 3065 }, { "epoch": 1.4428235294117648, "grad_norm": 1.3101371501513364, "learning_rate": 0.00029277989897450867, "loss": 1.1797, "step": 3066 }, { "epoch": 1.4432941176470588, "grad_norm": 1.0616697521844434, "learning_rate": 0.0002927664564544545, "loss": 0.7383, "step": 3067 }, { "epoch": 1.4437647058823528, "grad_norm": 1.3002729682996543, "learning_rate": 0.00029275300174144464, "loss": 1.25, "step": 3068 }, { "epoch": 1.444235294117647, "grad_norm": 0.7571393799786175, "learning_rate": 0.0002927395348366282, "loss": 0.5312, "step": 3069 }, { "epoch": 1.4447058823529413, "grad_norm": 1.0100691050911175, "learning_rate": 0.00029272605574115544, "loss": 0.6953, "step": 3070 }, { "epoch": 1.4451764705882353, "grad_norm": 2.158381324601665, "learning_rate": 0.00029271256445617737, "loss": 2.0156, "step": 3071 }, { "epoch": 1.4456470588235293, "grad_norm": 1.3322012407828396, "learning_rate": 0.0002926990609828463, "loss": 1.2109, "step": 3072 }, { "epoch": 1.4461176470588235, "grad_norm": 1.8501869015032066, "learning_rate": 0.00029268554532231556, "loss": 1.4297, "step": 3073 }, { "epoch": 1.4465882352941177, "grad_norm": 0.6916577701006162, "learning_rate": 0.00029267201747573934, "loss": 0.4355, "step": 3074 }, { "epoch": 1.4470588235294117, "grad_norm": 1.038202725812564, "learning_rate": 0.00029265847744427303, "loss": 0.4082, "step": 3075 }, { "epoch": 1.447529411764706, "grad_norm": 1.4610231626285497, "learning_rate": 0.000292644925229073, "loss": 1.3594, "step": 3076 }, { "epoch": 1.448, "grad_norm": 1.5096605753219259, "learning_rate": 0.00029263136083129666, "loss": 1.3828, "step": 3077 }, { "epoch": 1.4484705882352942, "grad_norm": 1.0624284352001796, "learning_rate": 0.00029261778425210256, "loss": 0.8555, "step": 3078 }, { "epoch": 1.4489411764705882, "grad_norm": 1.5580781597299793, "learning_rate": 0.0002926041954926501, "loss": 1.25, "step": 3079 }, { "epoch": 1.4494117647058824, "grad_norm": 1.691421728158993, "learning_rate": 0.00029259059455409987, "loss": 1.5, "step": 3080 }, { "epoch": 1.4498823529411764, "grad_norm": 1.4732860850204472, "learning_rate": 0.00029257698143761354, "loss": 1.8594, "step": 3081 }, { "epoch": 1.4503529411764706, "grad_norm": 0.8788462779221605, "learning_rate": 0.00029256335614435356, "loss": 0.4121, "step": 3082 }, { "epoch": 1.4508235294117646, "grad_norm": 1.1245662043003952, "learning_rate": 0.00029254971867548373, "loss": 1.0703, "step": 3083 }, { "epoch": 1.4512941176470588, "grad_norm": 1.4914149646019643, "learning_rate": 0.00029253606903216874, "loss": 0.7461, "step": 3084 }, { "epoch": 1.4517647058823528, "grad_norm": 1.405086279458228, "learning_rate": 0.00029252240721557433, "loss": 1.7812, "step": 3085 }, { "epoch": 1.452235294117647, "grad_norm": 1.4191668943922977, "learning_rate": 0.00029250873322686727, "loss": 0.8633, "step": 3086 }, { "epoch": 1.4527058823529413, "grad_norm": 1.5904522737537599, "learning_rate": 0.00029249504706721534, "loss": 1.3438, "step": 3087 }, { "epoch": 1.4531764705882353, "grad_norm": 1.0416592598611971, "learning_rate": 0.00029248134873778754, "loss": 0.8398, "step": 3088 }, { "epoch": 1.4536470588235293, "grad_norm": 1.8865803878354868, "learning_rate": 0.0002924676382397536, "loss": 1.4062, "step": 3089 }, { "epoch": 1.4541176470588235, "grad_norm": 1.1102419795688423, "learning_rate": 0.0002924539155742846, "loss": 0.6758, "step": 3090 }, { "epoch": 1.4545882352941177, "grad_norm": 0.6397450531117448, "learning_rate": 0.0002924401807425525, "loss": 0.3867, "step": 3091 }, { "epoch": 1.4550588235294117, "grad_norm": 1.6969912566982093, "learning_rate": 0.0002924264337457303, "loss": 1.6562, "step": 3092 }, { "epoch": 1.455529411764706, "grad_norm": 1.0960956794348435, "learning_rate": 0.000292412674584992, "loss": 1.0469, "step": 3093 }, { "epoch": 1.456, "grad_norm": 1.451126822846482, "learning_rate": 0.0002923989032615128, "loss": 1.5391, "step": 3094 }, { "epoch": 1.4564705882352942, "grad_norm": 0.8533033556463064, "learning_rate": 0.0002923851197764688, "loss": 0.4375, "step": 3095 }, { "epoch": 1.4569411764705882, "grad_norm": 1.4334511541333539, "learning_rate": 0.00029237132413103715, "loss": 1.0391, "step": 3096 }, { "epoch": 1.4574117647058824, "grad_norm": 1.3272179685572623, "learning_rate": 0.00029235751632639617, "loss": 1.375, "step": 3097 }, { "epoch": 1.4578823529411764, "grad_norm": 1.5260533927814848, "learning_rate": 0.00029234369636372496, "loss": 1.3828, "step": 3098 }, { "epoch": 1.4583529411764706, "grad_norm": 1.0576823523179493, "learning_rate": 0.0002923298642442039, "loss": 0.8594, "step": 3099 }, { "epoch": 1.4588235294117646, "grad_norm": 1.2942925740203919, "learning_rate": 0.00029231601996901433, "loss": 1.2734, "step": 3100 }, { "epoch": 1.4592941176470589, "grad_norm": 1.2168596148932211, "learning_rate": 0.0002923021635393386, "loss": 1.5078, "step": 3101 }, { "epoch": 1.4597647058823529, "grad_norm": 0.7353604972394514, "learning_rate": 0.0002922882949563601, "loss": 0.3652, "step": 3102 }, { "epoch": 1.460235294117647, "grad_norm": 1.2669443934074645, "learning_rate": 0.0002922744142212633, "loss": 1.1953, "step": 3103 }, { "epoch": 1.4607058823529413, "grad_norm": 1.2818151698669191, "learning_rate": 0.00029226052133523363, "loss": 1.1953, "step": 3104 }, { "epoch": 1.4611764705882353, "grad_norm": 1.0087811774774662, "learning_rate": 0.0002922466162994577, "loss": 1.0156, "step": 3105 }, { "epoch": 1.4616470588235293, "grad_norm": 1.2330227632643562, "learning_rate": 0.000292232699115123, "loss": 1.9531, "step": 3106 }, { "epoch": 1.4621176470588235, "grad_norm": 1.4210524271548859, "learning_rate": 0.0002922187697834182, "loss": 1.6875, "step": 3107 }, { "epoch": 1.4625882352941177, "grad_norm": 1.408079276023658, "learning_rate": 0.00029220482830553284, "loss": 1.3906, "step": 3108 }, { "epoch": 1.4630588235294117, "grad_norm": 1.0382182935643045, "learning_rate": 0.0002921908746826576, "loss": 0.9023, "step": 3109 }, { "epoch": 1.4635294117647057, "grad_norm": 0.9412463548278491, "learning_rate": 0.0002921769089159843, "loss": 0.7578, "step": 3110 }, { "epoch": 1.464, "grad_norm": 1.2770682234435804, "learning_rate": 0.00029216293100670554, "loss": 1.3281, "step": 3111 }, { "epoch": 1.4644705882352942, "grad_norm": 1.1466572884915631, "learning_rate": 0.0002921489409560152, "loss": 0.7305, "step": 3112 }, { "epoch": 1.4649411764705882, "grad_norm": 1.4559324609531086, "learning_rate": 0.00029213493876510807, "loss": 1.8906, "step": 3113 }, { "epoch": 1.4654117647058824, "grad_norm": 0.359794022285543, "learning_rate": 0.0002921209244351799, "loss": 0.0957, "step": 3114 }, { "epoch": 1.4658823529411764, "grad_norm": 1.0800422815694866, "learning_rate": 0.00029210689796742783, "loss": 0.8711, "step": 3115 }, { "epoch": 1.4663529411764706, "grad_norm": 1.6829137131360294, "learning_rate": 0.00029209285936304956, "loss": 0.2949, "step": 3116 }, { "epoch": 1.4668235294117646, "grad_norm": 1.2336608848729826, "learning_rate": 0.0002920788086232442, "loss": 0.8125, "step": 3117 }, { "epoch": 1.4672941176470589, "grad_norm": 1.113145657947733, "learning_rate": 0.00029206474574921165, "loss": 1.0547, "step": 3118 }, { "epoch": 1.4677647058823529, "grad_norm": 1.016197586022966, "learning_rate": 0.0002920506707421529, "loss": 0.6328, "step": 3119 }, { "epoch": 1.468235294117647, "grad_norm": 1.0255156371907923, "learning_rate": 0.0002920365836032703, "loss": 0.8867, "step": 3120 }, { "epoch": 1.4687058823529413, "grad_norm": 0.8474771081679701, "learning_rate": 0.00029202248433376663, "loss": 0.6523, "step": 3121 }, { "epoch": 1.4691764705882353, "grad_norm": 1.152137674814497, "learning_rate": 0.00029200837293484627, "loss": 1.2734, "step": 3122 }, { "epoch": 1.4696470588235293, "grad_norm": 1.0415675299096434, "learning_rate": 0.0002919942494077142, "loss": 1.3203, "step": 3123 }, { "epoch": 1.4701176470588235, "grad_norm": 0.7260806035623613, "learning_rate": 0.0002919801137535769, "loss": 0.3574, "step": 3124 }, { "epoch": 1.4705882352941178, "grad_norm": 1.2317621176436762, "learning_rate": 0.0002919659659736414, "loss": 1.1797, "step": 3125 }, { "epoch": 1.4710588235294118, "grad_norm": 1.4681513006077833, "learning_rate": 0.00029195180606911606, "loss": 1.6562, "step": 3126 }, { "epoch": 1.4715294117647058, "grad_norm": 1.1440082173452644, "learning_rate": 0.0002919376340412102, "loss": 0.9375, "step": 3127 }, { "epoch": 1.472, "grad_norm": 1.3344144645005072, "learning_rate": 0.00029192344989113425, "loss": 1.1016, "step": 3128 }, { "epoch": 1.4724705882352942, "grad_norm": 0.6754226826924249, "learning_rate": 0.0002919092536200995, "loss": 0.2109, "step": 3129 }, { "epoch": 1.4729411764705882, "grad_norm": 1.189701091410305, "learning_rate": 0.00029189504522931847, "loss": 0.7578, "step": 3130 }, { "epoch": 1.4734117647058824, "grad_norm": 1.3693342160706508, "learning_rate": 0.00029188082472000453, "loss": 1.5547, "step": 3131 }, { "epoch": 1.4738823529411764, "grad_norm": 1.4116906148417128, "learning_rate": 0.0002918665920933722, "loss": 1.2734, "step": 3132 }, { "epoch": 1.4743529411764706, "grad_norm": 1.444364809287446, "learning_rate": 0.0002918523473506372, "loss": 1.9766, "step": 3133 }, { "epoch": 1.4748235294117646, "grad_norm": 0.8779769050319691, "learning_rate": 0.00029183809049301585, "loss": 0.5, "step": 3134 }, { "epoch": 1.4752941176470589, "grad_norm": 1.4316116174956177, "learning_rate": 0.0002918238215217259, "loss": 1.4453, "step": 3135 }, { "epoch": 1.4757647058823529, "grad_norm": 1.0304990733705457, "learning_rate": 0.00029180954043798595, "loss": 0.8047, "step": 3136 }, { "epoch": 1.476235294117647, "grad_norm": 1.076886779907902, "learning_rate": 0.00029179524724301563, "loss": 0.6367, "step": 3137 }, { "epoch": 1.4767058823529413, "grad_norm": 1.0273408658964809, "learning_rate": 0.00029178094193803576, "loss": 0.8867, "step": 3138 }, { "epoch": 1.4771764705882353, "grad_norm": 1.4017555904345385, "learning_rate": 0.00029176662452426796, "loss": 1.3984, "step": 3139 }, { "epoch": 1.4776470588235293, "grad_norm": 0.8732014214651629, "learning_rate": 0.0002917522950029351, "loss": 1.0, "step": 3140 }, { "epoch": 1.4781176470588235, "grad_norm": 1.2156556877477942, "learning_rate": 0.00029173795337526093, "loss": 1.0234, "step": 3141 }, { "epoch": 1.4785882352941178, "grad_norm": 0.67928110242861, "learning_rate": 0.0002917235996424703, "loss": 0.2148, "step": 3142 }, { "epoch": 1.4790588235294118, "grad_norm": 1.3008017000351846, "learning_rate": 0.0002917092338057892, "loss": 0.8789, "step": 3143 }, { "epoch": 1.4795294117647058, "grad_norm": 1.2064518252736303, "learning_rate": 0.00029169485586644434, "loss": 0.8242, "step": 3144 }, { "epoch": 1.48, "grad_norm": 1.1120324109569457, "learning_rate": 0.00029168046582566385, "loss": 1.1016, "step": 3145 }, { "epoch": 1.4804705882352942, "grad_norm": 1.0263271953883575, "learning_rate": 0.00029166606368467667, "loss": 0.8008, "step": 3146 }, { "epoch": 1.4809411764705882, "grad_norm": 1.5613605512476436, "learning_rate": 0.0002916516494447127, "loss": 1.5156, "step": 3147 }, { "epoch": 1.4814117647058824, "grad_norm": 1.633968720056689, "learning_rate": 0.0002916372231070031, "loss": 1.4922, "step": 3148 }, { "epoch": 1.4818823529411764, "grad_norm": 1.465101996983228, "learning_rate": 0.00029162278467277994, "loss": 1.5469, "step": 3149 }, { "epoch": 1.4823529411764707, "grad_norm": 1.0730925416976655, "learning_rate": 0.0002916083341432763, "loss": 0.9336, "step": 3150 }, { "epoch": 1.4828235294117647, "grad_norm": 1.510306946802089, "learning_rate": 0.0002915938715197264, "loss": 1.4609, "step": 3151 }, { "epoch": 1.4832941176470589, "grad_norm": 0.6804738229626259, "learning_rate": 0.00029157939680336527, "loss": 0.3477, "step": 3152 }, { "epoch": 1.4837647058823529, "grad_norm": 0.48966132460620937, "learning_rate": 0.00029156490999542923, "loss": 0.1963, "step": 3153 }, { "epoch": 1.484235294117647, "grad_norm": 1.45638031111648, "learning_rate": 0.0002915504110971555, "loss": 2.1094, "step": 3154 }, { "epoch": 1.484705882352941, "grad_norm": 1.2129697505129686, "learning_rate": 0.0002915359001097824, "loss": 1.3047, "step": 3155 }, { "epoch": 1.4851764705882353, "grad_norm": 1.1424171947518784, "learning_rate": 0.0002915213770345492, "loss": 0.8828, "step": 3156 }, { "epoch": 1.4856470588235293, "grad_norm": 1.5482986114106545, "learning_rate": 0.0002915068418726962, "loss": 1.1484, "step": 3157 }, { "epoch": 1.4861176470588235, "grad_norm": 0.8943832092714721, "learning_rate": 0.00029149229462546483, "loss": 0.4941, "step": 3158 }, { "epoch": 1.4865882352941178, "grad_norm": 0.7815455251644731, "learning_rate": 0.0002914777352940975, "loss": 0.3125, "step": 3159 }, { "epoch": 1.4870588235294118, "grad_norm": 1.035223399185829, "learning_rate": 0.0002914631638798376, "loss": 0.9102, "step": 3160 }, { "epoch": 1.4875294117647058, "grad_norm": 1.6115685435498126, "learning_rate": 0.00029144858038392964, "loss": 1.2969, "step": 3161 }, { "epoch": 1.488, "grad_norm": 1.1091494736046004, "learning_rate": 0.00029143398480761915, "loss": 0.582, "step": 3162 }, { "epoch": 1.4884705882352942, "grad_norm": 0.5999607052758089, "learning_rate": 0.00029141937715215256, "loss": 0.4688, "step": 3163 }, { "epoch": 1.4889411764705882, "grad_norm": 1.4148083582514754, "learning_rate": 0.0002914047574187776, "loss": 1.875, "step": 3164 }, { "epoch": 1.4894117647058824, "grad_norm": 1.14910359868458, "learning_rate": 0.0002913901256087427, "loss": 0.7344, "step": 3165 }, { "epoch": 1.4898823529411764, "grad_norm": 1.2488594722334259, "learning_rate": 0.00029137548172329756, "loss": 1.7344, "step": 3166 }, { "epoch": 1.4903529411764707, "grad_norm": 1.3218723202805736, "learning_rate": 0.0002913608257636929, "loss": 1.5859, "step": 3167 }, { "epoch": 1.4908235294117647, "grad_norm": 1.3419494316942324, "learning_rate": 0.0002913461577311803, "loss": 1.8359, "step": 3168 }, { "epoch": 1.4912941176470589, "grad_norm": 0.7505016610105492, "learning_rate": 0.00029133147762701246, "loss": 0.5391, "step": 3169 }, { "epoch": 1.4917647058823529, "grad_norm": 1.2733508471417494, "learning_rate": 0.00029131678545244333, "loss": 0.457, "step": 3170 }, { "epoch": 1.492235294117647, "grad_norm": 0.7460218124187395, "learning_rate": 0.00029130208120872747, "loss": 0.3105, "step": 3171 }, { "epoch": 1.492705882352941, "grad_norm": 1.2682424268271426, "learning_rate": 0.00029128736489712087, "loss": 0.9453, "step": 3172 }, { "epoch": 1.4931764705882353, "grad_norm": 1.2300379030047621, "learning_rate": 0.0002912726365188803, "loss": 1.1016, "step": 3173 }, { "epoch": 1.4936470588235293, "grad_norm": 0.8247140273711437, "learning_rate": 0.0002912578960752636, "loss": 0.4316, "step": 3174 }, { "epoch": 1.4941176470588236, "grad_norm": 0.7869836520095008, "learning_rate": 0.00029124314356752967, "loss": 0.3809, "step": 3175 }, { "epoch": 1.4945882352941178, "grad_norm": 1.1449124092149372, "learning_rate": 0.00029122837899693847, "loss": 0.9023, "step": 3176 }, { "epoch": 1.4950588235294118, "grad_norm": 2.022944376627076, "learning_rate": 0.0002912136023647511, "loss": 1.8359, "step": 3177 }, { "epoch": 1.4955294117647058, "grad_norm": 1.188261703454458, "learning_rate": 0.0002911988136722293, "loss": 1.4141, "step": 3178 }, { "epoch": 1.496, "grad_norm": 1.6617566449235053, "learning_rate": 0.00029118401292063634, "loss": 1.9531, "step": 3179 }, { "epoch": 1.4964705882352942, "grad_norm": 0.814044106151879, "learning_rate": 0.00029116920011123615, "loss": 0.3984, "step": 3180 }, { "epoch": 1.4969411764705882, "grad_norm": 1.5108172401899913, "learning_rate": 0.00029115437524529383, "loss": 1.2812, "step": 3181 }, { "epoch": 1.4974117647058822, "grad_norm": 1.3378465789312852, "learning_rate": 0.0002911395383240755, "loss": 1.4141, "step": 3182 }, { "epoch": 1.4978823529411764, "grad_norm": 1.0441899824976868, "learning_rate": 0.0002911246893488483, "loss": 1.5469, "step": 3183 }, { "epoch": 1.4983529411764707, "grad_norm": 0.6266151545861601, "learning_rate": 0.0002911098283208805, "loss": 0.3477, "step": 3184 }, { "epoch": 1.4988235294117647, "grad_norm": 1.2409173797146908, "learning_rate": 0.0002910949552414411, "loss": 1.4219, "step": 3185 }, { "epoch": 1.4992941176470589, "grad_norm": 0.7438258759513503, "learning_rate": 0.0002910800701118005, "loss": 0.2314, "step": 3186 }, { "epoch": 1.4997647058823529, "grad_norm": 1.0056703307886115, "learning_rate": 0.00029106517293323, "loss": 0.8242, "step": 3187 }, { "epoch": 1.5002352941176471, "grad_norm": 0.8110934306495002, "learning_rate": 0.0002910502637070017, "loss": 0.4531, "step": 3188 }, { "epoch": 1.5007058823529413, "grad_norm": 1.3256306915721434, "learning_rate": 0.0002910353424343891, "loss": 1.1953, "step": 3189 }, { "epoch": 1.5011764705882353, "grad_norm": 1.372970009166801, "learning_rate": 0.00029102040911666645, "loss": 1.3203, "step": 3190 }, { "epoch": 1.5016470588235293, "grad_norm": 1.202023162415183, "learning_rate": 0.0002910054637551092, "loss": 0.5117, "step": 3191 }, { "epoch": 1.5021176470588236, "grad_norm": 1.4256264755744175, "learning_rate": 0.0002909905063509936, "loss": 1.4531, "step": 3192 }, { "epoch": 1.5025882352941178, "grad_norm": 0.9377517541916558, "learning_rate": 0.0002909755369055973, "loss": 1.1562, "step": 3193 }, { "epoch": 1.5030588235294118, "grad_norm": 1.5621454572964186, "learning_rate": 0.00029096055542019865, "loss": 1.6094, "step": 3194 }, { "epoch": 1.5035294117647058, "grad_norm": 1.2293514005742017, "learning_rate": 0.0002909455618960772, "loss": 1.3125, "step": 3195 }, { "epoch": 1.504, "grad_norm": 1.2023066974735808, "learning_rate": 0.0002909305563345134, "loss": 1.4375, "step": 3196 }, { "epoch": 1.5044705882352942, "grad_norm": 0.9184622659009078, "learning_rate": 0.0002909155387367888, "loss": 0.6602, "step": 3197 }, { "epoch": 1.5049411764705882, "grad_norm": 0.9781885854916376, "learning_rate": 0.0002909005091041861, "loss": 0.6406, "step": 3198 }, { "epoch": 1.5054117647058822, "grad_norm": 1.3145034264156001, "learning_rate": 0.00029088546743798866, "loss": 0.5664, "step": 3199 }, { "epoch": 1.5058823529411764, "grad_norm": 1.268886409340166, "learning_rate": 0.00029087041373948135, "loss": 0.9023, "step": 3200 }, { "epoch": 1.5063529411764707, "grad_norm": 0.3128199844948891, "learning_rate": 0.00029085534800994976, "loss": 0.1807, "step": 3201 }, { "epoch": 1.5068235294117647, "grad_norm": 1.4757067620134123, "learning_rate": 0.0002908402702506805, "loss": 1.2656, "step": 3202 }, { "epoch": 1.5072941176470587, "grad_norm": 1.8770340732840909, "learning_rate": 0.0002908251804629614, "loss": 2.3594, "step": 3203 }, { "epoch": 1.507764705882353, "grad_norm": 1.1095408583551076, "learning_rate": 0.0002908100786480811, "loss": 0.9609, "step": 3204 }, { "epoch": 1.5082352941176471, "grad_norm": 2.5859152022714627, "learning_rate": 0.0002907949648073294, "loss": 1.1797, "step": 3205 }, { "epoch": 1.5087058823529413, "grad_norm": 1.5334724924434895, "learning_rate": 0.00029077983894199716, "loss": 1.6797, "step": 3206 }, { "epoch": 1.5091764705882353, "grad_norm": 0.5405259205232096, "learning_rate": 0.00029076470105337615, "loss": 0.3496, "step": 3207 }, { "epoch": 1.5096470588235293, "grad_norm": 0.6407121523954546, "learning_rate": 0.0002907495511427592, "loss": 0.2178, "step": 3208 }, { "epoch": 1.5101176470588236, "grad_norm": 0.9669950179216577, "learning_rate": 0.0002907343892114402, "loss": 1.6328, "step": 3209 }, { "epoch": 1.5105882352941178, "grad_norm": 1.0143141670619562, "learning_rate": 0.0002907192152607141, "loss": 1.1016, "step": 3210 }, { "epoch": 1.5110588235294118, "grad_norm": 1.4888787772658088, "learning_rate": 0.00029070402929187673, "loss": 1.5625, "step": 3211 }, { "epoch": 1.5115294117647058, "grad_norm": 1.0104607025780028, "learning_rate": 0.0002906888313062252, "loss": 0.5898, "step": 3212 }, { "epoch": 1.512, "grad_norm": 1.6112324617407074, "learning_rate": 0.00029067362130505734, "loss": 2.2656, "step": 3213 }, { "epoch": 1.5124705882352942, "grad_norm": 0.5673459746427842, "learning_rate": 0.00029065839928967224, "loss": 0.3594, "step": 3214 }, { "epoch": 1.5129411764705882, "grad_norm": 1.0619030276489443, "learning_rate": 0.0002906431652613699, "loss": 1.2656, "step": 3215 }, { "epoch": 1.5134117647058822, "grad_norm": 1.0009733977804025, "learning_rate": 0.0002906279192214514, "loss": 0.9414, "step": 3216 }, { "epoch": 1.5138823529411765, "grad_norm": 1.1372111780364913, "learning_rate": 0.0002906126611712188, "loss": 0.9258, "step": 3217 }, { "epoch": 1.5143529411764707, "grad_norm": 1.2840452179388664, "learning_rate": 0.00029059739111197524, "loss": 1.3438, "step": 3218 }, { "epoch": 1.5148235294117647, "grad_norm": 1.1907957034593062, "learning_rate": 0.0002905821090450249, "loss": 1.0469, "step": 3219 }, { "epoch": 1.5152941176470587, "grad_norm": 1.3162820036177323, "learning_rate": 0.00029056681497167285, "loss": 1.3359, "step": 3220 }, { "epoch": 1.515764705882353, "grad_norm": 1.2539165022001202, "learning_rate": 0.00029055150889322536, "loss": 1.4922, "step": 3221 }, { "epoch": 1.5162352941176471, "grad_norm": 1.199673534644106, "learning_rate": 0.0002905361908109896, "loss": 1.125, "step": 3222 }, { "epoch": 1.5167058823529411, "grad_norm": 0.8195427499493232, "learning_rate": 0.0002905208607262738, "loss": 0.3672, "step": 3223 }, { "epoch": 1.5171764705882353, "grad_norm": 1.5889400101225362, "learning_rate": 0.0002905055186403873, "loss": 2.5312, "step": 3224 }, { "epoch": 1.5176470588235293, "grad_norm": 1.2115025463956124, "learning_rate": 0.0002904901645546403, "loss": 1.3984, "step": 3225 }, { "epoch": 1.5181176470588236, "grad_norm": 1.0945913500527809, "learning_rate": 0.0002904747984703441, "loss": 0.6992, "step": 3226 }, { "epoch": 1.5185882352941178, "grad_norm": 1.0325681371630788, "learning_rate": 0.00029045942038881113, "loss": 1.0547, "step": 3227 }, { "epoch": 1.5190588235294118, "grad_norm": 1.8106556870495878, "learning_rate": 0.00029044403031135466, "loss": 1.3594, "step": 3228 }, { "epoch": 1.5195294117647058, "grad_norm": 1.0191016566683817, "learning_rate": 0.0002904286282392892, "loss": 0.8672, "step": 3229 }, { "epoch": 1.52, "grad_norm": 1.6788133636082674, "learning_rate": 0.00029041321417393006, "loss": 1.8984, "step": 3230 }, { "epoch": 1.5204705882352942, "grad_norm": 1.1851519408273656, "learning_rate": 0.0002903977881165937, "loss": 0.9609, "step": 3231 }, { "epoch": 1.5209411764705882, "grad_norm": 0.3437297132719872, "learning_rate": 0.0002903823500685976, "loss": 0.062, "step": 3232 }, { "epoch": 1.5214117647058822, "grad_norm": 0.569123340952562, "learning_rate": 0.0002903669000312602, "loss": 0.2832, "step": 3233 }, { "epoch": 1.5218823529411765, "grad_norm": 0.3011529204932546, "learning_rate": 0.0002903514380059011, "loss": 0.0308, "step": 3234 }, { "epoch": 1.5223529411764707, "grad_norm": 1.1077034117697988, "learning_rate": 0.0002903359639938407, "loss": 0.8086, "step": 3235 }, { "epoch": 1.5228235294117647, "grad_norm": 1.3689615061659257, "learning_rate": 0.0002903204779964007, "loss": 1.5156, "step": 3236 }, { "epoch": 1.5232941176470587, "grad_norm": 1.3033260141104601, "learning_rate": 0.00029030498001490356, "loss": 1.0078, "step": 3237 }, { "epoch": 1.523764705882353, "grad_norm": 0.9174093584408168, "learning_rate": 0.00029028947005067297, "loss": 0.4336, "step": 3238 }, { "epoch": 1.5242352941176471, "grad_norm": 1.2743920963369695, "learning_rate": 0.00029027394810503346, "loss": 1.5469, "step": 3239 }, { "epoch": 1.5247058823529411, "grad_norm": 1.2719172850276488, "learning_rate": 0.0002902584141793108, "loss": 1.7344, "step": 3240 }, { "epoch": 1.5251764705882351, "grad_norm": 0.740914799680839, "learning_rate": 0.0002902428682748316, "loss": 0.3125, "step": 3241 }, { "epoch": 1.5256470588235294, "grad_norm": 1.1055601672183153, "learning_rate": 0.0002902273103929236, "loss": 1.8984, "step": 3242 }, { "epoch": 1.5261176470588236, "grad_norm": 0.7477229951996668, "learning_rate": 0.00029021174053491547, "loss": 0.3184, "step": 3243 }, { "epoch": 1.5265882352941178, "grad_norm": 0.7454423518453975, "learning_rate": 0.0002901961587021369, "loss": 0.6758, "step": 3244 }, { "epoch": 1.5270588235294118, "grad_norm": 1.2975569194111058, "learning_rate": 0.00029018056489591874, "loss": 1.6094, "step": 3245 }, { "epoch": 1.5275294117647058, "grad_norm": 0.8499796530094308, "learning_rate": 0.0002901649591175928, "loss": 0.3496, "step": 3246 }, { "epoch": 1.528, "grad_norm": 1.0014261367587427, "learning_rate": 0.0002901493413684918, "loss": 1.0234, "step": 3247 }, { "epoch": 1.5284705882352942, "grad_norm": 1.7307774517525882, "learning_rate": 0.0002901337116499496, "loss": 2.125, "step": 3248 }, { "epoch": 1.5289411764705882, "grad_norm": 0.8327616761302349, "learning_rate": 0.0002901180699633011, "loss": 0.3691, "step": 3249 }, { "epoch": 1.5294117647058822, "grad_norm": 1.383356236180371, "learning_rate": 0.00029010241630988217, "loss": 1.4141, "step": 3250 }, { "epoch": 1.5298823529411765, "grad_norm": 1.1682566611017267, "learning_rate": 0.0002900867506910297, "loss": 0.9922, "step": 3251 }, { "epoch": 1.5303529411764707, "grad_norm": 0.9469729627469765, "learning_rate": 0.00029007107310808153, "loss": 0.6055, "step": 3252 }, { "epoch": 1.5308235294117647, "grad_norm": 0.5070624681546987, "learning_rate": 0.00029005538356237674, "loss": 0.1177, "step": 3253 }, { "epoch": 1.5312941176470587, "grad_norm": 1.0952302367897124, "learning_rate": 0.00029003968205525523, "loss": 1.2344, "step": 3254 }, { "epoch": 1.531764705882353, "grad_norm": 1.4104270330369542, "learning_rate": 0.00029002396858805794, "loss": 1.4375, "step": 3255 }, { "epoch": 1.5322352941176471, "grad_norm": 1.3365431264975627, "learning_rate": 0.0002900082431621269, "loss": 1.0703, "step": 3256 }, { "epoch": 1.5327058823529411, "grad_norm": 1.1135413831688, "learning_rate": 0.0002899925057788052, "loss": 1.0391, "step": 3257 }, { "epoch": 1.5331764705882351, "grad_norm": 2.2602868941466427, "learning_rate": 0.00028997675643943686, "loss": 1.2812, "step": 3258 }, { "epoch": 1.5336470588235294, "grad_norm": 0.5721899620773664, "learning_rate": 0.0002899609951453669, "loss": 0.1797, "step": 3259 }, { "epoch": 1.5341176470588236, "grad_norm": 1.1862486204678164, "learning_rate": 0.0002899452218979415, "loss": 1.4688, "step": 3260 }, { "epoch": 1.5345882352941178, "grad_norm": 0.9614204163952478, "learning_rate": 0.0002899294366985076, "loss": 0.6641, "step": 3261 }, { "epoch": 1.5350588235294118, "grad_norm": 1.405197911074814, "learning_rate": 0.0002899136395484136, "loss": 1.6172, "step": 3262 }, { "epoch": 1.5355294117647058, "grad_norm": 1.471115172347349, "learning_rate": 0.00028989783044900845, "loss": 1.3906, "step": 3263 }, { "epoch": 1.536, "grad_norm": 1.2062165339658721, "learning_rate": 0.0002898820094016424, "loss": 1.8125, "step": 3264 }, { "epoch": 1.5364705882352943, "grad_norm": 1.1385980745186954, "learning_rate": 0.0002898661764076666, "loss": 0.8789, "step": 3265 }, { "epoch": 1.5369411764705883, "grad_norm": 0.8226633240514055, "learning_rate": 0.00028985033146843335, "loss": 0.377, "step": 3266 }, { "epoch": 1.5374117647058823, "grad_norm": 0.7531824261804233, "learning_rate": 0.00028983447458529575, "loss": 0.375, "step": 3267 }, { "epoch": 1.5378823529411765, "grad_norm": 1.219813929719997, "learning_rate": 0.0002898186057596082, "loss": 1.5391, "step": 3268 }, { "epoch": 1.5383529411764707, "grad_norm": 1.1143786980539334, "learning_rate": 0.0002898027249927259, "loss": 0.9297, "step": 3269 }, { "epoch": 1.5388235294117647, "grad_norm": 0.8338187273783813, "learning_rate": 0.00028978683228600514, "loss": 0.4922, "step": 3270 }, { "epoch": 1.5392941176470587, "grad_norm": 1.6173314472163425, "learning_rate": 0.0002897709276408032, "loss": 1.8281, "step": 3271 }, { "epoch": 1.539764705882353, "grad_norm": 1.5631589200346707, "learning_rate": 0.00028975501105847856, "loss": 2.0312, "step": 3272 }, { "epoch": 1.5402352941176471, "grad_norm": 0.7752436072828082, "learning_rate": 0.00028973908254039044, "loss": 0.5703, "step": 3273 }, { "epoch": 1.5407058823529411, "grad_norm": 0.7480226032793996, "learning_rate": 0.00028972314208789926, "loss": 0.7266, "step": 3274 }, { "epoch": 1.5411764705882351, "grad_norm": 1.0389203229871928, "learning_rate": 0.00028970718970236644, "loss": 0.6055, "step": 3275 }, { "epoch": 1.5416470588235294, "grad_norm": 1.5996459111077266, "learning_rate": 0.00028969122538515435, "loss": 2.1875, "step": 3276 }, { "epoch": 1.5421176470588236, "grad_norm": 1.2201831181924123, "learning_rate": 0.00028967524913762636, "loss": 1.1875, "step": 3277 }, { "epoch": 1.5425882352941176, "grad_norm": 0.6508266349899429, "learning_rate": 0.00028965926096114706, "loss": 0.5859, "step": 3278 }, { "epoch": 1.5430588235294118, "grad_norm": 0.9616379518357446, "learning_rate": 0.00028964326085708183, "loss": 1.0, "step": 3279 }, { "epoch": 1.5435294117647058, "grad_norm": 1.007235272806626, "learning_rate": 0.00028962724882679715, "loss": 0.9883, "step": 3280 }, { "epoch": 1.544, "grad_norm": 0.9496263402440769, "learning_rate": 0.00028961122487166063, "loss": 1.3594, "step": 3281 }, { "epoch": 1.5444705882352943, "grad_norm": 1.296698392560545, "learning_rate": 0.00028959518899304064, "loss": 1.4141, "step": 3282 }, { "epoch": 1.5449411764705883, "grad_norm": 1.0279215886374315, "learning_rate": 0.00028957914119230686, "loss": 0.5469, "step": 3283 }, { "epoch": 1.5454117647058823, "grad_norm": 1.0406608348547375, "learning_rate": 0.0002895630814708297, "loss": 0.9492, "step": 3284 }, { "epoch": 1.5458823529411765, "grad_norm": 1.1571324649794967, "learning_rate": 0.00028954700982998093, "loss": 1.2188, "step": 3285 }, { "epoch": 1.5463529411764707, "grad_norm": 0.5116507667939093, "learning_rate": 0.000289530926271133, "loss": 0.5312, "step": 3286 }, { "epoch": 1.5468235294117647, "grad_norm": 0.7538108111489458, "learning_rate": 0.0002895148307956596, "loss": 0.3613, "step": 3287 }, { "epoch": 1.5472941176470587, "grad_norm": 0.9410990339836189, "learning_rate": 0.00028949872340493525, "loss": 0.5859, "step": 3288 }, { "epoch": 1.547764705882353, "grad_norm": 1.1279711427131716, "learning_rate": 0.00028948260410033576, "loss": 0.8594, "step": 3289 }, { "epoch": 1.5482352941176472, "grad_norm": 2.263410942987009, "learning_rate": 0.00028946647288323766, "loss": 2.4531, "step": 3290 }, { "epoch": 1.5487058823529412, "grad_norm": 1.1980093451424068, "learning_rate": 0.00028945032975501877, "loss": 0.9648, "step": 3291 }, { "epoch": 1.5491764705882352, "grad_norm": 0.7287798565877057, "learning_rate": 0.0002894341747170577, "loss": 0.4336, "step": 3292 }, { "epoch": 1.5496470588235294, "grad_norm": 1.2881252553711005, "learning_rate": 0.00028941800777073414, "loss": 0.9922, "step": 3293 }, { "epoch": 1.5501176470588236, "grad_norm": 0.8368702178281159, "learning_rate": 0.0002894018289174289, "loss": 0.6328, "step": 3294 }, { "epoch": 1.5505882352941176, "grad_norm": 1.3782469859265931, "learning_rate": 0.0002893856381585237, "loss": 1.1172, "step": 3295 }, { "epoch": 1.5510588235294116, "grad_norm": 1.287012623257477, "learning_rate": 0.0002893694354954014, "loss": 0.5234, "step": 3296 }, { "epoch": 1.5515294117647058, "grad_norm": 0.798359190386967, "learning_rate": 0.00028935322092944564, "loss": 0.418, "step": 3297 }, { "epoch": 1.552, "grad_norm": 1.1065167040613642, "learning_rate": 0.00028933699446204126, "loss": 0.8945, "step": 3298 }, { "epoch": 1.5524705882352943, "grad_norm": 0.7116210147886916, "learning_rate": 0.0002893207560945742, "loss": 0.3281, "step": 3299 }, { "epoch": 1.5529411764705883, "grad_norm": 1.1985724387431969, "learning_rate": 0.0002893045058284311, "loss": 1.2031, "step": 3300 }, { "epoch": 1.5534117647058823, "grad_norm": 1.0546980996676656, "learning_rate": 0.000289288243665, "loss": 0.8203, "step": 3301 }, { "epoch": 1.5538823529411765, "grad_norm": 1.3744978663602632, "learning_rate": 0.00028927196960566964, "loss": 1.7188, "step": 3302 }, { "epoch": 1.5543529411764707, "grad_norm": 0.9211546397602036, "learning_rate": 0.00028925568365182995, "loss": 0.5156, "step": 3303 }, { "epoch": 1.5548235294117647, "grad_norm": 1.1731974888073717, "learning_rate": 0.00028923938580487183, "loss": 1.1484, "step": 3304 }, { "epoch": 1.5552941176470587, "grad_norm": 1.1664683098454633, "learning_rate": 0.0002892230760661872, "loss": 1.2891, "step": 3305 }, { "epoch": 1.555764705882353, "grad_norm": 1.3480694490552039, "learning_rate": 0.000289206754437169, "loss": 1.5156, "step": 3306 }, { "epoch": 1.5562352941176472, "grad_norm": 1.168941784229678, "learning_rate": 0.0002891904209192112, "loss": 1.4531, "step": 3307 }, { "epoch": 1.5567058823529412, "grad_norm": 1.1394977102066925, "learning_rate": 0.00028917407551370863, "loss": 0.7891, "step": 3308 }, { "epoch": 1.5571764705882352, "grad_norm": 1.0061018959872063, "learning_rate": 0.00028915771822205747, "loss": 1.0, "step": 3309 }, { "epoch": 1.5576470588235294, "grad_norm": 1.2787683486709054, "learning_rate": 0.00028914134904565455, "loss": 1.6719, "step": 3310 }, { "epoch": 1.5581176470588236, "grad_norm": 1.7642989591369007, "learning_rate": 0.00028912496798589796, "loss": 2.5469, "step": 3311 }, { "epoch": 1.5585882352941176, "grad_norm": 1.412216250138776, "learning_rate": 0.0002891085750441867, "loss": 1.5, "step": 3312 }, { "epoch": 1.5590588235294116, "grad_norm": 0.8883912316770207, "learning_rate": 0.0002890921702219208, "loss": 1.1406, "step": 3313 }, { "epoch": 1.5595294117647058, "grad_norm": 1.1682495554874113, "learning_rate": 0.00028907575352050133, "loss": 1.4766, "step": 3314 }, { "epoch": 1.56, "grad_norm": 1.1306636623180932, "learning_rate": 0.00028905932494133034, "loss": 0.9805, "step": 3315 }, { "epoch": 1.5604705882352943, "grad_norm": 1.0241455839872005, "learning_rate": 0.0002890428844858109, "loss": 0.6914, "step": 3316 }, { "epoch": 1.5609411764705883, "grad_norm": 1.6332360957371295, "learning_rate": 0.0002890264321553471, "loss": 2.5781, "step": 3317 }, { "epoch": 1.5614117647058823, "grad_norm": 1.4284706667004834, "learning_rate": 0.00028900996795134414, "loss": 1.8438, "step": 3318 }, { "epoch": 1.5618823529411765, "grad_norm": 1.0082119635512485, "learning_rate": 0.000288993491875208, "loss": 0.7695, "step": 3319 }, { "epoch": 1.5623529411764707, "grad_norm": 1.7043842770622017, "learning_rate": 0.000288977003928346, "loss": 1.5781, "step": 3320 }, { "epoch": 1.5628235294117647, "grad_norm": 0.9126623183859818, "learning_rate": 0.0002889605041121661, "loss": 0.5703, "step": 3321 }, { "epoch": 1.5632941176470587, "grad_norm": 0.6445961776708059, "learning_rate": 0.0002889439924280776, "loss": 0.2988, "step": 3322 }, { "epoch": 1.563764705882353, "grad_norm": 1.0663302923674156, "learning_rate": 0.0002889274688774906, "loss": 0.8867, "step": 3323 }, { "epoch": 1.5642352941176472, "grad_norm": 0.7323734502052709, "learning_rate": 0.00028891093346181633, "loss": 0.4023, "step": 3324 }, { "epoch": 1.5647058823529412, "grad_norm": 1.229440703937815, "learning_rate": 0.000288894386182467, "loss": 1.1562, "step": 3325 }, { "epoch": 1.5651764705882352, "grad_norm": 1.1869606346817325, "learning_rate": 0.0002888778270408558, "loss": 0.9258, "step": 3326 }, { "epoch": 1.5656470588235294, "grad_norm": 1.6799062013511197, "learning_rate": 0.000288861256038397, "loss": 2.1094, "step": 3327 }, { "epoch": 1.5661176470588236, "grad_norm": 0.7483410009497952, "learning_rate": 0.0002888446731765058, "loss": 0.4004, "step": 3328 }, { "epoch": 1.5665882352941176, "grad_norm": 0.9287247592995695, "learning_rate": 0.00028882807845659846, "loss": 0.5781, "step": 3329 }, { "epoch": 1.5670588235294116, "grad_norm": 0.7640505288237496, "learning_rate": 0.00028881147188009233, "loss": 0.6484, "step": 3330 }, { "epoch": 1.5675294117647058, "grad_norm": 1.1451803627143464, "learning_rate": 0.00028879485344840563, "loss": 1.1953, "step": 3331 }, { "epoch": 1.568, "grad_norm": 0.9191651123985278, "learning_rate": 0.00028877822316295763, "loss": 0.543, "step": 3332 }, { "epoch": 1.5684705882352943, "grad_norm": 2.439424302279623, "learning_rate": 0.0002887615810251687, "loss": 1.3594, "step": 3333 }, { "epoch": 1.5689411764705883, "grad_norm": 0.9207622620569935, "learning_rate": 0.0002887449270364601, "loss": 0.9805, "step": 3334 }, { "epoch": 1.5694117647058823, "grad_norm": 1.2667529896368053, "learning_rate": 0.0002887282611982542, "loss": 1.5234, "step": 3335 }, { "epoch": 1.5698823529411765, "grad_norm": 1.2882426467860482, "learning_rate": 0.0002887115835119743, "loss": 1.4922, "step": 3336 }, { "epoch": 1.5703529411764707, "grad_norm": 0.9649487607736283, "learning_rate": 0.00028869489397904486, "loss": 1.0703, "step": 3337 }, { "epoch": 1.5708235294117647, "grad_norm": 1.0882945856341117, "learning_rate": 0.00028867819260089113, "loss": 1.0859, "step": 3338 }, { "epoch": 1.5712941176470587, "grad_norm": 0.8975531040646164, "learning_rate": 0.0002886614793789396, "loss": 0.7148, "step": 3339 }, { "epoch": 1.571764705882353, "grad_norm": 0.783502565056499, "learning_rate": 0.0002886447543146175, "loss": 0.6719, "step": 3340 }, { "epoch": 1.5722352941176472, "grad_norm": 1.3645738707945902, "learning_rate": 0.0002886280174093534, "loss": 0.8242, "step": 3341 }, { "epoch": 1.5727058823529412, "grad_norm": 1.3098148859162806, "learning_rate": 0.0002886112686645766, "loss": 1.8516, "step": 3342 }, { "epoch": 1.5731764705882352, "grad_norm": 1.0876425565703718, "learning_rate": 0.00028859450808171763, "loss": 0.9062, "step": 3343 }, { "epoch": 1.5736470588235294, "grad_norm": 0.9710968793281927, "learning_rate": 0.00028857773566220784, "loss": 0.875, "step": 3344 }, { "epoch": 1.5741176470588236, "grad_norm": 1.1495292932994519, "learning_rate": 0.0002885609514074797, "loss": 1.1641, "step": 3345 }, { "epoch": 1.5745882352941176, "grad_norm": 1.00114994305625, "learning_rate": 0.0002885441553189667, "loss": 1.2031, "step": 3346 }, { "epoch": 1.5750588235294116, "grad_norm": 1.0491733372988035, "learning_rate": 0.00028852734739810324, "loss": 1.2812, "step": 3347 }, { "epoch": 1.5755294117647058, "grad_norm": 0.7303518155454053, "learning_rate": 0.00028851052764632487, "loss": 0.3535, "step": 3348 }, { "epoch": 1.576, "grad_norm": 0.8236195024398005, "learning_rate": 0.00028849369606506807, "loss": 0.4648, "step": 3349 }, { "epoch": 1.576470588235294, "grad_norm": 0.7538503038260304, "learning_rate": 0.0002884768526557703, "loss": 0.3398, "step": 3350 }, { "epoch": 1.5769411764705883, "grad_norm": 1.3343926173392764, "learning_rate": 0.00028845999741987003, "loss": 1.3828, "step": 3351 }, { "epoch": 1.5774117647058823, "grad_norm": 1.7035490260066641, "learning_rate": 0.0002884431303588069, "loss": 2.6094, "step": 3352 }, { "epoch": 1.5778823529411765, "grad_norm": 1.15614697209277, "learning_rate": 0.00028842625147402134, "loss": 1.7422, "step": 3353 }, { "epoch": 1.5783529411764707, "grad_norm": 1.0746873411195172, "learning_rate": 0.00028840936076695493, "loss": 0.707, "step": 3354 }, { "epoch": 1.5788235294117647, "grad_norm": 1.3497442573575802, "learning_rate": 0.0002883924582390503, "loss": 1.4297, "step": 3355 }, { "epoch": 1.5792941176470587, "grad_norm": 0.6188159652813795, "learning_rate": 0.0002883755438917508, "loss": 0.3965, "step": 3356 }, { "epoch": 1.579764705882353, "grad_norm": 0.8566831197195792, "learning_rate": 0.0002883586177265012, "loss": 0.5, "step": 3357 }, { "epoch": 1.5802352941176472, "grad_norm": 1.5513833750152197, "learning_rate": 0.000288341679744747, "loss": 1.7734, "step": 3358 }, { "epoch": 1.5807058823529412, "grad_norm": 0.9157335055868998, "learning_rate": 0.0002883247299479348, "loss": 0.8438, "step": 3359 }, { "epoch": 1.5811764705882352, "grad_norm": 0.683786958034836, "learning_rate": 0.0002883077683375121, "loss": 0.5312, "step": 3360 }, { "epoch": 1.5816470588235294, "grad_norm": 1.0157189019626722, "learning_rate": 0.0002882907949149277, "loss": 1.4219, "step": 3361 }, { "epoch": 1.5821176470588236, "grad_norm": 1.208279662477191, "learning_rate": 0.000288273809681631, "loss": 1.3594, "step": 3362 }, { "epoch": 1.5825882352941176, "grad_norm": 0.7317229018621265, "learning_rate": 0.00028825681263907276, "loss": 0.6445, "step": 3363 }, { "epoch": 1.5830588235294116, "grad_norm": 1.1489698447778267, "learning_rate": 0.00028823980378870457, "loss": 0.6641, "step": 3364 }, { "epoch": 1.5835294117647059, "grad_norm": 1.0644044966079196, "learning_rate": 0.0002882227831319791, "loss": 1.0078, "step": 3365 }, { "epoch": 1.584, "grad_norm": 1.4431030668596212, "learning_rate": 0.00028820575067035, "loss": 1.3516, "step": 3366 }, { "epoch": 1.584470588235294, "grad_norm": 1.680849023309813, "learning_rate": 0.00028818870640527183, "loss": 1.5938, "step": 3367 }, { "epoch": 1.584941176470588, "grad_norm": 1.221353855465346, "learning_rate": 0.0002881716503382004, "loss": 1.4922, "step": 3368 }, { "epoch": 1.5854117647058823, "grad_norm": 1.4380615656183804, "learning_rate": 0.0002881545824705922, "loss": 1.3672, "step": 3369 }, { "epoch": 1.5858823529411765, "grad_norm": 1.7380565799001237, "learning_rate": 0.00028813750280390506, "loss": 1.1562, "step": 3370 }, { "epoch": 1.5863529411764707, "grad_norm": 0.8660395805362158, "learning_rate": 0.00028812041133959764, "loss": 0.3848, "step": 3371 }, { "epoch": 1.5868235294117647, "grad_norm": 1.7830670530649197, "learning_rate": 0.0002881033080791296, "loss": 2.1875, "step": 3372 }, { "epoch": 1.5872941176470587, "grad_norm": 1.153671328708436, "learning_rate": 0.0002880861930239616, "loss": 1.0859, "step": 3373 }, { "epoch": 1.587764705882353, "grad_norm": 1.3096715486693458, "learning_rate": 0.0002880690661755555, "loss": 1.7109, "step": 3374 }, { "epoch": 1.5882352941176472, "grad_norm": 1.0870751401580239, "learning_rate": 0.00028805192753537386, "loss": 0.8828, "step": 3375 }, { "epoch": 1.5887058823529412, "grad_norm": 1.3802361801653946, "learning_rate": 0.00028803477710488055, "loss": 1.0859, "step": 3376 }, { "epoch": 1.5891764705882352, "grad_norm": 1.5094090776340368, "learning_rate": 0.00028801761488554017, "loss": 1.375, "step": 3377 }, { "epoch": 1.5896470588235294, "grad_norm": 1.0008939124153367, "learning_rate": 0.0002880004408788185, "loss": 0.9805, "step": 3378 }, { "epoch": 1.5901176470588236, "grad_norm": 1.2971104807869964, "learning_rate": 0.0002879832550861823, "loss": 1.6094, "step": 3379 }, { "epoch": 1.5905882352941176, "grad_norm": 1.0219590903639324, "learning_rate": 0.0002879660575090993, "loss": 0.8164, "step": 3380 }, { "epoch": 1.5910588235294116, "grad_norm": 1.1641657524410896, "learning_rate": 0.0002879488481490382, "loss": 1.9609, "step": 3381 }, { "epoch": 1.5915294117647059, "grad_norm": 0.9375671568764572, "learning_rate": 0.0002879316270074689, "loss": 0.9141, "step": 3382 }, { "epoch": 1.592, "grad_norm": 0.798662729731861, "learning_rate": 0.00028791439408586215, "loss": 0.6562, "step": 3383 }, { "epoch": 1.592470588235294, "grad_norm": 0.729055787364016, "learning_rate": 0.00028789714938568965, "loss": 0.2168, "step": 3384 }, { "epoch": 1.592941176470588, "grad_norm": 1.0251503876247887, "learning_rate": 0.0002878798929084242, "loss": 1.0312, "step": 3385 }, { "epoch": 1.5934117647058823, "grad_norm": 1.2749799045749608, "learning_rate": 0.00028786262465553954, "loss": 1.7031, "step": 3386 }, { "epoch": 1.5938823529411765, "grad_norm": 1.3024182411136116, "learning_rate": 0.00028784534462851063, "loss": 0.9336, "step": 3387 }, { "epoch": 1.5943529411764708, "grad_norm": 1.109644338829964, "learning_rate": 0.0002878280528288131, "loss": 0.6211, "step": 3388 }, { "epoch": 1.5948235294117648, "grad_norm": 1.478938614080935, "learning_rate": 0.00028781074925792377, "loss": 1.2031, "step": 3389 }, { "epoch": 1.5952941176470588, "grad_norm": 0.6667945975639898, "learning_rate": 0.00028779343391732053, "loss": 0.2598, "step": 3390 }, { "epoch": 1.595764705882353, "grad_norm": 1.5853906971346394, "learning_rate": 0.0002877761068084822, "loss": 1.3359, "step": 3391 }, { "epoch": 1.5962352941176472, "grad_norm": 1.1149173433644102, "learning_rate": 0.00028775876793288855, "loss": 0.9531, "step": 3392 }, { "epoch": 1.5967058823529412, "grad_norm": 1.1690037105898738, "learning_rate": 0.0002877414172920204, "loss": 1.0078, "step": 3393 }, { "epoch": 1.5971764705882352, "grad_norm": 1.105512990270017, "learning_rate": 0.00028772405488735963, "loss": 1.5391, "step": 3394 }, { "epoch": 1.5976470588235294, "grad_norm": 1.5927332039164086, "learning_rate": 0.000287706680720389, "loss": 1.7422, "step": 3395 }, { "epoch": 1.5981176470588236, "grad_norm": 1.0673375185419964, "learning_rate": 0.0002876892947925924, "loss": 1.0547, "step": 3396 }, { "epoch": 1.5985882352941176, "grad_norm": 0.6327168702060045, "learning_rate": 0.0002876718971054547, "loss": 0.4375, "step": 3397 }, { "epoch": 1.5990588235294116, "grad_norm": 0.49400098650707747, "learning_rate": 0.00028765448766046163, "loss": 0.1758, "step": 3398 }, { "epoch": 1.5995294117647059, "grad_norm": 0.8071543116777122, "learning_rate": 0.00028763706645910014, "loss": 0.4395, "step": 3399 }, { "epoch": 1.6, "grad_norm": 1.0552461427183641, "learning_rate": 0.0002876196335028581, "loss": 0.4414, "step": 3400 }, { "epoch": 1.600470588235294, "grad_norm": 1.079911299038372, "learning_rate": 0.00028760218879322436, "loss": 0.5508, "step": 3401 }, { "epoch": 1.600941176470588, "grad_norm": 0.8334199793602362, "learning_rate": 0.00028758473233168874, "loss": 0.7734, "step": 3402 }, { "epoch": 1.6014117647058823, "grad_norm": 1.1908294066141756, "learning_rate": 0.0002875672641197421, "loss": 1.75, "step": 3403 }, { "epoch": 1.6018823529411765, "grad_norm": 2.0422888509413286, "learning_rate": 0.00028754978415887635, "loss": 1.5781, "step": 3404 }, { "epoch": 1.6023529411764705, "grad_norm": 0.7203609034531324, "learning_rate": 0.0002875322924505844, "loss": 0.2314, "step": 3405 }, { "epoch": 1.6028235294117648, "grad_norm": 1.5834576332389345, "learning_rate": 0.00028751478899636, "loss": 1.3828, "step": 3406 }, { "epoch": 1.6032941176470588, "grad_norm": 1.0441156355437267, "learning_rate": 0.00028749727379769817, "loss": 0.9883, "step": 3407 }, { "epoch": 1.603764705882353, "grad_norm": 1.2972062603409564, "learning_rate": 0.0002874797468560948, "loss": 0.7383, "step": 3408 }, { "epoch": 1.6042352941176472, "grad_norm": 0.8219413963616918, "learning_rate": 0.0002874622081730466, "loss": 0.4199, "step": 3409 }, { "epoch": 1.6047058823529412, "grad_norm": 0.4892169089337716, "learning_rate": 0.0002874446577500516, "loss": 0.3926, "step": 3410 }, { "epoch": 1.6051764705882352, "grad_norm": 0.8226509664471707, "learning_rate": 0.0002874270955886087, "loss": 0.4883, "step": 3411 }, { "epoch": 1.6056470588235294, "grad_norm": 1.2423698712202054, "learning_rate": 0.0002874095216902178, "loss": 1.7109, "step": 3412 }, { "epoch": 1.6061176470588236, "grad_norm": 0.8020010987842695, "learning_rate": 0.00028739193605637967, "loss": 0.3926, "step": 3413 }, { "epoch": 1.6065882352941177, "grad_norm": 0.6760001823378071, "learning_rate": 0.00028737433868859634, "loss": 0.3828, "step": 3414 }, { "epoch": 1.6070588235294117, "grad_norm": 0.7140654442965239, "learning_rate": 0.0002873567295883707, "loss": 0.6875, "step": 3415 }, { "epoch": 1.6075294117647059, "grad_norm": 1.1654116920410433, "learning_rate": 0.00028733910875720654, "loss": 1.6719, "step": 3416 }, { "epoch": 1.608, "grad_norm": 1.036410416386269, "learning_rate": 0.00028732147619660883, "loss": 0.6328, "step": 3417 }, { "epoch": 1.608470588235294, "grad_norm": 0.644884933739989, "learning_rate": 0.00028730383190808364, "loss": 0.4336, "step": 3418 }, { "epoch": 1.608941176470588, "grad_norm": 1.3952256390008833, "learning_rate": 0.0002872861758931376, "loss": 1.9922, "step": 3419 }, { "epoch": 1.6094117647058823, "grad_norm": 1.0655965691146259, "learning_rate": 0.0002872685081532788, "loss": 1.0703, "step": 3420 }, { "epoch": 1.6098823529411765, "grad_norm": 1.0680601477484744, "learning_rate": 0.00028725082869001615, "loss": 0.9453, "step": 3421 }, { "epoch": 1.6103529411764705, "grad_norm": 0.7051479092189299, "learning_rate": 0.00028723313750485946, "loss": 0.459, "step": 3422 }, { "epoch": 1.6108235294117645, "grad_norm": 1.0062813295909216, "learning_rate": 0.0002872154345993197, "loss": 1.1953, "step": 3423 }, { "epoch": 1.6112941176470588, "grad_norm": 1.1244538549184235, "learning_rate": 0.0002871977199749088, "loss": 1.0234, "step": 3424 }, { "epoch": 1.611764705882353, "grad_norm": 0.8256081972740854, "learning_rate": 0.00028717999363313967, "loss": 0.457, "step": 3425 }, { "epoch": 1.6122352941176472, "grad_norm": 0.9089364127817361, "learning_rate": 0.0002871622555755262, "loss": 0.5977, "step": 3426 }, { "epoch": 1.6127058823529412, "grad_norm": 1.901470823398659, "learning_rate": 0.00028714450580358334, "loss": 0.5039, "step": 3427 }, { "epoch": 1.6131764705882352, "grad_norm": 1.2080779952419487, "learning_rate": 0.000287126744318827, "loss": 0.9961, "step": 3428 }, { "epoch": 1.6136470588235294, "grad_norm": 1.6280063737711503, "learning_rate": 0.0002871089711227741, "loss": 2.5312, "step": 3429 }, { "epoch": 1.6141176470588237, "grad_norm": 1.5608716304153876, "learning_rate": 0.0002870911862169426, "loss": 1.8672, "step": 3430 }, { "epoch": 1.6145882352941177, "grad_norm": 0.8684381028226487, "learning_rate": 0.0002870733896028513, "loss": 1.0625, "step": 3431 }, { "epoch": 1.6150588235294117, "grad_norm": 1.1472305878870162, "learning_rate": 0.00028705558128202017, "loss": 0.9375, "step": 3432 }, { "epoch": 1.6155294117647059, "grad_norm": 1.0359642688693165, "learning_rate": 0.0002870377612559702, "loss": 1.0469, "step": 3433 }, { "epoch": 1.616, "grad_norm": 0.699558980202277, "learning_rate": 0.0002870199295262232, "loss": 0.6992, "step": 3434 }, { "epoch": 1.616470588235294, "grad_norm": 1.54719814945137, "learning_rate": 0.00028700208609430217, "loss": 1.1094, "step": 3435 }, { "epoch": 1.616941176470588, "grad_norm": 1.4201599076897382, "learning_rate": 0.00028698423096173097, "loss": 1.3047, "step": 3436 }, { "epoch": 1.6174117647058823, "grad_norm": 0.6504101537228164, "learning_rate": 0.0002869663641300346, "loss": 0.2598, "step": 3437 }, { "epoch": 1.6178823529411765, "grad_norm": 1.4990624820574934, "learning_rate": 0.00028694848560073885, "loss": 1.3594, "step": 3438 }, { "epoch": 1.6183529411764705, "grad_norm": 1.2390158264413356, "learning_rate": 0.00028693059537537077, "loss": 1.375, "step": 3439 }, { "epoch": 1.6188235294117646, "grad_norm": 0.8424261853502741, "learning_rate": 0.00028691269345545816, "loss": 0.5156, "step": 3440 }, { "epoch": 1.6192941176470588, "grad_norm": 1.3109238901422062, "learning_rate": 0.00028689477984252996, "loss": 1.7656, "step": 3441 }, { "epoch": 1.619764705882353, "grad_norm": 1.2345312779108655, "learning_rate": 0.00028687685453811614, "loss": 1.3984, "step": 3442 }, { "epoch": 1.6202352941176472, "grad_norm": 1.0202379837041835, "learning_rate": 0.00028685891754374755, "loss": 0.3203, "step": 3443 }, { "epoch": 1.6207058823529412, "grad_norm": 1.0186683951061177, "learning_rate": 0.0002868409688609561, "loss": 0.6797, "step": 3444 }, { "epoch": 1.6211764705882352, "grad_norm": 0.8415144125707139, "learning_rate": 0.00028682300849127465, "loss": 0.4844, "step": 3445 }, { "epoch": 1.6216470588235294, "grad_norm": 0.8218562952367983, "learning_rate": 0.0002868050364362373, "loss": 0.5742, "step": 3446 }, { "epoch": 1.6221176470588237, "grad_norm": 1.6874721415458898, "learning_rate": 0.00028678705269737873, "loss": 1.875, "step": 3447 }, { "epoch": 1.6225882352941177, "grad_norm": 0.7086116365030123, "learning_rate": 0.0002867690572762349, "loss": 0.4258, "step": 3448 }, { "epoch": 1.6230588235294117, "grad_norm": 2.121856226278949, "learning_rate": 0.0002867510501743428, "loss": 2.1562, "step": 3449 }, { "epoch": 1.6235294117647059, "grad_norm": 0.7907143796084302, "learning_rate": 0.0002867330313932402, "loss": 0.3613, "step": 3450 }, { "epoch": 1.624, "grad_norm": 1.3205225733229118, "learning_rate": 0.00028671500093446604, "loss": 1.4297, "step": 3451 }, { "epoch": 1.624470588235294, "grad_norm": 1.36409138082202, "learning_rate": 0.0002866969587995602, "loss": 1.0391, "step": 3452 }, { "epoch": 1.6249411764705881, "grad_norm": 1.3314632411660332, "learning_rate": 0.00028667890499006364, "loss": 1.3594, "step": 3453 }, { "epoch": 1.6254117647058823, "grad_norm": 0.6204217659614593, "learning_rate": 0.00028666083950751824, "loss": 0.2949, "step": 3454 }, { "epoch": 1.6258823529411766, "grad_norm": 1.2589020774114488, "learning_rate": 0.0002866427623534667, "loss": 1.2656, "step": 3455 }, { "epoch": 1.6263529411764706, "grad_norm": 1.2638419310051188, "learning_rate": 0.0002866246735294531, "loss": 1.6875, "step": 3456 }, { "epoch": 1.6268235294117646, "grad_norm": 0.9636825507125486, "learning_rate": 0.00028660657303702225, "loss": 0.6055, "step": 3457 }, { "epoch": 1.6272941176470588, "grad_norm": 1.0692767670133503, "learning_rate": 0.00028658846087771994, "loss": 0.8164, "step": 3458 }, { "epoch": 1.627764705882353, "grad_norm": 0.7505388761579918, "learning_rate": 0.0002865703370530932, "loss": 0.4902, "step": 3459 }, { "epoch": 1.6282352941176472, "grad_norm": 1.2342976869838616, "learning_rate": 0.0002865522015646897, "loss": 1.6562, "step": 3460 }, { "epoch": 1.6287058823529412, "grad_norm": 0.822212514374463, "learning_rate": 0.0002865340544140585, "loss": 0.5781, "step": 3461 }, { "epoch": 1.6291764705882352, "grad_norm": 0.9516360373428117, "learning_rate": 0.00028651589560274934, "loss": 0.8398, "step": 3462 }, { "epoch": 1.6296470588235294, "grad_norm": 0.4598076353703868, "learning_rate": 0.00028649772513231306, "loss": 0.3438, "step": 3463 }, { "epoch": 1.6301176470588237, "grad_norm": 1.3857703459178576, "learning_rate": 0.00028647954300430154, "loss": 1.5859, "step": 3464 }, { "epoch": 1.6305882352941177, "grad_norm": 1.0047907660331126, "learning_rate": 0.0002864613492202676, "loss": 1.1094, "step": 3465 }, { "epoch": 1.6310588235294117, "grad_norm": 1.1296835235791356, "learning_rate": 0.00028644314378176513, "loss": 0.9336, "step": 3466 }, { "epoch": 1.631529411764706, "grad_norm": 1.0008360134637926, "learning_rate": 0.00028642492669034894, "loss": 0.7422, "step": 3467 }, { "epoch": 1.6320000000000001, "grad_norm": 1.2405634001156538, "learning_rate": 0.0002864066979475748, "loss": 1.3984, "step": 3468 }, { "epoch": 1.6324705882352941, "grad_norm": 0.7910970651373193, "learning_rate": 0.00028638845755499964, "loss": 0.4902, "step": 3469 }, { "epoch": 1.6329411764705881, "grad_norm": 1.2123714270310009, "learning_rate": 0.00028637020551418123, "loss": 1.6016, "step": 3470 }, { "epoch": 1.6334117647058823, "grad_norm": 1.2945345878096997, "learning_rate": 0.00028635194182667834, "loss": 1.7656, "step": 3471 }, { "epoch": 1.6338823529411766, "grad_norm": 0.8983765702132877, "learning_rate": 0.0002863336664940508, "loss": 1.1016, "step": 3472 }, { "epoch": 1.6343529411764706, "grad_norm": 1.252949992359584, "learning_rate": 0.00028631537951785945, "loss": 1.2109, "step": 3473 }, { "epoch": 1.6348235294117646, "grad_norm": 0.7741716921402995, "learning_rate": 0.000286297080899666, "loss": 0.6289, "step": 3474 }, { "epoch": 1.6352941176470588, "grad_norm": 0.8081400649114484, "learning_rate": 0.0002862787706410334, "loss": 0.416, "step": 3475 }, { "epoch": 1.635764705882353, "grad_norm": 1.107868667562854, "learning_rate": 0.00028626044874352534, "loss": 1.1953, "step": 3476 }, { "epoch": 1.636235294117647, "grad_norm": 1.1877780816650025, "learning_rate": 0.0002862421152087066, "loss": 0.8594, "step": 3477 }, { "epoch": 1.636705882352941, "grad_norm": 0.9325485942338974, "learning_rate": 0.00028622377003814295, "loss": 1.1484, "step": 3478 }, { "epoch": 1.6371764705882352, "grad_norm": 0.8307157522428088, "learning_rate": 0.00028620541323340115, "loss": 0.5312, "step": 3479 }, { "epoch": 1.6376470588235295, "grad_norm": 0.8957322724893804, "learning_rate": 0.00028618704479604895, "loss": 0.457, "step": 3480 }, { "epoch": 1.6381176470588237, "grad_norm": 4.075719505679159, "learning_rate": 0.00028616866472765514, "loss": 1.1953, "step": 3481 }, { "epoch": 1.6385882352941177, "grad_norm": 0.8258538603839347, "learning_rate": 0.00028615027302978947, "loss": 0.4375, "step": 3482 }, { "epoch": 1.6390588235294117, "grad_norm": 0.8380955823421997, "learning_rate": 0.00028613186970402274, "loss": 0.5859, "step": 3483 }, { "epoch": 1.639529411764706, "grad_norm": 1.4790279780244355, "learning_rate": 0.0002861134547519265, "loss": 1.1719, "step": 3484 }, { "epoch": 1.6400000000000001, "grad_norm": 0.9677784578751296, "learning_rate": 0.0002860950281750737, "loss": 0.6328, "step": 3485 }, { "epoch": 1.6404705882352941, "grad_norm": 0.8445375231010562, "learning_rate": 0.00028607658997503786, "loss": 0.5039, "step": 3486 }, { "epoch": 1.6409411764705881, "grad_norm": 1.0092540126218392, "learning_rate": 0.00028605814015339383, "loss": 0.4629, "step": 3487 }, { "epoch": 1.6414117647058823, "grad_norm": 1.3183343264905998, "learning_rate": 0.0002860396787117173, "loss": 1.2969, "step": 3488 }, { "epoch": 1.6418823529411766, "grad_norm": 1.2934241019176869, "learning_rate": 0.00028602120565158486, "loss": 1.1562, "step": 3489 }, { "epoch": 1.6423529411764706, "grad_norm": 1.3213545269126694, "learning_rate": 0.0002860027209745743, "loss": 1.2266, "step": 3490 }, { "epoch": 1.6428235294117646, "grad_norm": 1.0285506052405764, "learning_rate": 0.00028598422468226425, "loss": 0.9102, "step": 3491 }, { "epoch": 1.6432941176470588, "grad_norm": 1.2662805508948232, "learning_rate": 0.0002859657167762345, "loss": 1.4922, "step": 3492 }, { "epoch": 1.643764705882353, "grad_norm": 1.144501351542776, "learning_rate": 0.00028594719725806556, "loss": 0.7344, "step": 3493 }, { "epoch": 1.644235294117647, "grad_norm": 0.9742553488798253, "learning_rate": 0.0002859286661293392, "loss": 0.8945, "step": 3494 }, { "epoch": 1.644705882352941, "grad_norm": 1.0957877591683673, "learning_rate": 0.00028591012339163806, "loss": 0.8789, "step": 3495 }, { "epoch": 1.6451764705882352, "grad_norm": 1.2161927740163883, "learning_rate": 0.00028589156904654565, "loss": 1.2578, "step": 3496 }, { "epoch": 1.6456470588235295, "grad_norm": 0.6074758550708667, "learning_rate": 0.00028587300309564675, "loss": 0.3008, "step": 3497 }, { "epoch": 1.6461176470588237, "grad_norm": 0.799445383516067, "learning_rate": 0.0002858544255405269, "loss": 0.3418, "step": 3498 }, { "epoch": 1.6465882352941177, "grad_norm": 1.6782408137719482, "learning_rate": 0.0002858358363827728, "loss": 1.2578, "step": 3499 }, { "epoch": 1.6470588235294117, "grad_norm": 0.740642770787842, "learning_rate": 0.000285817235623972, "loss": 0.2715, "step": 3500 }, { "epoch": 1.647529411764706, "grad_norm": 1.2516420371327497, "learning_rate": 0.00028579862326571305, "loss": 1.2109, "step": 3501 }, { "epoch": 1.6480000000000001, "grad_norm": 1.2980323020583555, "learning_rate": 0.0002857799993095856, "loss": 1.7422, "step": 3502 }, { "epoch": 1.6484705882352941, "grad_norm": 1.7073069651693962, "learning_rate": 0.0002857613637571802, "loss": 1.0547, "step": 3503 }, { "epoch": 1.6489411764705881, "grad_norm": 1.708155632441864, "learning_rate": 0.00028574271661008845, "loss": 2.2031, "step": 3504 }, { "epoch": 1.6494117647058824, "grad_norm": 1.282525294056839, "learning_rate": 0.00028572405786990294, "loss": 1.2422, "step": 3505 }, { "epoch": 1.6498823529411766, "grad_norm": 1.6621253211595293, "learning_rate": 0.00028570538753821704, "loss": 1.6016, "step": 3506 }, { "epoch": 1.6503529411764706, "grad_norm": 1.1709795541506156, "learning_rate": 0.0002856867056166255, "loss": 1.1719, "step": 3507 }, { "epoch": 1.6508235294117646, "grad_norm": 0.7221685077547462, "learning_rate": 0.00028566801210672376, "loss": 0.2969, "step": 3508 }, { "epoch": 1.6512941176470588, "grad_norm": 0.5817452334748048, "learning_rate": 0.0002856493070101084, "loss": 0.1187, "step": 3509 }, { "epoch": 1.651764705882353, "grad_norm": 1.1559931719772711, "learning_rate": 0.0002856305903283767, "loss": 0.9961, "step": 3510 }, { "epoch": 1.652235294117647, "grad_norm": 1.084868581091792, "learning_rate": 0.00028561186206312744, "loss": 1.1875, "step": 3511 }, { "epoch": 1.652705882352941, "grad_norm": 1.0487159012416847, "learning_rate": 0.00028559312221595996, "loss": 1.0625, "step": 3512 }, { "epoch": 1.6531764705882352, "grad_norm": 1.3722554666725941, "learning_rate": 0.00028557437078847477, "loss": 1.4688, "step": 3513 }, { "epoch": 1.6536470588235295, "grad_norm": 1.297786929232221, "learning_rate": 0.0002855556077822733, "loss": 1.9922, "step": 3514 }, { "epoch": 1.6541176470588237, "grad_norm": 1.1542020841395328, "learning_rate": 0.00028553683319895804, "loss": 0.9453, "step": 3515 }, { "epoch": 1.6545882352941177, "grad_norm": 1.0011641148348065, "learning_rate": 0.00028551804704013244, "loss": 0.5117, "step": 3516 }, { "epoch": 1.6550588235294117, "grad_norm": 1.2782114728743805, "learning_rate": 0.0002854992493074009, "loss": 1.25, "step": 3517 }, { "epoch": 1.655529411764706, "grad_norm": 1.448713738588688, "learning_rate": 0.00028548044000236886, "loss": 1.25, "step": 3518 }, { "epoch": 1.6560000000000001, "grad_norm": 1.0384140466229816, "learning_rate": 0.0002854616191266427, "loss": 1.0391, "step": 3519 }, { "epoch": 1.6564705882352941, "grad_norm": 0.8553278843913488, "learning_rate": 0.0002854427866818298, "loss": 0.5898, "step": 3520 }, { "epoch": 1.6569411764705881, "grad_norm": 1.7700447374240638, "learning_rate": 0.00028542394266953865, "loss": 2.2188, "step": 3521 }, { "epoch": 1.6574117647058824, "grad_norm": 1.2295522225859736, "learning_rate": 0.00028540508709137846, "loss": 1.25, "step": 3522 }, { "epoch": 1.6578823529411766, "grad_norm": 1.3239210984956429, "learning_rate": 0.00028538621994895976, "loss": 1.2891, "step": 3523 }, { "epoch": 1.6583529411764706, "grad_norm": 0.8286234207811816, "learning_rate": 0.00028536734124389373, "loss": 0.4062, "step": 3524 }, { "epoch": 1.6588235294117646, "grad_norm": 0.9662204964382357, "learning_rate": 0.0002853484509777928, "loss": 1.0078, "step": 3525 }, { "epoch": 1.6592941176470588, "grad_norm": 1.0021690208057086, "learning_rate": 0.00028532954915227033, "loss": 1.0859, "step": 3526 }, { "epoch": 1.659764705882353, "grad_norm": 1.4251249920241802, "learning_rate": 0.0002853106357689405, "loss": 1.6406, "step": 3527 }, { "epoch": 1.660235294117647, "grad_norm": 0.9517262711229195, "learning_rate": 0.0002852917108294187, "loss": 0.9336, "step": 3528 }, { "epoch": 1.660705882352941, "grad_norm": 1.324506976311272, "learning_rate": 0.0002852727743353213, "loss": 1.4453, "step": 3529 }, { "epoch": 1.6611764705882353, "grad_norm": 0.9052573810433855, "learning_rate": 0.00028525382628826536, "loss": 0.6875, "step": 3530 }, { "epoch": 1.6616470588235295, "grad_norm": 1.3700300797453455, "learning_rate": 0.0002852348666898692, "loss": 1.4453, "step": 3531 }, { "epoch": 1.6621176470588235, "grad_norm": 0.7987300318502398, "learning_rate": 0.0002852158955417522, "loss": 0.4473, "step": 3532 }, { "epoch": 1.6625882352941177, "grad_norm": 1.0129762860937228, "learning_rate": 0.0002851969128455345, "loss": 0.6797, "step": 3533 }, { "epoch": 1.6630588235294117, "grad_norm": 0.5455779515349396, "learning_rate": 0.0002851779186028372, "loss": 0.2197, "step": 3534 }, { "epoch": 1.663529411764706, "grad_norm": 0.7649898585356101, "learning_rate": 0.0002851589128152827, "loss": 0.3281, "step": 3535 }, { "epoch": 1.6640000000000001, "grad_norm": 1.8056210280081637, "learning_rate": 0.000285139895484494, "loss": 1.9688, "step": 3536 }, { "epoch": 1.6644705882352941, "grad_norm": 0.8529525649703112, "learning_rate": 0.0002851208666120955, "loss": 1.1953, "step": 3537 }, { "epoch": 1.6649411764705881, "grad_norm": 1.6073022409368591, "learning_rate": 0.0002851018261997121, "loss": 1.8359, "step": 3538 }, { "epoch": 1.6654117647058824, "grad_norm": 0.48015050692173145, "learning_rate": 0.00028508277424897015, "loss": 0.1816, "step": 3539 }, { "epoch": 1.6658823529411766, "grad_norm": 0.9966715737545386, "learning_rate": 0.00028506371076149665, "loss": 0.8516, "step": 3540 }, { "epoch": 1.6663529411764706, "grad_norm": 0.7404165977332645, "learning_rate": 0.0002850446357389198, "loss": 0.416, "step": 3541 }, { "epoch": 1.6668235294117646, "grad_norm": 0.8982877982246982, "learning_rate": 0.00028502554918286867, "loss": 0.5781, "step": 3542 }, { "epoch": 1.6672941176470588, "grad_norm": 0.9852515756682172, "learning_rate": 0.0002850064510949733, "loss": 0.7656, "step": 3543 }, { "epoch": 1.667764705882353, "grad_norm": 1.030459968167281, "learning_rate": 0.00028498734147686485, "loss": 0.6875, "step": 3544 }, { "epoch": 1.668235294117647, "grad_norm": 1.6691693900625422, "learning_rate": 0.0002849682203301753, "loss": 1.4375, "step": 3545 }, { "epoch": 1.668705882352941, "grad_norm": 1.2919134346694046, "learning_rate": 0.0002849490876565378, "loss": 1.1953, "step": 3546 }, { "epoch": 1.6691764705882353, "grad_norm": 1.4598088343203186, "learning_rate": 0.0002849299434575862, "loss": 1.6094, "step": 3547 }, { "epoch": 1.6696470588235295, "grad_norm": 1.3218008920337134, "learning_rate": 0.00028491078773495564, "loss": 2.125, "step": 3548 }, { "epoch": 1.6701176470588235, "grad_norm": 1.3946385440987794, "learning_rate": 0.00028489162049028205, "loss": 1.5234, "step": 3549 }, { "epoch": 1.6705882352941175, "grad_norm": 0.6481901441006536, "learning_rate": 0.00028487244172520246, "loss": 0.2891, "step": 3550 }, { "epoch": 1.6710588235294117, "grad_norm": 1.4188270113475165, "learning_rate": 0.0002848532514413548, "loss": 1.8438, "step": 3551 }, { "epoch": 1.671529411764706, "grad_norm": 1.153547254606109, "learning_rate": 0.00028483404964037794, "loss": 0.8906, "step": 3552 }, { "epoch": 1.6720000000000002, "grad_norm": 0.8853640214313718, "learning_rate": 0.000284814836323912, "loss": 0.4727, "step": 3553 }, { "epoch": 1.6724705882352942, "grad_norm": 1.0737882269466348, "learning_rate": 0.0002847956114935977, "loss": 0.7578, "step": 3554 }, { "epoch": 1.6729411764705882, "grad_norm": 0.5831980629149603, "learning_rate": 0.000284776375151077, "loss": 0.3086, "step": 3555 }, { "epoch": 1.6734117647058824, "grad_norm": 1.283541535008559, "learning_rate": 0.00028475712729799285, "loss": 1.3281, "step": 3556 }, { "epoch": 1.6738823529411766, "grad_norm": 0.9795995937778874, "learning_rate": 0.000284737867935989, "loss": 0.6211, "step": 3557 }, { "epoch": 1.6743529411764706, "grad_norm": 1.2035488308925413, "learning_rate": 0.0002847185970667104, "loss": 1.7344, "step": 3558 }, { "epoch": 1.6748235294117646, "grad_norm": 1.0836396015027003, "learning_rate": 0.00028469931469180275, "loss": 1.0859, "step": 3559 }, { "epoch": 1.6752941176470588, "grad_norm": 1.2586052222477107, "learning_rate": 0.00028468002081291296, "loss": 1.8281, "step": 3560 }, { "epoch": 1.675764705882353, "grad_norm": 1.270298709173978, "learning_rate": 0.0002846607154316888, "loss": 1.5, "step": 3561 }, { "epoch": 1.676235294117647, "grad_norm": 0.485969346052031, "learning_rate": 0.000284641398549779, "loss": 0.1621, "step": 3562 }, { "epoch": 1.676705882352941, "grad_norm": 1.4377536372895212, "learning_rate": 0.00028462207016883336, "loss": 2.4688, "step": 3563 }, { "epoch": 1.6771764705882353, "grad_norm": 1.0861903535749322, "learning_rate": 0.0002846027302905026, "loss": 1.4141, "step": 3564 }, { "epoch": 1.6776470588235295, "grad_norm": 0.7585286677061739, "learning_rate": 0.0002845833789164385, "loss": 0.5156, "step": 3565 }, { "epoch": 1.6781176470588235, "grad_norm": 1.1645966042301075, "learning_rate": 0.00028456401604829367, "loss": 1.1641, "step": 3566 }, { "epoch": 1.6785882352941175, "grad_norm": 0.707020655134454, "learning_rate": 0.00028454464168772186, "loss": 0.4473, "step": 3567 }, { "epoch": 1.6790588235294117, "grad_norm": 1.48009638940367, "learning_rate": 0.00028452525583637774, "loss": 1.5703, "step": 3568 }, { "epoch": 1.679529411764706, "grad_norm": 0.8362654759252873, "learning_rate": 0.0002845058584959169, "loss": 0.3457, "step": 3569 }, { "epoch": 1.6800000000000002, "grad_norm": 4.037255296489893, "learning_rate": 0.00028448644966799595, "loss": 1.4062, "step": 3570 }, { "epoch": 1.6804705882352942, "grad_norm": 1.875439185190569, "learning_rate": 0.0002844670293542726, "loss": 1.6328, "step": 3571 }, { "epoch": 1.6809411764705882, "grad_norm": 1.1150843165320392, "learning_rate": 0.00028444759755640535, "loss": 1.6094, "step": 3572 }, { "epoch": 1.6814117647058824, "grad_norm": 1.2657196584796924, "learning_rate": 0.0002844281542760539, "loss": 1.625, "step": 3573 }, { "epoch": 1.6818823529411766, "grad_norm": 0.7584774371348717, "learning_rate": 0.0002844086995148786, "loss": 0.5898, "step": 3574 }, { "epoch": 1.6823529411764706, "grad_norm": 0.9960404692622411, "learning_rate": 0.00028438923327454114, "loss": 0.4824, "step": 3575 }, { "epoch": 1.6828235294117646, "grad_norm": 1.3590679918680093, "learning_rate": 0.000284369755556704, "loss": 1.3906, "step": 3576 }, { "epoch": 1.6832941176470588, "grad_norm": 1.1658009630081325, "learning_rate": 0.0002843502663630306, "loss": 0.7891, "step": 3577 }, { "epoch": 1.683764705882353, "grad_norm": 0.38956088166977587, "learning_rate": 0.0002843307656951855, "loss": 0.4805, "step": 3578 }, { "epoch": 1.684235294117647, "grad_norm": 1.6008387710528897, "learning_rate": 0.0002843112535548341, "loss": 2.1406, "step": 3579 }, { "epoch": 1.684705882352941, "grad_norm": 1.1549330615001934, "learning_rate": 0.0002842917299436429, "loss": 1.5625, "step": 3580 }, { "epoch": 1.6851764705882353, "grad_norm": 0.6570298861745713, "learning_rate": 0.00028427219486327923, "loss": 1.0469, "step": 3581 }, { "epoch": 1.6856470588235295, "grad_norm": 0.9422518643347099, "learning_rate": 0.00028425264831541153, "loss": 0.8359, "step": 3582 }, { "epoch": 1.6861176470588235, "grad_norm": 1.405546118763393, "learning_rate": 0.0002842330903017091, "loss": 2.2188, "step": 3583 }, { "epoch": 1.6865882352941175, "grad_norm": 1.5097289094028263, "learning_rate": 0.00028421352082384245, "loss": 1.4219, "step": 3584 }, { "epoch": 1.6870588235294117, "grad_norm": 0.9892069695377054, "learning_rate": 0.00028419393988348275, "loss": 0.6641, "step": 3585 }, { "epoch": 1.687529411764706, "grad_norm": 1.2927435780087069, "learning_rate": 0.00028417434748230237, "loss": 1.0156, "step": 3586 }, { "epoch": 1.688, "grad_norm": 0.6285843659750612, "learning_rate": 0.0002841547436219746, "loss": 0.291, "step": 3587 }, { "epoch": 1.6884705882352942, "grad_norm": 0.9064857836192962, "learning_rate": 0.0002841351283041737, "loss": 1.0234, "step": 3588 }, { "epoch": 1.6889411764705882, "grad_norm": 0.50716042184443, "learning_rate": 0.0002841155015305749, "loss": 0.4395, "step": 3589 }, { "epoch": 1.6894117647058824, "grad_norm": 0.9587718040697267, "learning_rate": 0.00028409586330285444, "loss": 1.2578, "step": 3590 }, { "epoch": 1.6898823529411766, "grad_norm": 1.5805748672571436, "learning_rate": 0.00028407621362268955, "loss": 1.5938, "step": 3591 }, { "epoch": 1.6903529411764706, "grad_norm": 1.4573658234351559, "learning_rate": 0.0002840565524917583, "loss": 1.9141, "step": 3592 }, { "epoch": 1.6908235294117646, "grad_norm": 0.837792900669292, "learning_rate": 0.00028403687991174003, "loss": 0.3242, "step": 3593 }, { "epoch": 1.6912941176470588, "grad_norm": 0.693276600987391, "learning_rate": 0.0002840171958843147, "loss": 0.3574, "step": 3594 }, { "epoch": 1.691764705882353, "grad_norm": 1.401769334336846, "learning_rate": 0.0002839975004111635, "loss": 1.7266, "step": 3595 }, { "epoch": 1.692235294117647, "grad_norm": 1.189208787208301, "learning_rate": 0.00028397779349396854, "loss": 1.125, "step": 3596 }, { "epoch": 1.692705882352941, "grad_norm": 0.9301240061168975, "learning_rate": 0.0002839580751344128, "loss": 0.7109, "step": 3597 }, { "epoch": 1.6931764705882353, "grad_norm": 1.483513482912213, "learning_rate": 0.00028393834533418043, "loss": 1.7812, "step": 3598 }, { "epoch": 1.6936470588235295, "grad_norm": 1.196591673365265, "learning_rate": 0.0002839186040949564, "loss": 1.2109, "step": 3599 }, { "epoch": 1.6941176470588235, "grad_norm": 1.11912103831786, "learning_rate": 0.0002838988514184267, "loss": 1.0391, "step": 3600 }, { "epoch": 1.6945882352941175, "grad_norm": 0.5898852778329197, "learning_rate": 0.0002838790873062784, "loss": 0.208, "step": 3601 }, { "epoch": 1.6950588235294117, "grad_norm": 1.0419697271013202, "learning_rate": 0.0002838593117601992, "loss": 0.6875, "step": 3602 }, { "epoch": 1.695529411764706, "grad_norm": 1.028157782287379, "learning_rate": 0.0002838395247818783, "loss": 0.6836, "step": 3603 }, { "epoch": 1.696, "grad_norm": 0.7279744600678143, "learning_rate": 0.0002838197263730055, "loss": 0.3027, "step": 3604 }, { "epoch": 1.696470588235294, "grad_norm": 0.8078891317952864, "learning_rate": 0.0002837999165352717, "loss": 0.3691, "step": 3605 }, { "epoch": 1.6969411764705882, "grad_norm": 1.1503414241750112, "learning_rate": 0.0002837800952703687, "loss": 0.8828, "step": 3606 }, { "epoch": 1.6974117647058824, "grad_norm": 0.7899075427827442, "learning_rate": 0.00028376026257998947, "loss": 0.4297, "step": 3607 }, { "epoch": 1.6978823529411766, "grad_norm": 1.560294484264563, "learning_rate": 0.0002837404184658276, "loss": 0.9727, "step": 3608 }, { "epoch": 1.6983529411764706, "grad_norm": 1.3898730440953881, "learning_rate": 0.00028372056292957806, "loss": 1.3281, "step": 3609 }, { "epoch": 1.6988235294117646, "grad_norm": 0.6818636629811742, "learning_rate": 0.0002837006959729366, "loss": 0.3262, "step": 3610 }, { "epoch": 1.6992941176470588, "grad_norm": 1.2717622741863204, "learning_rate": 0.00028368081759759983, "loss": 1.7344, "step": 3611 }, { "epoch": 1.699764705882353, "grad_norm": 1.480957038685064, "learning_rate": 0.00028366092780526554, "loss": 1.6016, "step": 3612 }, { "epoch": 1.700235294117647, "grad_norm": 1.162706092460937, "learning_rate": 0.00028364102659763247, "loss": 1.3594, "step": 3613 }, { "epoch": 1.700705882352941, "grad_norm": 1.862181441430524, "learning_rate": 0.00028362111397640016, "loss": 1.4766, "step": 3614 }, { "epoch": 1.7011764705882353, "grad_norm": 1.301842105940089, "learning_rate": 0.00028360118994326936, "loss": 1.0, "step": 3615 }, { "epoch": 1.7016470588235295, "grad_norm": 1.1758118710619332, "learning_rate": 0.00028358125449994156, "loss": 0.9883, "step": 3616 }, { "epoch": 1.7021176470588235, "grad_norm": 1.355816421745529, "learning_rate": 0.0002835613076481195, "loss": 1.3203, "step": 3617 }, { "epoch": 1.7025882352941175, "grad_norm": 2.675791271634324, "learning_rate": 0.0002835413493895066, "loss": 1.7109, "step": 3618 }, { "epoch": 1.7030588235294117, "grad_norm": 0.7547989889663707, "learning_rate": 0.00028352137972580745, "loss": 0.5117, "step": 3619 }, { "epoch": 1.703529411764706, "grad_norm": 0.8615921702727467, "learning_rate": 0.0002835013986587276, "loss": 0.7539, "step": 3620 }, { "epoch": 1.704, "grad_norm": 1.0539528919031507, "learning_rate": 0.0002834814061899734, "loss": 0.6211, "step": 3621 }, { "epoch": 1.704470588235294, "grad_norm": 1.06031758323146, "learning_rate": 0.0002834614023212525, "loss": 1.1875, "step": 3622 }, { "epoch": 1.7049411764705882, "grad_norm": 1.3269591754943877, "learning_rate": 0.0002834413870542731, "loss": 1.5078, "step": 3623 }, { "epoch": 1.7054117647058824, "grad_norm": 0.8801615014738176, "learning_rate": 0.00028342136039074476, "loss": 0.668, "step": 3624 }, { "epoch": 1.7058823529411766, "grad_norm": 0.9435789835547197, "learning_rate": 0.0002834013223323778, "loss": 0.9375, "step": 3625 }, { "epoch": 1.7063529411764706, "grad_norm": 1.574008677473829, "learning_rate": 0.0002833812728808836, "loss": 1.7031, "step": 3626 }, { "epoch": 1.7068235294117646, "grad_norm": 0.49415495906431767, "learning_rate": 0.00028336121203797446, "loss": 0.2041, "step": 3627 }, { "epoch": 1.7072941176470589, "grad_norm": 1.296120490684884, "learning_rate": 0.0002833411398053637, "loss": 0.8594, "step": 3628 }, { "epoch": 1.707764705882353, "grad_norm": 0.7187347688143146, "learning_rate": 0.00028332105618476556, "loss": 0.8203, "step": 3629 }, { "epoch": 1.708235294117647, "grad_norm": 0.7783068382416698, "learning_rate": 0.0002833009611778953, "loss": 0.7734, "step": 3630 }, { "epoch": 1.708705882352941, "grad_norm": 0.5935914824266418, "learning_rate": 0.0002832808547864691, "loss": 0.2871, "step": 3631 }, { "epoch": 1.7091764705882353, "grad_norm": 1.4224099262440673, "learning_rate": 0.00028326073701220416, "loss": 1.5078, "step": 3632 }, { "epoch": 1.7096470588235295, "grad_norm": 1.276001951989634, "learning_rate": 0.0002832406078568186, "loss": 1.2109, "step": 3633 }, { "epoch": 1.7101176470588235, "grad_norm": 1.1584776622812765, "learning_rate": 0.00028322046732203165, "loss": 1.9766, "step": 3634 }, { "epoch": 1.7105882352941175, "grad_norm": 1.1913355576366627, "learning_rate": 0.0002832003154095633, "loss": 1.3828, "step": 3635 }, { "epoch": 1.7110588235294117, "grad_norm": 1.4592730353327787, "learning_rate": 0.0002831801521211347, "loss": 1.1172, "step": 3636 }, { "epoch": 1.711529411764706, "grad_norm": 0.8111712450490681, "learning_rate": 0.0002831599774584679, "loss": 0.2598, "step": 3637 }, { "epoch": 1.712, "grad_norm": 1.1945016932531958, "learning_rate": 0.0002831397914232858, "loss": 1.4297, "step": 3638 }, { "epoch": 1.712470588235294, "grad_norm": 1.3343121103449287, "learning_rate": 0.0002831195940173125, "loss": 1.2578, "step": 3639 }, { "epoch": 1.7129411764705882, "grad_norm": 0.6691423781885534, "learning_rate": 0.00028309938524227296, "loss": 0.3711, "step": 3640 }, { "epoch": 1.7134117647058824, "grad_norm": 1.3694391556086927, "learning_rate": 0.000283079165099893, "loss": 0.6328, "step": 3641 }, { "epoch": 1.7138823529411766, "grad_norm": 1.3437247063596858, "learning_rate": 0.0002830589335918997, "loss": 1.5625, "step": 3642 }, { "epoch": 1.7143529411764706, "grad_norm": 0.7378296654382152, "learning_rate": 0.0002830386907200207, "loss": 0.3027, "step": 3643 }, { "epoch": 1.7148235294117646, "grad_norm": 0.7661662126088422, "learning_rate": 0.000283018436485985, "loss": 0.3438, "step": 3644 }, { "epoch": 1.7152941176470589, "grad_norm": 1.0107217038385363, "learning_rate": 0.0002829981708915224, "loss": 1.0469, "step": 3645 }, { "epoch": 1.715764705882353, "grad_norm": 0.9285799268002364, "learning_rate": 0.0002829778939383636, "loss": 0.6328, "step": 3646 }, { "epoch": 1.716235294117647, "grad_norm": 1.4742599984493385, "learning_rate": 0.0002829576056282405, "loss": 1.3438, "step": 3647 }, { "epoch": 1.716705882352941, "grad_norm": 1.2638801128663537, "learning_rate": 0.00028293730596288565, "loss": 1.2656, "step": 3648 }, { "epoch": 1.7171764705882353, "grad_norm": 0.9297044955948284, "learning_rate": 0.00028291699494403284, "loss": 0.5938, "step": 3649 }, { "epoch": 1.7176470588235295, "grad_norm": 0.7080563120742024, "learning_rate": 0.0002828966725734167, "loss": 0.3828, "step": 3650 }, { "epoch": 1.7181176470588235, "grad_norm": 1.0010679421670525, "learning_rate": 0.0002828763388527729, "loss": 1.0703, "step": 3651 }, { "epoch": 1.7185882352941175, "grad_norm": 0.7441324940356474, "learning_rate": 0.000282855993783838, "loss": 0.4961, "step": 3652 }, { "epoch": 1.7190588235294118, "grad_norm": 1.526379940302879, "learning_rate": 0.0002828356373683495, "loss": 1.9375, "step": 3653 }, { "epoch": 1.719529411764706, "grad_norm": 0.5794110954912411, "learning_rate": 0.0002828152696080461, "loss": 0.1992, "step": 3654 }, { "epoch": 1.72, "grad_norm": 0.4513905970833155, "learning_rate": 0.0002827948905046672, "loss": 0.1816, "step": 3655 }, { "epoch": 1.720470588235294, "grad_norm": 0.5048450042370174, "learning_rate": 0.00028277450005995326, "loss": 0.125, "step": 3656 }, { "epoch": 1.7209411764705882, "grad_norm": 0.9653645787580295, "learning_rate": 0.0002827540982756458, "loss": 0.707, "step": 3657 }, { "epoch": 1.7214117647058824, "grad_norm": 1.2407737140579238, "learning_rate": 0.00028273368515348713, "loss": 1.1328, "step": 3658 }, { "epoch": 1.7218823529411764, "grad_norm": 0.8980244096825967, "learning_rate": 0.00028271326069522076, "loss": 0.5039, "step": 3659 }, { "epoch": 1.7223529411764706, "grad_norm": 1.0710487172725014, "learning_rate": 0.0002826928249025909, "loss": 0.9844, "step": 3660 }, { "epoch": 1.7228235294117646, "grad_norm": 2.300575042219852, "learning_rate": 0.000282672377777343, "loss": 2.1406, "step": 3661 }, { "epoch": 1.7232941176470589, "grad_norm": 0.7715329931752312, "learning_rate": 0.0002826519193212233, "loss": 0.416, "step": 3662 }, { "epoch": 1.723764705882353, "grad_norm": 1.1277930389051165, "learning_rate": 0.000282631449535979, "loss": 0.9453, "step": 3663 }, { "epoch": 1.724235294117647, "grad_norm": 1.007636869395814, "learning_rate": 0.00028261096842335836, "loss": 1.1094, "step": 3664 }, { "epoch": 1.724705882352941, "grad_norm": 2.0915486759409423, "learning_rate": 0.00028259047598511053, "loss": 1.4141, "step": 3665 }, { "epoch": 1.7251764705882353, "grad_norm": 1.3416783742019505, "learning_rate": 0.0002825699722229858, "loss": 1.4297, "step": 3666 }, { "epoch": 1.7256470588235295, "grad_norm": 0.4710233787469075, "learning_rate": 0.0002825494571387351, "loss": 0.4512, "step": 3667 }, { "epoch": 1.7261176470588235, "grad_norm": 1.0078509878076447, "learning_rate": 0.00028252893073411065, "loss": 0.7344, "step": 3668 }, { "epoch": 1.7265882352941175, "grad_norm": 1.4586782019506856, "learning_rate": 0.00028250839301086544, "loss": 1.6875, "step": 3669 }, { "epoch": 1.7270588235294118, "grad_norm": 0.7830686927427988, "learning_rate": 0.0002824878439707535, "loss": 0.6953, "step": 3670 }, { "epoch": 1.727529411764706, "grad_norm": 2.431065994761122, "learning_rate": 0.0002824672836155299, "loss": 1.1719, "step": 3671 }, { "epoch": 1.728, "grad_norm": 1.3157261554329323, "learning_rate": 0.0002824467119469505, "loss": 1.1875, "step": 3672 }, { "epoch": 1.728470588235294, "grad_norm": 1.230800284279754, "learning_rate": 0.0002824261289667722, "loss": 1.0859, "step": 3673 }, { "epoch": 1.7289411764705882, "grad_norm": 0.8975289837411964, "learning_rate": 0.000282405534676753, "loss": 0.7461, "step": 3674 }, { "epoch": 1.7294117647058824, "grad_norm": 1.3725406702106846, "learning_rate": 0.0002823849290786517, "loss": 1.5547, "step": 3675 }, { "epoch": 1.7298823529411764, "grad_norm": 1.0392890966911268, "learning_rate": 0.0002823643121742281, "loss": 0.9297, "step": 3676 }, { "epoch": 1.7303529411764704, "grad_norm": 0.6089508765452621, "learning_rate": 0.000282343683965243, "loss": 0.2119, "step": 3677 }, { "epoch": 1.7308235294117646, "grad_norm": 1.3138189809468133, "learning_rate": 0.0002823230444534581, "loss": 1.2266, "step": 3678 }, { "epoch": 1.7312941176470589, "grad_norm": 1.2510748893517412, "learning_rate": 0.0002823023936406362, "loss": 1.1797, "step": 3679 }, { "epoch": 1.731764705882353, "grad_norm": 1.1897194216843063, "learning_rate": 0.0002822817315285409, "loss": 1.0547, "step": 3680 }, { "epoch": 1.732235294117647, "grad_norm": 1.2099868146362203, "learning_rate": 0.000282261058118937, "loss": 1.4453, "step": 3681 }, { "epoch": 1.732705882352941, "grad_norm": 1.5413768163213246, "learning_rate": 0.00028224037341358993, "loss": 1.3125, "step": 3682 }, { "epoch": 1.7331764705882353, "grad_norm": 1.1261542364019042, "learning_rate": 0.0002822196774142663, "loss": 1.8438, "step": 3683 }, { "epoch": 1.7336470588235295, "grad_norm": 0.5482174566764402, "learning_rate": 0.0002821989701227337, "loss": 0.168, "step": 3684 }, { "epoch": 1.7341176470588235, "grad_norm": 1.3388907854405825, "learning_rate": 0.00028217825154076067, "loss": 1.3438, "step": 3685 }, { "epoch": 1.7345882352941175, "grad_norm": 1.1588104406513537, "learning_rate": 0.0002821575216701166, "loss": 0.9375, "step": 3686 }, { "epoch": 1.7350588235294118, "grad_norm": 0.8825524084453782, "learning_rate": 0.00028213678051257194, "loss": 0.9688, "step": 3687 }, { "epoch": 1.735529411764706, "grad_norm": 0.7328742975775814, "learning_rate": 0.00028211602806989814, "loss": 0.375, "step": 3688 }, { "epoch": 1.736, "grad_norm": 1.0096450629961842, "learning_rate": 0.00028209526434386744, "loss": 0.9648, "step": 3689 }, { "epoch": 1.736470588235294, "grad_norm": 1.0554670608772767, "learning_rate": 0.0002820744893362533, "loss": 0.7383, "step": 3690 }, { "epoch": 1.7369411764705882, "grad_norm": 1.0099582702925605, "learning_rate": 0.0002820537030488299, "loss": 1.0703, "step": 3691 }, { "epoch": 1.7374117647058824, "grad_norm": 1.2215157397974696, "learning_rate": 0.00028203290548337257, "loss": 1.3125, "step": 3692 }, { "epoch": 1.7378823529411764, "grad_norm": 2.2260190861200098, "learning_rate": 0.0002820120966416575, "loss": 1.7266, "step": 3693 }, { "epoch": 1.7383529411764704, "grad_norm": 1.0330572342143607, "learning_rate": 0.0002819912765254619, "loss": 0.7461, "step": 3694 }, { "epoch": 1.7388235294117647, "grad_norm": 1.0620374950900224, "learning_rate": 0.0002819704451365638, "loss": 1.0234, "step": 3695 }, { "epoch": 1.7392941176470589, "grad_norm": 0.8229108007329828, "learning_rate": 0.0002819496024767424, "loss": 0.4629, "step": 3696 }, { "epoch": 1.739764705882353, "grad_norm": 0.33299893279838155, "learning_rate": 0.0002819287485477778, "loss": 0.1553, "step": 3697 }, { "epoch": 1.740235294117647, "grad_norm": 0.8872678722237204, "learning_rate": 0.0002819078833514509, "loss": 0.3945, "step": 3698 }, { "epoch": 1.740705882352941, "grad_norm": 1.1746002166339842, "learning_rate": 0.0002818870068895437, "loss": 0.8164, "step": 3699 }, { "epoch": 1.7411764705882353, "grad_norm": 1.234854945949633, "learning_rate": 0.0002818661191638393, "loss": 1.0859, "step": 3700 }, { "epoch": 1.7416470588235295, "grad_norm": 0.8903091287182623, "learning_rate": 0.0002818452201761215, "loss": 0.3555, "step": 3701 }, { "epoch": 1.7421176470588235, "grad_norm": 1.116214066998584, "learning_rate": 0.0002818243099281752, "loss": 1.2266, "step": 3702 }, { "epoch": 1.7425882352941175, "grad_norm": 1.1785618159576237, "learning_rate": 0.00028180338842178623, "loss": 1.1172, "step": 3703 }, { "epoch": 1.7430588235294118, "grad_norm": 1.0054129534700498, "learning_rate": 0.0002817824556587414, "loss": 0.5391, "step": 3704 }, { "epoch": 1.743529411764706, "grad_norm": 0.9425169120953193, "learning_rate": 0.0002817615116408285, "loss": 1.0781, "step": 3705 }, { "epoch": 1.744, "grad_norm": 1.3902024255705714, "learning_rate": 0.0002817405563698362, "loss": 1.1328, "step": 3706 }, { "epoch": 1.744470588235294, "grad_norm": 1.4479129046186392, "learning_rate": 0.0002817195898475542, "loss": 1.3359, "step": 3707 }, { "epoch": 1.7449411764705882, "grad_norm": 0.7133888980419012, "learning_rate": 0.00028169861207577313, "loss": 0.498, "step": 3708 }, { "epoch": 1.7454117647058824, "grad_norm": 0.8925595564878971, "learning_rate": 0.0002816776230562846, "loss": 0.4668, "step": 3709 }, { "epoch": 1.7458823529411764, "grad_norm": 1.2312167507232827, "learning_rate": 0.0002816566227908812, "loss": 1.2422, "step": 3710 }, { "epoch": 1.7463529411764704, "grad_norm": 0.7938023535684879, "learning_rate": 0.00028163561128135645, "loss": 0.3848, "step": 3711 }, { "epoch": 1.7468235294117647, "grad_norm": 1.2203137987060744, "learning_rate": 0.0002816145885295048, "loss": 1.3281, "step": 3712 }, { "epoch": 1.7472941176470589, "grad_norm": 1.133087931908552, "learning_rate": 0.0002815935545371218, "loss": 1.125, "step": 3713 }, { "epoch": 1.7477647058823529, "grad_norm": 0.8346561989658025, "learning_rate": 0.00028157250930600374, "loss": 0.5508, "step": 3714 }, { "epoch": 1.748235294117647, "grad_norm": 1.2581154156778511, "learning_rate": 0.00028155145283794797, "loss": 1.1328, "step": 3715 }, { "epoch": 1.748705882352941, "grad_norm": 1.6100964131969269, "learning_rate": 0.0002815303851347529, "loss": 1.4531, "step": 3716 }, { "epoch": 1.7491764705882353, "grad_norm": 1.732575615216736, "learning_rate": 0.0002815093061982178, "loss": 0.7227, "step": 3717 }, { "epoch": 1.7496470588235296, "grad_norm": 1.291088147586963, "learning_rate": 0.0002814882160301429, "loss": 0.707, "step": 3718 }, { "epoch": 1.7501176470588236, "grad_norm": 1.4295878088114755, "learning_rate": 0.00028146711463232943, "loss": 1.5156, "step": 3719 }, { "epoch": 1.7505882352941176, "grad_norm": 0.6971234580075115, "learning_rate": 0.0002814460020065795, "loss": 0.2988, "step": 3720 }, { "epoch": 1.7510588235294118, "grad_norm": 1.6321150663296828, "learning_rate": 0.00028142487815469627, "loss": 2.125, "step": 3721 }, { "epoch": 1.751529411764706, "grad_norm": 2.0614745539140773, "learning_rate": 0.00028140374307848384, "loss": 2.2812, "step": 3722 }, { "epoch": 1.752, "grad_norm": 1.4698387452223811, "learning_rate": 0.00028138259677974714, "loss": 1.5938, "step": 3723 }, { "epoch": 1.752470588235294, "grad_norm": 0.6962597028444227, "learning_rate": 0.0002813614392602923, "loss": 0.2471, "step": 3724 }, { "epoch": 1.7529411764705882, "grad_norm": 1.117868549223829, "learning_rate": 0.00028134027052192624, "loss": 1.1172, "step": 3725 }, { "epoch": 1.7534117647058824, "grad_norm": 1.2690518598334797, "learning_rate": 0.0002813190905664568, "loss": 1.2969, "step": 3726 }, { "epoch": 1.7538823529411764, "grad_norm": 1.313504837902196, "learning_rate": 0.0002812978993956929, "loss": 0.9414, "step": 3727 }, { "epoch": 1.7543529411764704, "grad_norm": 1.475212223942656, "learning_rate": 0.0002812766970114445, "loss": 1.2656, "step": 3728 }, { "epoch": 1.7548235294117647, "grad_norm": 0.913869839604244, "learning_rate": 0.00028125548341552213, "loss": 0.8281, "step": 3729 }, { "epoch": 1.755294117647059, "grad_norm": 1.562832286609521, "learning_rate": 0.00028123425860973773, "loss": 1.9922, "step": 3730 }, { "epoch": 1.755764705882353, "grad_norm": 1.4124486749751688, "learning_rate": 0.000281213022595904, "loss": 1.125, "step": 3731 }, { "epoch": 1.756235294117647, "grad_norm": 1.3695669558545134, "learning_rate": 0.0002811917753758345, "loss": 1.4375, "step": 3732 }, { "epoch": 1.7567058823529411, "grad_norm": 9.390617463579265, "learning_rate": 0.00028117051695134384, "loss": 0.7031, "step": 3733 }, { "epoch": 1.7571764705882353, "grad_norm": 0.7781298828311032, "learning_rate": 0.0002811492473242477, "loss": 0.4434, "step": 3734 }, { "epoch": 1.7576470588235296, "grad_norm": 1.2278748468114005, "learning_rate": 0.00028112796649636254, "loss": 1.2891, "step": 3735 }, { "epoch": 1.7581176470588236, "grad_norm": 1.3393321710483408, "learning_rate": 0.0002811066744695059, "loss": 2.0156, "step": 3736 }, { "epoch": 1.7585882352941176, "grad_norm": 1.1675988527196044, "learning_rate": 0.0002810853712454961, "loss": 1.0781, "step": 3737 }, { "epoch": 1.7590588235294118, "grad_norm": 0.5201527753302431, "learning_rate": 0.00028106405682615265, "loss": 0.1641, "step": 3738 }, { "epoch": 1.759529411764706, "grad_norm": 3.017120235809889, "learning_rate": 0.0002810427312132959, "loss": 1.6406, "step": 3739 }, { "epoch": 1.76, "grad_norm": 1.5125308169997123, "learning_rate": 0.00028102139440874715, "loss": 1.8125, "step": 3740 }, { "epoch": 1.760470588235294, "grad_norm": 1.2371298450034882, "learning_rate": 0.0002810000464143287, "loss": 1.4766, "step": 3741 }, { "epoch": 1.7609411764705882, "grad_norm": 0.8335857794324746, "learning_rate": 0.0002809786872318637, "loss": 0.8242, "step": 3742 }, { "epoch": 1.7614117647058825, "grad_norm": 1.158274826497748, "learning_rate": 0.00028095731686317635, "loss": 1.2891, "step": 3743 }, { "epoch": 1.7618823529411765, "grad_norm": 1.0769442927239934, "learning_rate": 0.0002809359353100918, "loss": 0.9141, "step": 3744 }, { "epoch": 1.7623529411764705, "grad_norm": 1.3189354554737618, "learning_rate": 0.0002809145425744361, "loss": 1.2656, "step": 3745 }, { "epoch": 1.7628235294117647, "grad_norm": 0.5957520195504623, "learning_rate": 0.00028089313865803644, "loss": 0.2217, "step": 3746 }, { "epoch": 1.763294117647059, "grad_norm": 0.7919066793042138, "learning_rate": 0.0002808717235627206, "loss": 0.3379, "step": 3747 }, { "epoch": 1.763764705882353, "grad_norm": 1.082944586499714, "learning_rate": 0.0002808502972903177, "loss": 0.7461, "step": 3748 }, { "epoch": 1.764235294117647, "grad_norm": 1.0477240077771157, "learning_rate": 0.0002808288598426576, "loss": 0.8594, "step": 3749 }, { "epoch": 1.7647058823529411, "grad_norm": 0.48199389314575786, "learning_rate": 0.0002808074112215711, "loss": 0.4551, "step": 3750 }, { "epoch": 1.7651764705882353, "grad_norm": 0.5388944512384878, "learning_rate": 0.00028078595142889013, "loss": 0.3809, "step": 3751 }, { "epoch": 1.7656470588235296, "grad_norm": 1.190228560246262, "learning_rate": 0.0002807644804664474, "loss": 1.0234, "step": 3752 }, { "epoch": 1.7661176470588236, "grad_norm": 1.01591106082813, "learning_rate": 0.0002807429983360766, "loss": 0.7734, "step": 3753 }, { "epoch": 1.7665882352941176, "grad_norm": 0.8665949486474949, "learning_rate": 0.0002807215050396125, "loss": 0.6016, "step": 3754 }, { "epoch": 1.7670588235294118, "grad_norm": 1.1284486840155101, "learning_rate": 0.0002807000005788907, "loss": 1.2109, "step": 3755 }, { "epoch": 1.767529411764706, "grad_norm": 0.9931967494077303, "learning_rate": 0.00028067848495574773, "loss": 1.2578, "step": 3756 }, { "epoch": 1.768, "grad_norm": 1.0260259855936609, "learning_rate": 0.0002806569581720212, "loss": 0.9609, "step": 3757 }, { "epoch": 1.768470588235294, "grad_norm": 0.5414071491723659, "learning_rate": 0.00028063542022954956, "loss": 0.6133, "step": 3758 }, { "epoch": 1.7689411764705882, "grad_norm": 0.8988099423964734, "learning_rate": 0.00028061387113017226, "loss": 0.6836, "step": 3759 }, { "epoch": 1.7694117647058825, "grad_norm": 0.9342330419824313, "learning_rate": 0.0002805923108757297, "loss": 0.4707, "step": 3760 }, { "epoch": 1.7698823529411765, "grad_norm": 0.9064383103497264, "learning_rate": 0.00028057073946806323, "loss": 0.3965, "step": 3761 }, { "epoch": 1.7703529411764705, "grad_norm": 0.8401173564499435, "learning_rate": 0.00028054915690901524, "loss": 0.6367, "step": 3762 }, { "epoch": 1.7708235294117647, "grad_norm": 1.139344837364803, "learning_rate": 0.0002805275632004288, "loss": 1.1016, "step": 3763 }, { "epoch": 1.771294117647059, "grad_norm": 1.257934836095151, "learning_rate": 0.00028050595834414835, "loss": 1.0859, "step": 3764 }, { "epoch": 1.771764705882353, "grad_norm": 1.3468202144915693, "learning_rate": 0.00028048434234201883, "loss": 1.6953, "step": 3765 }, { "epoch": 1.772235294117647, "grad_norm": 1.0362026252672765, "learning_rate": 0.00028046271519588644, "loss": 1.4688, "step": 3766 }, { "epoch": 1.7727058823529411, "grad_norm": 1.1243257469647137, "learning_rate": 0.0002804410769075984, "loss": 1.4375, "step": 3767 }, { "epoch": 1.7731764705882354, "grad_norm": 0.9474230619144299, "learning_rate": 0.0002804194274790024, "loss": 0.5625, "step": 3768 }, { "epoch": 1.7736470588235296, "grad_norm": 1.138958661570981, "learning_rate": 0.00028039776691194773, "loss": 1.2891, "step": 3769 }, { "epoch": 1.7741176470588236, "grad_norm": 0.7742295420116861, "learning_rate": 0.0002803760952082841, "loss": 0.625, "step": 3770 }, { "epoch": 1.7745882352941176, "grad_norm": 1.2114957117860508, "learning_rate": 0.0002803544123698624, "loss": 0.9805, "step": 3771 }, { "epoch": 1.7750588235294118, "grad_norm": 0.8709941662559175, "learning_rate": 0.0002803327183985346, "loss": 0.7422, "step": 3772 }, { "epoch": 1.775529411764706, "grad_norm": 0.8449028708192848, "learning_rate": 0.00028031101329615334, "loss": 0.3848, "step": 3773 }, { "epoch": 1.776, "grad_norm": 0.6644035624414363, "learning_rate": 0.00028028929706457235, "loss": 0.2285, "step": 3774 }, { "epoch": 1.776470588235294, "grad_norm": 1.1722291788290042, "learning_rate": 0.0002802675697056463, "loss": 1.1875, "step": 3775 }, { "epoch": 1.7769411764705882, "grad_norm": 1.0039462769605643, "learning_rate": 0.00028024583122123094, "loss": 0.7344, "step": 3776 }, { "epoch": 1.7774117647058825, "grad_norm": 1.322291675731079, "learning_rate": 0.00028022408161318266, "loss": 1.5156, "step": 3777 }, { "epoch": 1.7778823529411765, "grad_norm": 0.9567922673926137, "learning_rate": 0.00028020232088335907, "loss": 0.8281, "step": 3778 }, { "epoch": 1.7783529411764705, "grad_norm": 0.9599661266905466, "learning_rate": 0.00028018054903361864, "loss": 0.5938, "step": 3779 }, { "epoch": 1.7788235294117647, "grad_norm": 1.1847887276051599, "learning_rate": 0.00028015876606582083, "loss": 1.2656, "step": 3780 }, { "epoch": 1.779294117647059, "grad_norm": 0.9109546380993292, "learning_rate": 0.0002801369719818259, "loss": 0.4766, "step": 3781 }, { "epoch": 1.779764705882353, "grad_norm": 1.2834747750529838, "learning_rate": 0.0002801151667834953, "loss": 0.9805, "step": 3782 }, { "epoch": 1.780235294117647, "grad_norm": 0.767703421572144, "learning_rate": 0.00028009335047269126, "loss": 0.3848, "step": 3783 }, { "epoch": 1.7807058823529411, "grad_norm": 0.9630781744091759, "learning_rate": 0.00028007152305127694, "loss": 0.7031, "step": 3784 }, { "epoch": 1.7811764705882354, "grad_norm": 0.8771840735093096, "learning_rate": 0.0002800496845211166, "loss": 0.6406, "step": 3785 }, { "epoch": 1.7816470588235294, "grad_norm": 0.794881252477197, "learning_rate": 0.0002800278348840753, "loss": 0.7344, "step": 3786 }, { "epoch": 1.7821176470588236, "grad_norm": 0.7968273655274563, "learning_rate": 0.00028000597414201906, "loss": 0.5938, "step": 3787 }, { "epoch": 1.7825882352941176, "grad_norm": 1.0123604554596777, "learning_rate": 0.000279984102296815, "loss": 1.2266, "step": 3788 }, { "epoch": 1.7830588235294118, "grad_norm": 0.8778140486284498, "learning_rate": 0.000279962219350331, "loss": 0.3105, "step": 3789 }, { "epoch": 1.783529411764706, "grad_norm": 1.3191593809983562, "learning_rate": 0.00027994032530443606, "loss": 1.875, "step": 3790 }, { "epoch": 1.784, "grad_norm": 0.9622930698176055, "learning_rate": 0.00027991842016099994, "loss": 0.6914, "step": 3791 }, { "epoch": 1.784470588235294, "grad_norm": 0.8062249768377803, "learning_rate": 0.0002798965039218936, "loss": 0.6758, "step": 3792 }, { "epoch": 1.7849411764705883, "grad_norm": 1.0199736710370064, "learning_rate": 0.0002798745765889886, "loss": 0.707, "step": 3793 }, { "epoch": 1.7854117647058825, "grad_norm": 1.2059447617566779, "learning_rate": 0.00027985263816415767, "loss": 1.0469, "step": 3794 }, { "epoch": 1.7858823529411765, "grad_norm": 1.2143471576045486, "learning_rate": 0.0002798306886492746, "loss": 1.0078, "step": 3795 }, { "epoch": 1.7863529411764705, "grad_norm": 0.7204795235924276, "learning_rate": 0.00027980872804621385, "loss": 0.3535, "step": 3796 }, { "epoch": 1.7868235294117647, "grad_norm": 1.1396942693700807, "learning_rate": 0.0002797867563568511, "loss": 1.3828, "step": 3797 }, { "epoch": 1.787294117647059, "grad_norm": 1.5949012123181943, "learning_rate": 0.0002797647735830627, "loss": 1.375, "step": 3798 }, { "epoch": 1.787764705882353, "grad_norm": 1.1987932336676768, "learning_rate": 0.00027974277972672614, "loss": 2.0469, "step": 3799 }, { "epoch": 1.788235294117647, "grad_norm": 1.139478313256879, "learning_rate": 0.0002797207747897198, "loss": 0.5938, "step": 3800 }, { "epoch": 1.7887058823529411, "grad_norm": 0.8247986398631977, "learning_rate": 0.000279698758773923, "loss": 0.4551, "step": 3801 }, { "epoch": 1.7891764705882354, "grad_norm": 1.175006256608571, "learning_rate": 0.0002796767316812161, "loss": 1.5469, "step": 3802 }, { "epoch": 1.7896470588235294, "grad_norm": 0.8606230280069308, "learning_rate": 0.0002796546935134802, "loss": 0.9297, "step": 3803 }, { "epoch": 1.7901176470588234, "grad_norm": 0.9819513803593477, "learning_rate": 0.00027963264427259755, "loss": 0.625, "step": 3804 }, { "epoch": 1.7905882352941176, "grad_norm": 0.8711272025031451, "learning_rate": 0.00027961058396045124, "loss": 0.4512, "step": 3805 }, { "epoch": 1.7910588235294118, "grad_norm": 1.1364761732350501, "learning_rate": 0.00027958851257892526, "loss": 1.4062, "step": 3806 }, { "epoch": 1.791529411764706, "grad_norm": 1.173302148499028, "learning_rate": 0.0002795664301299047, "loss": 1.1797, "step": 3807 }, { "epoch": 1.792, "grad_norm": 1.2927036655684128, "learning_rate": 0.00027954433661527554, "loss": 1.1094, "step": 3808 }, { "epoch": 1.792470588235294, "grad_norm": 1.5765973806926308, "learning_rate": 0.0002795222320369246, "loss": 1.2344, "step": 3809 }, { "epoch": 1.7929411764705883, "grad_norm": 1.2510707496097604, "learning_rate": 0.00027950011639673976, "loss": 1.4219, "step": 3810 }, { "epoch": 1.7934117647058825, "grad_norm": 0.613869697177558, "learning_rate": 0.00027947798969660975, "loss": 0.2227, "step": 3811 }, { "epoch": 1.7938823529411765, "grad_norm": 1.3104038831558518, "learning_rate": 0.00027945585193842433, "loss": 1.0234, "step": 3812 }, { "epoch": 1.7943529411764705, "grad_norm": 1.3126144501290389, "learning_rate": 0.00027943370312407415, "loss": 1.4766, "step": 3813 }, { "epoch": 1.7948235294117647, "grad_norm": 1.3818706145373867, "learning_rate": 0.00027941154325545096, "loss": 1.5703, "step": 3814 }, { "epoch": 1.795294117647059, "grad_norm": 0.8202198018119391, "learning_rate": 0.00027938937233444713, "loss": 0.7148, "step": 3815 }, { "epoch": 1.795764705882353, "grad_norm": 1.494591078943053, "learning_rate": 0.00027936719036295623, "loss": 1.8438, "step": 3816 }, { "epoch": 1.796235294117647, "grad_norm": 1.6691433525812698, "learning_rate": 0.0002793449973428728, "loss": 1.2891, "step": 3817 }, { "epoch": 1.7967058823529412, "grad_norm": 1.1408319599937011, "learning_rate": 0.0002793227932760921, "loss": 1.0156, "step": 3818 }, { "epoch": 1.7971764705882354, "grad_norm": 1.1477002317540268, "learning_rate": 0.0002793005781645106, "loss": 1.4297, "step": 3819 }, { "epoch": 1.7976470588235294, "grad_norm": 1.1471151135764812, "learning_rate": 0.00027927835201002545, "loss": 0.8086, "step": 3820 }, { "epoch": 1.7981176470588234, "grad_norm": 0.7920352347218024, "learning_rate": 0.00027925611481453497, "loss": 0.6484, "step": 3821 }, { "epoch": 1.7985882352941176, "grad_norm": 0.9924732346598315, "learning_rate": 0.0002792338665799383, "loss": 0.6914, "step": 3822 }, { "epoch": 1.7990588235294118, "grad_norm": 0.9276402788311278, "learning_rate": 0.00027921160730813544, "loss": 0.9297, "step": 3823 }, { "epoch": 1.799529411764706, "grad_norm": 0.9356173011849364, "learning_rate": 0.0002791893370010276, "loss": 0.5938, "step": 3824 }, { "epoch": 1.8, "grad_norm": 1.273532047474981, "learning_rate": 0.00027916705566051676, "loss": 2.0938, "step": 3825 }, { "epoch": 1.800470588235294, "grad_norm": 1.4143904159784955, "learning_rate": 0.00027914476328850575, "loss": 1.8125, "step": 3826 }, { "epoch": 1.8009411764705883, "grad_norm": 0.7294873397334094, "learning_rate": 0.0002791224598868985, "loss": 0.3867, "step": 3827 }, { "epoch": 1.8014117647058825, "grad_norm": 1.5168201721644252, "learning_rate": 0.00027910014545759986, "loss": 1.9062, "step": 3828 }, { "epoch": 1.8018823529411765, "grad_norm": 0.7113357586704404, "learning_rate": 0.00027907782000251557, "loss": 0.3555, "step": 3829 }, { "epoch": 1.8023529411764705, "grad_norm": 1.5384762037307027, "learning_rate": 0.00027905548352355226, "loss": 2.0, "step": 3830 }, { "epoch": 1.8028235294117647, "grad_norm": 1.819336829791408, "learning_rate": 0.00027903313602261774, "loss": 1.6562, "step": 3831 }, { "epoch": 1.803294117647059, "grad_norm": 0.9340335216822069, "learning_rate": 0.00027901077750162043, "loss": 1.0078, "step": 3832 }, { "epoch": 1.803764705882353, "grad_norm": 0.9394108523055136, "learning_rate": 0.00027898840796246994, "loss": 0.7969, "step": 3833 }, { "epoch": 1.804235294117647, "grad_norm": 1.0551320174380467, "learning_rate": 0.00027896602740707673, "loss": 1.0625, "step": 3834 }, { "epoch": 1.8047058823529412, "grad_norm": 0.9414810593432485, "learning_rate": 0.0002789436358373522, "loss": 0.7891, "step": 3835 }, { "epoch": 1.8051764705882354, "grad_norm": 1.613304523472492, "learning_rate": 0.0002789212332552088, "loss": 1.4922, "step": 3836 }, { "epoch": 1.8056470588235294, "grad_norm": 0.5734491363479559, "learning_rate": 0.0002788988196625596, "loss": 0.3633, "step": 3837 }, { "epoch": 1.8061176470588234, "grad_norm": 0.6586006319060989, "learning_rate": 0.000278876395061319, "loss": 0.4766, "step": 3838 }, { "epoch": 1.8065882352941176, "grad_norm": 1.262964283660364, "learning_rate": 0.0002788539594534021, "loss": 1.3438, "step": 3839 }, { "epoch": 1.8070588235294118, "grad_norm": 1.4846367454327567, "learning_rate": 0.0002788315128407251, "loss": 1.2578, "step": 3840 }, { "epoch": 1.8075294117647058, "grad_norm": 0.8306544469716924, "learning_rate": 0.0002788090552252049, "loss": 0.6602, "step": 3841 }, { "epoch": 1.808, "grad_norm": 1.1242689397516985, "learning_rate": 0.00027878658660875967, "loss": 0.9219, "step": 3842 }, { "epoch": 1.808470588235294, "grad_norm": 1.1682021647633563, "learning_rate": 0.0002787641069933082, "loss": 1.1172, "step": 3843 }, { "epoch": 1.8089411764705883, "grad_norm": 0.7279734573434959, "learning_rate": 0.0002787416163807704, "loss": 0.793, "step": 3844 }, { "epoch": 1.8094117647058825, "grad_norm": 0.8600394976585248, "learning_rate": 0.0002787191147730671, "loss": 0.6445, "step": 3845 }, { "epoch": 1.8098823529411765, "grad_norm": 1.0137396014225621, "learning_rate": 0.00027869660217212005, "loss": 0.957, "step": 3846 }, { "epoch": 1.8103529411764705, "grad_norm": 1.2908325332024473, "learning_rate": 0.0002786740785798519, "loss": 1.8047, "step": 3847 }, { "epoch": 1.8108235294117647, "grad_norm": 1.1394073483978713, "learning_rate": 0.00027865154399818627, "loss": 0.8945, "step": 3848 }, { "epoch": 1.811294117647059, "grad_norm": 1.2135838935211243, "learning_rate": 0.00027862899842904783, "loss": 0.9883, "step": 3849 }, { "epoch": 1.811764705882353, "grad_norm": 0.9264753483481628, "learning_rate": 0.00027860644187436195, "loss": 0.5781, "step": 3850 }, { "epoch": 1.812235294117647, "grad_norm": 1.0121584907849617, "learning_rate": 0.0002785838743360551, "loss": 1.2031, "step": 3851 }, { "epoch": 1.8127058823529412, "grad_norm": 1.0741594504535408, "learning_rate": 0.00027856129581605474, "loss": 1.1797, "step": 3852 }, { "epoch": 1.8131764705882354, "grad_norm": 1.2396705062564886, "learning_rate": 0.0002785387063162891, "loss": 1.2812, "step": 3853 }, { "epoch": 1.8136470588235294, "grad_norm": 1.1135378962227267, "learning_rate": 0.00027851610583868745, "loss": 1.2734, "step": 3854 }, { "epoch": 1.8141176470588234, "grad_norm": 1.0399294915106247, "learning_rate": 0.00027849349438518003, "loss": 1.1641, "step": 3855 }, { "epoch": 1.8145882352941176, "grad_norm": 1.1216352217446546, "learning_rate": 0.0002784708719576979, "loss": 0.6289, "step": 3856 }, { "epoch": 1.8150588235294118, "grad_norm": 1.0844730985165099, "learning_rate": 0.00027844823855817315, "loss": 1.2969, "step": 3857 }, { "epoch": 1.8155294117647058, "grad_norm": 1948.614226279199, "learning_rate": 0.0002784255941885388, "loss": 0.5859, "step": 3858 }, { "epoch": 1.8159999999999998, "grad_norm": 1.4208839768008392, "learning_rate": 0.0002784029388507288, "loss": 2.1719, "step": 3859 }, { "epoch": 1.816470588235294, "grad_norm": 1.3879710827000942, "learning_rate": 0.000278380272546678, "loss": 1.6953, "step": 3860 }, { "epoch": 1.8169411764705883, "grad_norm": 0.5207612676371248, "learning_rate": 0.00027835759527832226, "loss": 0.168, "step": 3861 }, { "epoch": 1.8174117647058825, "grad_norm": 0.7407545025476766, "learning_rate": 0.0002783349070475982, "loss": 0.543, "step": 3862 }, { "epoch": 1.8178823529411765, "grad_norm": 0.650666836666707, "learning_rate": 0.00027831220785644366, "loss": 0.2109, "step": 3863 }, { "epoch": 1.8183529411764705, "grad_norm": 0.6653589891133295, "learning_rate": 0.0002782894977067972, "loss": 0.2891, "step": 3864 }, { "epoch": 1.8188235294117647, "grad_norm": 0.9703856171294039, "learning_rate": 0.0002782667766005984, "loss": 1.125, "step": 3865 }, { "epoch": 1.819294117647059, "grad_norm": 0.6758784134584407, "learning_rate": 0.0002782440445397877, "loss": 0.3574, "step": 3866 }, { "epoch": 1.819764705882353, "grad_norm": 0.5679138605986319, "learning_rate": 0.00027822130152630657, "loss": 0.1855, "step": 3867 }, { "epoch": 1.820235294117647, "grad_norm": 1.4777899185854126, "learning_rate": 0.00027819854756209734, "loss": 2.5781, "step": 3868 }, { "epoch": 1.8207058823529412, "grad_norm": 0.7852877884625926, "learning_rate": 0.00027817578264910343, "loss": 0.582, "step": 3869 }, { "epoch": 1.8211764705882354, "grad_norm": 0.6764371879177331, "learning_rate": 0.00027815300678926886, "loss": 0.498, "step": 3870 }, { "epoch": 1.8216470588235294, "grad_norm": 0.9536604509501599, "learning_rate": 0.000278130219984539, "loss": 1.2891, "step": 3871 }, { "epoch": 1.8221176470588234, "grad_norm": 1.170017351723571, "learning_rate": 0.00027810742223685983, "loss": 0.3926, "step": 3872 }, { "epoch": 1.8225882352941176, "grad_norm": 1.3786295646584141, "learning_rate": 0.00027808461354817847, "loss": 1.1328, "step": 3873 }, { "epoch": 1.8230588235294118, "grad_norm": 1.1435556738291828, "learning_rate": 0.00027806179392044286, "loss": 1.8438, "step": 3874 }, { "epoch": 1.8235294117647058, "grad_norm": 1.0055670754601096, "learning_rate": 0.0002780389633556019, "loss": 0.5234, "step": 3875 }, { "epoch": 1.8239999999999998, "grad_norm": 1.1147529198447097, "learning_rate": 0.00027801612185560554, "loss": 1.3203, "step": 3876 }, { "epoch": 1.824470588235294, "grad_norm": 0.8788722142459789, "learning_rate": 0.00027799326942240434, "loss": 0.3223, "step": 3877 }, { "epoch": 1.8249411764705883, "grad_norm": 0.8652185394235574, "learning_rate": 0.0002779704060579501, "loss": 0.4609, "step": 3878 }, { "epoch": 1.8254117647058825, "grad_norm": 0.6418304289133533, "learning_rate": 0.00027794753176419557, "loss": 0.3203, "step": 3879 }, { "epoch": 1.8258823529411765, "grad_norm": 1.4135007228054224, "learning_rate": 0.00027792464654309427, "loss": 1.9062, "step": 3880 }, { "epoch": 1.8263529411764705, "grad_norm": 0.8157307221340098, "learning_rate": 0.0002779017503966006, "loss": 0.5234, "step": 3881 }, { "epoch": 1.8268235294117647, "grad_norm": 1.2291202952392066, "learning_rate": 0.0002778788433266702, "loss": 1.6406, "step": 3882 }, { "epoch": 1.827294117647059, "grad_norm": 1.272825918962426, "learning_rate": 0.00027785592533525924, "loss": 1.9453, "step": 3883 }, { "epoch": 1.827764705882353, "grad_norm": 0.7181792035311427, "learning_rate": 0.00027783299642432515, "loss": 0.957, "step": 3884 }, { "epoch": 1.828235294117647, "grad_norm": 1.325312226887178, "learning_rate": 0.0002778100565958262, "loss": 0.7656, "step": 3885 }, { "epoch": 1.8287058823529412, "grad_norm": 0.9991274946843524, "learning_rate": 0.0002777871058517215, "loss": 1.3281, "step": 3886 }, { "epoch": 1.8291764705882354, "grad_norm": 0.8837159466224397, "learning_rate": 0.00027776414419397106, "loss": 0.6328, "step": 3887 }, { "epoch": 1.8296470588235294, "grad_norm": 1.0508503611254438, "learning_rate": 0.00027774117162453617, "loss": 1.0469, "step": 3888 }, { "epoch": 1.8301176470588234, "grad_norm": 1.1052163050041903, "learning_rate": 0.00027771818814537855, "loss": 1.8516, "step": 3889 }, { "epoch": 1.8305882352941176, "grad_norm": 0.7866222563549252, "learning_rate": 0.00027769519375846126, "loss": 0.8047, "step": 3890 }, { "epoch": 1.8310588235294118, "grad_norm": 0.9450475577909251, "learning_rate": 0.0002776721884657481, "loss": 0.6641, "step": 3891 }, { "epoch": 1.8315294117647059, "grad_norm": 2.4130964804701023, "learning_rate": 0.00027764917226920377, "loss": 1.8984, "step": 3892 }, { "epoch": 1.8319999999999999, "grad_norm": 1.3641163852885785, "learning_rate": 0.000277626145170794, "loss": 1.5859, "step": 3893 }, { "epoch": 1.832470588235294, "grad_norm": 1.124379453408754, "learning_rate": 0.0002776031071724855, "loss": 1.7266, "step": 3894 }, { "epoch": 1.8329411764705883, "grad_norm": 1.1650917078055723, "learning_rate": 0.0002775800582762457, "loss": 1.2266, "step": 3895 }, { "epoch": 1.8334117647058825, "grad_norm": 0.8323220473234992, "learning_rate": 0.00027755699848404315, "loss": 0.7617, "step": 3896 }, { "epoch": 1.8338823529411765, "grad_norm": 1.8205667707715152, "learning_rate": 0.0002775339277978473, "loss": 1.7031, "step": 3897 }, { "epoch": 1.8343529411764705, "grad_norm": 0.9885736204131998, "learning_rate": 0.0002775108462196284, "loss": 0.8281, "step": 3898 }, { "epoch": 1.8348235294117647, "grad_norm": 0.7322484368701117, "learning_rate": 0.0002774877537513579, "loss": 0.293, "step": 3899 }, { "epoch": 1.835294117647059, "grad_norm": 1.0800066087341198, "learning_rate": 0.0002774646503950078, "loss": 0.875, "step": 3900 }, { "epoch": 1.835764705882353, "grad_norm": 1.0195437244510268, "learning_rate": 0.0002774415361525514, "loss": 0.7656, "step": 3901 }, { "epoch": 1.836235294117647, "grad_norm": 2.5240467358539003, "learning_rate": 0.00027741841102596276, "loss": 0.8867, "step": 3902 }, { "epoch": 1.8367058823529412, "grad_norm": 1.0639380776534506, "learning_rate": 0.0002773952750172168, "loss": 1.3594, "step": 3903 }, { "epoch": 1.8371764705882354, "grad_norm": 0.736658658842821, "learning_rate": 0.0002773721281282894, "loss": 0.2451, "step": 3904 }, { "epoch": 1.8376470588235294, "grad_norm": 1.272870611850447, "learning_rate": 0.0002773489703611576, "loss": 1.3125, "step": 3905 }, { "epoch": 1.8381176470588234, "grad_norm": 1.1685788176837735, "learning_rate": 0.00027732580171779906, "loss": 1.0234, "step": 3906 }, { "epoch": 1.8385882352941176, "grad_norm": 0.7877872614889363, "learning_rate": 0.0002773026222001925, "loss": 0.5703, "step": 3907 }, { "epoch": 1.8390588235294119, "grad_norm": 1.2134016577456808, "learning_rate": 0.00027727943181031766, "loss": 1.2969, "step": 3908 }, { "epoch": 1.8395294117647059, "grad_norm": 0.5336080613103154, "learning_rate": 0.00027725623055015495, "loss": 0.2773, "step": 3909 }, { "epoch": 1.8399999999999999, "grad_norm": 1.151715088143522, "learning_rate": 0.000277233018421686, "loss": 1.3516, "step": 3910 }, { "epoch": 1.840470588235294, "grad_norm": 1.1866713811175789, "learning_rate": 0.00027720979542689324, "loss": 1.7969, "step": 3911 }, { "epoch": 1.8409411764705883, "grad_norm": 0.9563638543993727, "learning_rate": 0.0002771865615677599, "loss": 0.5977, "step": 3912 }, { "epoch": 1.8414117647058823, "grad_norm": 1.8775115354556628, "learning_rate": 0.00027716331684627044, "loss": 2.4062, "step": 3913 }, { "epoch": 1.8418823529411765, "grad_norm": 1.4605923001150702, "learning_rate": 0.00027714006126441, "loss": 1.1094, "step": 3914 }, { "epoch": 1.8423529411764705, "grad_norm": 1.2716809285660498, "learning_rate": 0.00027711679482416465, "loss": 1.4297, "step": 3915 }, { "epoch": 1.8428235294117647, "grad_norm": 1.4022219067124602, "learning_rate": 0.0002770935175275215, "loss": 1.5312, "step": 3916 }, { "epoch": 1.843294117647059, "grad_norm": 1.181912707724503, "learning_rate": 0.0002770702293764686, "loss": 1.2109, "step": 3917 }, { "epoch": 1.843764705882353, "grad_norm": 0.8685503841893785, "learning_rate": 0.0002770469303729949, "loss": 0.5352, "step": 3918 }, { "epoch": 1.844235294117647, "grad_norm": 1.7897491801235896, "learning_rate": 0.0002770236205190901, "loss": 1.6484, "step": 3919 }, { "epoch": 1.8447058823529412, "grad_norm": 0.8274285911114915, "learning_rate": 0.00027700029981674506, "loss": 0.7617, "step": 3920 }, { "epoch": 1.8451764705882354, "grad_norm": 1.1406944613706074, "learning_rate": 0.0002769769682679515, "loss": 1.125, "step": 3921 }, { "epoch": 1.8456470588235294, "grad_norm": 2.015269973617527, "learning_rate": 0.00027695362587470203, "loss": 1.5469, "step": 3922 }, { "epoch": 1.8461176470588234, "grad_norm": 1.1868241816462146, "learning_rate": 0.0002769302726389902, "loss": 0.8477, "step": 3923 }, { "epoch": 1.8465882352941176, "grad_norm": 0.592150265207981, "learning_rate": 0.00027690690856281054, "loss": 0.3984, "step": 3924 }, { "epoch": 1.8470588235294119, "grad_norm": 1.3506515318317378, "learning_rate": 0.00027688353364815834, "loss": 1.0781, "step": 3925 }, { "epoch": 1.8475294117647059, "grad_norm": 0.3104478928371177, "learning_rate": 0.0002768601478970301, "loss": 0.1089, "step": 3926 }, { "epoch": 1.8479999999999999, "grad_norm": 0.9008479727697377, "learning_rate": 0.00027683675131142294, "loss": 0.6719, "step": 3927 }, { "epoch": 1.848470588235294, "grad_norm": 1.1711110891999905, "learning_rate": 0.000276813343893335, "loss": 1.5625, "step": 3928 }, { "epoch": 1.8489411764705883, "grad_norm": 1.3021186489590848, "learning_rate": 0.0002767899256447656, "loss": 1.0547, "step": 3929 }, { "epoch": 1.8494117647058823, "grad_norm": 0.5808011547069302, "learning_rate": 0.0002767664965677146, "loss": 0.1621, "step": 3930 }, { "epoch": 1.8498823529411763, "grad_norm": 1.2815463300638477, "learning_rate": 0.000276743056664183, "loss": 1.4219, "step": 3931 }, { "epoch": 1.8503529411764705, "grad_norm": 0.8824776099764124, "learning_rate": 0.0002767196059361727, "loss": 0.7148, "step": 3932 }, { "epoch": 1.8508235294117648, "grad_norm": 0.5638005956634919, "learning_rate": 0.0002766961443856865, "loss": 0.2617, "step": 3933 }, { "epoch": 1.851294117647059, "grad_norm": 0.7915443558686164, "learning_rate": 0.00027667267201472815, "loss": 0.3066, "step": 3934 }, { "epoch": 1.851764705882353, "grad_norm": 1.3623895297634918, "learning_rate": 0.00027664918882530225, "loss": 1.7656, "step": 3935 }, { "epoch": 1.852235294117647, "grad_norm": 0.9311907516592738, "learning_rate": 0.0002766256948194144, "loss": 0.8711, "step": 3936 }, { "epoch": 1.8527058823529412, "grad_norm": 1.0661456698449752, "learning_rate": 0.0002766021899990712, "loss": 1.0938, "step": 3937 }, { "epoch": 1.8531764705882354, "grad_norm": 1.0196404305686533, "learning_rate": 0.0002765786743662799, "loss": 1.3672, "step": 3938 }, { "epoch": 1.8536470588235294, "grad_norm": 1.142882218335924, "learning_rate": 0.0002765551479230489, "loss": 0.8477, "step": 3939 }, { "epoch": 1.8541176470588234, "grad_norm": 0.9371834779004327, "learning_rate": 0.00027653161067138764, "loss": 1.1172, "step": 3940 }, { "epoch": 1.8545882352941176, "grad_norm": 1.3365389995846004, "learning_rate": 0.00027650806261330616, "loss": 0.9844, "step": 3941 }, { "epoch": 1.8550588235294119, "grad_norm": 1.1169907704115516, "learning_rate": 0.0002764845037508156, "loss": 1.2109, "step": 3942 }, { "epoch": 1.8555294117647059, "grad_norm": 0.770453173073951, "learning_rate": 0.00027646093408592804, "loss": 0.5938, "step": 3943 }, { "epoch": 1.8559999999999999, "grad_norm": 1.1351809031122038, "learning_rate": 0.0002764373536206564, "loss": 1.2812, "step": 3944 }, { "epoch": 1.856470588235294, "grad_norm": 1.2897700176720834, "learning_rate": 0.0002764137623570146, "loss": 1.5781, "step": 3945 }, { "epoch": 1.8569411764705883, "grad_norm": 1.287183922383122, "learning_rate": 0.0002763901602970174, "loss": 1.4688, "step": 3946 }, { "epoch": 1.8574117647058823, "grad_norm": 0.6137700652737201, "learning_rate": 0.0002763665474426806, "loss": 0.3047, "step": 3947 }, { "epoch": 1.8578823529411763, "grad_norm": 1.1396294858920135, "learning_rate": 0.00027634292379602086, "loss": 1.3438, "step": 3948 }, { "epoch": 1.8583529411764705, "grad_norm": 0.9521390799238447, "learning_rate": 0.00027631928935905567, "loss": 0.8555, "step": 3949 }, { "epoch": 1.8588235294117648, "grad_norm": 0.9733551041794756, "learning_rate": 0.0002762956441338036, "loss": 1.0469, "step": 3950 }, { "epoch": 1.859294117647059, "grad_norm": 1.067571151136008, "learning_rate": 0.00027627198812228405, "loss": 0.5781, "step": 3951 }, { "epoch": 1.859764705882353, "grad_norm": 0.539766204528749, "learning_rate": 0.0002762483213265173, "loss": 0.1533, "step": 3952 }, { "epoch": 1.860235294117647, "grad_norm": 1.1524360281688246, "learning_rate": 0.0002762246437485247, "loss": 1.3906, "step": 3953 }, { "epoch": 1.8607058823529412, "grad_norm": 1.1057347958387516, "learning_rate": 0.0002762009553903284, "loss": 0.8555, "step": 3954 }, { "epoch": 1.8611764705882354, "grad_norm": 1.0795667322147, "learning_rate": 0.0002761772562539515, "loss": 1.4609, "step": 3955 }, { "epoch": 1.8616470588235294, "grad_norm": 1.1527766821421301, "learning_rate": 0.0002761535463414181, "loss": 0.957, "step": 3956 }, { "epoch": 1.8621176470588234, "grad_norm": 1.0253279534563087, "learning_rate": 0.0002761298256547529, "loss": 0.6719, "step": 3957 }, { "epoch": 1.8625882352941177, "grad_norm": 1.443780432745679, "learning_rate": 0.00027610609419598204, "loss": 1.7344, "step": 3958 }, { "epoch": 1.8630588235294119, "grad_norm": 1.077590622598761, "learning_rate": 0.0002760823519671321, "loss": 1.5781, "step": 3959 }, { "epoch": 1.8635294117647059, "grad_norm": 1.246100735301359, "learning_rate": 0.0002760585989702309, "loss": 0.9297, "step": 3960 }, { "epoch": 1.8639999999999999, "grad_norm": 1.5625420565221033, "learning_rate": 0.0002760348352073071, "loss": 0.9062, "step": 3961 }, { "epoch": 1.864470588235294, "grad_norm": 0.7396264041665347, "learning_rate": 0.0002760110606803901, "loss": 0.4766, "step": 3962 }, { "epoch": 1.8649411764705883, "grad_norm": 0.6815269425746313, "learning_rate": 0.00027598727539151054, "loss": 0.4355, "step": 3963 }, { "epoch": 1.8654117647058823, "grad_norm": 0.5816956965039513, "learning_rate": 0.0002759634793426996, "loss": 0.2695, "step": 3964 }, { "epoch": 1.8658823529411763, "grad_norm": 0.9720981349141483, "learning_rate": 0.00027593967253598974, "loss": 0.6875, "step": 3965 }, { "epoch": 1.8663529411764705, "grad_norm": 1.3189305310269979, "learning_rate": 0.00027591585497341406, "loss": 1.1641, "step": 3966 }, { "epoch": 1.8668235294117648, "grad_norm": 0.870830579257363, "learning_rate": 0.00027589202665700676, "loss": 0.7109, "step": 3967 }, { "epoch": 1.8672941176470588, "grad_norm": 1.315677240924655, "learning_rate": 0.0002758681875888029, "loss": 1.0391, "step": 3968 }, { "epoch": 1.867764705882353, "grad_norm": 1.5818429488821362, "learning_rate": 0.00027584433777083845, "loss": 1.8594, "step": 3969 }, { "epoch": 1.868235294117647, "grad_norm": 0.6199540319262106, "learning_rate": 0.0002758204772051503, "loss": 0.4824, "step": 3970 }, { "epoch": 1.8687058823529412, "grad_norm": 0.30870872645016467, "learning_rate": 0.00027579660589377626, "loss": 0.1006, "step": 3971 }, { "epoch": 1.8691764705882354, "grad_norm": 0.9609377597396405, "learning_rate": 0.00027577272383875497, "loss": 0.9297, "step": 3972 }, { "epoch": 1.8696470588235294, "grad_norm": 0.8876330880691432, "learning_rate": 0.0002757488310421262, "loss": 0.582, "step": 3973 }, { "epoch": 1.8701176470588234, "grad_norm": 1.220550820164612, "learning_rate": 0.0002757249275059305, "loss": 1.6328, "step": 3974 }, { "epoch": 1.8705882352941177, "grad_norm": 1.0326462662470446, "learning_rate": 0.00027570101323220927, "loss": 0.957, "step": 3975 }, { "epoch": 1.8710588235294119, "grad_norm": 1.0750446869524974, "learning_rate": 0.00027567708822300496, "loss": 0.7734, "step": 3976 }, { "epoch": 1.8715294117647059, "grad_norm": 1.079583446660548, "learning_rate": 0.0002756531524803609, "loss": 0.9961, "step": 3977 }, { "epoch": 1.8719999999999999, "grad_norm": 1.0332124578667918, "learning_rate": 0.00027562920600632127, "loss": 1.0547, "step": 3978 }, { "epoch": 1.872470588235294, "grad_norm": 0.9767461896193359, "learning_rate": 0.0002756052488029312, "loss": 1.0156, "step": 3979 }, { "epoch": 1.8729411764705883, "grad_norm": 1.2489370695831543, "learning_rate": 0.0002755812808722368, "loss": 1.5469, "step": 3980 }, { "epoch": 1.8734117647058823, "grad_norm": 1.1411302020734166, "learning_rate": 0.0002755573022162851, "loss": 0.7344, "step": 3981 }, { "epoch": 1.8738823529411763, "grad_norm": 0.5936659356704228, "learning_rate": 0.0002755333128371239, "loss": 0.1406, "step": 3982 }, { "epoch": 1.8743529411764706, "grad_norm": 1.220537790930372, "learning_rate": 0.00027550931273680207, "loss": 1.3906, "step": 3983 }, { "epoch": 1.8748235294117648, "grad_norm": 0.917339346428009, "learning_rate": 0.00027548530191736926, "loss": 0.6328, "step": 3984 }, { "epoch": 1.8752941176470588, "grad_norm": 0.7124196469284502, "learning_rate": 0.0002754612803808762, "loss": 0.4902, "step": 3985 }, { "epoch": 1.8757647058823528, "grad_norm": 0.8980050345610032, "learning_rate": 0.00027543724812937443, "loss": 0.6172, "step": 3986 }, { "epoch": 1.876235294117647, "grad_norm": 0.685451315702988, "learning_rate": 0.0002754132051649164, "loss": 0.248, "step": 3987 }, { "epoch": 1.8767058823529412, "grad_norm": 1.6645642687580664, "learning_rate": 0.00027538915148955545, "loss": 1.875, "step": 3988 }, { "epoch": 1.8771764705882354, "grad_norm": 1.030540395859696, "learning_rate": 0.000275365087105346, "loss": 0.6016, "step": 3989 }, { "epoch": 1.8776470588235294, "grad_norm": 1.4453424355452225, "learning_rate": 0.00027534101201434315, "loss": 2.0625, "step": 3990 }, { "epoch": 1.8781176470588234, "grad_norm": 1.0341844168469105, "learning_rate": 0.00027531692621860307, "loss": 1.1016, "step": 3991 }, { "epoch": 1.8785882352941177, "grad_norm": 0.7849319432226691, "learning_rate": 0.00027529282972018287, "loss": 0.8438, "step": 3992 }, { "epoch": 1.879058823529412, "grad_norm": 0.8111238720276953, "learning_rate": 0.0002752687225211404, "loss": 0.4922, "step": 3993 }, { "epoch": 1.879529411764706, "grad_norm": 1.4606105623445311, "learning_rate": 0.0002752446046235346, "loss": 2.0781, "step": 3994 }, { "epoch": 1.88, "grad_norm": 0.640157563590919, "learning_rate": 0.00027522047602942525, "loss": 0.2461, "step": 3995 }, { "epoch": 1.8804705882352941, "grad_norm": 2.0171669728060597, "learning_rate": 0.00027519633674087305, "loss": 1.9922, "step": 3996 }, { "epoch": 1.8809411764705883, "grad_norm": 1.0888671372194314, "learning_rate": 0.0002751721867599396, "loss": 0.9375, "step": 3997 }, { "epoch": 1.8814117647058823, "grad_norm": 1.1380816181862055, "learning_rate": 0.00027514802608868743, "loss": 0.8984, "step": 3998 }, { "epoch": 1.8818823529411763, "grad_norm": 1.5050803177729881, "learning_rate": 0.00027512385472918, "loss": 1.8359, "step": 3999 }, { "epoch": 1.8823529411764706, "grad_norm": 0.6666948606662128, "learning_rate": 0.0002750996726834817, "loss": 0.2197, "step": 4000 }, { "epoch": 1.8828235294117648, "grad_norm": 1.238481898075931, "learning_rate": 0.00027507547995365766, "loss": 0.9883, "step": 4001 }, { "epoch": 1.8832941176470588, "grad_norm": 1.437109692534243, "learning_rate": 0.00027505127654177413, "loss": 1.6797, "step": 4002 }, { "epoch": 1.8837647058823528, "grad_norm": 1.2865182127281942, "learning_rate": 0.0002750270624498983, "loss": 1.8906, "step": 4003 }, { "epoch": 1.884235294117647, "grad_norm": 1.4385164253218332, "learning_rate": 0.000275002837680098, "loss": 0.8242, "step": 4004 }, { "epoch": 1.8847058823529412, "grad_norm": 2.6861035214168996, "learning_rate": 0.0002749786022344422, "loss": 3.4219, "step": 4005 }, { "epoch": 1.8851764705882355, "grad_norm": 1.3844254134494802, "learning_rate": 0.0002749543561150009, "loss": 1.4531, "step": 4006 }, { "epoch": 1.8856470588235295, "grad_norm": 0.8377506449858438, "learning_rate": 0.00027493009932384456, "loss": 0.6211, "step": 4007 }, { "epoch": 1.8861176470588235, "grad_norm": 0.9692191086909653, "learning_rate": 0.000274905831863045, "loss": 1.3359, "step": 4008 }, { "epoch": 1.8865882352941177, "grad_norm": 1.3150044506413052, "learning_rate": 0.00027488155373467474, "loss": 1.5312, "step": 4009 }, { "epoch": 1.887058823529412, "grad_norm": 0.9857685426061567, "learning_rate": 0.00027485726494080724, "loss": 1.2109, "step": 4010 }, { "epoch": 1.887529411764706, "grad_norm": 1.1097692423207322, "learning_rate": 0.0002748329654835169, "loss": 1.4141, "step": 4011 }, { "epoch": 1.888, "grad_norm": 1.2425884840314791, "learning_rate": 0.000274808655364879, "loss": 1.7891, "step": 4012 }, { "epoch": 1.8884705882352941, "grad_norm": 1.021926929949127, "learning_rate": 0.00027478433458696976, "loss": 0.4434, "step": 4013 }, { "epoch": 1.8889411764705883, "grad_norm": 1.179565901642856, "learning_rate": 0.0002747600031518663, "loss": 0.6094, "step": 4014 }, { "epoch": 1.8894117647058823, "grad_norm": 1.0646332690179776, "learning_rate": 0.0002747356610616466, "loss": 1.6953, "step": 4015 }, { "epoch": 1.8898823529411763, "grad_norm": 1.483031608850844, "learning_rate": 0.0002747113083183896, "loss": 1.9453, "step": 4016 }, { "epoch": 1.8903529411764706, "grad_norm": 1.3497328791758196, "learning_rate": 0.00027468694492417515, "loss": 1.8984, "step": 4017 }, { "epoch": 1.8908235294117648, "grad_norm": 0.4898616027226316, "learning_rate": 0.0002746625708810841, "loss": 0.3516, "step": 4018 }, { "epoch": 1.8912941176470588, "grad_norm": 0.8169938301051637, "learning_rate": 0.00027463818619119794, "loss": 0.6953, "step": 4019 }, { "epoch": 1.8917647058823528, "grad_norm": 0.9796947782859968, "learning_rate": 0.0002746137908565994, "loss": 1.0781, "step": 4020 }, { "epoch": 1.892235294117647, "grad_norm": 1.1017152842971065, "learning_rate": 0.00027458938487937186, "loss": 1.4609, "step": 4021 }, { "epoch": 1.8927058823529412, "grad_norm": 0.9923847572591938, "learning_rate": 0.0002745649682615998, "loss": 0.9766, "step": 4022 }, { "epoch": 1.8931764705882355, "grad_norm": 1.5647893896759522, "learning_rate": 0.00027454054100536836, "loss": 1.0781, "step": 4023 }, { "epoch": 1.8936470588235295, "grad_norm": 1.1677187200968338, "learning_rate": 0.00027451610311276397, "loss": 1.4688, "step": 4024 }, { "epoch": 1.8941176470588235, "grad_norm": 0.5135848344358352, "learning_rate": 0.00027449165458587354, "loss": 0.1719, "step": 4025 }, { "epoch": 1.8945882352941177, "grad_norm": 1.0477414068643411, "learning_rate": 0.0002744671954267852, "loss": 0.6914, "step": 4026 }, { "epoch": 1.895058823529412, "grad_norm": 1.189302022834611, "learning_rate": 0.00027444272563758784, "loss": 0.9492, "step": 4027 }, { "epoch": 1.895529411764706, "grad_norm": 1.1236380147349705, "learning_rate": 0.00027441824522037137, "loss": 0.7305, "step": 4028 }, { "epoch": 1.896, "grad_norm": 0.7794076728065203, "learning_rate": 0.0002743937541772265, "loss": 0.8164, "step": 4029 }, { "epoch": 1.8964705882352941, "grad_norm": 0.6688056862696329, "learning_rate": 0.0002743692525102448, "loss": 0.8047, "step": 4030 }, { "epoch": 1.8969411764705884, "grad_norm": 0.16611516960295927, "learning_rate": 0.000274344740221519, "loss": 0.127, "step": 4031 }, { "epoch": 1.8974117647058824, "grad_norm": 1.0467896229897662, "learning_rate": 0.0002743202173131424, "loss": 1.0781, "step": 4032 }, { "epoch": 1.8978823529411764, "grad_norm": 0.9437099062148324, "learning_rate": 0.00027429568378720945, "loss": 0.7148, "step": 4033 }, { "epoch": 1.8983529411764706, "grad_norm": 1.7827394727015415, "learning_rate": 0.0002742711396458155, "loss": 2.2812, "step": 4034 }, { "epoch": 1.8988235294117648, "grad_norm": 1.3145093596428086, "learning_rate": 0.00027424658489105667, "loss": 1.0469, "step": 4035 }, { "epoch": 1.8992941176470588, "grad_norm": 0.6701465154015328, "learning_rate": 0.00027422201952503005, "loss": 0.625, "step": 4036 }, { "epoch": 1.8997647058823528, "grad_norm": 1.0192433118703828, "learning_rate": 0.00027419744354983364, "loss": 0.7305, "step": 4037 }, { "epoch": 1.900235294117647, "grad_norm": 1.401014519882843, "learning_rate": 0.00027417285696756634, "loss": 1.9766, "step": 4038 }, { "epoch": 1.9007058823529412, "grad_norm": 0.8727533750456085, "learning_rate": 0.00027414825978032805, "loss": 1.2188, "step": 4039 }, { "epoch": 1.9011764705882352, "grad_norm": 0.930850896120296, "learning_rate": 0.0002741236519902194, "loss": 0.5781, "step": 4040 }, { "epoch": 1.9016470588235292, "grad_norm": 1.8235076477576355, "learning_rate": 0.0002740990335993421, "loss": 2.2188, "step": 4041 }, { "epoch": 1.9021176470588235, "grad_norm": 0.8878537375394169, "learning_rate": 0.00027407440460979854, "loss": 0.543, "step": 4042 }, { "epoch": 1.9025882352941177, "grad_norm": 1.3140723243051986, "learning_rate": 0.0002740497650236923, "loss": 1.625, "step": 4043 }, { "epoch": 1.903058823529412, "grad_norm": 0.5840276519048074, "learning_rate": 0.00027402511484312764, "loss": 0.5938, "step": 4044 }, { "epoch": 1.903529411764706, "grad_norm": 0.7431254056323915, "learning_rate": 0.00027400045407020986, "loss": 0.793, "step": 4045 }, { "epoch": 1.904, "grad_norm": 0.6811627561060747, "learning_rate": 0.00027397578270704504, "loss": 0.3516, "step": 4046 }, { "epoch": 1.9044705882352941, "grad_norm": 1.5562147411891594, "learning_rate": 0.0002739511007557403, "loss": 0.9766, "step": 4047 }, { "epoch": 1.9049411764705884, "grad_norm": 1.099933371443813, "learning_rate": 0.00027392640821840365, "loss": 1.0625, "step": 4048 }, { "epoch": 1.9054117647058824, "grad_norm": 1.5251266060163577, "learning_rate": 0.0002739017050971438, "loss": 1.8203, "step": 4049 }, { "epoch": 1.9058823529411764, "grad_norm": 1.1076090293379586, "learning_rate": 0.0002738769913940706, "loss": 1.25, "step": 4050 }, { "epoch": 1.9063529411764706, "grad_norm": 0.970288164556195, "learning_rate": 0.00027385226711129477, "loss": 1.2188, "step": 4051 }, { "epoch": 1.9068235294117648, "grad_norm": 1.4737268251113456, "learning_rate": 0.00027382753225092777, "loss": 1.4062, "step": 4052 }, { "epoch": 1.9072941176470588, "grad_norm": 0.6798023096709718, "learning_rate": 0.0002738027868150822, "loss": 0.7109, "step": 4053 }, { "epoch": 1.9077647058823528, "grad_norm": 0.6938244154030225, "learning_rate": 0.0002737780308058714, "loss": 0.5586, "step": 4054 }, { "epoch": 1.908235294117647, "grad_norm": 1.1404294747622477, "learning_rate": 0.00027375326422540957, "loss": 1.0391, "step": 4055 }, { "epoch": 1.9087058823529413, "grad_norm": 1.1231540890298908, "learning_rate": 0.000273728487075812, "loss": 0.9492, "step": 4056 }, { "epoch": 1.9091764705882353, "grad_norm": 0.6142215993351777, "learning_rate": 0.0002737036993591948, "loss": 0.3281, "step": 4057 }, { "epoch": 1.9096470588235293, "grad_norm": 1.2231550171957077, "learning_rate": 0.0002736789010776749, "loss": 0.9922, "step": 4058 }, { "epoch": 1.9101176470588235, "grad_norm": 0.7400206700037266, "learning_rate": 0.00027365409223337016, "loss": 0.6953, "step": 4059 }, { "epoch": 1.9105882352941177, "grad_norm": 0.8128146407537298, "learning_rate": 0.0002736292728283994, "loss": 0.957, "step": 4060 }, { "epoch": 1.911058823529412, "grad_norm": 1.0897742892355204, "learning_rate": 0.0002736044428648824, "loss": 1.4219, "step": 4061 }, { "epoch": 1.911529411764706, "grad_norm": 1.0369320341696504, "learning_rate": 0.0002735796023449397, "loss": 0.8125, "step": 4062 }, { "epoch": 1.912, "grad_norm": 0.7200489144338281, "learning_rate": 0.00027355475127069275, "loss": 0.334, "step": 4063 }, { "epoch": 1.9124705882352941, "grad_norm": 0.5643611751296522, "learning_rate": 0.0002735298896442641, "loss": 0.2275, "step": 4064 }, { "epoch": 1.9129411764705884, "grad_norm": 0.8211363510163983, "learning_rate": 0.0002735050174677769, "loss": 0.6172, "step": 4065 }, { "epoch": 1.9134117647058824, "grad_norm": 1.2591400767637477, "learning_rate": 0.00027348013474335546, "loss": 1.3828, "step": 4066 }, { "epoch": 1.9138823529411764, "grad_norm": 1.1199244615804163, "learning_rate": 0.0002734552414731248, "loss": 0.7383, "step": 4067 }, { "epoch": 1.9143529411764706, "grad_norm": 1.1456184552915998, "learning_rate": 0.000273430337659211, "loss": 1.5391, "step": 4068 }, { "epoch": 1.9148235294117648, "grad_norm": 0.9270370626831935, "learning_rate": 0.0002734054233037409, "loss": 0.5156, "step": 4069 }, { "epoch": 1.9152941176470588, "grad_norm": 1.3079860916862194, "learning_rate": 0.00027338049840884245, "loss": 0.9531, "step": 4070 }, { "epoch": 1.9157647058823528, "grad_norm": 0.9394758985082979, "learning_rate": 0.00027335556297664417, "loss": 0.9258, "step": 4071 }, { "epoch": 1.916235294117647, "grad_norm": 1.0254360620761902, "learning_rate": 0.00027333061700927577, "loss": 0.8477, "step": 4072 }, { "epoch": 1.9167058823529413, "grad_norm": 0.6951827055494006, "learning_rate": 0.00027330566050886776, "loss": 0.3164, "step": 4073 }, { "epoch": 1.9171764705882353, "grad_norm": 0.6727676042704157, "learning_rate": 0.0002732806934775515, "loss": 0.3086, "step": 4074 }, { "epoch": 1.9176470588235293, "grad_norm": 1.3152318472172335, "learning_rate": 0.0002732557159174594, "loss": 1.6797, "step": 4075 }, { "epoch": 1.9181176470588235, "grad_norm": 0.850984806717762, "learning_rate": 0.0002732307278307246, "loss": 0.6367, "step": 4076 }, { "epoch": 1.9185882352941177, "grad_norm": 0.9371598374460121, "learning_rate": 0.0002732057292194811, "loss": 0.6484, "step": 4077 }, { "epoch": 1.919058823529412, "grad_norm": 0.7386227499655545, "learning_rate": 0.000273180720085864, "loss": 0.3848, "step": 4078 }, { "epoch": 1.919529411764706, "grad_norm": 1.1026166626040896, "learning_rate": 0.0002731557004320093, "loss": 0.6797, "step": 4079 }, { "epoch": 1.92, "grad_norm": 1.2688159489277275, "learning_rate": 0.0002731306702600537, "loss": 1.5781, "step": 4080 }, { "epoch": 1.9204705882352942, "grad_norm": 0.5787130261221346, "learning_rate": 0.00027310562957213487, "loss": 0.4336, "step": 4081 }, { "epoch": 1.9209411764705884, "grad_norm": 1.580728212024198, "learning_rate": 0.00027308057837039144, "loss": 1.0, "step": 4082 }, { "epoch": 1.9214117647058824, "grad_norm": 0.7615779624253786, "learning_rate": 0.0002730555166569629, "loss": 0.6875, "step": 4083 }, { "epoch": 1.9218823529411764, "grad_norm": 1.1723608742550573, "learning_rate": 0.00027303044443398967, "loss": 1.4766, "step": 4084 }, { "epoch": 1.9223529411764706, "grad_norm": 1.200170913771378, "learning_rate": 0.00027300536170361307, "loss": 1.4609, "step": 4085 }, { "epoch": 1.9228235294117648, "grad_norm": 1.2471831046675743, "learning_rate": 0.00027298026846797515, "loss": 1.8906, "step": 4086 }, { "epoch": 1.9232941176470588, "grad_norm": 1.2593660807134623, "learning_rate": 0.0002729551647292191, "loss": 1.25, "step": 4087 }, { "epoch": 1.9237647058823528, "grad_norm": 1.0110029239506682, "learning_rate": 0.000272930050489489, "loss": 0.6719, "step": 4088 }, { "epoch": 1.924235294117647, "grad_norm": 1.0299277134944071, "learning_rate": 0.0002729049257509295, "loss": 1.5078, "step": 4089 }, { "epoch": 1.9247058823529413, "grad_norm": 1.0230299544057408, "learning_rate": 0.0002728797905156866, "loss": 1.1484, "step": 4090 }, { "epoch": 1.9251764705882353, "grad_norm": 0.7074653546810896, "learning_rate": 0.0002728546447859068, "loss": 0.4043, "step": 4091 }, { "epoch": 1.9256470588235293, "grad_norm": 0.614125657528015, "learning_rate": 0.00027282948856373774, "loss": 0.2139, "step": 4092 }, { "epoch": 1.9261176470588235, "grad_norm": 1.1585026271266465, "learning_rate": 0.0002728043218513279, "loss": 1.3906, "step": 4093 }, { "epoch": 1.9265882352941177, "grad_norm": 1.0485012316772366, "learning_rate": 0.0002727791446508266, "loss": 0.957, "step": 4094 }, { "epoch": 1.9270588235294117, "grad_norm": 0.7970608458245548, "learning_rate": 0.00027275395696438413, "loss": 0.3047, "step": 4095 }, { "epoch": 1.927529411764706, "grad_norm": 1.6631019357972219, "learning_rate": 0.0002727287587941516, "loss": 1.4922, "step": 4096 }, { "epoch": 1.928, "grad_norm": 0.6903341182434755, "learning_rate": 0.00027270355014228113, "loss": 0.332, "step": 4097 }, { "epoch": 1.9284705882352942, "grad_norm": 1.6398031116418095, "learning_rate": 0.00027267833101092564, "loss": 1.9531, "step": 4098 }, { "epoch": 1.9289411764705884, "grad_norm": 1.3067145298989369, "learning_rate": 0.0002726531014022389, "loss": 1.8359, "step": 4099 }, { "epoch": 1.9294117647058824, "grad_norm": 1.2170355793401277, "learning_rate": 0.00027262786131837573, "loss": 1.5859, "step": 4100 }, { "epoch": 1.9298823529411764, "grad_norm": 1.458819975910836, "learning_rate": 0.00027260261076149166, "loss": 1.2188, "step": 4101 }, { "epoch": 1.9303529411764706, "grad_norm": 1.4715657003315639, "learning_rate": 0.00027257734973374335, "loss": 1.875, "step": 4102 }, { "epoch": 1.9308235294117648, "grad_norm": 0.8211117480287785, "learning_rate": 0.0002725520782372881, "loss": 0.5273, "step": 4103 }, { "epoch": 1.9312941176470588, "grad_norm": 0.653419533824702, "learning_rate": 0.00027252679627428423, "loss": 0.3516, "step": 4104 }, { "epoch": 1.9317647058823528, "grad_norm": 1.0934120360694706, "learning_rate": 0.00027250150384689094, "loss": 1.3438, "step": 4105 }, { "epoch": 1.932235294117647, "grad_norm": 1.3062487142609251, "learning_rate": 0.0002724762009572684, "loss": 1.4609, "step": 4106 }, { "epoch": 1.9327058823529413, "grad_norm": 1.1001882202213396, "learning_rate": 0.0002724508876075776, "loss": 1.3281, "step": 4107 }, { "epoch": 1.9331764705882353, "grad_norm": 0.679361003123569, "learning_rate": 0.0002724255637999803, "loss": 0.3359, "step": 4108 }, { "epoch": 1.9336470588235293, "grad_norm": 0.43381667309947824, "learning_rate": 0.0002724002295366394, "loss": 0.2334, "step": 4109 }, { "epoch": 1.9341176470588235, "grad_norm": 0.7598191401382939, "learning_rate": 0.0002723748848197185, "loss": 0.4863, "step": 4110 }, { "epoch": 1.9345882352941177, "grad_norm": 0.9002644875062408, "learning_rate": 0.00027234952965138223, "loss": 1.0469, "step": 4111 }, { "epoch": 1.9350588235294117, "grad_norm": 1.1112306134371628, "learning_rate": 0.000272324164033796, "loss": 1.2266, "step": 4112 }, { "epoch": 1.9355294117647057, "grad_norm": 1.5143725949950224, "learning_rate": 0.0002722987879691261, "loss": 1.9219, "step": 4113 }, { "epoch": 1.936, "grad_norm": 1.06771927633018, "learning_rate": 0.00027227340145953994, "loss": 1.1797, "step": 4114 }, { "epoch": 1.9364705882352942, "grad_norm": 0.6656119398619964, "learning_rate": 0.0002722480045072055, "loss": 0.2471, "step": 4115 }, { "epoch": 1.9369411764705884, "grad_norm": 0.9751489155666987, "learning_rate": 0.0002722225971142918, "loss": 1.0859, "step": 4116 }, { "epoch": 1.9374117647058824, "grad_norm": 1.0824124925065048, "learning_rate": 0.00027219717928296885, "loss": 0.8008, "step": 4117 }, { "epoch": 1.9378823529411764, "grad_norm": 0.9447352076589771, "learning_rate": 0.0002721717510154074, "loss": 0.875, "step": 4118 }, { "epoch": 1.9383529411764706, "grad_norm": 1.0005002061879558, "learning_rate": 0.0002721463123137792, "loss": 0.8008, "step": 4119 }, { "epoch": 1.9388235294117648, "grad_norm": 0.6640893766353761, "learning_rate": 0.0002721208631802568, "loss": 0.2471, "step": 4120 }, { "epoch": 1.9392941176470588, "grad_norm": 1.0065405748085499, "learning_rate": 0.00027209540361701365, "loss": 1.2266, "step": 4121 }, { "epoch": 1.9397647058823528, "grad_norm": 0.8525298573346426, "learning_rate": 0.0002720699336262242, "loss": 0.7812, "step": 4122 }, { "epoch": 1.940235294117647, "grad_norm": 1.165372503345957, "learning_rate": 0.00027204445321006365, "loss": 1.6094, "step": 4123 }, { "epoch": 1.9407058823529413, "grad_norm": 2.021738164277387, "learning_rate": 0.00027201896237070815, "loss": 0.7656, "step": 4124 }, { "epoch": 1.9411764705882353, "grad_norm": 0.8287382856465225, "learning_rate": 0.0002719934611103348, "loss": 0.3809, "step": 4125 }, { "epoch": 1.9416470588235293, "grad_norm": 1.1240228032030726, "learning_rate": 0.00027196794943112147, "loss": 1.9922, "step": 4126 }, { "epoch": 1.9421176470588235, "grad_norm": 1.104664908218759, "learning_rate": 0.000271942427335247, "loss": 1.2344, "step": 4127 }, { "epoch": 1.9425882352941177, "grad_norm": 2.2471642666256355, "learning_rate": 0.00027191689482489114, "loss": 1.2656, "step": 4128 }, { "epoch": 1.9430588235294117, "grad_norm": 0.5805783454511215, "learning_rate": 0.0002718913519022344, "loss": 0.2773, "step": 4129 }, { "epoch": 1.9435294117647057, "grad_norm": 0.5476611252173547, "learning_rate": 0.00027186579856945843, "loss": 0.1973, "step": 4130 }, { "epoch": 1.944, "grad_norm": 0.8889036150874698, "learning_rate": 0.00027184023482874545, "loss": 0.9766, "step": 4131 }, { "epoch": 1.9444705882352942, "grad_norm": 0.8048032322417554, "learning_rate": 0.00027181466068227885, "loss": 0.8594, "step": 4132 }, { "epoch": 1.9449411764705884, "grad_norm": 1.3962188582492665, "learning_rate": 0.00027178907613224264, "loss": 0.9961, "step": 4133 }, { "epoch": 1.9454117647058824, "grad_norm": 0.6974275151565515, "learning_rate": 0.00027176348118082203, "loss": 0.4277, "step": 4134 }, { "epoch": 1.9458823529411764, "grad_norm": 0.559673353366001, "learning_rate": 0.0002717378758302029, "loss": 0.4668, "step": 4135 }, { "epoch": 1.9463529411764706, "grad_norm": 1.1220039044558416, "learning_rate": 0.000271712260082572, "loss": 1.1953, "step": 4136 }, { "epoch": 1.9468235294117648, "grad_norm": 1.0379052001512905, "learning_rate": 0.0002716866339401171, "loss": 1.3125, "step": 4137 }, { "epoch": 1.9472941176470588, "grad_norm": 0.5695371006714197, "learning_rate": 0.0002716609974050268, "loss": 0.2178, "step": 4138 }, { "epoch": 1.9477647058823528, "grad_norm": 1.3530142823055553, "learning_rate": 0.0002716353504794906, "loss": 1.4922, "step": 4139 }, { "epoch": 1.948235294117647, "grad_norm": 1.0773764462442308, "learning_rate": 0.0002716096931656988, "loss": 0.9648, "step": 4140 }, { "epoch": 1.9487058823529413, "grad_norm": 1.1915611323807584, "learning_rate": 0.0002715840254658427, "loss": 1.3281, "step": 4141 }, { "epoch": 1.9491764705882353, "grad_norm": 0.8094228851586135, "learning_rate": 0.0002715583473821145, "loss": 0.5195, "step": 4142 }, { "epoch": 1.9496470588235293, "grad_norm": 0.8796327090935224, "learning_rate": 0.0002715326589167072, "loss": 1.125, "step": 4143 }, { "epoch": 1.9501176470588235, "grad_norm": 0.8516493958182232, "learning_rate": 0.0002715069600718147, "loss": 0.3418, "step": 4144 }, { "epoch": 1.9505882352941177, "grad_norm": 1.0546212965777368, "learning_rate": 0.0002714812508496319, "loss": 1.2578, "step": 4145 }, { "epoch": 1.9510588235294117, "grad_norm": 1.1760533433338511, "learning_rate": 0.0002714555312523543, "loss": 1.4062, "step": 4146 }, { "epoch": 1.9515294117647057, "grad_norm": 0.9464356758466309, "learning_rate": 0.00027142980128217864, "loss": 1.125, "step": 4147 }, { "epoch": 1.952, "grad_norm": 0.6215812715301833, "learning_rate": 0.00027140406094130234, "loss": 0.4395, "step": 4148 }, { "epoch": 1.9524705882352942, "grad_norm": 1.6549681157898617, "learning_rate": 0.0002713783102319237, "loss": 2.5, "step": 4149 }, { "epoch": 1.9529411764705882, "grad_norm": 0.8853357268223936, "learning_rate": 0.0002713525491562421, "loss": 1.1406, "step": 4150 }, { "epoch": 1.9534117647058824, "grad_norm": 0.6669558971225102, "learning_rate": 0.0002713267777164575, "loss": 0.7969, "step": 4151 }, { "epoch": 1.9538823529411764, "grad_norm": 0.727553611549762, "learning_rate": 0.00027130099591477104, "loss": 0.3926, "step": 4152 }, { "epoch": 1.9543529411764706, "grad_norm": 0.7975431398352464, "learning_rate": 0.0002712752037533845, "loss": 0.4648, "step": 4153 }, { "epoch": 1.9548235294117648, "grad_norm": 1.3062784741022104, "learning_rate": 0.00027124940123450075, "loss": 1.5312, "step": 4154 }, { "epoch": 1.9552941176470588, "grad_norm": 0.986869995667415, "learning_rate": 0.00027122358836032337, "loss": 1.5234, "step": 4155 }, { "epoch": 1.9557647058823528, "grad_norm": 0.9838735903931651, "learning_rate": 0.000271197765133057, "loss": 1.2891, "step": 4156 }, { "epoch": 1.956235294117647, "grad_norm": 0.9619247374653455, "learning_rate": 0.00027117193155490695, "loss": 1.1875, "step": 4157 }, { "epoch": 1.9567058823529413, "grad_norm": 1.3893407245984528, "learning_rate": 0.00027114608762807965, "loss": 1.9375, "step": 4158 }, { "epoch": 1.9571764705882353, "grad_norm": 0.5901271614248901, "learning_rate": 0.0002711202333547822, "loss": 0.1963, "step": 4159 }, { "epoch": 1.9576470588235293, "grad_norm": 1.616090368379424, "learning_rate": 0.00027109436873722277, "loss": 0.2432, "step": 4160 }, { "epoch": 1.9581176470588235, "grad_norm": 0.905406215996713, "learning_rate": 0.00027106849377761023, "loss": 0.7852, "step": 4161 }, { "epoch": 1.9585882352941177, "grad_norm": 0.9523095153453118, "learning_rate": 0.00027104260847815457, "loss": 1.1094, "step": 4162 }, { "epoch": 1.9590588235294117, "grad_norm": 1.1535591093966775, "learning_rate": 0.0002710167128410664, "loss": 1.4297, "step": 4163 }, { "epoch": 1.9595294117647057, "grad_norm": 0.662749524225992, "learning_rate": 0.00027099080686855735, "loss": 0.5938, "step": 4164 }, { "epoch": 1.96, "grad_norm": 1.5272886353539896, "learning_rate": 0.00027096489056283995, "loss": 1.6562, "step": 4165 }, { "epoch": 1.9604705882352942, "grad_norm": 0.8985063554820772, "learning_rate": 0.00027093896392612753, "loss": 0.6289, "step": 4166 }, { "epoch": 1.9609411764705882, "grad_norm": 1.2238883340384235, "learning_rate": 0.0002709130269606345, "loss": 1.0625, "step": 4167 }, { "epoch": 1.9614117647058822, "grad_norm": 1.2828546905229643, "learning_rate": 0.00027088707966857576, "loss": 1.6719, "step": 4168 }, { "epoch": 1.9618823529411764, "grad_norm": 0.9867687103380545, "learning_rate": 0.0002708611220521675, "loss": 1.3516, "step": 4169 }, { "epoch": 1.9623529411764706, "grad_norm": 1.0874805748364949, "learning_rate": 0.0002708351541136267, "loss": 1.0469, "step": 4170 }, { "epoch": 1.9628235294117649, "grad_norm": 1.1769901170911132, "learning_rate": 0.0002708091758551709, "loss": 1.6484, "step": 4171 }, { "epoch": 1.9632941176470589, "grad_norm": 0.6933517419035122, "learning_rate": 0.000270783187279019, "loss": 0.4746, "step": 4172 }, { "epoch": 1.9637647058823529, "grad_norm": 1.8594029681213922, "learning_rate": 0.00027075718838739043, "loss": 2.1094, "step": 4173 }, { "epoch": 1.964235294117647, "grad_norm": 0.9570414770830294, "learning_rate": 0.0002707311791825057, "loss": 0.832, "step": 4174 }, { "epoch": 1.9647058823529413, "grad_norm": 0.9627439043011924, "learning_rate": 0.00027070515966658604, "loss": 0.668, "step": 4175 }, { "epoch": 1.9651764705882353, "grad_norm": 1.3035672581953128, "learning_rate": 0.00027067912984185376, "loss": 1.0938, "step": 4176 }, { "epoch": 1.9656470588235293, "grad_norm": 0.7823798374683083, "learning_rate": 0.0002706530897105318, "loss": 0.9844, "step": 4177 }, { "epoch": 1.9661176470588235, "grad_norm": 0.9720308199694988, "learning_rate": 0.0002706270392748442, "loss": 0.8125, "step": 4178 }, { "epoch": 1.9665882352941177, "grad_norm": 0.7815279924797023, "learning_rate": 0.00027060097853701576, "loss": 0.5, "step": 4179 }, { "epoch": 1.9670588235294117, "grad_norm": 1.0403263364592423, "learning_rate": 0.0002705749074992722, "loss": 0.7188, "step": 4180 }, { "epoch": 1.9675294117647057, "grad_norm": 1.0790857564283864, "learning_rate": 0.00027054882616384015, "loss": 0.7461, "step": 4181 }, { "epoch": 1.968, "grad_norm": 1.0613975204684776, "learning_rate": 0.0002705227345329471, "loss": 1.0078, "step": 4182 }, { "epoch": 1.9684705882352942, "grad_norm": 1.1473597065321464, "learning_rate": 0.00027049663260882124, "loss": 1.1172, "step": 4183 }, { "epoch": 1.9689411764705882, "grad_norm": 0.9464141610827302, "learning_rate": 0.00027047052039369204, "loss": 0.7188, "step": 4184 }, { "epoch": 1.9694117647058822, "grad_norm": 1.5932824862178125, "learning_rate": 0.0002704443978897894, "loss": 1.7734, "step": 4185 }, { "epoch": 1.9698823529411764, "grad_norm": 1.4491897332443093, "learning_rate": 0.0002704182650993445, "loss": 0.9102, "step": 4186 }, { "epoch": 1.9703529411764706, "grad_norm": 1.0148497154940803, "learning_rate": 0.00027039212202458905, "loss": 1.2734, "step": 4187 }, { "epoch": 1.9708235294117649, "grad_norm": 0.935368542881783, "learning_rate": 0.0002703659686677559, "loss": 0.625, "step": 4188 }, { "epoch": 1.9712941176470589, "grad_norm": 0.44429570494370096, "learning_rate": 0.00027033980503107856, "loss": 0.1118, "step": 4189 }, { "epoch": 1.9717647058823529, "grad_norm": 0.9491770197200504, "learning_rate": 0.00027031363111679164, "loss": 0.8594, "step": 4190 }, { "epoch": 1.972235294117647, "grad_norm": 1.209455324844697, "learning_rate": 0.0002702874469271305, "loss": 1.4688, "step": 4191 }, { "epoch": 1.9727058823529413, "grad_norm": 0.9452607495771826, "learning_rate": 0.00027026125246433144, "loss": 0.6172, "step": 4192 }, { "epoch": 1.9731764705882353, "grad_norm": 0.8219420609620143, "learning_rate": 0.0002702350477306315, "loss": 0.457, "step": 4193 }, { "epoch": 1.9736470588235293, "grad_norm": 1.1728200694174, "learning_rate": 0.0002702088327282687, "loss": 1.3906, "step": 4194 }, { "epoch": 1.9741176470588235, "grad_norm": 0.8284975740862921, "learning_rate": 0.00027018260745948196, "loss": 0.4141, "step": 4195 }, { "epoch": 1.9745882352941178, "grad_norm": 1.1899851794748937, "learning_rate": 0.0002701563719265111, "loss": 1.2891, "step": 4196 }, { "epoch": 1.9750588235294118, "grad_norm": 0.696102164013969, "learning_rate": 0.0002701301261315967, "loss": 0.4434, "step": 4197 }, { "epoch": 1.9755294117647058, "grad_norm": 1.8796542464138766, "learning_rate": 0.0002701038700769803, "loss": 2.1406, "step": 4198 }, { "epoch": 1.976, "grad_norm": 0.5375158208225269, "learning_rate": 0.00027007760376490425, "loss": 0.2598, "step": 4199 }, { "epoch": 1.9764705882352942, "grad_norm": 0.5809129579926741, "learning_rate": 0.0002700513271976119, "loss": 0.4961, "step": 4200 }, { "epoch": 1.9769411764705882, "grad_norm": 0.7604265873189162, "learning_rate": 0.0002700250403773473, "loss": 0.6094, "step": 4201 }, { "epoch": 1.9774117647058822, "grad_norm": 0.852800903065788, "learning_rate": 0.0002699987433063556, "loss": 0.5195, "step": 4202 }, { "epoch": 1.9778823529411764, "grad_norm": 1.0575907535498763, "learning_rate": 0.0002699724359868826, "loss": 1.3828, "step": 4203 }, { "epoch": 1.9783529411764706, "grad_norm": 0.7385303022326929, "learning_rate": 0.0002699461184211751, "loss": 0.3027, "step": 4204 }, { "epoch": 1.9788235294117649, "grad_norm": 0.6225437506384154, "learning_rate": 0.00026991979061148077, "loss": 0.3164, "step": 4205 }, { "epoch": 1.9792941176470589, "grad_norm": 1.0910658619388605, "learning_rate": 0.00026989345256004814, "loss": 1.0156, "step": 4206 }, { "epoch": 1.9797647058823529, "grad_norm": 1.1232674413400854, "learning_rate": 0.00026986710426912656, "loss": 1.6016, "step": 4207 }, { "epoch": 1.980235294117647, "grad_norm": 0.8717275360055781, "learning_rate": 0.0002698407457409663, "loss": 0.6055, "step": 4208 }, { "epoch": 1.9807058823529413, "grad_norm": 0.9787043503180378, "learning_rate": 0.00026981437697781854, "loss": 0.6797, "step": 4209 }, { "epoch": 1.9811764705882353, "grad_norm": 0.9771449037991996, "learning_rate": 0.0002697879979819354, "loss": 1.1797, "step": 4210 }, { "epoch": 1.9816470588235293, "grad_norm": 0.8955985782803344, "learning_rate": 0.0002697616087555696, "loss": 0.8711, "step": 4211 }, { "epoch": 1.9821176470588235, "grad_norm": 0.9791588699693676, "learning_rate": 0.00026973520930097494, "loss": 0.8789, "step": 4212 }, { "epoch": 1.9825882352941178, "grad_norm": 0.600398086901933, "learning_rate": 0.0002697087996204062, "loss": 0.249, "step": 4213 }, { "epoch": 1.9830588235294118, "grad_norm": 0.9538657300405488, "learning_rate": 0.0002696823797161187, "loss": 1.0703, "step": 4214 }, { "epoch": 1.9835294117647058, "grad_norm": 1.0364795908527764, "learning_rate": 0.00026965594959036906, "loss": 0.8711, "step": 4215 }, { "epoch": 1.984, "grad_norm": 1.012918144782757, "learning_rate": 0.00026962950924541435, "loss": 1.1641, "step": 4216 }, { "epoch": 1.9844705882352942, "grad_norm": 0.8854120915164454, "learning_rate": 0.0002696030586835128, "loss": 0.6758, "step": 4217 }, { "epoch": 1.9849411764705882, "grad_norm": 0.8363269888643803, "learning_rate": 0.00026957659790692335, "loss": 1.0859, "step": 4218 }, { "epoch": 1.9854117647058822, "grad_norm": 0.790325930279655, "learning_rate": 0.0002695501269179059, "loss": 0.4219, "step": 4219 }, { "epoch": 1.9858823529411764, "grad_norm": 0.7117121431562414, "learning_rate": 0.00026952364571872125, "loss": 0.2334, "step": 4220 }, { "epoch": 1.9863529411764707, "grad_norm": 0.5034666007859006, "learning_rate": 0.000269497154311631, "loss": 0.3613, "step": 4221 }, { "epoch": 1.9868235294117647, "grad_norm": 0.9909788432566874, "learning_rate": 0.00026947065269889765, "loss": 0.7305, "step": 4222 }, { "epoch": 1.9872941176470589, "grad_norm": 1.066909001437766, "learning_rate": 0.00026944414088278453, "loss": 1.1953, "step": 4223 }, { "epoch": 1.9877647058823529, "grad_norm": 0.4734513706823109, "learning_rate": 0.00026941761886555595, "loss": 0.1592, "step": 4224 }, { "epoch": 1.988235294117647, "grad_norm": 1.424926794041235, "learning_rate": 0.00026939108664947694, "loss": 2.0156, "step": 4225 }, { "epoch": 1.9887058823529413, "grad_norm": 0.8166245764571286, "learning_rate": 0.0002693645442368135, "loss": 0.5781, "step": 4226 }, { "epoch": 1.9891764705882353, "grad_norm": 1.03079110195453, "learning_rate": 0.0002693379916298326, "loss": 1.2109, "step": 4227 }, { "epoch": 1.9896470588235293, "grad_norm": 0.7025796765600762, "learning_rate": 0.00026931142883080176, "loss": 0.416, "step": 4228 }, { "epoch": 1.9901176470588235, "grad_norm": 1.04920635226721, "learning_rate": 0.00026928485584198973, "loss": 0.4844, "step": 4229 }, { "epoch": 1.9905882352941178, "grad_norm": 1.2815078782693885, "learning_rate": 0.00026925827266566595, "loss": 1.5547, "step": 4230 }, { "epoch": 1.9910588235294118, "grad_norm": 0.6314142728244282, "learning_rate": 0.0002692316793041007, "loss": 0.3711, "step": 4231 }, { "epoch": 1.9915294117647058, "grad_norm": 0.8196152514069673, "learning_rate": 0.0002692050757595652, "loss": 0.4473, "step": 4232 }, { "epoch": 1.992, "grad_norm": 1.237492128082269, "learning_rate": 0.00026917846203433155, "loss": 0.9453, "step": 4233 }, { "epoch": 1.9924705882352942, "grad_norm": 1.5156382474256758, "learning_rate": 0.00026915183813067264, "loss": 1.375, "step": 4234 }, { "epoch": 1.9929411764705882, "grad_norm": 1.0771577514156572, "learning_rate": 0.0002691252040508623, "loss": 1.1172, "step": 4235 }, { "epoch": 1.9934117647058822, "grad_norm": 0.8491323824316365, "learning_rate": 0.0002690985597971753, "loss": 0.3281, "step": 4236 }, { "epoch": 1.9938823529411764, "grad_norm": 1.1786726757847674, "learning_rate": 0.00026907190537188707, "loss": 1.2578, "step": 4237 }, { "epoch": 1.9943529411764707, "grad_norm": 1.2775364037224517, "learning_rate": 0.0002690452407772741, "loss": 0.7891, "step": 4238 }, { "epoch": 1.9948235294117647, "grad_norm": 0.986962298365457, "learning_rate": 0.0002690185660156136, "loss": 0.7188, "step": 4239 }, { "epoch": 1.9952941176470587, "grad_norm": 0.8213967102656758, "learning_rate": 0.00026899188108918387, "loss": 0.668, "step": 4240 }, { "epoch": 1.9957647058823529, "grad_norm": 1.3010480076116202, "learning_rate": 0.00026896518600026377, "loss": 1.5859, "step": 4241 }, { "epoch": 1.996235294117647, "grad_norm": 1.2590636176377354, "learning_rate": 0.00026893848075113327, "loss": 1.75, "step": 4242 }, { "epoch": 1.9967058823529413, "grad_norm": 1.0059305637001617, "learning_rate": 0.00026891176534407315, "loss": 1.7344, "step": 4243 }, { "epoch": 1.9971764705882353, "grad_norm": 1.1634138145625268, "learning_rate": 0.00026888503978136506, "loss": 1.2891, "step": 4244 }, { "epoch": 1.9976470588235293, "grad_norm": 0.7753494372070615, "learning_rate": 0.00026885830406529136, "loss": 0.707, "step": 4245 }, { "epoch": 1.9981176470588236, "grad_norm": 1.0101041999741047, "learning_rate": 0.0002688315581981355, "loss": 0.7969, "step": 4246 }, { "epoch": 1.9985882352941178, "grad_norm": 1.071396473448802, "learning_rate": 0.0002688048021821817, "loss": 1.4141, "step": 4247 }, { "epoch": 1.9990588235294118, "grad_norm": 1.188883912500722, "learning_rate": 0.00026877803601971516, "loss": 1.6406, "step": 4248 }, { "epoch": 1.9995294117647058, "grad_norm": 1.3680762140511027, "learning_rate": 0.00026875125971302167, "loss": 1.4453, "step": 4249 }, { "epoch": 2.0, "grad_norm": 1.1387793118980234, "learning_rate": 0.0002687244732643881, "loss": 1.0703, "step": 4250 }, { "epoch": 2.0, "eval_loss": 0.9545533657073975, "eval_runtime": 36.9874, "eval_samples_per_second": 20.277, "eval_steps_per_second": 10.139, "step": 4250 }, { "epoch": 2.0004705882352942, "grad_norm": 0.9273105415169562, "learning_rate": 0.0002686976766761022, "loss": 0.5625, "step": 4251 }, { "epoch": 2.0009411764705884, "grad_norm": 0.7804777045065583, "learning_rate": 0.0002686708699504525, "loss": 0.6211, "step": 4252 }, { "epoch": 2.0014117647058822, "grad_norm": 0.798376268531579, "learning_rate": 0.0002686440530897284, "loss": 0.4629, "step": 4253 }, { "epoch": 2.0018823529411764, "grad_norm": 1.2905712462940668, "learning_rate": 0.00026861722609622027, "loss": 1.4141, "step": 4254 }, { "epoch": 2.0023529411764707, "grad_norm": 0.557288583496407, "learning_rate": 0.0002685903889722192, "loss": 0.5742, "step": 4255 }, { "epoch": 2.002823529411765, "grad_norm": 0.7708605400616314, "learning_rate": 0.00026856354172001724, "loss": 0.7578, "step": 4256 }, { "epoch": 2.0032941176470587, "grad_norm": 1.3559512017533504, "learning_rate": 0.0002685366843419073, "loss": 1.4609, "step": 4257 }, { "epoch": 2.003764705882353, "grad_norm": 1.2686669874265506, "learning_rate": 0.000268509816840183, "loss": 0.7031, "step": 4258 }, { "epoch": 2.004235294117647, "grad_norm": 0.7493663022021233, "learning_rate": 0.00026848293921713913, "loss": 0.5352, "step": 4259 }, { "epoch": 2.0047058823529413, "grad_norm": 0.9402386626103859, "learning_rate": 0.0002684560514750711, "loss": 0.5508, "step": 4260 }, { "epoch": 2.005176470588235, "grad_norm": 1.5556734041972833, "learning_rate": 0.00026842915361627523, "loss": 1.2656, "step": 4261 }, { "epoch": 2.0056470588235293, "grad_norm": 0.8962448847456577, "learning_rate": 0.0002684022456430487, "loss": 0.9961, "step": 4262 }, { "epoch": 2.0061176470588236, "grad_norm": 0.7223198656320702, "learning_rate": 0.0002683753275576897, "loss": 0.3555, "step": 4263 }, { "epoch": 2.006588235294118, "grad_norm": 1.1609722837142744, "learning_rate": 0.0002683483993624971, "loss": 1.5156, "step": 4264 }, { "epoch": 2.0070588235294116, "grad_norm": 0.6646741570467576, "learning_rate": 0.0002683214610597706, "loss": 0.3164, "step": 4265 }, { "epoch": 2.007529411764706, "grad_norm": 0.8694116533440814, "learning_rate": 0.0002682945126518111, "loss": 0.6562, "step": 4266 }, { "epoch": 2.008, "grad_norm": 1.0325178415185288, "learning_rate": 0.00026826755414091993, "loss": 1.3594, "step": 4267 }, { "epoch": 2.0084705882352942, "grad_norm": 1.2982585987331128, "learning_rate": 0.00026824058552939954, "loss": 0.9102, "step": 4268 }, { "epoch": 2.0089411764705885, "grad_norm": 1.0472333578978263, "learning_rate": 0.00026821360681955316, "loss": 0.9141, "step": 4269 }, { "epoch": 2.0094117647058822, "grad_norm": 0.8180975892740585, "learning_rate": 0.00026818661801368494, "loss": 1.0312, "step": 4270 }, { "epoch": 2.0098823529411765, "grad_norm": 1.0826756264451385, "learning_rate": 0.0002681596191140998, "loss": 1.0938, "step": 4271 }, { "epoch": 2.0103529411764707, "grad_norm": 0.4008685713820498, "learning_rate": 0.00026813261012310365, "loss": 0.1211, "step": 4272 }, { "epoch": 2.010823529411765, "grad_norm": 1.170614510319163, "learning_rate": 0.00026810559104300315, "loss": 1.3125, "step": 4273 }, { "epoch": 2.0112941176470587, "grad_norm": 1.2327286914330673, "learning_rate": 0.0002680785618761059, "loss": 1.4062, "step": 4274 }, { "epoch": 2.011764705882353, "grad_norm": 0.7000169755985868, "learning_rate": 0.00026805152262472025, "loss": 0.3457, "step": 4275 }, { "epoch": 2.012235294117647, "grad_norm": 1.0899925701970328, "learning_rate": 0.00026802447329115557, "loss": 0.8516, "step": 4276 }, { "epoch": 2.0127058823529413, "grad_norm": 1.4256300603989336, "learning_rate": 0.00026799741387772193, "loss": 2.0781, "step": 4277 }, { "epoch": 2.013176470588235, "grad_norm": 0.9253375827625867, "learning_rate": 0.0002679703443867304, "loss": 1.1953, "step": 4278 }, { "epoch": 2.0136470588235293, "grad_norm": 0.5736947256419033, "learning_rate": 0.0002679432648204928, "loss": 0.4375, "step": 4279 }, { "epoch": 2.0141176470588236, "grad_norm": 0.8012300207584382, "learning_rate": 0.00026791617518132186, "loss": 0.6016, "step": 4280 }, { "epoch": 2.014588235294118, "grad_norm": 1.3439908567769887, "learning_rate": 0.00026788907547153126, "loss": 1.6406, "step": 4281 }, { "epoch": 2.0150588235294116, "grad_norm": 0.8593747058694745, "learning_rate": 0.00026786196569343525, "loss": 0.7734, "step": 4282 }, { "epoch": 2.015529411764706, "grad_norm": 1.1946596384480381, "learning_rate": 0.0002678348458493494, "loss": 0.7266, "step": 4283 }, { "epoch": 2.016, "grad_norm": 1.128806820141306, "learning_rate": 0.00026780771594158966, "loss": 1.0312, "step": 4284 }, { "epoch": 2.0164705882352942, "grad_norm": 0.939757415814214, "learning_rate": 0.0002677805759724731, "loss": 1.0781, "step": 4285 }, { "epoch": 2.016941176470588, "grad_norm": 0.7594247132265924, "learning_rate": 0.0002677534259443177, "loss": 0.6445, "step": 4286 }, { "epoch": 2.0174117647058822, "grad_norm": 0.9418505895807374, "learning_rate": 0.00026772626585944213, "loss": 0.5195, "step": 4287 }, { "epoch": 2.0178823529411765, "grad_norm": 0.8968668519436285, "learning_rate": 0.00026769909572016605, "loss": 1.2969, "step": 4288 }, { "epoch": 2.0183529411764707, "grad_norm": 0.6316999023893773, "learning_rate": 0.0002676719155288098, "loss": 0.2559, "step": 4289 }, { "epoch": 2.018823529411765, "grad_norm": 0.8951699261146948, "learning_rate": 0.00026764472528769485, "loss": 0.5508, "step": 4290 }, { "epoch": 2.0192941176470587, "grad_norm": 1.46298857703543, "learning_rate": 0.0002676175249991433, "loss": 0.377, "step": 4291 }, { "epoch": 2.019764705882353, "grad_norm": 0.9168785150400779, "learning_rate": 0.0002675903146654782, "loss": 0.9258, "step": 4292 }, { "epoch": 2.020235294117647, "grad_norm": 1.1516516165805903, "learning_rate": 0.00026756309428902343, "loss": 1.7422, "step": 4293 }, { "epoch": 2.0207058823529414, "grad_norm": 0.9287180273800295, "learning_rate": 0.0002675358638721038, "loss": 1.3203, "step": 4294 }, { "epoch": 2.021176470588235, "grad_norm": 0.8012961221789908, "learning_rate": 0.0002675086234170449, "loss": 0.9648, "step": 4295 }, { "epoch": 2.0216470588235294, "grad_norm": 0.5883282916413748, "learning_rate": 0.0002674813729261731, "loss": 0.3086, "step": 4296 }, { "epoch": 2.0221176470588236, "grad_norm": 1.4530174023377913, "learning_rate": 0.0002674541124018159, "loss": 1.2656, "step": 4297 }, { "epoch": 2.022588235294118, "grad_norm": 1.1256963970322902, "learning_rate": 0.00026742684184630135, "loss": 1.6641, "step": 4298 }, { "epoch": 2.0230588235294116, "grad_norm": 0.9243164193547798, "learning_rate": 0.00026739956126195855, "loss": 0.8828, "step": 4299 }, { "epoch": 2.023529411764706, "grad_norm": 0.707366072170919, "learning_rate": 0.0002673722706511174, "loss": 0.4023, "step": 4300 }, { "epoch": 2.024, "grad_norm": 0.6791002000033823, "learning_rate": 0.00026734497001610854, "loss": 0.165, "step": 4301 }, { "epoch": 2.0244705882352942, "grad_norm": 0.645129398511877, "learning_rate": 0.00026731765935926374, "loss": 0.2441, "step": 4302 }, { "epoch": 2.024941176470588, "grad_norm": 1.1697993663095676, "learning_rate": 0.00026729033868291537, "loss": 0.957, "step": 4303 }, { "epoch": 2.0254117647058822, "grad_norm": 0.8738626418791997, "learning_rate": 0.00026726300798939675, "loss": 0.6953, "step": 4304 }, { "epoch": 2.0258823529411765, "grad_norm": 0.6300845829982723, "learning_rate": 0.0002672356672810421, "loss": 0.4688, "step": 4305 }, { "epoch": 2.0263529411764707, "grad_norm": 1.5166328542596095, "learning_rate": 0.00026720831656018644, "loss": 1.0391, "step": 4306 }, { "epoch": 2.026823529411765, "grad_norm": 1.0623057586987275, "learning_rate": 0.0002671809558291656, "loss": 0.9766, "step": 4307 }, { "epoch": 2.0272941176470587, "grad_norm": 0.7832845824994547, "learning_rate": 0.00026715358509031643, "loss": 1.0312, "step": 4308 }, { "epoch": 2.027764705882353, "grad_norm": 1.1521919281037305, "learning_rate": 0.00026712620434597636, "loss": 1.3047, "step": 4309 }, { "epoch": 2.028235294117647, "grad_norm": 0.9071742259256441, "learning_rate": 0.00026709881359848406, "loss": 0.8633, "step": 4310 }, { "epoch": 2.0287058823529414, "grad_norm": 0.7829013928845885, "learning_rate": 0.0002670714128501786, "loss": 0.3965, "step": 4311 }, { "epoch": 2.029176470588235, "grad_norm": 0.9814727418918688, "learning_rate": 0.0002670440021034003, "loss": 0.9961, "step": 4312 }, { "epoch": 2.0296470588235294, "grad_norm": 0.8042294985198036, "learning_rate": 0.00026701658136049015, "loss": 0.4785, "step": 4313 }, { "epoch": 2.0301176470588236, "grad_norm": 1.2822911830046377, "learning_rate": 0.00026698915062378994, "loss": 1.75, "step": 4314 }, { "epoch": 2.030588235294118, "grad_norm": 1.3371848278184788, "learning_rate": 0.00026696170989564245, "loss": 1.7891, "step": 4315 }, { "epoch": 2.0310588235294116, "grad_norm": 1.0347274041110743, "learning_rate": 0.0002669342591783913, "loss": 0.6602, "step": 4316 }, { "epoch": 2.031529411764706, "grad_norm": 0.7867928719923435, "learning_rate": 0.00026690679847438077, "loss": 0.707, "step": 4317 }, { "epoch": 2.032, "grad_norm": 0.9424469170215022, "learning_rate": 0.0002668793277859563, "loss": 1.1016, "step": 4318 }, { "epoch": 2.0324705882352943, "grad_norm": 0.3718850268157863, "learning_rate": 0.0002668518471154639, "loss": 0.0859, "step": 4319 }, { "epoch": 2.032941176470588, "grad_norm": 0.8846006115024732, "learning_rate": 0.0002668243564652506, "loss": 0.8477, "step": 4320 }, { "epoch": 2.0334117647058823, "grad_norm": 0.6323680589996192, "learning_rate": 0.0002667968558376642, "loss": 0.3438, "step": 4321 }, { "epoch": 2.0338823529411765, "grad_norm": 0.8718643195447897, "learning_rate": 0.0002667693452350535, "loss": 0.6406, "step": 4322 }, { "epoch": 2.0343529411764707, "grad_norm": 0.8484640784600813, "learning_rate": 0.0002667418246597679, "loss": 0.5312, "step": 4323 }, { "epoch": 2.034823529411765, "grad_norm": 0.6990536058073006, "learning_rate": 0.0002667142941141579, "loss": 0.543, "step": 4324 }, { "epoch": 2.0352941176470587, "grad_norm": 1.5648552370039763, "learning_rate": 0.0002666867536005747, "loss": 1.2031, "step": 4325 }, { "epoch": 2.035764705882353, "grad_norm": 1.3165770194306745, "learning_rate": 0.00026665920312137045, "loss": 1.1406, "step": 4326 }, { "epoch": 2.036235294117647, "grad_norm": 6.250408944480102, "learning_rate": 0.000266631642678898, "loss": 1.1328, "step": 4327 }, { "epoch": 2.0367058823529414, "grad_norm": 1.3976328462389, "learning_rate": 0.00026660407227551113, "loss": 1.2344, "step": 4328 }, { "epoch": 2.037176470588235, "grad_norm": 0.6328270371540015, "learning_rate": 0.0002665764919135646, "loss": 0.375, "step": 4329 }, { "epoch": 2.0376470588235294, "grad_norm": 0.22711448908174764, "learning_rate": 0.00026654890159541386, "loss": 0.0581, "step": 4330 }, { "epoch": 2.0381176470588236, "grad_norm": 0.622657296686105, "learning_rate": 0.0002665213013234153, "loss": 0.4609, "step": 4331 }, { "epoch": 2.038588235294118, "grad_norm": 0.8470803872580279, "learning_rate": 0.00026649369109992596, "loss": 0.5703, "step": 4332 }, { "epoch": 2.0390588235294116, "grad_norm": 1.210236399993847, "learning_rate": 0.0002664660709273041, "loss": 1.1406, "step": 4333 }, { "epoch": 2.039529411764706, "grad_norm": 1.1558798801581036, "learning_rate": 0.0002664384408079085, "loss": 1.3828, "step": 4334 }, { "epoch": 2.04, "grad_norm": 1.2236205605080996, "learning_rate": 0.0002664108007440989, "loss": 1.2109, "step": 4335 }, { "epoch": 2.0404705882352943, "grad_norm": 1.0820357147569661, "learning_rate": 0.00026638315073823594, "loss": 1.2344, "step": 4336 }, { "epoch": 2.040941176470588, "grad_norm": 1.268433779967261, "learning_rate": 0.00026635549079268104, "loss": 1.3594, "step": 4337 }, { "epoch": 2.0414117647058823, "grad_norm": 0.7386195044861547, "learning_rate": 0.00026632782090979653, "loss": 0.4922, "step": 4338 }, { "epoch": 2.0418823529411765, "grad_norm": 0.694352274516402, "learning_rate": 0.0002663001410919455, "loss": 0.2637, "step": 4339 }, { "epoch": 2.0423529411764707, "grad_norm": 0.8978688375545429, "learning_rate": 0.000266272451341492, "loss": 0.9414, "step": 4340 }, { "epoch": 2.0428235294117645, "grad_norm": 1.1477410890856206, "learning_rate": 0.0002662447516608008, "loss": 1.3516, "step": 4341 }, { "epoch": 2.0432941176470587, "grad_norm": 1.441501510374838, "learning_rate": 0.0002662170420522377, "loss": 1.6797, "step": 4342 }, { "epoch": 2.043764705882353, "grad_norm": 1.38479762934438, "learning_rate": 0.0002661893225181691, "loss": 1.7578, "step": 4343 }, { "epoch": 2.044235294117647, "grad_norm": 0.9380122285422529, "learning_rate": 0.0002661615930609625, "loss": 1.0078, "step": 4344 }, { "epoch": 2.0447058823529414, "grad_norm": 1.021177397497106, "learning_rate": 0.000266133853682986, "loss": 1.2734, "step": 4345 }, { "epoch": 2.045176470588235, "grad_norm": 0.5388417087462388, "learning_rate": 0.0002661061043866089, "loss": 0.1777, "step": 4346 }, { "epoch": 2.0456470588235294, "grad_norm": 1.0564035431292969, "learning_rate": 0.00026607834517420085, "loss": 1.0156, "step": 4347 }, { "epoch": 2.0461176470588236, "grad_norm": 0.7593672280707475, "learning_rate": 0.00026605057604813285, "loss": 0.6406, "step": 4348 }, { "epoch": 2.046588235294118, "grad_norm": 1.536745280658045, "learning_rate": 0.00026602279701077637, "loss": 1.1797, "step": 4349 }, { "epoch": 2.0470588235294116, "grad_norm": 1.1931862558250292, "learning_rate": 0.000265995008064504, "loss": 0.668, "step": 4350 }, { "epoch": 2.047529411764706, "grad_norm": 0.907331879601742, "learning_rate": 0.00026596720921168897, "loss": 0.9297, "step": 4351 }, { "epoch": 2.048, "grad_norm": 0.40968777862915995, "learning_rate": 0.00026593940045470545, "loss": 0.252, "step": 4352 }, { "epoch": 2.0484705882352943, "grad_norm": 1.221757049441879, "learning_rate": 0.00026591158179592846, "loss": 1.125, "step": 4353 }, { "epoch": 2.048941176470588, "grad_norm": 0.7362280437228528, "learning_rate": 0.0002658837532377338, "loss": 0.5195, "step": 4354 }, { "epoch": 2.0494117647058823, "grad_norm": 1.134277079123518, "learning_rate": 0.00026585591478249824, "loss": 1.0078, "step": 4355 }, { "epoch": 2.0498823529411765, "grad_norm": 1.1823927902181575, "learning_rate": 0.0002658280664325993, "loss": 1.4688, "step": 4356 }, { "epoch": 2.0503529411764707, "grad_norm": 1.205910784038373, "learning_rate": 0.00026580020819041533, "loss": 1.2266, "step": 4357 }, { "epoch": 2.0508235294117645, "grad_norm": 0.7912707470724627, "learning_rate": 0.0002657723400583256, "loss": 0.5781, "step": 4358 }, { "epoch": 2.0512941176470587, "grad_norm": 0.6601017971157865, "learning_rate": 0.00026574446203871013, "loss": 0.4746, "step": 4359 }, { "epoch": 2.051764705882353, "grad_norm": 1.0309485001446372, "learning_rate": 0.0002657165741339499, "loss": 1.3203, "step": 4360 }, { "epoch": 2.052235294117647, "grad_norm": 0.6039970177690536, "learning_rate": 0.00026568867634642665, "loss": 0.2314, "step": 4361 }, { "epoch": 2.0527058823529414, "grad_norm": 0.4741772313631315, "learning_rate": 0.00026566076867852304, "loss": 0.4043, "step": 4362 }, { "epoch": 2.053176470588235, "grad_norm": 1.1901200778839853, "learning_rate": 0.0002656328511326224, "loss": 1.8516, "step": 4363 }, { "epoch": 2.0536470588235294, "grad_norm": 1.7266079979823687, "learning_rate": 0.00026560492371110913, "loss": 1.0469, "step": 4364 }, { "epoch": 2.0541176470588236, "grad_norm": 0.8676415514170952, "learning_rate": 0.00026557698641636835, "loss": 0.5312, "step": 4365 }, { "epoch": 2.054588235294118, "grad_norm": 0.9048053330906429, "learning_rate": 0.000265549039250786, "loss": 0.9336, "step": 4366 }, { "epoch": 2.0550588235294116, "grad_norm": 3.1626243369924696, "learning_rate": 0.000265521082216749, "loss": 0.7188, "step": 4367 }, { "epoch": 2.055529411764706, "grad_norm": 0.5995021939445913, "learning_rate": 0.0002654931153166449, "loss": 0.3906, "step": 4368 }, { "epoch": 2.056, "grad_norm": 0.9417872025086018, "learning_rate": 0.0002654651385528622, "loss": 0.6289, "step": 4369 }, { "epoch": 2.0564705882352943, "grad_norm": 1.2558021158531256, "learning_rate": 0.00026543715192779043, "loss": 1.375, "step": 4370 }, { "epoch": 2.056941176470588, "grad_norm": 1.1975048130493098, "learning_rate": 0.0002654091554438196, "loss": 0.9766, "step": 4371 }, { "epoch": 2.0574117647058823, "grad_norm": 1.2201406095184033, "learning_rate": 0.0002653811491033408, "loss": 1.5391, "step": 4372 }, { "epoch": 2.0578823529411765, "grad_norm": 0.8273546430392886, "learning_rate": 0.00026535313290874593, "loss": 0.5742, "step": 4373 }, { "epoch": 2.0583529411764707, "grad_norm": 0.7806248536694098, "learning_rate": 0.0002653251068624278, "loss": 0.6328, "step": 4374 }, { "epoch": 2.0588235294117645, "grad_norm": 1.4619399254875936, "learning_rate": 0.00026529707096677977, "loss": 1.8047, "step": 4375 }, { "epoch": 2.0592941176470587, "grad_norm": 0.7822815943077999, "learning_rate": 0.00026526902522419637, "loss": 0.6016, "step": 4376 }, { "epoch": 2.059764705882353, "grad_norm": 0.5532857452256006, "learning_rate": 0.0002652409696370729, "loss": 0.3379, "step": 4377 }, { "epoch": 2.060235294117647, "grad_norm": 0.9535574058191837, "learning_rate": 0.0002652129042078053, "loss": 0.8984, "step": 4378 }, { "epoch": 2.0607058823529414, "grad_norm": 0.9631220992653039, "learning_rate": 0.00026518482893879056, "loss": 0.4785, "step": 4379 }, { "epoch": 2.061176470588235, "grad_norm": 0.744507595995828, "learning_rate": 0.0002651567438324265, "loss": 0.4648, "step": 4380 }, { "epoch": 2.0616470588235294, "grad_norm": 1.3523036287210408, "learning_rate": 0.00026512864889111163, "loss": 1.5938, "step": 4381 }, { "epoch": 2.0621176470588236, "grad_norm": 0.6971350048412018, "learning_rate": 0.00026510054411724545, "loss": 0.3027, "step": 4382 }, { "epoch": 2.062588235294118, "grad_norm": 0.800764988826089, "learning_rate": 0.00026507242951322825, "loss": 0.957, "step": 4383 }, { "epoch": 2.0630588235294116, "grad_norm": 1.1709997206202138, "learning_rate": 0.00026504430508146116, "loss": 1.0391, "step": 4384 }, { "epoch": 2.063529411764706, "grad_norm": 1.3395846690723772, "learning_rate": 0.00026501617082434607, "loss": 1.0312, "step": 4385 }, { "epoch": 2.064, "grad_norm": 1.5802457479893322, "learning_rate": 0.0002649880267442859, "loss": 1.4062, "step": 4386 }, { "epoch": 2.0644705882352943, "grad_norm": 1.6614429659636603, "learning_rate": 0.00026495987284368415, "loss": 1.7734, "step": 4387 }, { "epoch": 2.064941176470588, "grad_norm": 1.1030651378243455, "learning_rate": 0.0002649317091249455, "loss": 1.3516, "step": 4388 }, { "epoch": 2.0654117647058823, "grad_norm": 0.981065973988778, "learning_rate": 0.00026490353559047505, "loss": 1.0469, "step": 4389 }, { "epoch": 2.0658823529411765, "grad_norm": 0.3400821357602996, "learning_rate": 0.0002648753522426791, "loss": 0.0327, "step": 4390 }, { "epoch": 2.0663529411764707, "grad_norm": 0.8717358926420637, "learning_rate": 0.0002648471590839646, "loss": 0.4102, "step": 4391 }, { "epoch": 2.0668235294117645, "grad_norm": 1.4984936195635052, "learning_rate": 0.00026481895611673944, "loss": 1.7578, "step": 4392 }, { "epoch": 2.0672941176470587, "grad_norm": 0.698237916618865, "learning_rate": 0.0002647907433434122, "loss": 0.4668, "step": 4393 }, { "epoch": 2.067764705882353, "grad_norm": 0.8092693635608411, "learning_rate": 0.00026476252076639245, "loss": 0.4258, "step": 4394 }, { "epoch": 2.068235294117647, "grad_norm": 0.5045846811507013, "learning_rate": 0.0002647342883880905, "loss": 0.2227, "step": 4395 }, { "epoch": 2.0687058823529414, "grad_norm": 1.5577798024824487, "learning_rate": 0.0002647060462109176, "loss": 1.8203, "step": 4396 }, { "epoch": 2.069176470588235, "grad_norm": 0.5784945313965696, "learning_rate": 0.00026467779423728564, "loss": 0.1992, "step": 4397 }, { "epoch": 2.0696470588235294, "grad_norm": 1.0550430651773128, "learning_rate": 0.0002646495324696076, "loss": 1.0469, "step": 4398 }, { "epoch": 2.0701176470588236, "grad_norm": 0.663592513177661, "learning_rate": 0.0002646212609102972, "loss": 0.25, "step": 4399 }, { "epoch": 2.070588235294118, "grad_norm": 1.073622867653523, "learning_rate": 0.00026459297956176885, "loss": 0.9336, "step": 4400 }, { "epoch": 2.0710588235294116, "grad_norm": 1.081456420787269, "learning_rate": 0.000264564688426438, "loss": 1.0625, "step": 4401 }, { "epoch": 2.071529411764706, "grad_norm": 1.266249460536294, "learning_rate": 0.0002645363875067208, "loss": 2.0, "step": 4402 }, { "epoch": 2.072, "grad_norm": 0.8769909904473123, "learning_rate": 0.0002645080768050342, "loss": 0.5391, "step": 4403 }, { "epoch": 2.0724705882352943, "grad_norm": 0.8248105663707117, "learning_rate": 0.00026447975632379637, "loss": 0.7734, "step": 4404 }, { "epoch": 2.072941176470588, "grad_norm": 1.4687803452263735, "learning_rate": 0.0002644514260654257, "loss": 0.832, "step": 4405 }, { "epoch": 2.0734117647058823, "grad_norm": 1.2425560149970847, "learning_rate": 0.0002644230860323419, "loss": 0.9336, "step": 4406 }, { "epoch": 2.0738823529411765, "grad_norm": 1.5258907498687102, "learning_rate": 0.00026439473622696534, "loss": 2.3281, "step": 4407 }, { "epoch": 2.0743529411764707, "grad_norm": 1.4574559577800668, "learning_rate": 0.0002643663766517172, "loss": 1.0703, "step": 4408 }, { "epoch": 2.0748235294117645, "grad_norm": 0.9891049027984262, "learning_rate": 0.00026433800730901947, "loss": 0.6797, "step": 4409 }, { "epoch": 2.0752941176470587, "grad_norm": 0.8557911898375805, "learning_rate": 0.00026430962820129513, "loss": 0.8047, "step": 4410 }, { "epoch": 2.075764705882353, "grad_norm": 0.654906062060763, "learning_rate": 0.00026428123933096786, "loss": 0.3516, "step": 4411 }, { "epoch": 2.076235294117647, "grad_norm": 1.0595912063263326, "learning_rate": 0.0002642528407004622, "loss": 1.3438, "step": 4412 }, { "epoch": 2.0767058823529414, "grad_norm": 0.811721615178816, "learning_rate": 0.00026422443231220353, "loss": 0.6328, "step": 4413 }, { "epoch": 2.077176470588235, "grad_norm": 1.181210658365522, "learning_rate": 0.0002641960141686181, "loss": 1.1797, "step": 4414 }, { "epoch": 2.0776470588235294, "grad_norm": 1.4065932745322298, "learning_rate": 0.0002641675862721329, "loss": 1.7656, "step": 4415 }, { "epoch": 2.0781176470588236, "grad_norm": 0.4623627090904533, "learning_rate": 0.00026413914862517585, "loss": 0.1699, "step": 4416 }, { "epoch": 2.078588235294118, "grad_norm": 1.0296003494656245, "learning_rate": 0.00026411070123017567, "loss": 1.2344, "step": 4417 }, { "epoch": 2.0790588235294116, "grad_norm": 0.6541719544340778, "learning_rate": 0.0002640822440895619, "loss": 0.2422, "step": 4418 }, { "epoch": 2.079529411764706, "grad_norm": 0.9157004019948148, "learning_rate": 0.0002640537772057649, "loss": 0.668, "step": 4419 }, { "epoch": 2.08, "grad_norm": 1.0115789053688085, "learning_rate": 0.0002640253005812159, "loss": 0.7773, "step": 4420 }, { "epoch": 2.0804705882352943, "grad_norm": 0.7332627265073152, "learning_rate": 0.00026399681421834694, "loss": 0.2949, "step": 4421 }, { "epoch": 2.080941176470588, "grad_norm": 0.9874921046525262, "learning_rate": 0.0002639683181195909, "loss": 1.1641, "step": 4422 }, { "epoch": 2.0814117647058823, "grad_norm": 0.727397751252144, "learning_rate": 0.00026393981228738144, "loss": 0.8711, "step": 4423 }, { "epoch": 2.0818823529411765, "grad_norm": 1.5193854433144303, "learning_rate": 0.0002639112967241532, "loss": 1.5391, "step": 4424 }, { "epoch": 2.0823529411764707, "grad_norm": 0.6823527750134749, "learning_rate": 0.00026388277143234146, "loss": 0.4961, "step": 4425 }, { "epoch": 2.0828235294117645, "grad_norm": 3.349394351071865, "learning_rate": 0.0002638542364143825, "loss": 0.05, "step": 4426 }, { "epoch": 2.0832941176470587, "grad_norm": 1.0939977213331524, "learning_rate": 0.00026382569167271317, "loss": 0.9648, "step": 4427 }, { "epoch": 2.083764705882353, "grad_norm": 0.960994374046938, "learning_rate": 0.00026379713720977156, "loss": 0.8828, "step": 4428 }, { "epoch": 2.084235294117647, "grad_norm": 0.4386766647733364, "learning_rate": 0.0002637685730279962, "loss": 0.1719, "step": 4429 }, { "epoch": 2.084705882352941, "grad_norm": 1.0934423557615316, "learning_rate": 0.00026373999912982676, "loss": 1.0469, "step": 4430 }, { "epoch": 2.085176470588235, "grad_norm": 0.47194318349126985, "learning_rate": 0.00026371141551770344, "loss": 0.209, "step": 4431 }, { "epoch": 2.0856470588235294, "grad_norm": 0.7650321765876615, "learning_rate": 0.0002636828221940675, "loss": 0.4629, "step": 4432 }, { "epoch": 2.0861176470588236, "grad_norm": 1.653584695937243, "learning_rate": 0.00026365421916136093, "loss": 0.9805, "step": 4433 }, { "epoch": 2.086588235294118, "grad_norm": 0.6539288172009212, "learning_rate": 0.00026362560642202655, "loss": 0.293, "step": 4434 }, { "epoch": 2.0870588235294116, "grad_norm": 1.324152439401627, "learning_rate": 0.0002635969839785081, "loss": 1.3828, "step": 4435 }, { "epoch": 2.087529411764706, "grad_norm": 1.0745764385424061, "learning_rate": 0.00026356835183325, "loss": 1.1875, "step": 4436 }, { "epoch": 2.088, "grad_norm": 0.5687784743885828, "learning_rate": 0.00026353970998869763, "loss": 0.2119, "step": 4437 }, { "epoch": 2.0884705882352943, "grad_norm": 1.0976967179781056, "learning_rate": 0.0002635110584472971, "loss": 1.1641, "step": 4438 }, { "epoch": 2.088941176470588, "grad_norm": 1.1057766710729018, "learning_rate": 0.00026348239721149535, "loss": 1.0156, "step": 4439 }, { "epoch": 2.0894117647058823, "grad_norm": 0.6441320033370154, "learning_rate": 0.0002634537262837403, "loss": 0.4746, "step": 4440 }, { "epoch": 2.0898823529411765, "grad_norm": 0.5692247667186215, "learning_rate": 0.00026342504566648055, "loss": 0.2256, "step": 4441 }, { "epoch": 2.0903529411764707, "grad_norm": 0.980566809142366, "learning_rate": 0.0002633963553621655, "loss": 0.7227, "step": 4442 }, { "epoch": 2.0908235294117645, "grad_norm": 1.4358292646838353, "learning_rate": 0.0002633676553732456, "loss": 1.7656, "step": 4443 }, { "epoch": 2.0912941176470587, "grad_norm": 0.6628044640621001, "learning_rate": 0.0002633389457021718, "loss": 0.4824, "step": 4444 }, { "epoch": 2.091764705882353, "grad_norm": 0.9353204382369261, "learning_rate": 0.0002633102263513961, "loss": 0.8984, "step": 4445 }, { "epoch": 2.092235294117647, "grad_norm": 0.8857581722166536, "learning_rate": 0.0002632814973233713, "loss": 0.9766, "step": 4446 }, { "epoch": 2.092705882352941, "grad_norm": 0.8879874083646694, "learning_rate": 0.00026325275862055104, "loss": 0.6836, "step": 4447 }, { "epoch": 2.093176470588235, "grad_norm": 1.0120350543292118, "learning_rate": 0.0002632240102453896, "loss": 1.2578, "step": 4448 }, { "epoch": 2.0936470588235294, "grad_norm": 0.6779374167744349, "learning_rate": 0.0002631952522003424, "loss": 0.8789, "step": 4449 }, { "epoch": 2.0941176470588236, "grad_norm": 1.3068710725649866, "learning_rate": 0.00026316648448786536, "loss": 1.8828, "step": 4450 }, { "epoch": 2.094588235294118, "grad_norm": 1.0906941327765662, "learning_rate": 0.00026313770711041556, "loss": 1.125, "step": 4451 }, { "epoch": 2.0950588235294116, "grad_norm": 0.4384288444474091, "learning_rate": 0.0002631089200704506, "loss": 0.1719, "step": 4452 }, { "epoch": 2.095529411764706, "grad_norm": 0.8239329528175522, "learning_rate": 0.00026308012337042906, "loss": 0.8086, "step": 4453 }, { "epoch": 2.096, "grad_norm": 0.6087252920092863, "learning_rate": 0.0002630513170128103, "loss": 0.5586, "step": 4454 }, { "epoch": 2.0964705882352943, "grad_norm": 0.9334796254110108, "learning_rate": 0.0002630225010000546, "loss": 0.8945, "step": 4455 }, { "epoch": 2.096941176470588, "grad_norm": 1.530483184457855, "learning_rate": 0.0002629936753346229, "loss": 2.1094, "step": 4456 }, { "epoch": 2.0974117647058823, "grad_norm": 1.0600757791957889, "learning_rate": 0.00026296484001897715, "loss": 1.1328, "step": 4457 }, { "epoch": 2.0978823529411765, "grad_norm": 0.7240947054915864, "learning_rate": 0.00026293599505558, "loss": 0.5703, "step": 4458 }, { "epoch": 2.0983529411764708, "grad_norm": 0.9968543070517879, "learning_rate": 0.00026290714044689483, "loss": 1.3828, "step": 4459 }, { "epoch": 2.0988235294117645, "grad_norm": 0.6613154606308214, "learning_rate": 0.0002628782761953861, "loss": 0.373, "step": 4460 }, { "epoch": 2.0992941176470588, "grad_norm": 1.2311734423560992, "learning_rate": 0.0002628494023035189, "loss": 0.5703, "step": 4461 }, { "epoch": 2.099764705882353, "grad_norm": 0.9389279839973852, "learning_rate": 0.00026282051877375923, "loss": 0.9219, "step": 4462 }, { "epoch": 2.100235294117647, "grad_norm": 0.7903058948628569, "learning_rate": 0.0002627916256085739, "loss": 0.6211, "step": 4463 }, { "epoch": 2.100705882352941, "grad_norm": 0.9332550379663195, "learning_rate": 0.0002627627228104305, "loss": 1.1641, "step": 4464 }, { "epoch": 2.101176470588235, "grad_norm": 0.8117872727542343, "learning_rate": 0.00026273381038179744, "loss": 0.6289, "step": 4465 }, { "epoch": 2.1016470588235294, "grad_norm": 0.9727085323645414, "learning_rate": 0.000262704888325144, "loss": 1.4219, "step": 4466 }, { "epoch": 2.1021176470588236, "grad_norm": 0.7654153570010818, "learning_rate": 0.0002626759566429404, "loss": 0.3926, "step": 4467 }, { "epoch": 2.1025882352941174, "grad_norm": 0.8628613099841282, "learning_rate": 0.0002626470153376573, "loss": 0.3516, "step": 4468 }, { "epoch": 2.1030588235294116, "grad_norm": 0.9526790287616498, "learning_rate": 0.00026261806441176666, "loss": 0.9883, "step": 4469 }, { "epoch": 2.103529411764706, "grad_norm": 1.763731072786062, "learning_rate": 0.0002625891038677408, "loss": 2.1719, "step": 4470 }, { "epoch": 2.104, "grad_norm": 1.233104652903361, "learning_rate": 0.0002625601337080533, "loss": 0.9336, "step": 4471 }, { "epoch": 2.1044705882352943, "grad_norm": 0.7128539428233102, "learning_rate": 0.00026253115393517834, "loss": 0.3047, "step": 4472 }, { "epoch": 2.104941176470588, "grad_norm": 0.7145990609592188, "learning_rate": 0.0002625021645515908, "loss": 0.4258, "step": 4473 }, { "epoch": 2.1054117647058823, "grad_norm": 1.0302804451797383, "learning_rate": 0.00026247316555976657, "loss": 1.3672, "step": 4474 }, { "epoch": 2.1058823529411765, "grad_norm": 2.3911338801515685, "learning_rate": 0.00026244415696218237, "loss": 1.0156, "step": 4475 }, { "epoch": 2.1063529411764708, "grad_norm": 0.6555892327277703, "learning_rate": 0.00026241513876131564, "loss": 0.6562, "step": 4476 }, { "epoch": 2.1068235294117645, "grad_norm": 0.944622230251236, "learning_rate": 0.00026238611095964466, "loss": 1.0156, "step": 4477 }, { "epoch": 2.1072941176470588, "grad_norm": 1.099946288096382, "learning_rate": 0.0002623570735596485, "loss": 0.8633, "step": 4478 }, { "epoch": 2.107764705882353, "grad_norm": 1.0637448849673858, "learning_rate": 0.0002623280265638072, "loss": 1.0703, "step": 4479 }, { "epoch": 2.108235294117647, "grad_norm": 1.226684524698901, "learning_rate": 0.0002622989699746014, "loss": 1.1953, "step": 4480 }, { "epoch": 2.108705882352941, "grad_norm": 1.1628265574654277, "learning_rate": 0.00026226990379451284, "loss": 1.7266, "step": 4481 }, { "epoch": 2.109176470588235, "grad_norm": 1.5938915213653002, "learning_rate": 0.00026224082802602373, "loss": 1.5312, "step": 4482 }, { "epoch": 2.1096470588235294, "grad_norm": 0.513368846722974, "learning_rate": 0.00026221174267161744, "loss": 0.8086, "step": 4483 }, { "epoch": 2.1101176470588237, "grad_norm": 0.9919970014980161, "learning_rate": 0.0002621826477337779, "loss": 1.0156, "step": 4484 }, { "epoch": 2.1105882352941174, "grad_norm": 0.9903216624023803, "learning_rate": 0.00026215354321498997, "loss": 0.668, "step": 4485 }, { "epoch": 2.1110588235294117, "grad_norm": 0.9714619354958505, "learning_rate": 0.00026212442911773936, "loss": 0.6445, "step": 4486 }, { "epoch": 2.111529411764706, "grad_norm": 0.8186731398239544, "learning_rate": 0.0002620953054445125, "loss": 0.5117, "step": 4487 }, { "epoch": 2.112, "grad_norm": 0.6692810679102815, "learning_rate": 0.0002620661721977968, "loss": 0.2617, "step": 4488 }, { "epoch": 2.1124705882352943, "grad_norm": 0.9616773935177011, "learning_rate": 0.00026203702938008025, "loss": 0.8164, "step": 4489 }, { "epoch": 2.112941176470588, "grad_norm": 1.1413823787625061, "learning_rate": 0.00026200787699385186, "loss": 0.7969, "step": 4490 }, { "epoch": 2.1134117647058823, "grad_norm": 1.1311766751380745, "learning_rate": 0.0002619787150416014, "loss": 1.6719, "step": 4491 }, { "epoch": 2.1138823529411765, "grad_norm": 0.8952132985230079, "learning_rate": 0.00026194954352581944, "loss": 0.5977, "step": 4492 }, { "epoch": 2.1143529411764708, "grad_norm": 1.3379309793124476, "learning_rate": 0.0002619203624489973, "loss": 2.0469, "step": 4493 }, { "epoch": 2.1148235294117645, "grad_norm": 1.2661680578489731, "learning_rate": 0.00026189117181362733, "loss": 1.6641, "step": 4494 }, { "epoch": 2.1152941176470588, "grad_norm": 0.647914429773346, "learning_rate": 0.0002618619716222024, "loss": 0.7227, "step": 4495 }, { "epoch": 2.115764705882353, "grad_norm": 1.2439251146068508, "learning_rate": 0.00026183276187721635, "loss": 1.9062, "step": 4496 }, { "epoch": 2.116235294117647, "grad_norm": 1.0667790233747718, "learning_rate": 0.000261803542581164, "loss": 1.0547, "step": 4497 }, { "epoch": 2.116705882352941, "grad_norm": 1.0704202824684155, "learning_rate": 0.0002617743137365407, "loss": 1.0703, "step": 4498 }, { "epoch": 2.117176470588235, "grad_norm": 1.0535834922609244, "learning_rate": 0.0002617450753458427, "loss": 1.0859, "step": 4499 }, { "epoch": 2.1176470588235294, "grad_norm": 0.8089560286987806, "learning_rate": 0.00026171582741156725, "loss": 0.4688, "step": 4500 }, { "epoch": 2.1181176470588237, "grad_norm": 0.8412956033110409, "learning_rate": 0.0002616865699362121, "loss": 0.6211, "step": 4501 }, { "epoch": 2.1185882352941174, "grad_norm": 0.9364504541836288, "learning_rate": 0.0002616573029222761, "loss": 1.0234, "step": 4502 }, { "epoch": 2.1190588235294117, "grad_norm": 1.166721045698818, "learning_rate": 0.0002616280263722588, "loss": 0.9805, "step": 4503 }, { "epoch": 2.119529411764706, "grad_norm": 0.6535897001782753, "learning_rate": 0.00026159874028866044, "loss": 0.5391, "step": 4504 }, { "epoch": 2.12, "grad_norm": 0.9346522116491606, "learning_rate": 0.0002615694446739823, "loss": 0.5781, "step": 4505 }, { "epoch": 2.1204705882352943, "grad_norm": 1.0994074719115776, "learning_rate": 0.0002615401395307263, "loss": 0.6875, "step": 4506 }, { "epoch": 2.120941176470588, "grad_norm": 0.8827626969558228, "learning_rate": 0.00026151082486139534, "loss": 0.8594, "step": 4507 }, { "epoch": 2.1214117647058823, "grad_norm": 0.9333488057915419, "learning_rate": 0.0002614815006684929, "loss": 0.7656, "step": 4508 }, { "epoch": 2.1218823529411766, "grad_norm": 0.9071520203293277, "learning_rate": 0.0002614521669545236, "loss": 0.9414, "step": 4509 }, { "epoch": 2.1223529411764708, "grad_norm": 0.5698813171662069, "learning_rate": 0.0002614228237219925, "loss": 0.2393, "step": 4510 }, { "epoch": 2.1228235294117646, "grad_norm": 0.9115156191126274, "learning_rate": 0.00026139347097340574, "loss": 0.9609, "step": 4511 }, { "epoch": 2.1232941176470588, "grad_norm": 1.6695496335496487, "learning_rate": 0.00026136410871127017, "loss": 2.0156, "step": 4512 }, { "epoch": 2.123764705882353, "grad_norm": 0.9081612105575733, "learning_rate": 0.0002613347369380935, "loss": 1.3047, "step": 4513 }, { "epoch": 2.124235294117647, "grad_norm": 1.134525434062352, "learning_rate": 0.00026130535565638415, "loss": 0.918, "step": 4514 }, { "epoch": 2.124705882352941, "grad_norm": 0.7960744070227787, "learning_rate": 0.00026127596486865155, "loss": 0.4922, "step": 4515 }, { "epoch": 2.1251764705882352, "grad_norm": 1.1753210604859428, "learning_rate": 0.00026124656457740574, "loss": 1.2812, "step": 4516 }, { "epoch": 2.1256470588235294, "grad_norm": 0.9286919292262336, "learning_rate": 0.00026121715478515757, "loss": 0.9883, "step": 4517 }, { "epoch": 2.1261176470588237, "grad_norm": 0.9384251742069908, "learning_rate": 0.0002611877354944189, "loss": 1.1016, "step": 4518 }, { "epoch": 2.1265882352941174, "grad_norm": 1.3607722329854932, "learning_rate": 0.00026115830670770225, "loss": 1.9922, "step": 4519 }, { "epoch": 2.1270588235294117, "grad_norm": 1.0613534073180937, "learning_rate": 0.00026112886842752104, "loss": 1.4062, "step": 4520 }, { "epoch": 2.127529411764706, "grad_norm": 1.603271073349348, "learning_rate": 0.0002610994206563893, "loss": 2.2188, "step": 4521 }, { "epoch": 2.128, "grad_norm": 1.1004680517768386, "learning_rate": 0.0002610699633968222, "loss": 0.832, "step": 4522 }, { "epoch": 2.1284705882352943, "grad_norm": 1.0839748258156081, "learning_rate": 0.00026104049665133533, "loss": 1.3438, "step": 4523 }, { "epoch": 2.128941176470588, "grad_norm": 1.11554286193655, "learning_rate": 0.00026101102042244543, "loss": 1.3828, "step": 4524 }, { "epoch": 2.1294117647058823, "grad_norm": 0.9278380375981693, "learning_rate": 0.0002609815347126699, "loss": 0.7539, "step": 4525 }, { "epoch": 2.1298823529411766, "grad_norm": 0.9533648696839063, "learning_rate": 0.00026095203952452694, "loss": 0.7266, "step": 4526 }, { "epoch": 2.130352941176471, "grad_norm": 1.0249406460713644, "learning_rate": 0.00026092253486053555, "loss": 0.8203, "step": 4527 }, { "epoch": 2.1308235294117646, "grad_norm": 0.7957757687998367, "learning_rate": 0.0002608930207232156, "loss": 0.5547, "step": 4528 }, { "epoch": 2.131294117647059, "grad_norm": 1.3028552123570956, "learning_rate": 0.0002608634971150878, "loss": 1.2109, "step": 4529 }, { "epoch": 2.131764705882353, "grad_norm": 1.1804807715914902, "learning_rate": 0.0002608339640386735, "loss": 1.0781, "step": 4530 }, { "epoch": 2.1322352941176472, "grad_norm": 1.2102779782639588, "learning_rate": 0.00026080442149649516, "loss": 1.7578, "step": 4531 }, { "epoch": 2.132705882352941, "grad_norm": 0.9824355887856994, "learning_rate": 0.0002607748694910756, "loss": 1.1406, "step": 4532 }, { "epoch": 2.1331764705882352, "grad_norm": 1.2470120164376934, "learning_rate": 0.0002607453080249389, "loss": 1.7656, "step": 4533 }, { "epoch": 2.1336470588235295, "grad_norm": 1.1817955443202321, "learning_rate": 0.0002607157371006097, "loss": 1.0703, "step": 4534 }, { "epoch": 2.1341176470588237, "grad_norm": 0.7540801285350771, "learning_rate": 0.00026068615672061353, "loss": 0.6758, "step": 4535 }, { "epoch": 2.1345882352941175, "grad_norm": 0.6772891906492492, "learning_rate": 0.0002606565668874766, "loss": 0.4473, "step": 4536 }, { "epoch": 2.1350588235294117, "grad_norm": 1.1027160850830677, "learning_rate": 0.00026062696760372606, "loss": 1.4297, "step": 4537 }, { "epoch": 2.135529411764706, "grad_norm": 1.2182660818939157, "learning_rate": 0.0002605973588718899, "loss": 1.375, "step": 4538 }, { "epoch": 2.136, "grad_norm": 0.8321183980350417, "learning_rate": 0.00026056774069449685, "loss": 0.582, "step": 4539 }, { "epoch": 2.1364705882352943, "grad_norm": 1.4719486585336545, "learning_rate": 0.00026053811307407635, "loss": 2.3594, "step": 4540 }, { "epoch": 2.136941176470588, "grad_norm": 0.6070867848788377, "learning_rate": 0.00026050847601315886, "loss": 0.5391, "step": 4541 }, { "epoch": 2.1374117647058823, "grad_norm": 1.2724857916401031, "learning_rate": 0.00026047882951427543, "loss": 0.9102, "step": 4542 }, { "epoch": 2.1378823529411766, "grad_norm": 1.1417141178970112, "learning_rate": 0.00026044917357995813, "loss": 1.125, "step": 4543 }, { "epoch": 2.138352941176471, "grad_norm": 2.9717723676917522, "learning_rate": 0.00026041950821273965, "loss": 1.375, "step": 4544 }, { "epoch": 2.1388235294117646, "grad_norm": 0.9171597201549635, "learning_rate": 0.0002603898334151535, "loss": 0.7461, "step": 4545 }, { "epoch": 2.139294117647059, "grad_norm": 0.9251621921261014, "learning_rate": 0.00026036014918973414, "loss": 0.6523, "step": 4546 }, { "epoch": 2.139764705882353, "grad_norm": 0.45276663163289305, "learning_rate": 0.00026033045553901674, "loss": 0.2969, "step": 4547 }, { "epoch": 2.1402352941176472, "grad_norm": 1.0717046449203216, "learning_rate": 0.0002603007524655373, "loss": 1.0859, "step": 4548 }, { "epoch": 2.140705882352941, "grad_norm": 0.8646829635008322, "learning_rate": 0.0002602710399718325, "loss": 0.9062, "step": 4549 }, { "epoch": 2.1411764705882352, "grad_norm": 0.8133344187792729, "learning_rate": 0.0002602413180604401, "loss": 0.543, "step": 4550 }, { "epoch": 2.1416470588235295, "grad_norm": 0.6781633295286169, "learning_rate": 0.0002602115867338983, "loss": 0.3516, "step": 4551 }, { "epoch": 2.1421176470588237, "grad_norm": 0.6768851838216113, "learning_rate": 0.0002601818459947465, "loss": 1.0156, "step": 4552 }, { "epoch": 2.1425882352941175, "grad_norm": 1.238883545718429, "learning_rate": 0.0002601520958455246, "loss": 1.1797, "step": 4553 }, { "epoch": 2.1430588235294117, "grad_norm": 1.0355114504253835, "learning_rate": 0.00026012233628877336, "loss": 1.1797, "step": 4554 }, { "epoch": 2.143529411764706, "grad_norm": 1.425084714023041, "learning_rate": 0.0002600925673270345, "loss": 1.8438, "step": 4555 }, { "epoch": 2.144, "grad_norm": 1.2952693467080267, "learning_rate": 0.0002600627889628504, "loss": 1.8125, "step": 4556 }, { "epoch": 2.1444705882352944, "grad_norm": 0.9302626782762002, "learning_rate": 0.00026003300119876424, "loss": 0.6875, "step": 4557 }, { "epoch": 2.144941176470588, "grad_norm": 0.992996340995214, "learning_rate": 0.0002600032040373201, "loss": 1.5547, "step": 4558 }, { "epoch": 2.1454117647058824, "grad_norm": 1.272323647354472, "learning_rate": 0.00025997339748106275, "loss": 1.2266, "step": 4559 }, { "epoch": 2.1458823529411766, "grad_norm": 0.8292205636779169, "learning_rate": 0.00025994358153253784, "loss": 1.0312, "step": 4560 }, { "epoch": 2.1463529411764704, "grad_norm": 1.0191216040986986, "learning_rate": 0.0002599137561942918, "loss": 1.4609, "step": 4561 }, { "epoch": 2.1468235294117646, "grad_norm": 0.4674033694722173, "learning_rate": 0.00025988392146887183, "loss": 0.1592, "step": 4562 }, { "epoch": 2.147294117647059, "grad_norm": 0.3825183627365742, "learning_rate": 0.00025985407735882606, "loss": 0.0737, "step": 4563 }, { "epoch": 2.147764705882353, "grad_norm": 0.8664983566547044, "learning_rate": 0.0002598242238667032, "loss": 0.8867, "step": 4564 }, { "epoch": 2.1482352941176472, "grad_norm": 1.023114268997391, "learning_rate": 0.00025979436099505296, "loss": 1.1484, "step": 4565 }, { "epoch": 2.148705882352941, "grad_norm": 0.8917123601115161, "learning_rate": 0.0002597644887464257, "loss": 0.5469, "step": 4566 }, { "epoch": 2.1491764705882352, "grad_norm": 1.2577748959636497, "learning_rate": 0.00025973460712337275, "loss": 1.6797, "step": 4567 }, { "epoch": 2.1496470588235295, "grad_norm": 0.8738247087175014, "learning_rate": 0.0002597047161284461, "loss": 0.6562, "step": 4568 }, { "epoch": 2.1501176470588237, "grad_norm": 0.8193417480514903, "learning_rate": 0.00025967481576419864, "loss": 0.4824, "step": 4569 }, { "epoch": 2.1505882352941175, "grad_norm": 1.6319071102739162, "learning_rate": 0.0002596449060331839, "loss": 1.7891, "step": 4570 }, { "epoch": 2.1510588235294117, "grad_norm": 0.9726390219514014, "learning_rate": 0.0002596149869379564, "loss": 1.9219, "step": 4571 }, { "epoch": 2.151529411764706, "grad_norm": 1.0866741215835272, "learning_rate": 0.00025958505848107133, "loss": 1.0703, "step": 4572 }, { "epoch": 2.152, "grad_norm": 0.32476610225524666, "learning_rate": 0.00025955512066508477, "loss": 0.1992, "step": 4573 }, { "epoch": 2.152470588235294, "grad_norm": 1.3143469246167818, "learning_rate": 0.0002595251734925536, "loss": 1.3125, "step": 4574 }, { "epoch": 2.152941176470588, "grad_norm": 0.962404352783964, "learning_rate": 0.00025949521696603536, "loss": 0.7188, "step": 4575 }, { "epoch": 2.1534117647058824, "grad_norm": 0.8128600642301709, "learning_rate": 0.0002594652510880885, "loss": 0.4473, "step": 4576 }, { "epoch": 2.1538823529411766, "grad_norm": 1.3909125532470719, "learning_rate": 0.0002594352758612723, "loss": 1.7812, "step": 4577 }, { "epoch": 2.1543529411764704, "grad_norm": 0.7329125380259901, "learning_rate": 0.0002594052912881467, "loss": 0.4434, "step": 4578 }, { "epoch": 2.1548235294117646, "grad_norm": 0.9598404952695576, "learning_rate": 0.00025937529737127263, "loss": 0.9453, "step": 4579 }, { "epoch": 2.155294117647059, "grad_norm": 1.086597277907555, "learning_rate": 0.0002593452941132117, "loss": 1.3438, "step": 4580 }, { "epoch": 2.155764705882353, "grad_norm": 1.44337991697202, "learning_rate": 0.0002593152815165263, "loss": 0.8125, "step": 4581 }, { "epoch": 2.1562352941176472, "grad_norm": 0.9748811489833482, "learning_rate": 0.0002592852595837797, "loss": 0.6445, "step": 4582 }, { "epoch": 2.156705882352941, "grad_norm": 0.8874987367191645, "learning_rate": 0.00025925522831753584, "loss": 0.6055, "step": 4583 }, { "epoch": 2.1571764705882353, "grad_norm": 0.9387647687030903, "learning_rate": 0.00025922518772035963, "loss": 1.3125, "step": 4584 }, { "epoch": 2.1576470588235295, "grad_norm": 0.9565634689830869, "learning_rate": 0.0002591951377948166, "loss": 0.4707, "step": 4585 }, { "epoch": 2.1581176470588237, "grad_norm": 1.217771642369785, "learning_rate": 0.00025916507854347327, "loss": 1.75, "step": 4586 }, { "epoch": 2.1585882352941175, "grad_norm": 1.2137981899518664, "learning_rate": 0.00025913500996889677, "loss": 1.2969, "step": 4587 }, { "epoch": 2.1590588235294117, "grad_norm": 0.9110705285520068, "learning_rate": 0.00025910493207365504, "loss": 0.9414, "step": 4588 }, { "epoch": 2.159529411764706, "grad_norm": 0.9829422675737283, "learning_rate": 0.000259074844860317, "loss": 1.3125, "step": 4589 }, { "epoch": 2.16, "grad_norm": 0.6092089021896014, "learning_rate": 0.00025904474833145225, "loss": 0.2695, "step": 4590 }, { "epoch": 2.160470588235294, "grad_norm": 0.9392248526904919, "learning_rate": 0.00025901464248963106, "loss": 1.4766, "step": 4591 }, { "epoch": 2.160941176470588, "grad_norm": 1.1542392777699804, "learning_rate": 0.00025898452733742473, "loss": 1.5312, "step": 4592 }, { "epoch": 2.1614117647058824, "grad_norm": 1.0146239024436057, "learning_rate": 0.00025895440287740514, "loss": 0.8203, "step": 4593 }, { "epoch": 2.1618823529411766, "grad_norm": 0.9940417743243695, "learning_rate": 0.00025892426911214517, "loss": 1.1875, "step": 4594 }, { "epoch": 2.1623529411764704, "grad_norm": 0.608895829922113, "learning_rate": 0.0002588941260442183, "loss": 0.3301, "step": 4595 }, { "epoch": 2.1628235294117646, "grad_norm": 0.8731576323969306, "learning_rate": 0.00025886397367619904, "loss": 0.5781, "step": 4596 }, { "epoch": 2.163294117647059, "grad_norm": 1.1973058188301557, "learning_rate": 0.00025883381201066236, "loss": 1.6484, "step": 4597 }, { "epoch": 2.163764705882353, "grad_norm": 1.009175226651507, "learning_rate": 0.0002588036410501844, "loss": 0.9805, "step": 4598 }, { "epoch": 2.1642352941176473, "grad_norm": 0.8112863594403296, "learning_rate": 0.00025877346079734175, "loss": 0.459, "step": 4599 }, { "epoch": 2.164705882352941, "grad_norm": 0.93336885582583, "learning_rate": 0.000258743271254712, "loss": 0.8828, "step": 4600 }, { "epoch": 2.1651764705882353, "grad_norm": 1.1085140166483491, "learning_rate": 0.00025871307242487354, "loss": 0.8242, "step": 4601 }, { "epoch": 2.1656470588235295, "grad_norm": 0.6961649663511117, "learning_rate": 0.00025868286431040546, "loss": 0.4004, "step": 4602 }, { "epoch": 2.1661176470588237, "grad_norm": 1.359479370457749, "learning_rate": 0.0002586526469138877, "loss": 2.3125, "step": 4603 }, { "epoch": 2.1665882352941175, "grad_norm": 0.6325853281572532, "learning_rate": 0.0002586224202379009, "loss": 0.5312, "step": 4604 }, { "epoch": 2.1670588235294117, "grad_norm": 0.9221248037793284, "learning_rate": 0.00025859218428502667, "loss": 0.9766, "step": 4605 }, { "epoch": 2.167529411764706, "grad_norm": 0.8038106960417912, "learning_rate": 0.00025856193905784725, "loss": 0.5078, "step": 4606 }, { "epoch": 2.168, "grad_norm": 1.0557324998084607, "learning_rate": 0.00025853168455894566, "loss": 0.9023, "step": 4607 }, { "epoch": 2.168470588235294, "grad_norm": 0.9072283857527239, "learning_rate": 0.000258501420790906, "loss": 0.8828, "step": 4608 }, { "epoch": 2.168941176470588, "grad_norm": 0.7101640023362772, "learning_rate": 0.0002584711477563127, "loss": 0.4199, "step": 4609 }, { "epoch": 2.1694117647058824, "grad_norm": 0.8684735929375111, "learning_rate": 0.0002584408654577514, "loss": 0.8945, "step": 4610 }, { "epoch": 2.1698823529411766, "grad_norm": 1.3021717117080804, "learning_rate": 0.0002584105738978083, "loss": 1.25, "step": 4611 }, { "epoch": 2.1703529411764704, "grad_norm": 0.9669292919328835, "learning_rate": 0.00025838027307907037, "loss": 0.8047, "step": 4612 }, { "epoch": 2.1708235294117646, "grad_norm": 1.1944706509950698, "learning_rate": 0.00025834996300412556, "loss": 1.0078, "step": 4613 }, { "epoch": 2.171294117647059, "grad_norm": 0.9391222793480852, "learning_rate": 0.0002583196436755624, "loss": 1.4531, "step": 4614 }, { "epoch": 2.171764705882353, "grad_norm": 0.6643485199352693, "learning_rate": 0.00025828931509597045, "loss": 0.6445, "step": 4615 }, { "epoch": 2.1722352941176473, "grad_norm": 1.0218839875164192, "learning_rate": 0.0002582589772679398, "loss": 0.9961, "step": 4616 }, { "epoch": 2.172705882352941, "grad_norm": 0.4579787059695191, "learning_rate": 0.0002582286301940615, "loss": 0.1729, "step": 4617 }, { "epoch": 2.1731764705882353, "grad_norm": 0.6303563501043228, "learning_rate": 0.00025819827387692733, "loss": 0.6016, "step": 4618 }, { "epoch": 2.1736470588235295, "grad_norm": 0.5314187358114503, "learning_rate": 0.0002581679083191298, "loss": 0.3965, "step": 4619 }, { "epoch": 2.1741176470588237, "grad_norm": 1.148176503178279, "learning_rate": 0.0002581375335232624, "loss": 1.1641, "step": 4620 }, { "epoch": 2.1745882352941175, "grad_norm": 0.7784327932337622, "learning_rate": 0.00025810714949191916, "loss": 0.457, "step": 4621 }, { "epoch": 2.1750588235294117, "grad_norm": 2.306196020184743, "learning_rate": 0.00025807675622769514, "loss": 0.3867, "step": 4622 }, { "epoch": 2.175529411764706, "grad_norm": 1.2700101819924485, "learning_rate": 0.00025804635373318603, "loss": 2.3281, "step": 4623 }, { "epoch": 2.176, "grad_norm": 0.9848347311629871, "learning_rate": 0.0002580159420109883, "loss": 0.6914, "step": 4624 }, { "epoch": 2.176470588235294, "grad_norm": 0.6184017076397341, "learning_rate": 0.00025798552106369937, "loss": 0.3887, "step": 4625 }, { "epoch": 2.176941176470588, "grad_norm": 0.8815224593042061, "learning_rate": 0.00025795509089391726, "loss": 0.9453, "step": 4626 }, { "epoch": 2.1774117647058824, "grad_norm": 1.2663965531610815, "learning_rate": 0.00025792465150424083, "loss": 0.9219, "step": 4627 }, { "epoch": 2.1778823529411766, "grad_norm": 0.5953546020494601, "learning_rate": 0.00025789420289726983, "loss": 0.2598, "step": 4628 }, { "epoch": 2.1783529411764704, "grad_norm": 0.9407461147615289, "learning_rate": 0.00025786374507560467, "loss": 0.5977, "step": 4629 }, { "epoch": 2.1788235294117646, "grad_norm": 1.0876762689446122, "learning_rate": 0.0002578332780418466, "loss": 0.9453, "step": 4630 }, { "epoch": 2.179294117647059, "grad_norm": 1.5883663453247239, "learning_rate": 0.0002578028017985977, "loss": 2.7344, "step": 4631 }, { "epoch": 2.179764705882353, "grad_norm": 1.2482636801068276, "learning_rate": 0.00025777231634846074, "loss": 1.4766, "step": 4632 }, { "epoch": 2.1802352941176473, "grad_norm": 1.064021251914367, "learning_rate": 0.00025774182169403933, "loss": 1.3594, "step": 4633 }, { "epoch": 2.180705882352941, "grad_norm": 1.2585057237581676, "learning_rate": 0.00025771131783793787, "loss": 0.9062, "step": 4634 }, { "epoch": 2.1811764705882353, "grad_norm": 1.0479365791805781, "learning_rate": 0.00025768080478276163, "loss": 0.6797, "step": 4635 }, { "epoch": 2.1816470588235295, "grad_norm": 0.8762454415946993, "learning_rate": 0.0002576502825311164, "loss": 1.1406, "step": 4636 }, { "epoch": 2.1821176470588237, "grad_norm": 1.1005822729365087, "learning_rate": 0.000257619751085609, "loss": 1.5312, "step": 4637 }, { "epoch": 2.1825882352941175, "grad_norm": 0.7889945204852077, "learning_rate": 0.00025758921044884706, "loss": 0.9961, "step": 4638 }, { "epoch": 2.1830588235294117, "grad_norm": 1.01741881575218, "learning_rate": 0.0002575586606234388, "loss": 0.9609, "step": 4639 }, { "epoch": 2.183529411764706, "grad_norm": 0.9292854908200472, "learning_rate": 0.0002575281016119933, "loss": 0.7734, "step": 4640 }, { "epoch": 2.184, "grad_norm": 0.5862046812135413, "learning_rate": 0.00025749753341712056, "loss": 0.2832, "step": 4641 }, { "epoch": 2.184470588235294, "grad_norm": 0.9453791482579851, "learning_rate": 0.0002574669560414312, "loss": 1.1641, "step": 4642 }, { "epoch": 2.184941176470588, "grad_norm": 1.6106322471186545, "learning_rate": 0.0002574363694875367, "loss": 1.3281, "step": 4643 }, { "epoch": 2.1854117647058824, "grad_norm": 0.7852862250815396, "learning_rate": 0.0002574057737580492, "loss": 0.5977, "step": 4644 }, { "epoch": 2.1858823529411766, "grad_norm": 0.8593195922922667, "learning_rate": 0.0002573751688555818, "loss": 0.9648, "step": 4645 }, { "epoch": 2.1863529411764704, "grad_norm": 1.04279607353707, "learning_rate": 0.0002573445547827484, "loss": 1.2031, "step": 4646 }, { "epoch": 2.1868235294117646, "grad_norm": 1.2401362009944217, "learning_rate": 0.0002573139315421634, "loss": 1.3047, "step": 4647 }, { "epoch": 2.187294117647059, "grad_norm": 0.9630391538214461, "learning_rate": 0.0002572832991364423, "loss": 1.3438, "step": 4648 }, { "epoch": 2.187764705882353, "grad_norm": 1.3350945663702916, "learning_rate": 0.00025725265756820125, "loss": 1.875, "step": 4649 }, { "epoch": 2.1882352941176473, "grad_norm": 0.7088292782173865, "learning_rate": 0.00025722200684005715, "loss": 0.5352, "step": 4650 }, { "epoch": 2.188705882352941, "grad_norm": 1.0709461752149096, "learning_rate": 0.0002571913469546277, "loss": 0.7227, "step": 4651 }, { "epoch": 2.1891764705882353, "grad_norm": 1.0370391058941715, "learning_rate": 0.00025716067791453157, "loss": 1.4375, "step": 4652 }, { "epoch": 2.1896470588235295, "grad_norm": 0.8270989176729027, "learning_rate": 0.0002571299997223878, "loss": 0.9219, "step": 4653 }, { "epoch": 2.1901176470588237, "grad_norm": 0.993370098312364, "learning_rate": 0.0002570993123808167, "loss": 0.7148, "step": 4654 }, { "epoch": 2.1905882352941175, "grad_norm": 0.6615751655323253, "learning_rate": 0.00025706861589243894, "loss": 0.3535, "step": 4655 }, { "epoch": 2.1910588235294117, "grad_norm": 1.1324736237247637, "learning_rate": 0.0002570379102598762, "loss": 0.9844, "step": 4656 }, { "epoch": 2.191529411764706, "grad_norm": 1.2736754302664408, "learning_rate": 0.000257007195485751, "loss": 1.9219, "step": 4657 }, { "epoch": 2.192, "grad_norm": 1.3470287882974379, "learning_rate": 0.00025697647157268635, "loss": 0.8438, "step": 4658 }, { "epoch": 2.192470588235294, "grad_norm": 0.9390243151451704, "learning_rate": 0.0002569457385233063, "loss": 1.0, "step": 4659 }, { "epoch": 2.192941176470588, "grad_norm": 1.0994963395331585, "learning_rate": 0.00025691499634023564, "loss": 0.875, "step": 4660 }, { "epoch": 2.1934117647058824, "grad_norm": 0.9536173336323508, "learning_rate": 0.0002568842450260999, "loss": 0.9531, "step": 4661 }, { "epoch": 2.1938823529411766, "grad_norm": 1.0470947448665875, "learning_rate": 0.0002568534845835253, "loss": 0.7188, "step": 4662 }, { "epoch": 2.1943529411764704, "grad_norm": 0.9941386347754247, "learning_rate": 0.0002568227150151391, "loss": 0.9336, "step": 4663 }, { "epoch": 2.1948235294117646, "grad_norm": 0.975489730827976, "learning_rate": 0.000256791936323569, "loss": 0.9609, "step": 4664 }, { "epoch": 2.195294117647059, "grad_norm": 0.7488523954128206, "learning_rate": 0.00025676114851144375, "loss": 0.3965, "step": 4665 }, { "epoch": 2.195764705882353, "grad_norm": 1.035251775762179, "learning_rate": 0.0002567303515813928, "loss": 1.2031, "step": 4666 }, { "epoch": 2.1962352941176473, "grad_norm": 1.5730137924549532, "learning_rate": 0.0002566995455360463, "loss": 1.9375, "step": 4667 }, { "epoch": 2.196705882352941, "grad_norm": 0.8183161210251616, "learning_rate": 0.00025666873037803517, "loss": 0.7852, "step": 4668 }, { "epoch": 2.1971764705882353, "grad_norm": 0.9439211507100855, "learning_rate": 0.00025663790610999134, "loss": 0.7383, "step": 4669 }, { "epoch": 2.1976470588235295, "grad_norm": 1.2199456376405844, "learning_rate": 0.0002566070727345473, "loss": 1.1719, "step": 4670 }, { "epoch": 2.1981176470588237, "grad_norm": 1.2869911439744444, "learning_rate": 0.00025657623025433627, "loss": 0.8203, "step": 4671 }, { "epoch": 2.1985882352941175, "grad_norm": 1.0737247827111767, "learning_rate": 0.0002565453786719924, "loss": 1.4609, "step": 4672 }, { "epoch": 2.1990588235294117, "grad_norm": 1.157034460250095, "learning_rate": 0.00025651451799015067, "loss": 0.9258, "step": 4673 }, { "epoch": 2.199529411764706, "grad_norm": 0.9745187110019622, "learning_rate": 0.0002564836482114466, "loss": 1.3047, "step": 4674 }, { "epoch": 2.2, "grad_norm": 1.2398436994263227, "learning_rate": 0.00025645276933851667, "loss": 1.125, "step": 4675 }, { "epoch": 2.200470588235294, "grad_norm": 0.9822491691133454, "learning_rate": 0.00025642188137399815, "loss": 0.8086, "step": 4676 }, { "epoch": 2.200941176470588, "grad_norm": 0.7245516640080643, "learning_rate": 0.00025639098432052895, "loss": 0.4395, "step": 4677 }, { "epoch": 2.2014117647058824, "grad_norm": 1.261629260488872, "learning_rate": 0.0002563600781807478, "loss": 1.25, "step": 4678 }, { "epoch": 2.2018823529411766, "grad_norm": 0.6101792531159717, "learning_rate": 0.0002563291629572943, "loss": 0.3027, "step": 4679 }, { "epoch": 2.2023529411764704, "grad_norm": 0.5789601527600132, "learning_rate": 0.00025629823865280874, "loss": 0.1953, "step": 4680 }, { "epoch": 2.2028235294117646, "grad_norm": 0.7973975206098003, "learning_rate": 0.00025626730526993223, "loss": 0.5664, "step": 4681 }, { "epoch": 2.203294117647059, "grad_norm": 0.9414553077015629, "learning_rate": 0.00025623636281130657, "loss": 0.6289, "step": 4682 }, { "epoch": 2.203764705882353, "grad_norm": 1.214906190825305, "learning_rate": 0.0002562054112795745, "loss": 1.2031, "step": 4683 }, { "epoch": 2.2042352941176473, "grad_norm": 1.1928020221196722, "learning_rate": 0.00025617445067737933, "loss": 0.8555, "step": 4684 }, { "epoch": 2.204705882352941, "grad_norm": 0.6550222760441122, "learning_rate": 0.0002561434810073653, "loss": 0.9219, "step": 4685 }, { "epoch": 2.2051764705882353, "grad_norm": 1.8136786048724562, "learning_rate": 0.00025611250227217737, "loss": 0.75, "step": 4686 }, { "epoch": 2.2056470588235295, "grad_norm": 0.9753777320910167, "learning_rate": 0.0002560815144744612, "loss": 1.125, "step": 4687 }, { "epoch": 2.2061176470588233, "grad_norm": 1.3972118774002846, "learning_rate": 0.00025605051761686345, "loss": 2.1562, "step": 4688 }, { "epoch": 2.2065882352941175, "grad_norm": 0.7138741727762303, "learning_rate": 0.0002560195117020313, "loss": 0.3945, "step": 4689 }, { "epoch": 2.2070588235294117, "grad_norm": 0.595307245932658, "learning_rate": 0.0002559884967326128, "loss": 0.1943, "step": 4690 }, { "epoch": 2.207529411764706, "grad_norm": 1.002123980395028, "learning_rate": 0.0002559574727112569, "loss": 1.4141, "step": 4691 }, { "epoch": 2.208, "grad_norm": 0.9180258485343653, "learning_rate": 0.000255926439640613, "loss": 0.7422, "step": 4692 }, { "epoch": 2.208470588235294, "grad_norm": 1.1470650096529185, "learning_rate": 0.00025589539752333165, "loss": 1.4688, "step": 4693 }, { "epoch": 2.208941176470588, "grad_norm": 0.7064927507451083, "learning_rate": 0.0002558643463620639, "loss": 0.6172, "step": 4694 }, { "epoch": 2.2094117647058824, "grad_norm": 1.1128311927934602, "learning_rate": 0.0002558332861594617, "loss": 0.7461, "step": 4695 }, { "epoch": 2.2098823529411766, "grad_norm": 0.5671680933101576, "learning_rate": 0.0002558022169181778, "loss": 0.3867, "step": 4696 }, { "epoch": 2.2103529411764704, "grad_norm": 0.8444255705120068, "learning_rate": 0.00025577113864086557, "loss": 0.4863, "step": 4697 }, { "epoch": 2.2108235294117646, "grad_norm": 0.8368681976155107, "learning_rate": 0.0002557400513301794, "loss": 0.8398, "step": 4698 }, { "epoch": 2.211294117647059, "grad_norm": 1.049623453456353, "learning_rate": 0.00025570895498877404, "loss": 1.0859, "step": 4699 }, { "epoch": 2.211764705882353, "grad_norm": 0.5293459667313856, "learning_rate": 0.00025567784961930546, "loss": 0.1641, "step": 4700 }, { "epoch": 2.212235294117647, "grad_norm": 1.618931078285277, "learning_rate": 0.0002556467352244302, "loss": 1.6016, "step": 4701 }, { "epoch": 2.212705882352941, "grad_norm": 0.56372690706859, "learning_rate": 0.00025561561180680563, "loss": 0.2051, "step": 4702 }, { "epoch": 2.2131764705882353, "grad_norm": 0.9586990318076587, "learning_rate": 0.0002555844793690897, "loss": 1.2266, "step": 4703 }, { "epoch": 2.2136470588235295, "grad_norm": 0.9390651054406343, "learning_rate": 0.0002555533379139414, "loss": 0.5273, "step": 4704 }, { "epoch": 2.2141176470588233, "grad_norm": 1.136836915415018, "learning_rate": 0.00025552218744402026, "loss": 1.4141, "step": 4705 }, { "epoch": 2.2145882352941175, "grad_norm": 0.9606181996893471, "learning_rate": 0.00025549102796198673, "loss": 0.7461, "step": 4706 }, { "epoch": 2.2150588235294117, "grad_norm": 1.0693386928566906, "learning_rate": 0.00025545985947050197, "loss": 1.0156, "step": 4707 }, { "epoch": 2.215529411764706, "grad_norm": 0.7670021381909895, "learning_rate": 0.00025542868197222797, "loss": 0.4082, "step": 4708 }, { "epoch": 2.216, "grad_norm": 0.9343712406586102, "learning_rate": 0.00025539749546982736, "loss": 0.6992, "step": 4709 }, { "epoch": 2.216470588235294, "grad_norm": 1.0474346194450215, "learning_rate": 0.00025536629996596367, "loss": 1.125, "step": 4710 }, { "epoch": 2.216941176470588, "grad_norm": 1.1091783226594492, "learning_rate": 0.0002553350954633012, "loss": 1.0234, "step": 4711 }, { "epoch": 2.2174117647058824, "grad_norm": 1.149643837900485, "learning_rate": 0.00025530388196450487, "loss": 0.9492, "step": 4712 }, { "epoch": 2.2178823529411766, "grad_norm": 1.0530735544433811, "learning_rate": 0.0002552726594722405, "loss": 0.8086, "step": 4713 }, { "epoch": 2.2183529411764704, "grad_norm": 0.8735845546163288, "learning_rate": 0.00025524142798917466, "loss": 0.7227, "step": 4714 }, { "epoch": 2.2188235294117646, "grad_norm": 1.3321715547684634, "learning_rate": 0.00025521018751797464, "loss": 0.9531, "step": 4715 }, { "epoch": 2.219294117647059, "grad_norm": 0.9073441189569719, "learning_rate": 0.0002551789380613086, "loss": 1.2656, "step": 4716 }, { "epoch": 2.219764705882353, "grad_norm": 1.2067490811947983, "learning_rate": 0.00025514767962184535, "loss": 1.375, "step": 4717 }, { "epoch": 2.220235294117647, "grad_norm": 1.1842331545150584, "learning_rate": 0.0002551164122022545, "loss": 1.6719, "step": 4718 }, { "epoch": 2.220705882352941, "grad_norm": 0.4935591599630188, "learning_rate": 0.00025508513580520644, "loss": 0.3789, "step": 4719 }, { "epoch": 2.2211764705882353, "grad_norm": 1.0721059940253106, "learning_rate": 0.00025505385043337236, "loss": 0.9062, "step": 4720 }, { "epoch": 2.2216470588235295, "grad_norm": 1.0909492505375182, "learning_rate": 0.0002550225560894241, "loss": 1.6875, "step": 4721 }, { "epoch": 2.2221176470588233, "grad_norm": 0.5373677408153206, "learning_rate": 0.00025499125277603446, "loss": 0.2314, "step": 4722 }, { "epoch": 2.2225882352941175, "grad_norm": 0.3605463740748955, "learning_rate": 0.0002549599404958769, "loss": 0.0233, "step": 4723 }, { "epoch": 2.2230588235294118, "grad_norm": 0.8751090731577836, "learning_rate": 0.0002549286192516256, "loss": 0.6797, "step": 4724 }, { "epoch": 2.223529411764706, "grad_norm": 1.096848015391958, "learning_rate": 0.0002548972890459555, "loss": 1.6875, "step": 4725 }, { "epoch": 2.224, "grad_norm": 0.8517328181283702, "learning_rate": 0.00025486594988154235, "loss": 0.9531, "step": 4726 }, { "epoch": 2.224470588235294, "grad_norm": 0.7444126334056449, "learning_rate": 0.0002548346017610628, "loss": 0.4746, "step": 4727 }, { "epoch": 2.224941176470588, "grad_norm": 0.5920188822576539, "learning_rate": 0.000254803244687194, "loss": 0.1689, "step": 4728 }, { "epoch": 2.2254117647058824, "grad_norm": 0.9117650074359253, "learning_rate": 0.0002547718786626141, "loss": 0.6992, "step": 4729 }, { "epoch": 2.2258823529411766, "grad_norm": 1.424468464628264, "learning_rate": 0.0002547405036900018, "loss": 1.9141, "step": 4730 }, { "epoch": 2.2263529411764704, "grad_norm": 1.0198710521393963, "learning_rate": 0.0002547091197720368, "loss": 0.8477, "step": 4731 }, { "epoch": 2.2268235294117646, "grad_norm": 0.6161755702054327, "learning_rate": 0.0002546777269113993, "loss": 0.2832, "step": 4732 }, { "epoch": 2.227294117647059, "grad_norm": 1.0771035694361566, "learning_rate": 0.0002546463251107705, "loss": 0.9375, "step": 4733 }, { "epoch": 2.227764705882353, "grad_norm": 1.0111271333107832, "learning_rate": 0.0002546149143728323, "loss": 0.793, "step": 4734 }, { "epoch": 2.228235294117647, "grad_norm": 0.4307673926353734, "learning_rate": 0.0002545834947002672, "loss": 0.252, "step": 4735 }, { "epoch": 2.228705882352941, "grad_norm": 0.5947199264454215, "learning_rate": 0.00025455206609575876, "loss": 0.4551, "step": 4736 }, { "epoch": 2.2291764705882353, "grad_norm": 0.8957498945511192, "learning_rate": 0.000254520628561991, "loss": 0.9844, "step": 4737 }, { "epoch": 2.2296470588235295, "grad_norm": 0.42246689362566026, "learning_rate": 0.00025448918210164886, "loss": 0.1128, "step": 4738 }, { "epoch": 2.2301176470588233, "grad_norm": 0.7906388315858808, "learning_rate": 0.00025445772671741816, "loss": 0.6992, "step": 4739 }, { "epoch": 2.2305882352941175, "grad_norm": 0.9487643487411601, "learning_rate": 0.00025442626241198513, "loss": 1.0938, "step": 4740 }, { "epoch": 2.2310588235294118, "grad_norm": 0.7190021470573285, "learning_rate": 0.00025439478918803715, "loss": 0.3691, "step": 4741 }, { "epoch": 2.231529411764706, "grad_norm": 0.8497020529954952, "learning_rate": 0.00025436330704826205, "loss": 0.8359, "step": 4742 }, { "epoch": 2.232, "grad_norm": 1.1313974930995414, "learning_rate": 0.0002543318159953487, "loss": 1.2188, "step": 4743 }, { "epoch": 2.232470588235294, "grad_norm": 0.6799188433742553, "learning_rate": 0.00025430031603198644, "loss": 0.832, "step": 4744 }, { "epoch": 2.232941176470588, "grad_norm": 0.9767926897356375, "learning_rate": 0.0002542688071608657, "loss": 1.2109, "step": 4745 }, { "epoch": 2.2334117647058824, "grad_norm": 0.9619698177133517, "learning_rate": 0.0002542372893846773, "loss": 0.9453, "step": 4746 }, { "epoch": 2.2338823529411767, "grad_norm": 0.8064768318494413, "learning_rate": 0.00025420576270611313, "loss": 0.4883, "step": 4747 }, { "epoch": 2.2343529411764704, "grad_norm": 1.2808096298358231, "learning_rate": 0.00025417422712786575, "loss": 1.25, "step": 4748 }, { "epoch": 2.2348235294117647, "grad_norm": 1.0498502843926958, "learning_rate": 0.00025414268265262833, "loss": 0.9414, "step": 4749 }, { "epoch": 2.235294117647059, "grad_norm": 1.0432996205746274, "learning_rate": 0.0002541111292830951, "loss": 0.9883, "step": 4750 }, { "epoch": 2.235764705882353, "grad_norm": 0.924097250091921, "learning_rate": 0.0002540795670219606, "loss": 0.7969, "step": 4751 }, { "epoch": 2.236235294117647, "grad_norm": 1.0753279205542583, "learning_rate": 0.0002540479958719207, "loss": 0.832, "step": 4752 }, { "epoch": 2.236705882352941, "grad_norm": 1.161234441852189, "learning_rate": 0.0002540164158356716, "loss": 0.8633, "step": 4753 }, { "epoch": 2.2371764705882353, "grad_norm": 0.936416502802337, "learning_rate": 0.0002539848269159104, "loss": 0.7617, "step": 4754 }, { "epoch": 2.2376470588235295, "grad_norm": 1.191743674940116, "learning_rate": 0.00025395322911533484, "loss": 2.1875, "step": 4755 }, { "epoch": 2.2381176470588233, "grad_norm": 1.2569574164488027, "learning_rate": 0.0002539216224366437, "loss": 1.5469, "step": 4756 }, { "epoch": 2.2385882352941175, "grad_norm": 1.1154536037077631, "learning_rate": 0.00025389000688253626, "loss": 1.4531, "step": 4757 }, { "epoch": 2.2390588235294118, "grad_norm": 0.9187365455092522, "learning_rate": 0.00025385838245571264, "loss": 0.8047, "step": 4758 }, { "epoch": 2.239529411764706, "grad_norm": 0.7486266661225649, "learning_rate": 0.00025382674915887374, "loss": 0.5312, "step": 4759 }, { "epoch": 2.24, "grad_norm": 0.9638876030287697, "learning_rate": 0.0002537951069947212, "loss": 0.6836, "step": 4760 }, { "epoch": 2.240470588235294, "grad_norm": 1.0347331232499382, "learning_rate": 0.00025376345596595743, "loss": 0.6641, "step": 4761 }, { "epoch": 2.240941176470588, "grad_norm": 0.8664185345424803, "learning_rate": 0.0002537317960752855, "loss": 0.6992, "step": 4762 }, { "epoch": 2.2414117647058824, "grad_norm": 0.8314349294901421, "learning_rate": 0.0002537001273254094, "loss": 0.668, "step": 4763 }, { "epoch": 2.2418823529411767, "grad_norm": 0.8865681837299746, "learning_rate": 0.00025366844971903385, "loss": 0.9844, "step": 4764 }, { "epoch": 2.2423529411764704, "grad_norm": 1.0346945757240436, "learning_rate": 0.00025363676325886416, "loss": 1.1094, "step": 4765 }, { "epoch": 2.2428235294117647, "grad_norm": 0.8346000208196765, "learning_rate": 0.00025360506794760653, "loss": 0.5039, "step": 4766 }, { "epoch": 2.243294117647059, "grad_norm": 0.9179909067833257, "learning_rate": 0.00025357336378796795, "loss": 0.4512, "step": 4767 }, { "epoch": 2.243764705882353, "grad_norm": 1.0645248294919578, "learning_rate": 0.0002535416507826561, "loss": 1.5781, "step": 4768 }, { "epoch": 2.244235294117647, "grad_norm": 1.1093640405948355, "learning_rate": 0.0002535099289343793, "loss": 1.1406, "step": 4769 }, { "epoch": 2.244705882352941, "grad_norm": 0.7532848774714904, "learning_rate": 0.00025347819824584695, "loss": 1.0859, "step": 4770 }, { "epoch": 2.2451764705882353, "grad_norm": 1.3339711674008665, "learning_rate": 0.0002534464587197689, "loss": 1.0391, "step": 4771 }, { "epoch": 2.2456470588235296, "grad_norm": 1.2294622845626313, "learning_rate": 0.00025341471035885587, "loss": 1.4141, "step": 4772 }, { "epoch": 2.2461176470588233, "grad_norm": 0.8919371462029327, "learning_rate": 0.0002533829531658193, "loss": 0.7656, "step": 4773 }, { "epoch": 2.2465882352941176, "grad_norm": 0.8821336612342855, "learning_rate": 0.00025335118714337146, "loss": 0.7891, "step": 4774 }, { "epoch": 2.2470588235294118, "grad_norm": 0.8137086512186271, "learning_rate": 0.0002533194122942253, "loss": 1.0859, "step": 4775 }, { "epoch": 2.247529411764706, "grad_norm": 0.8792701338266056, "learning_rate": 0.00025328762862109457, "loss": 1.4219, "step": 4776 }, { "epoch": 2.248, "grad_norm": 1.1305716173196394, "learning_rate": 0.00025325583612669373, "loss": 1.1484, "step": 4777 }, { "epoch": 2.248470588235294, "grad_norm": 1.0074418387220707, "learning_rate": 0.00025322403481373797, "loss": 1.1328, "step": 4778 }, { "epoch": 2.248941176470588, "grad_norm": 0.8883094237891346, "learning_rate": 0.00025319222468494335, "loss": 0.918, "step": 4779 }, { "epoch": 2.2494117647058824, "grad_norm": 0.7660400762035329, "learning_rate": 0.00025316040574302663, "loss": 0.5195, "step": 4780 }, { "epoch": 2.2498823529411767, "grad_norm": 2.699207658441135, "learning_rate": 0.00025312857799070524, "loss": 1.1016, "step": 4781 }, { "epoch": 2.2503529411764704, "grad_norm": 0.9086295729440912, "learning_rate": 0.0002530967414306974, "loss": 0.8633, "step": 4782 }, { "epoch": 2.2508235294117647, "grad_norm": 1.0909389938234089, "learning_rate": 0.0002530648960657222, "loss": 1.2422, "step": 4783 }, { "epoch": 2.251294117647059, "grad_norm": 1.44963057230376, "learning_rate": 0.0002530330418984993, "loss": 2.1719, "step": 4784 }, { "epoch": 2.251764705882353, "grad_norm": 0.735971254209444, "learning_rate": 0.00025300117893174934, "loss": 0.375, "step": 4785 }, { "epoch": 2.252235294117647, "grad_norm": 1.0811877032896242, "learning_rate": 0.0002529693071681934, "loss": 0.918, "step": 4786 }, { "epoch": 2.252705882352941, "grad_norm": 0.923388782164127, "learning_rate": 0.00025293742661055364, "loss": 0.5195, "step": 4787 }, { "epoch": 2.2531764705882353, "grad_norm": 0.46243518148659574, "learning_rate": 0.0002529055372615527, "loss": 0.3848, "step": 4788 }, { "epoch": 2.2536470588235296, "grad_norm": 0.972842093097096, "learning_rate": 0.00025287363912391403, "loss": 0.8867, "step": 4789 }, { "epoch": 2.2541176470588233, "grad_norm": 1.137193864733918, "learning_rate": 0.00025284173220036215, "loss": 1.3125, "step": 4790 }, { "epoch": 2.2545882352941176, "grad_norm": 0.6757753985621232, "learning_rate": 0.00025280981649362184, "loss": 0.4121, "step": 4791 }, { "epoch": 2.255058823529412, "grad_norm": 1.0631639434532798, "learning_rate": 0.0002527778920064189, "loss": 1.375, "step": 4792 }, { "epoch": 2.255529411764706, "grad_norm": 1.0759523507210327, "learning_rate": 0.00025274595874147984, "loss": 0.9336, "step": 4793 }, { "epoch": 2.2560000000000002, "grad_norm": 0.7314902020601146, "learning_rate": 0.000252714016701532, "loss": 0.416, "step": 4794 }, { "epoch": 2.256470588235294, "grad_norm": 1.069271624601881, "learning_rate": 0.0002526820658893033, "loss": 1.8984, "step": 4795 }, { "epoch": 2.2569411764705882, "grad_norm": 0.7407631933917025, "learning_rate": 0.0002526501063075225, "loss": 0.3984, "step": 4796 }, { "epoch": 2.2574117647058825, "grad_norm": 0.5729723643447842, "learning_rate": 0.00025261813795891916, "loss": 0.2334, "step": 4797 }, { "epoch": 2.2578823529411762, "grad_norm": 0.9111115669778305, "learning_rate": 0.00025258616084622345, "loss": 0.6914, "step": 4798 }, { "epoch": 2.2583529411764705, "grad_norm": 1.104393656716363, "learning_rate": 0.00025255417497216647, "loss": 0.9375, "step": 4799 }, { "epoch": 2.2588235294117647, "grad_norm": 0.3847531703845102, "learning_rate": 0.00025252218033947993, "loss": 0.1045, "step": 4800 }, { "epoch": 2.259294117647059, "grad_norm": 0.6409810441317356, "learning_rate": 0.0002524901769508963, "loss": 0.2324, "step": 4801 }, { "epoch": 2.259764705882353, "grad_norm": 1.2212677992350724, "learning_rate": 0.0002524581648091488, "loss": 1.2969, "step": 4802 }, { "epoch": 2.260235294117647, "grad_norm": 0.27874351167110833, "learning_rate": 0.00025242614391697153, "loss": 0.0664, "step": 4803 }, { "epoch": 2.260705882352941, "grad_norm": 0.7551394811591472, "learning_rate": 0.00025239411427709915, "loss": 0.3945, "step": 4804 }, { "epoch": 2.2611764705882353, "grad_norm": 1.1208134901281461, "learning_rate": 0.0002523620758922672, "loss": 1.3906, "step": 4805 }, { "epoch": 2.2616470588235296, "grad_norm": 0.7746816064439453, "learning_rate": 0.00025233002876521183, "loss": 0.9219, "step": 4806 }, { "epoch": 2.2621176470588233, "grad_norm": 0.5459166221099111, "learning_rate": 0.0002522979728986701, "loss": 0.1807, "step": 4807 }, { "epoch": 2.2625882352941176, "grad_norm": 1.3075489068481418, "learning_rate": 0.00025226590829537977, "loss": 1.6875, "step": 4808 }, { "epoch": 2.263058823529412, "grad_norm": 0.9292075003451366, "learning_rate": 0.0002522338349580792, "loss": 0.6641, "step": 4809 }, { "epoch": 2.263529411764706, "grad_norm": 1.2279967350849799, "learning_rate": 0.00025220175288950765, "loss": 1.0547, "step": 4810 }, { "epoch": 2.2640000000000002, "grad_norm": 1.286643596010485, "learning_rate": 0.00025216966209240516, "loss": 1.3047, "step": 4811 }, { "epoch": 2.264470588235294, "grad_norm": 0.9661648167259965, "learning_rate": 0.00025213756256951234, "loss": 0.7578, "step": 4812 }, { "epoch": 2.2649411764705882, "grad_norm": 1.3864837828392866, "learning_rate": 0.00025210545432357076, "loss": 0.334, "step": 4813 }, { "epoch": 2.2654117647058825, "grad_norm": 1.106768805795677, "learning_rate": 0.0002520733373573225, "loss": 0.9141, "step": 4814 }, { "epoch": 2.2658823529411762, "grad_norm": 0.9278211487680724, "learning_rate": 0.0002520412116735106, "loss": 0.7344, "step": 4815 }, { "epoch": 2.2663529411764705, "grad_norm": 1.4035268171228001, "learning_rate": 0.0002520090772748787, "loss": 2.7344, "step": 4816 }, { "epoch": 2.2668235294117647, "grad_norm": 0.7533744224523894, "learning_rate": 0.00025197693416417126, "loss": 0.4316, "step": 4817 }, { "epoch": 2.267294117647059, "grad_norm": 0.9576815627044973, "learning_rate": 0.0002519447823441334, "loss": 0.6211, "step": 4818 }, { "epoch": 2.267764705882353, "grad_norm": 0.8316789761500132, "learning_rate": 0.00025191262181751113, "loss": 0.7656, "step": 4819 }, { "epoch": 2.268235294117647, "grad_norm": 0.890284768201419, "learning_rate": 0.00025188045258705107, "loss": 0.832, "step": 4820 }, { "epoch": 2.268705882352941, "grad_norm": 0.9692082695142644, "learning_rate": 0.0002518482746555007, "loss": 1.1875, "step": 4821 }, { "epoch": 2.2691764705882354, "grad_norm": 0.8227669147147166, "learning_rate": 0.000251816088025608, "loss": 0.3418, "step": 4822 }, { "epoch": 2.2696470588235296, "grad_norm": 0.7751639972248111, "learning_rate": 0.00025178389270012205, "loss": 0.4199, "step": 4823 }, { "epoch": 2.2701176470588234, "grad_norm": 0.7511036098794464, "learning_rate": 0.0002517516886817924, "loss": 0.6133, "step": 4824 }, { "epoch": 2.2705882352941176, "grad_norm": 0.7771802641170119, "learning_rate": 0.00025171947597336946, "loss": 0.3926, "step": 4825 }, { "epoch": 2.271058823529412, "grad_norm": 0.9105576226422947, "learning_rate": 0.00025168725457760434, "loss": 0.7578, "step": 4826 }, { "epoch": 2.271529411764706, "grad_norm": 1.243540110748871, "learning_rate": 0.0002516550244972489, "loss": 1.0312, "step": 4827 }, { "epoch": 2.2720000000000002, "grad_norm": 1.0272331281509233, "learning_rate": 0.00025162278573505575, "loss": 1.5312, "step": 4828 }, { "epoch": 2.272470588235294, "grad_norm": 0.6903704561010736, "learning_rate": 0.0002515905382937782, "loss": 0.416, "step": 4829 }, { "epoch": 2.2729411764705882, "grad_norm": 1.1287455016652932, "learning_rate": 0.0002515582821761704, "loss": 1.6094, "step": 4830 }, { "epoch": 2.2734117647058825, "grad_norm": 0.907486566648899, "learning_rate": 0.0002515260173849872, "loss": 0.6523, "step": 4831 }, { "epoch": 2.2738823529411762, "grad_norm": 1.1007475350995053, "learning_rate": 0.00025149374392298407, "loss": 1.1719, "step": 4832 }, { "epoch": 2.2743529411764705, "grad_norm": 1.0468212815695697, "learning_rate": 0.00025146146179291746, "loss": 0.6484, "step": 4833 }, { "epoch": 2.2748235294117647, "grad_norm": 1.0934254606651572, "learning_rate": 0.0002514291709975443, "loss": 1.1016, "step": 4834 }, { "epoch": 2.275294117647059, "grad_norm": 0.9209448033123498, "learning_rate": 0.00025139687153962246, "loss": 0.6484, "step": 4835 }, { "epoch": 2.275764705882353, "grad_norm": 0.6963311317765413, "learning_rate": 0.00025136456342191043, "loss": 0.4023, "step": 4836 }, { "epoch": 2.276235294117647, "grad_norm": 0.9183576682340464, "learning_rate": 0.0002513322466471675, "loss": 0.4375, "step": 4837 }, { "epoch": 2.276705882352941, "grad_norm": 1.142979534116264, "learning_rate": 0.00025129992121815365, "loss": 1.9453, "step": 4838 }, { "epoch": 2.2771764705882354, "grad_norm": 0.5776621506484936, "learning_rate": 0.00025126758713762964, "loss": 0.2217, "step": 4839 }, { "epoch": 2.2776470588235296, "grad_norm": 0.7522010437926935, "learning_rate": 0.000251235244408357, "loss": 0.5273, "step": 4840 }, { "epoch": 2.2781176470588234, "grad_norm": 0.8233654234382366, "learning_rate": 0.00025120289303309795, "loss": 0.3613, "step": 4841 }, { "epoch": 2.2785882352941176, "grad_norm": 1.4602375468903281, "learning_rate": 0.00025117053301461546, "loss": 1.7266, "step": 4842 }, { "epoch": 2.279058823529412, "grad_norm": 0.8657341279419867, "learning_rate": 0.00025113816435567315, "loss": 1.4453, "step": 4843 }, { "epoch": 2.279529411764706, "grad_norm": 1.0716021442741062, "learning_rate": 0.00025110578705903557, "loss": 1.6016, "step": 4844 }, { "epoch": 2.2800000000000002, "grad_norm": 1.0434418839841495, "learning_rate": 0.0002510734011274678, "loss": 1.7031, "step": 4845 }, { "epoch": 2.280470588235294, "grad_norm": 1.0355537273546778, "learning_rate": 0.00025104100656373585, "loss": 1.2109, "step": 4846 }, { "epoch": 2.2809411764705882, "grad_norm": 0.9544081093371427, "learning_rate": 0.0002510086033706063, "loss": 0.9766, "step": 4847 }, { "epoch": 2.2814117647058825, "grad_norm": 1.0903999691310884, "learning_rate": 0.0002509761915508466, "loss": 0.7344, "step": 4848 }, { "epoch": 2.2818823529411763, "grad_norm": 1.0154987562399431, "learning_rate": 0.0002509437711072249, "loss": 0.8906, "step": 4849 }, { "epoch": 2.2823529411764705, "grad_norm": 1.0796037946803199, "learning_rate": 0.00025091134204250997, "loss": 1.0312, "step": 4850 }, { "epoch": 2.2828235294117647, "grad_norm": 0.7701718402713927, "learning_rate": 0.0002508789043594715, "loss": 0.3887, "step": 4851 }, { "epoch": 2.283294117647059, "grad_norm": 0.7401800541897142, "learning_rate": 0.00025084645806087974, "loss": 0.4121, "step": 4852 }, { "epoch": 2.283764705882353, "grad_norm": 0.699823364994413, "learning_rate": 0.00025081400314950586, "loss": 0.5977, "step": 4853 }, { "epoch": 2.284235294117647, "grad_norm": 0.5882505911244983, "learning_rate": 0.0002507815396281216, "loss": 0.1963, "step": 4854 }, { "epoch": 2.284705882352941, "grad_norm": 0.9890119432618741, "learning_rate": 0.0002507490674994995, "loss": 1.1328, "step": 4855 }, { "epoch": 2.2851764705882354, "grad_norm": 0.9025923415277841, "learning_rate": 0.0002507165867664129, "loss": 0.6953, "step": 4856 }, { "epoch": 2.2856470588235296, "grad_norm": 1.0983559757410752, "learning_rate": 0.0002506840974316357, "loss": 0.9727, "step": 4857 }, { "epoch": 2.2861176470588234, "grad_norm": 0.7266575663844942, "learning_rate": 0.00025065159949794283, "loss": 0.4434, "step": 4858 }, { "epoch": 2.2865882352941176, "grad_norm": 1.0037640974425992, "learning_rate": 0.00025061909296810966, "loss": 1.2344, "step": 4859 }, { "epoch": 2.287058823529412, "grad_norm": 0.9058583356739072, "learning_rate": 0.0002505865778449124, "loss": 0.5742, "step": 4860 }, { "epoch": 2.287529411764706, "grad_norm": 1.0332808574588663, "learning_rate": 0.000250554054131128, "loss": 0.7695, "step": 4861 }, { "epoch": 2.288, "grad_norm": 0.6934503614738874, "learning_rate": 0.00025052152182953414, "loss": 0.3086, "step": 4862 }, { "epoch": 2.288470588235294, "grad_norm": 1.0205983899431277, "learning_rate": 0.0002504889809429093, "loss": 1.0156, "step": 4863 }, { "epoch": 2.2889411764705883, "grad_norm": 0.7782682485681374, "learning_rate": 0.0002504564314740326, "loss": 0.4512, "step": 4864 }, { "epoch": 2.2894117647058825, "grad_norm": 1.769017677502543, "learning_rate": 0.0002504238734256839, "loss": 1.5547, "step": 4865 }, { "epoch": 2.2898823529411763, "grad_norm": 1.0767305086347982, "learning_rate": 0.0002503913068006438, "loss": 1.3203, "step": 4866 }, { "epoch": 2.2903529411764705, "grad_norm": 1.3588177155052734, "learning_rate": 0.00025035873160169373, "loss": 1.5938, "step": 4867 }, { "epoch": 2.2908235294117647, "grad_norm": 0.7917322415406144, "learning_rate": 0.0002503261478316157, "loss": 0.5781, "step": 4868 }, { "epoch": 2.291294117647059, "grad_norm": 0.9420898516327205, "learning_rate": 0.0002502935554931926, "loss": 0.8633, "step": 4869 }, { "epoch": 2.291764705882353, "grad_norm": 2.5713918641370173, "learning_rate": 0.0002502609545892079, "loss": 2.5312, "step": 4870 }, { "epoch": 2.292235294117647, "grad_norm": 1.0389157468060286, "learning_rate": 0.00025022834512244583, "loss": 1.3203, "step": 4871 }, { "epoch": 2.292705882352941, "grad_norm": 0.8034963773780538, "learning_rate": 0.00025019572709569153, "loss": 0.5312, "step": 4872 }, { "epoch": 2.2931764705882354, "grad_norm": 1.2385257206960718, "learning_rate": 0.0002501631005117307, "loss": 1.3281, "step": 4873 }, { "epoch": 2.2936470588235296, "grad_norm": 0.9124780404646013, "learning_rate": 0.0002501304653733497, "loss": 0.7422, "step": 4874 }, { "epoch": 2.2941176470588234, "grad_norm": 1.0020018669474184, "learning_rate": 0.0002500978216833359, "loss": 1.0938, "step": 4875 }, { "epoch": 2.2945882352941176, "grad_norm": 0.8977299484325063, "learning_rate": 0.0002500651694444771, "loss": 1.1562, "step": 4876 }, { "epoch": 2.295058823529412, "grad_norm": 1.413547968568019, "learning_rate": 0.00025003250865956203, "loss": 1.9609, "step": 4877 }, { "epoch": 2.295529411764706, "grad_norm": 1.081049747015154, "learning_rate": 0.0002499998393313801, "loss": 0.8672, "step": 4878 }, { "epoch": 2.296, "grad_norm": 0.9101488994672817, "learning_rate": 0.00024996716146272137, "loss": 0.9883, "step": 4879 }, { "epoch": 2.296470588235294, "grad_norm": 1.146182869289728, "learning_rate": 0.0002499344750563767, "loss": 1.3359, "step": 4880 }, { "epoch": 2.2969411764705883, "grad_norm": 0.9483811619636722, "learning_rate": 0.0002499017801151377, "loss": 1.1016, "step": 4881 }, { "epoch": 2.2974117647058825, "grad_norm": 0.7586576457586973, "learning_rate": 0.0002498690766417967, "loss": 0.7227, "step": 4882 }, { "epoch": 2.2978823529411763, "grad_norm": 0.7442149895608388, "learning_rate": 0.00024983636463914665, "loss": 0.3125, "step": 4883 }, { "epoch": 2.2983529411764705, "grad_norm": 1.772295862611779, "learning_rate": 0.00024980364410998133, "loss": 1.8047, "step": 4884 }, { "epoch": 2.2988235294117647, "grad_norm": 0.6037140611601324, "learning_rate": 0.0002497709150570953, "loss": 0.6562, "step": 4885 }, { "epoch": 2.299294117647059, "grad_norm": 0.9757450907590584, "learning_rate": 0.0002497381774832838, "loss": 0.5469, "step": 4886 }, { "epoch": 2.299764705882353, "grad_norm": 1.087605358225399, "learning_rate": 0.00024970543139134275, "loss": 1.0469, "step": 4887 }, { "epoch": 2.300235294117647, "grad_norm": 0.6748709717451977, "learning_rate": 0.0002496726767840687, "loss": 0.2969, "step": 4888 }, { "epoch": 2.300705882352941, "grad_norm": 1.2723014277475206, "learning_rate": 0.0002496399136642592, "loss": 1.1094, "step": 4889 }, { "epoch": 2.3011764705882354, "grad_norm": 1.686054579286601, "learning_rate": 0.0002496071420347124, "loss": 1.8359, "step": 4890 }, { "epoch": 2.3016470588235296, "grad_norm": 1.162243900751989, "learning_rate": 0.00024957436189822706, "loss": 1.1641, "step": 4891 }, { "epoch": 2.3021176470588234, "grad_norm": 1.3368971351352947, "learning_rate": 0.0002495415732576028, "loss": 2.0469, "step": 4892 }, { "epoch": 2.3025882352941176, "grad_norm": 0.3866570081390993, "learning_rate": 0.00024950877611564, "loss": 0.1196, "step": 4893 }, { "epoch": 2.303058823529412, "grad_norm": 1.324783145596103, "learning_rate": 0.0002494759704751395, "loss": 1.3125, "step": 4894 }, { "epoch": 2.303529411764706, "grad_norm": 1.112699426436202, "learning_rate": 0.00024944315633890327, "loss": 1.7266, "step": 4895 }, { "epoch": 2.304, "grad_norm": 1.1802580157573634, "learning_rate": 0.0002494103337097337, "loss": 1.6719, "step": 4896 }, { "epoch": 2.304470588235294, "grad_norm": 0.6610063307120843, "learning_rate": 0.00024937750259043407, "loss": 0.3691, "step": 4897 }, { "epoch": 2.3049411764705883, "grad_norm": 0.988541240509504, "learning_rate": 0.00024934466298380824, "loss": 0.8633, "step": 4898 }, { "epoch": 2.3054117647058825, "grad_norm": 1.2819847036119616, "learning_rate": 0.0002493118148926609, "loss": 1.6875, "step": 4899 }, { "epoch": 2.3058823529411763, "grad_norm": 0.7712281616409294, "learning_rate": 0.00024927895831979745, "loss": 0.6719, "step": 4900 }, { "epoch": 2.3063529411764705, "grad_norm": 1.0838475049905498, "learning_rate": 0.000249246093268024, "loss": 1.2109, "step": 4901 }, { "epoch": 2.3068235294117647, "grad_norm": 0.5997371529966146, "learning_rate": 0.0002492132197401474, "loss": 0.3203, "step": 4902 }, { "epoch": 2.307294117647059, "grad_norm": 1.0930923768945837, "learning_rate": 0.0002491803377389752, "loss": 0.9492, "step": 4903 }, { "epoch": 2.307764705882353, "grad_norm": 0.6173671910122228, "learning_rate": 0.0002491474472673157, "loss": 0.6055, "step": 4904 }, { "epoch": 2.308235294117647, "grad_norm": 1.1945564391425296, "learning_rate": 0.00024911454832797784, "loss": 1.3594, "step": 4905 }, { "epoch": 2.308705882352941, "grad_norm": 1.2051098466828916, "learning_rate": 0.0002490816409237714, "loss": 0.6523, "step": 4906 }, { "epoch": 2.3091764705882354, "grad_norm": 0.75155510363424, "learning_rate": 0.0002490487250575069, "loss": 0.4434, "step": 4907 }, { "epoch": 2.309647058823529, "grad_norm": 0.7819642898821509, "learning_rate": 0.00024901580073199536, "loss": 0.4453, "step": 4908 }, { "epoch": 2.3101176470588234, "grad_norm": 0.7573750844327709, "learning_rate": 0.0002489828679500488, "loss": 0.5547, "step": 4909 }, { "epoch": 2.3105882352941176, "grad_norm": 0.8968589432795405, "learning_rate": 0.00024894992671447983, "loss": 0.7383, "step": 4910 }, { "epoch": 2.311058823529412, "grad_norm": 1.0002077021711597, "learning_rate": 0.0002489169770281018, "loss": 0.9375, "step": 4911 }, { "epoch": 2.311529411764706, "grad_norm": 1.0634859468064408, "learning_rate": 0.0002488840188937287, "loss": 0.9141, "step": 4912 }, { "epoch": 2.312, "grad_norm": 1.1099645306924975, "learning_rate": 0.00024885105231417536, "loss": 1.2188, "step": 4913 }, { "epoch": 2.312470588235294, "grad_norm": 1.0030046052379085, "learning_rate": 0.00024881807729225727, "loss": 1.2422, "step": 4914 }, { "epoch": 2.3129411764705883, "grad_norm": 0.9852508018532317, "learning_rate": 0.00024878509383079075, "loss": 0.7617, "step": 4915 }, { "epoch": 2.3134117647058825, "grad_norm": 0.8800988381133547, "learning_rate": 0.0002487521019325926, "loss": 0.9883, "step": 4916 }, { "epoch": 2.3138823529411763, "grad_norm": 0.9904905461634671, "learning_rate": 0.00024871910160048063, "loss": 1.125, "step": 4917 }, { "epoch": 2.3143529411764705, "grad_norm": 1.0524706688507723, "learning_rate": 0.00024868609283727316, "loss": 1.4375, "step": 4918 }, { "epoch": 2.3148235294117647, "grad_norm": 1.2273000952852582, "learning_rate": 0.00024865307564578927, "loss": 1.3516, "step": 4919 }, { "epoch": 2.315294117647059, "grad_norm": 1.2674281583829545, "learning_rate": 0.00024862005002884886, "loss": 1.4922, "step": 4920 }, { "epoch": 2.315764705882353, "grad_norm": 1.2124460661309595, "learning_rate": 0.0002485870159892724, "loss": 1.8438, "step": 4921 }, { "epoch": 2.316235294117647, "grad_norm": 0.9906538061620349, "learning_rate": 0.00024855397352988127, "loss": 1.6719, "step": 4922 }, { "epoch": 2.316705882352941, "grad_norm": 1.1681456855984815, "learning_rate": 0.0002485209226534974, "loss": 0.8594, "step": 4923 }, { "epoch": 2.3171764705882354, "grad_norm": 0.8042287770961768, "learning_rate": 0.00024848786336294346, "loss": 0.6484, "step": 4924 }, { "epoch": 2.317647058823529, "grad_norm": 0.7360232946663717, "learning_rate": 0.0002484547956610429, "loss": 0.6719, "step": 4925 }, { "epoch": 2.3181176470588234, "grad_norm": 1.3424309035083564, "learning_rate": 0.0002484217195506199, "loss": 1.8438, "step": 4926 }, { "epoch": 2.3185882352941176, "grad_norm": 0.9139100060177284, "learning_rate": 0.00024838863503449926, "loss": 0.918, "step": 4927 }, { "epoch": 2.319058823529412, "grad_norm": 1.215220359018926, "learning_rate": 0.00024835554211550664, "loss": 0.6992, "step": 4928 }, { "epoch": 2.319529411764706, "grad_norm": 0.6232505455304749, "learning_rate": 0.00024832244079646827, "loss": 0.293, "step": 4929 }, { "epoch": 2.32, "grad_norm": 1.171489633487046, "learning_rate": 0.00024828933108021116, "loss": 1.6484, "step": 4930 }, { "epoch": 2.320470588235294, "grad_norm": 0.6612524527139592, "learning_rate": 0.00024825621296956313, "loss": 0.5586, "step": 4931 }, { "epoch": 2.3209411764705883, "grad_norm": 1.1918255309285213, "learning_rate": 0.00024822308646735254, "loss": 1.5938, "step": 4932 }, { "epoch": 2.3214117647058825, "grad_norm": 0.6745735190388911, "learning_rate": 0.00024818995157640855, "loss": 0.5117, "step": 4933 }, { "epoch": 2.3218823529411763, "grad_norm": 0.5510646734659652, "learning_rate": 0.00024815680829956115, "loss": 0.707, "step": 4934 }, { "epoch": 2.3223529411764705, "grad_norm": 1.1436315955858658, "learning_rate": 0.0002481236566396408, "loss": 1.5781, "step": 4935 }, { "epoch": 2.3228235294117647, "grad_norm": 0.8976338831551743, "learning_rate": 0.0002480904965994789, "loss": 1.0547, "step": 4936 }, { "epoch": 2.323294117647059, "grad_norm": 1.1563725075428066, "learning_rate": 0.00024805732818190747, "loss": 1.2734, "step": 4937 }, { "epoch": 2.323764705882353, "grad_norm": 0.8242519139055263, "learning_rate": 0.00024802415138975924, "loss": 0.6367, "step": 4938 }, { "epoch": 2.324235294117647, "grad_norm": 0.8962149747569783, "learning_rate": 0.00024799096622586767, "loss": 0.957, "step": 4939 }, { "epoch": 2.324705882352941, "grad_norm": 0.9639735929598375, "learning_rate": 0.00024795777269306696, "loss": 0.9414, "step": 4940 }, { "epoch": 2.3251764705882354, "grad_norm": 1.31154466454838, "learning_rate": 0.0002479245707941919, "loss": 2.0312, "step": 4941 }, { "epoch": 2.325647058823529, "grad_norm": 1.287325048013954, "learning_rate": 0.0002478913605320783, "loss": 1.3672, "step": 4942 }, { "epoch": 2.3261176470588234, "grad_norm": 0.7720444409342125, "learning_rate": 0.0002478581419095623, "loss": 0.5039, "step": 4943 }, { "epoch": 2.3265882352941176, "grad_norm": 0.7699273966530237, "learning_rate": 0.0002478249149294809, "loss": 0.4785, "step": 4944 }, { "epoch": 2.327058823529412, "grad_norm": 1.3806129208624986, "learning_rate": 0.00024779167959467207, "loss": 1.2422, "step": 4945 }, { "epoch": 2.327529411764706, "grad_norm": 1.209280099359724, "learning_rate": 0.0002477584359079741, "loss": 1.5859, "step": 4946 }, { "epoch": 2.328, "grad_norm": 0.6370983681703254, "learning_rate": 0.00024772518387222616, "loss": 0.6055, "step": 4947 }, { "epoch": 2.328470588235294, "grad_norm": 0.6171372858623267, "learning_rate": 0.0002476919234902682, "loss": 0.3652, "step": 4948 }, { "epoch": 2.3289411764705883, "grad_norm": 0.7402071740330277, "learning_rate": 0.0002476586547649408, "loss": 0.4355, "step": 4949 }, { "epoch": 2.3294117647058825, "grad_norm": 1.1534710309468272, "learning_rate": 0.00024762537769908535, "loss": 1.75, "step": 4950 }, { "epoch": 2.3298823529411763, "grad_norm": 0.9406693646692896, "learning_rate": 0.0002475920922955437, "loss": 1.0312, "step": 4951 }, { "epoch": 2.3303529411764705, "grad_norm": 0.8519116002147896, "learning_rate": 0.0002475587985571587, "loss": 0.6992, "step": 4952 }, { "epoch": 2.3308235294117647, "grad_norm": 0.730849870454697, "learning_rate": 0.0002475254964867738, "loss": 0.3535, "step": 4953 }, { "epoch": 2.331294117647059, "grad_norm": 1.2743537651251273, "learning_rate": 0.0002474921860872331, "loss": 1.2891, "step": 4954 }, { "epoch": 2.331764705882353, "grad_norm": 0.632878994995954, "learning_rate": 0.00024745886736138154, "loss": 0.3867, "step": 4955 }, { "epoch": 2.332235294117647, "grad_norm": 0.8811841433985346, "learning_rate": 0.0002474255403120647, "loss": 0.3867, "step": 4956 }, { "epoch": 2.332705882352941, "grad_norm": 0.975757277937653, "learning_rate": 0.0002473922049421288, "loss": 0.6562, "step": 4957 }, { "epoch": 2.3331764705882354, "grad_norm": 1.4753312635895657, "learning_rate": 0.0002473588612544209, "loss": 1.8984, "step": 4958 }, { "epoch": 2.333647058823529, "grad_norm": 0.9158090011482672, "learning_rate": 0.0002473255092517887, "loss": 1.3281, "step": 4959 }, { "epoch": 2.3341176470588234, "grad_norm": 1.037942306779119, "learning_rate": 0.0002472921489370807, "loss": 1.2031, "step": 4960 }, { "epoch": 2.3345882352941176, "grad_norm": 1.0156404986578658, "learning_rate": 0.0002472587803131459, "loss": 1.0625, "step": 4961 }, { "epoch": 2.335058823529412, "grad_norm": 1.1417140092935756, "learning_rate": 0.0002472254033828342, "loss": 0.7773, "step": 4962 }, { "epoch": 2.335529411764706, "grad_norm": 1.1783589213416934, "learning_rate": 0.00024719201814899617, "loss": 1.1094, "step": 4963 }, { "epoch": 2.336, "grad_norm": 0.8527958198348294, "learning_rate": 0.0002471586246144831, "loss": 0.5508, "step": 4964 }, { "epoch": 2.336470588235294, "grad_norm": 1.4145058192781799, "learning_rate": 0.0002471252227821469, "loss": 1.0781, "step": 4965 }, { "epoch": 2.3369411764705883, "grad_norm": 1.07957723401872, "learning_rate": 0.0002470918126548403, "loss": 1.1953, "step": 4966 }, { "epoch": 2.3374117647058825, "grad_norm": 0.6788899658894052, "learning_rate": 0.00024705839423541664, "loss": 0.2852, "step": 4967 }, { "epoch": 2.3378823529411763, "grad_norm": 0.6954730005976645, "learning_rate": 0.00024702496752673005, "loss": 0.5273, "step": 4968 }, { "epoch": 2.3383529411764705, "grad_norm": 0.23945256393036501, "learning_rate": 0.00024699153253163533, "loss": 0.0496, "step": 4969 }, { "epoch": 2.3388235294117647, "grad_norm": 1.0206614358870965, "learning_rate": 0.000246958089252988, "loss": 0.8203, "step": 4970 }, { "epoch": 2.339294117647059, "grad_norm": 0.9855145299670837, "learning_rate": 0.0002469246376936443, "loss": 1.125, "step": 4971 }, { "epoch": 2.339764705882353, "grad_norm": 1.5982023504936165, "learning_rate": 0.00024689117785646106, "loss": 1.5938, "step": 4972 }, { "epoch": 2.340235294117647, "grad_norm": 0.686249664570335, "learning_rate": 0.000246857709744296, "loss": 0.4766, "step": 4973 }, { "epoch": 2.340705882352941, "grad_norm": 1.0088073101425925, "learning_rate": 0.00024682423336000746, "loss": 0.5234, "step": 4974 }, { "epoch": 2.3411764705882354, "grad_norm": 1.3758199915699167, "learning_rate": 0.00024679074870645446, "loss": 1.9609, "step": 4975 }, { "epoch": 2.341647058823529, "grad_norm": 0.9205530244477036, "learning_rate": 0.0002467572557864968, "loss": 0.6328, "step": 4976 }, { "epoch": 2.3421176470588234, "grad_norm": 2.772277236266322, "learning_rate": 0.0002467237546029949, "loss": 0.7852, "step": 4977 }, { "epoch": 2.3425882352941176, "grad_norm": 0.7953066371029966, "learning_rate": 0.0002466902451588099, "loss": 0.4258, "step": 4978 }, { "epoch": 2.343058823529412, "grad_norm": 0.8245528102335458, "learning_rate": 0.0002466567274568037, "loss": 0.543, "step": 4979 }, { "epoch": 2.343529411764706, "grad_norm": 0.8817999982227446, "learning_rate": 0.0002466232014998389, "loss": 0.4355, "step": 4980 }, { "epoch": 2.344, "grad_norm": 1.0348229217965978, "learning_rate": 0.00024658966729077875, "loss": 1.1016, "step": 4981 }, { "epoch": 2.344470588235294, "grad_norm": 1.5918423245118536, "learning_rate": 0.00024655612483248727, "loss": 1.4844, "step": 4982 }, { "epoch": 2.3449411764705883, "grad_norm": 0.7657899769479112, "learning_rate": 0.00024652257412782904, "loss": 0.4395, "step": 4983 }, { "epoch": 2.3454117647058825, "grad_norm": 1.0748617096923079, "learning_rate": 0.0002464890151796696, "loss": 1.5625, "step": 4984 }, { "epoch": 2.3458823529411763, "grad_norm": 0.9953092008266656, "learning_rate": 0.000246455447990875, "loss": 1.1641, "step": 4985 }, { "epoch": 2.3463529411764705, "grad_norm": 2.7872128022592215, "learning_rate": 0.000246421872564312, "loss": 1.0234, "step": 4986 }, { "epoch": 2.3468235294117648, "grad_norm": 1.26128705346399, "learning_rate": 0.0002463882889028481, "loss": 1.7891, "step": 4987 }, { "epoch": 2.347294117647059, "grad_norm": 0.998610009444413, "learning_rate": 0.0002463546970093516, "loss": 1.1562, "step": 4988 }, { "epoch": 2.3477647058823528, "grad_norm": 1.1327959141426605, "learning_rate": 0.00024632109688669135, "loss": 1.2969, "step": 4989 }, { "epoch": 2.348235294117647, "grad_norm": 0.6894124865563214, "learning_rate": 0.000246287488537737, "loss": 0.6172, "step": 4990 }, { "epoch": 2.348705882352941, "grad_norm": 1.1066807960340599, "learning_rate": 0.00024625387196535885, "loss": 1.3359, "step": 4991 }, { "epoch": 2.3491764705882354, "grad_norm": 0.8728268354736496, "learning_rate": 0.0002462202471724278, "loss": 0.6211, "step": 4992 }, { "epoch": 2.349647058823529, "grad_norm": 0.681971023955021, "learning_rate": 0.0002461866141618158, "loss": 0.4512, "step": 4993 }, { "epoch": 2.3501176470588234, "grad_norm": 1.3565540409592038, "learning_rate": 0.0002461529729363951, "loss": 1.9141, "step": 4994 }, { "epoch": 2.3505882352941176, "grad_norm": 0.72960484833376, "learning_rate": 0.0002461193234990389, "loss": 0.4316, "step": 4995 }, { "epoch": 2.351058823529412, "grad_norm": 1.262069987601335, "learning_rate": 0.00024608566585262103, "loss": 1.4609, "step": 4996 }, { "epoch": 2.351529411764706, "grad_norm": 0.8537627023854467, "learning_rate": 0.000246052000000016, "loss": 0.8516, "step": 4997 }, { "epoch": 2.352, "grad_norm": 0.8273652441099452, "learning_rate": 0.00024601832594409905, "loss": 0.4785, "step": 4998 }, { "epoch": 2.352470588235294, "grad_norm": 0.6309487370066172, "learning_rate": 0.00024598464368774604, "loss": 0.3184, "step": 4999 }, { "epoch": 2.3529411764705883, "grad_norm": 0.9346872981295928, "learning_rate": 0.00024595095323383365, "loss": 0.7773, "step": 5000 }, { "epoch": 2.3534117647058825, "grad_norm": 1.032178831693569, "learning_rate": 0.0002459172545852393, "loss": 0.9648, "step": 5001 }, { "epoch": 2.3538823529411763, "grad_norm": 0.7109850573350969, "learning_rate": 0.0002458835477448408, "loss": 0.5, "step": 5002 }, { "epoch": 2.3543529411764705, "grad_norm": 1.054193951196541, "learning_rate": 0.00024584983271551717, "loss": 0.8906, "step": 5003 }, { "epoch": 2.3548235294117648, "grad_norm": 1.0551311094383249, "learning_rate": 0.0002458161095001476, "loss": 1.2734, "step": 5004 }, { "epoch": 2.355294117647059, "grad_norm": 0.8729212697916829, "learning_rate": 0.0002457823781016123, "loss": 0.7734, "step": 5005 }, { "epoch": 2.3557647058823528, "grad_norm": 0.8543666120931033, "learning_rate": 0.0002457486385227922, "loss": 0.75, "step": 5006 }, { "epoch": 2.356235294117647, "grad_norm": 1.1751762211987347, "learning_rate": 0.00024571489076656865, "loss": 1.2656, "step": 5007 }, { "epoch": 2.356705882352941, "grad_norm": 0.9647972117425356, "learning_rate": 0.0002456811348358239, "loss": 1.3594, "step": 5008 }, { "epoch": 2.3571764705882354, "grad_norm": 0.8904094106633949, "learning_rate": 0.000245647370733441, "loss": 0.6641, "step": 5009 }, { "epoch": 2.357647058823529, "grad_norm": 0.6974954089641767, "learning_rate": 0.0002456135984623034, "loss": 0.8359, "step": 5010 }, { "epoch": 2.3581176470588234, "grad_norm": 1.2432452175453017, "learning_rate": 0.0002455798180252956, "loss": 1.3828, "step": 5011 }, { "epoch": 2.3585882352941177, "grad_norm": 0.8355178538960719, "learning_rate": 0.0002455460294253025, "loss": 0.9492, "step": 5012 }, { "epoch": 2.359058823529412, "grad_norm": 0.5597139628432343, "learning_rate": 0.00024551223266520983, "loss": 0.375, "step": 5013 }, { "epoch": 2.359529411764706, "grad_norm": 0.9835097218566368, "learning_rate": 0.00024547842774790403, "loss": 1.1562, "step": 5014 }, { "epoch": 2.36, "grad_norm": 0.9560873605357788, "learning_rate": 0.0002454446146762721, "loss": 1.6562, "step": 5015 }, { "epoch": 2.360470588235294, "grad_norm": 0.8461066058075395, "learning_rate": 0.000245410793453202, "loss": 0.6953, "step": 5016 }, { "epoch": 2.3609411764705883, "grad_norm": 1.1472026675363425, "learning_rate": 0.00024537696408158215, "loss": 1.2266, "step": 5017 }, { "epoch": 2.3614117647058825, "grad_norm": 1.223278641083939, "learning_rate": 0.0002453431265643017, "loss": 1.0156, "step": 5018 }, { "epoch": 2.3618823529411763, "grad_norm": 1.0549550825784182, "learning_rate": 0.0002453092809042507, "loss": 1.6719, "step": 5019 }, { "epoch": 2.3623529411764705, "grad_norm": 1.1093928324758886, "learning_rate": 0.0002452754271043195, "loss": 0.9375, "step": 5020 }, { "epoch": 2.3628235294117648, "grad_norm": 1.13041731810004, "learning_rate": 0.0002452415651673995, "loss": 1.0078, "step": 5021 }, { "epoch": 2.363294117647059, "grad_norm": 0.8096529680072732, "learning_rate": 0.00024520769509638274, "loss": 0.6016, "step": 5022 }, { "epoch": 2.3637647058823528, "grad_norm": 0.8403094221996316, "learning_rate": 0.0002451738168941618, "loss": 1.1484, "step": 5023 }, { "epoch": 2.364235294117647, "grad_norm": 0.8666348415397884, "learning_rate": 0.0002451399305636301, "loss": 1.1797, "step": 5024 }, { "epoch": 2.364705882352941, "grad_norm": 1.0330435562881393, "learning_rate": 0.00024510603610768166, "loss": 1.3438, "step": 5025 }, { "epoch": 2.3651764705882354, "grad_norm": 0.7505626413984589, "learning_rate": 0.0002450721335292112, "loss": 0.4297, "step": 5026 }, { "epoch": 2.365647058823529, "grad_norm": 0.7822744759553881, "learning_rate": 0.00024503822283111423, "loss": 0.4727, "step": 5027 }, { "epoch": 2.3661176470588234, "grad_norm": 0.7409247065408648, "learning_rate": 0.0002450043040162869, "loss": 1.0312, "step": 5028 }, { "epoch": 2.3665882352941177, "grad_norm": 0.8775795779890008, "learning_rate": 0.000244970377087626, "loss": 0.6797, "step": 5029 }, { "epoch": 2.367058823529412, "grad_norm": 1.059823132620859, "learning_rate": 0.00024493644204802906, "loss": 1.8125, "step": 5030 }, { "epoch": 2.367529411764706, "grad_norm": 0.6271439005080143, "learning_rate": 0.0002449024989003943, "loss": 0.3887, "step": 5031 }, { "epoch": 2.368, "grad_norm": 0.9815211987824174, "learning_rate": 0.0002448685476476207, "loss": 1.25, "step": 5032 }, { "epoch": 2.368470588235294, "grad_norm": 0.652241083957951, "learning_rate": 0.0002448345882926077, "loss": 0.4688, "step": 5033 }, { "epoch": 2.3689411764705883, "grad_norm": 0.9780926564845607, "learning_rate": 0.0002448006208382558, "loss": 1.0859, "step": 5034 }, { "epoch": 2.369411764705882, "grad_norm": 1.1416420542531065, "learning_rate": 0.0002447666452874658, "loss": 1.4844, "step": 5035 }, { "epoch": 2.3698823529411763, "grad_norm": 0.8559358478012432, "learning_rate": 0.00024473266164313953, "loss": 0.5781, "step": 5036 }, { "epoch": 2.3703529411764706, "grad_norm": 0.8615395433905468, "learning_rate": 0.00024469866990817923, "loss": 0.8125, "step": 5037 }, { "epoch": 2.3708235294117648, "grad_norm": 0.8842627869650213, "learning_rate": 0.00024466467008548814, "loss": 0.6445, "step": 5038 }, { "epoch": 2.371294117647059, "grad_norm": 0.9477239100652051, "learning_rate": 0.0002446306621779698, "loss": 1.4609, "step": 5039 }, { "epoch": 2.3717647058823528, "grad_norm": 1.1995219943130837, "learning_rate": 0.00024459664618852877, "loss": 1.7188, "step": 5040 }, { "epoch": 2.372235294117647, "grad_norm": 1.0732627679079314, "learning_rate": 0.0002445626221200702, "loss": 0.6211, "step": 5041 }, { "epoch": 2.372705882352941, "grad_norm": 1.048285846501694, "learning_rate": 0.0002445285899754998, "loss": 0.6406, "step": 5042 }, { "epoch": 2.3731764705882354, "grad_norm": 1.1307561654743696, "learning_rate": 0.0002444945497577242, "loss": 1.5391, "step": 5043 }, { "epoch": 2.373647058823529, "grad_norm": 0.9555123155408044, "learning_rate": 0.0002444605014696506, "loss": 1.4297, "step": 5044 }, { "epoch": 2.3741176470588234, "grad_norm": 0.44048804898446065, "learning_rate": 0.0002444264451141868, "loss": 0.1631, "step": 5045 }, { "epoch": 2.3745882352941177, "grad_norm": 0.9201421229135756, "learning_rate": 0.0002443923806942414, "loss": 0.8438, "step": 5046 }, { "epoch": 2.375058823529412, "grad_norm": 0.7075894728193841, "learning_rate": 0.0002443583082127237, "loss": 0.5391, "step": 5047 }, { "epoch": 2.375529411764706, "grad_norm": 1.1617953761413438, "learning_rate": 0.0002443242276725437, "loss": 1.4844, "step": 5048 }, { "epoch": 2.376, "grad_norm": 1.286619028815187, "learning_rate": 0.00024429013907661197, "loss": 2.0938, "step": 5049 }, { "epoch": 2.376470588235294, "grad_norm": 0.7356336705328966, "learning_rate": 0.0002442560424278399, "loss": 0.4883, "step": 5050 }, { "epoch": 2.3769411764705883, "grad_norm": 1.1931511714200473, "learning_rate": 0.0002442219377291395, "loss": 1.2344, "step": 5051 }, { "epoch": 2.377411764705882, "grad_norm": 1.4310036692939934, "learning_rate": 0.00024418782498342336, "loss": 2.0781, "step": 5052 }, { "epoch": 2.3778823529411763, "grad_norm": 1.5855639370205619, "learning_rate": 0.00024415370419360506, "loss": 1.2656, "step": 5053 }, { "epoch": 2.3783529411764706, "grad_norm": 1.068141780105202, "learning_rate": 0.00024411957536259857, "loss": 1.4688, "step": 5054 }, { "epoch": 2.378823529411765, "grad_norm": 0.9723597973395663, "learning_rate": 0.0002440854384933187, "loss": 0.9375, "step": 5055 }, { "epoch": 2.379294117647059, "grad_norm": 0.9283259395930178, "learning_rate": 0.0002440512935886809, "loss": 0.5352, "step": 5056 }, { "epoch": 2.379764705882353, "grad_norm": 1.0883186034509154, "learning_rate": 0.0002440171406516013, "loss": 1.2812, "step": 5057 }, { "epoch": 2.380235294117647, "grad_norm": 0.7597014356886916, "learning_rate": 0.00024398297968499677, "loss": 0.7383, "step": 5058 }, { "epoch": 2.3807058823529412, "grad_norm": 1.034502437873498, "learning_rate": 0.00024394881069178477, "loss": 1.1562, "step": 5059 }, { "epoch": 2.3811764705882354, "grad_norm": 1.0579106276379873, "learning_rate": 0.00024391463367488348, "loss": 1.1641, "step": 5060 }, { "epoch": 2.3816470588235292, "grad_norm": 0.9768436626302842, "learning_rate": 0.00024388044863721188, "loss": 0.9961, "step": 5061 }, { "epoch": 2.3821176470588235, "grad_norm": 0.9101664546620813, "learning_rate": 0.00024384625558168948, "loss": 0.5625, "step": 5062 }, { "epoch": 2.3825882352941177, "grad_norm": 0.8449418186139809, "learning_rate": 0.0002438120545112365, "loss": 0.6914, "step": 5063 }, { "epoch": 2.383058823529412, "grad_norm": 0.6855705422052287, "learning_rate": 0.000243777845428774, "loss": 0.6055, "step": 5064 }, { "epoch": 2.383529411764706, "grad_norm": 0.7601138451004438, "learning_rate": 0.00024374362833722344, "loss": 0.4453, "step": 5065 }, { "epoch": 2.384, "grad_norm": 1.1797983990278305, "learning_rate": 0.00024370940323950724, "loss": 1.1484, "step": 5066 }, { "epoch": 2.384470588235294, "grad_norm": 1.107642499665735, "learning_rate": 0.0002436751701385484, "loss": 1.2734, "step": 5067 }, { "epoch": 2.3849411764705883, "grad_norm": 0.9404237860290966, "learning_rate": 0.0002436409290372705, "loss": 1.1016, "step": 5068 }, { "epoch": 2.385411764705882, "grad_norm": 1.0836029232117865, "learning_rate": 0.000243606679938598, "loss": 1.1484, "step": 5069 }, { "epoch": 2.3858823529411763, "grad_norm": 0.9846780335674884, "learning_rate": 0.00024357242284545587, "loss": 0.6055, "step": 5070 }, { "epoch": 2.3863529411764706, "grad_norm": 1.021169748394111, "learning_rate": 0.00024353815776076985, "loss": 0.7734, "step": 5071 }, { "epoch": 2.386823529411765, "grad_norm": 0.5637169467770946, "learning_rate": 0.00024350388468746635, "loss": 0.2461, "step": 5072 }, { "epoch": 2.387294117647059, "grad_norm": 0.6269304665694025, "learning_rate": 0.00024346960362847247, "loss": 0.1885, "step": 5073 }, { "epoch": 2.387764705882353, "grad_norm": 0.9883974153925108, "learning_rate": 0.000243435314586716, "loss": 1.0391, "step": 5074 }, { "epoch": 2.388235294117647, "grad_norm": 0.834491534698357, "learning_rate": 0.00024340101756512532, "loss": 0.6719, "step": 5075 }, { "epoch": 2.3887058823529412, "grad_norm": 0.5261927932224132, "learning_rate": 0.00024336671256662963, "loss": 0.1963, "step": 5076 }, { "epoch": 2.3891764705882355, "grad_norm": 0.8104320288840139, "learning_rate": 0.00024333239959415873, "loss": 0.334, "step": 5077 }, { "epoch": 2.3896470588235292, "grad_norm": 0.7122824071278221, "learning_rate": 0.0002432980786506431, "loss": 0.6953, "step": 5078 }, { "epoch": 2.3901176470588235, "grad_norm": 0.6311594778059657, "learning_rate": 0.00024326374973901395, "loss": 0.3418, "step": 5079 }, { "epoch": 2.3905882352941177, "grad_norm": 1.2214926323501643, "learning_rate": 0.0002432294128622031, "loss": 1.7188, "step": 5080 }, { "epoch": 2.391058823529412, "grad_norm": 0.5609163218174554, "learning_rate": 0.0002431950680231431, "loss": 0.3027, "step": 5081 }, { "epoch": 2.391529411764706, "grad_norm": 1.1469994717049672, "learning_rate": 0.00024316071522476718, "loss": 1.4375, "step": 5082 }, { "epoch": 2.392, "grad_norm": 1.0805165143537914, "learning_rate": 0.0002431263544700093, "loss": 0.8711, "step": 5083 }, { "epoch": 2.392470588235294, "grad_norm": 1.2350321093691419, "learning_rate": 0.00024309198576180388, "loss": 1.0391, "step": 5084 }, { "epoch": 2.3929411764705883, "grad_norm": 1.4548746818729401, "learning_rate": 0.00024305760910308627, "loss": 1.5703, "step": 5085 }, { "epoch": 2.393411764705882, "grad_norm": 1.1185683734880398, "learning_rate": 0.00024302322449679245, "loss": 1.7422, "step": 5086 }, { "epoch": 2.3938823529411764, "grad_norm": 0.5839781681648754, "learning_rate": 0.0002429888319458589, "loss": 0.1924, "step": 5087 }, { "epoch": 2.3943529411764706, "grad_norm": 0.9756631510388489, "learning_rate": 0.00024295443145322303, "loss": 1.2812, "step": 5088 }, { "epoch": 2.394823529411765, "grad_norm": 1.1944597255773888, "learning_rate": 0.0002429200230218228, "loss": 1.4062, "step": 5089 }, { "epoch": 2.395294117647059, "grad_norm": 1.1571459878166435, "learning_rate": 0.00024288560665459679, "loss": 1.3828, "step": 5090 }, { "epoch": 2.395764705882353, "grad_norm": 0.8923534476648999, "learning_rate": 0.0002428511823544844, "loss": 1.1484, "step": 5091 }, { "epoch": 2.396235294117647, "grad_norm": 0.8861895628140438, "learning_rate": 0.00024281675012442563, "loss": 0.7969, "step": 5092 }, { "epoch": 2.3967058823529412, "grad_norm": 1.091973924089993, "learning_rate": 0.00024278230996736109, "loss": 1.1562, "step": 5093 }, { "epoch": 2.3971764705882355, "grad_norm": 1.04867187171565, "learning_rate": 0.0002427478618862322, "loss": 0.9141, "step": 5094 }, { "epoch": 2.3976470588235292, "grad_norm": 0.9003262550258497, "learning_rate": 0.00024271340588398097, "loss": 1.1562, "step": 5095 }, { "epoch": 2.3981176470588235, "grad_norm": 1.1257206888163656, "learning_rate": 0.00024267894196355015, "loss": 1.2734, "step": 5096 }, { "epoch": 2.3985882352941177, "grad_norm": 1.0493781242000246, "learning_rate": 0.00024264447012788308, "loss": 1.2422, "step": 5097 }, { "epoch": 2.399058823529412, "grad_norm": 0.8193746497821695, "learning_rate": 0.00024260999037992384, "loss": 0.7109, "step": 5098 }, { "epoch": 2.3995294117647057, "grad_norm": 1.2800455942039402, "learning_rate": 0.00024257550272261714, "loss": 1.4141, "step": 5099 }, { "epoch": 2.4, "grad_norm": 1.1373049464825766, "learning_rate": 0.00024254100715890846, "loss": 2.2812, "step": 5100 }, { "epoch": 2.400470588235294, "grad_norm": 1.06358976328555, "learning_rate": 0.00024250650369174386, "loss": 1.0156, "step": 5101 }, { "epoch": 2.4009411764705884, "grad_norm": 1.299902623570041, "learning_rate": 0.0002424719923240701, "loss": 1.5938, "step": 5102 }, { "epoch": 2.401411764705882, "grad_norm": 0.6076259434206861, "learning_rate": 0.00024243747305883461, "loss": 0.3242, "step": 5103 }, { "epoch": 2.4018823529411764, "grad_norm": 0.79334575365562, "learning_rate": 0.00024240294589898554, "loss": 0.5234, "step": 5104 }, { "epoch": 2.4023529411764706, "grad_norm": 1.5717222330041958, "learning_rate": 0.00024236841084747165, "loss": 0.3945, "step": 5105 }, { "epoch": 2.402823529411765, "grad_norm": 1.2457897459756755, "learning_rate": 0.00024233386790724242, "loss": 0.793, "step": 5106 }, { "epoch": 2.403294117647059, "grad_norm": 0.8261530203016575, "learning_rate": 0.000242299317081248, "loss": 0.5586, "step": 5107 }, { "epoch": 2.403764705882353, "grad_norm": 0.9148363045729292, "learning_rate": 0.00024226475837243908, "loss": 0.5938, "step": 5108 }, { "epoch": 2.404235294117647, "grad_norm": 0.9541317495771106, "learning_rate": 0.00024223019178376732, "loss": 0.7695, "step": 5109 }, { "epoch": 2.4047058823529412, "grad_norm": 0.7264768818394658, "learning_rate": 0.00024219561731818475, "loss": 0.5703, "step": 5110 }, { "epoch": 2.4051764705882355, "grad_norm": 1.196937784032205, "learning_rate": 0.00024216103497864426, "loss": 1.0, "step": 5111 }, { "epoch": 2.4056470588235292, "grad_norm": 1.0072151641883837, "learning_rate": 0.00024212644476809934, "loss": 0.8281, "step": 5112 }, { "epoch": 2.4061176470588235, "grad_norm": 0.8437497328893439, "learning_rate": 0.0002420918466895042, "loss": 0.5195, "step": 5113 }, { "epoch": 2.4065882352941177, "grad_norm": 0.533027223761972, "learning_rate": 0.0002420572407458136, "loss": 0.1934, "step": 5114 }, { "epoch": 2.407058823529412, "grad_norm": 1.0432344297457041, "learning_rate": 0.0002420226269399831, "loss": 0.918, "step": 5115 }, { "epoch": 2.4075294117647057, "grad_norm": 1.3281497349735931, "learning_rate": 0.00024198800527496889, "loss": 1.4219, "step": 5116 }, { "epoch": 2.408, "grad_norm": 0.6309242368799401, "learning_rate": 0.00024195337575372788, "loss": 0.7773, "step": 5117 }, { "epoch": 2.408470588235294, "grad_norm": 1.0495610613020698, "learning_rate": 0.00024191873837921748, "loss": 0.9062, "step": 5118 }, { "epoch": 2.4089411764705884, "grad_norm": 0.5751783538482438, "learning_rate": 0.000241884093154396, "loss": 0.3613, "step": 5119 }, { "epoch": 2.409411764705882, "grad_norm": 0.9645871064211045, "learning_rate": 0.00024184944008222227, "loss": 0.4473, "step": 5120 }, { "epoch": 2.4098823529411764, "grad_norm": 0.3955211802339647, "learning_rate": 0.00024181477916565582, "loss": 0.1787, "step": 5121 }, { "epoch": 2.4103529411764706, "grad_norm": 0.9784381459094592, "learning_rate": 0.0002417801104076569, "loss": 1.0, "step": 5122 }, { "epoch": 2.410823529411765, "grad_norm": 1.181596745582703, "learning_rate": 0.00024174543381118638, "loss": 1.4844, "step": 5123 }, { "epoch": 2.411294117647059, "grad_norm": 1.0410836713477059, "learning_rate": 0.00024171074937920579, "loss": 0.8203, "step": 5124 }, { "epoch": 2.411764705882353, "grad_norm": 0.8519746702664455, "learning_rate": 0.00024167605711467738, "loss": 0.6289, "step": 5125 }, { "epoch": 2.412235294117647, "grad_norm": 0.7440295825274075, "learning_rate": 0.000241641357020564, "loss": 0.4805, "step": 5126 }, { "epoch": 2.4127058823529413, "grad_norm": 0.4959622043095565, "learning_rate": 0.00024160664909982928, "loss": 0.25, "step": 5127 }, { "epoch": 2.4131764705882355, "grad_norm": 1.419683559185877, "learning_rate": 0.00024157193335543734, "loss": 2.3438, "step": 5128 }, { "epoch": 2.4136470588235293, "grad_norm": 0.9337450481279248, "learning_rate": 0.00024153720979035317, "loss": 0.8281, "step": 5129 }, { "epoch": 2.4141176470588235, "grad_norm": 0.7630132638824273, "learning_rate": 0.00024150247840754228, "loss": 0.4609, "step": 5130 }, { "epoch": 2.4145882352941177, "grad_norm": 1.0171832805829473, "learning_rate": 0.00024146773920997091, "loss": 0.9336, "step": 5131 }, { "epoch": 2.415058823529412, "grad_norm": 1.5618604346116713, "learning_rate": 0.000241432992200606, "loss": 1.25, "step": 5132 }, { "epoch": 2.4155294117647057, "grad_norm": 1.0869228182847666, "learning_rate": 0.000241398237382415, "loss": 0.9102, "step": 5133 }, { "epoch": 2.416, "grad_norm": 1.0230647912854196, "learning_rate": 0.00024136347475836633, "loss": 1.75, "step": 5134 }, { "epoch": 2.416470588235294, "grad_norm": 0.804776924125877, "learning_rate": 0.0002413287043314287, "loss": 0.6367, "step": 5135 }, { "epoch": 2.4169411764705884, "grad_norm": 0.7684983023296582, "learning_rate": 0.00024129392610457176, "loss": 0.8633, "step": 5136 }, { "epoch": 2.417411764705882, "grad_norm": 0.7590671535699477, "learning_rate": 0.0002412591400807658, "loss": 0.5859, "step": 5137 }, { "epoch": 2.4178823529411764, "grad_norm": 0.9074253121033878, "learning_rate": 0.0002412243462629816, "loss": 0.5898, "step": 5138 }, { "epoch": 2.4183529411764706, "grad_norm": 1.0231233394048211, "learning_rate": 0.0002411895446541908, "loss": 1.1562, "step": 5139 }, { "epoch": 2.418823529411765, "grad_norm": 0.6727401180283102, "learning_rate": 0.00024115473525736561, "loss": 0.4551, "step": 5140 }, { "epoch": 2.419294117647059, "grad_norm": 1.1763323555107252, "learning_rate": 0.0002411199180754789, "loss": 0.9297, "step": 5141 }, { "epoch": 2.419764705882353, "grad_norm": 0.7309511574898694, "learning_rate": 0.00024108509311150422, "loss": 0.5312, "step": 5142 }, { "epoch": 2.420235294117647, "grad_norm": 0.896608206590387, "learning_rate": 0.00024105026036841582, "loss": 0.7461, "step": 5143 }, { "epoch": 2.4207058823529413, "grad_norm": 1.3550747726343801, "learning_rate": 0.00024101541984918864, "loss": 1.4766, "step": 5144 }, { "epoch": 2.4211764705882355, "grad_norm": 0.8121955091956534, "learning_rate": 0.00024098057155679813, "loss": 0.5117, "step": 5145 }, { "epoch": 2.4216470588235293, "grad_norm": 0.8151463165279185, "learning_rate": 0.0002409457154942206, "loss": 0.5312, "step": 5146 }, { "epoch": 2.4221176470588235, "grad_norm": 1.051784307737891, "learning_rate": 0.00024091085166443285, "loss": 1.0469, "step": 5147 }, { "epoch": 2.4225882352941177, "grad_norm": 1.0129707780711044, "learning_rate": 0.00024087598007041243, "loss": 0.6992, "step": 5148 }, { "epoch": 2.423058823529412, "grad_norm": 1.012219269280455, "learning_rate": 0.0002408411007151376, "loss": 1.2109, "step": 5149 }, { "epoch": 2.4235294117647057, "grad_norm": 0.701610385066585, "learning_rate": 0.00024080621360158717, "loss": 0.3984, "step": 5150 }, { "epoch": 2.424, "grad_norm": 0.8477672471052683, "learning_rate": 0.00024077131873274072, "loss": 0.4043, "step": 5151 }, { "epoch": 2.424470588235294, "grad_norm": 0.5763363472875932, "learning_rate": 0.0002407364161115784, "loss": 0.2891, "step": 5152 }, { "epoch": 2.4249411764705884, "grad_norm": 0.8602174616827414, "learning_rate": 0.00024070150574108113, "loss": 0.4785, "step": 5153 }, { "epoch": 2.425411764705882, "grad_norm": 1.4465893824021223, "learning_rate": 0.00024066658762423034, "loss": 1.5, "step": 5154 }, { "epoch": 2.4258823529411764, "grad_norm": 0.9080139985583572, "learning_rate": 0.00024063166176400825, "loss": 0.582, "step": 5155 }, { "epoch": 2.4263529411764706, "grad_norm": 0.8387216161070362, "learning_rate": 0.0002405967281633977, "loss": 1.0234, "step": 5156 }, { "epoch": 2.426823529411765, "grad_norm": 1.1035226754473892, "learning_rate": 0.00024056178682538222, "loss": 1.6328, "step": 5157 }, { "epoch": 2.427294117647059, "grad_norm": 1.2003162036584878, "learning_rate": 0.00024052683775294592, "loss": 1.8672, "step": 5158 }, { "epoch": 2.427764705882353, "grad_norm": 0.6700669380796315, "learning_rate": 0.00024049188094907363, "loss": 0.3242, "step": 5159 }, { "epoch": 2.428235294117647, "grad_norm": 0.7077413939064635, "learning_rate": 0.00024045691641675086, "loss": 0.3906, "step": 5160 }, { "epoch": 2.4287058823529413, "grad_norm": 1.4475355317149636, "learning_rate": 0.00024042194415896376, "loss": 1.5547, "step": 5161 }, { "epoch": 2.429176470588235, "grad_norm": 1.1774995427341461, "learning_rate": 0.00024038696417869908, "loss": 1.3203, "step": 5162 }, { "epoch": 2.4296470588235293, "grad_norm": 0.9955871489997902, "learning_rate": 0.00024035197647894434, "loss": 1.0859, "step": 5163 }, { "epoch": 2.4301176470588235, "grad_norm": 1.0065801287327145, "learning_rate": 0.00024031698106268763, "loss": 0.918, "step": 5164 }, { "epoch": 2.4305882352941177, "grad_norm": 0.909393883290606, "learning_rate": 0.0002402819779329177, "loss": 0.6172, "step": 5165 }, { "epoch": 2.431058823529412, "grad_norm": 1.0432621235288662, "learning_rate": 0.00024024696709262406, "loss": 0.6914, "step": 5166 }, { "epoch": 2.4315294117647057, "grad_norm": 0.9082961197234911, "learning_rate": 0.00024021194854479674, "loss": 0.9102, "step": 5167 }, { "epoch": 2.432, "grad_norm": 0.6486487441112921, "learning_rate": 0.00024017692229242655, "loss": 0.3555, "step": 5168 }, { "epoch": 2.432470588235294, "grad_norm": 0.9944384982330051, "learning_rate": 0.00024014188833850486, "loss": 0.9414, "step": 5169 }, { "epoch": 2.4329411764705884, "grad_norm": 0.8701071441764152, "learning_rate": 0.0002401068466860238, "loss": 0.9219, "step": 5170 }, { "epoch": 2.433411764705882, "grad_norm": 1.1603846015847168, "learning_rate": 0.00024007179733797604, "loss": 1.0469, "step": 5171 }, { "epoch": 2.4338823529411764, "grad_norm": 1.4348190065442137, "learning_rate": 0.00024003674029735497, "loss": 1.5625, "step": 5172 }, { "epoch": 2.4343529411764706, "grad_norm": 0.6773741505086107, "learning_rate": 0.00024000167556715468, "loss": 0.4414, "step": 5173 }, { "epoch": 2.434823529411765, "grad_norm": 1.2053559844270016, "learning_rate": 0.00023996660315036986, "loss": 1.0469, "step": 5174 }, { "epoch": 2.435294117647059, "grad_norm": 0.8821171647268227, "learning_rate": 0.00023993152304999582, "loss": 1.4219, "step": 5175 }, { "epoch": 2.435764705882353, "grad_norm": 0.855942130545817, "learning_rate": 0.0002398964352690286, "loss": 0.9297, "step": 5176 }, { "epoch": 2.436235294117647, "grad_norm": 0.891738023147949, "learning_rate": 0.00023986133981046485, "loss": 0.7969, "step": 5177 }, { "epoch": 2.4367058823529413, "grad_norm": 1.121800083115239, "learning_rate": 0.00023982623667730195, "loss": 1.2031, "step": 5178 }, { "epoch": 2.437176470588235, "grad_norm": 1.0128750474694028, "learning_rate": 0.0002397911258725378, "loss": 0.9258, "step": 5179 }, { "epoch": 2.4376470588235293, "grad_norm": 0.40635708820695776, "learning_rate": 0.00023975600739917113, "loss": 0.3965, "step": 5180 }, { "epoch": 2.4381176470588235, "grad_norm": 0.7721366542612772, "learning_rate": 0.00023972088126020114, "loss": 0.4668, "step": 5181 }, { "epoch": 2.4385882352941177, "grad_norm": 1.0534375229967994, "learning_rate": 0.00023968574745862782, "loss": 0.5, "step": 5182 }, { "epoch": 2.439058823529412, "grad_norm": 0.8673637162680292, "learning_rate": 0.0002396506059974518, "loss": 0.7227, "step": 5183 }, { "epoch": 2.4395294117647057, "grad_norm": 1.135149247725162, "learning_rate": 0.00023961545687967422, "loss": 1.7344, "step": 5184 }, { "epoch": 2.44, "grad_norm": 0.8581792797263638, "learning_rate": 0.0002395803001082971, "loss": 0.6992, "step": 5185 }, { "epoch": 2.440470588235294, "grad_norm": 1.0791296878348438, "learning_rate": 0.00023954513568632296, "loss": 1.4531, "step": 5186 }, { "epoch": 2.4409411764705884, "grad_norm": 2.0081205218873452, "learning_rate": 0.000239509963616755, "loss": 0.3027, "step": 5187 }, { "epoch": 2.441411764705882, "grad_norm": 0.9564399742815168, "learning_rate": 0.00023947478390259712, "loss": 0.6133, "step": 5188 }, { "epoch": 2.4418823529411764, "grad_norm": 0.7394569121164518, "learning_rate": 0.00023943959654685383, "loss": 0.4004, "step": 5189 }, { "epoch": 2.4423529411764706, "grad_norm": 0.6171610966476491, "learning_rate": 0.00023940440155253024, "loss": 0.5938, "step": 5190 }, { "epoch": 2.442823529411765, "grad_norm": 1.057725392592135, "learning_rate": 0.0002393691989226323, "loss": 0.8203, "step": 5191 }, { "epoch": 2.443294117647059, "grad_norm": 1.3334670862746099, "learning_rate": 0.00023933398866016642, "loss": 1.25, "step": 5192 }, { "epoch": 2.443764705882353, "grad_norm": 0.66806195994832, "learning_rate": 0.0002392987707681397, "loss": 0.3789, "step": 5193 }, { "epoch": 2.444235294117647, "grad_norm": 1.3216926974225267, "learning_rate": 0.00023926354524955994, "loss": 1.2422, "step": 5194 }, { "epoch": 2.4447058823529413, "grad_norm": 1.2044653963275418, "learning_rate": 0.0002392283121074356, "loss": 1.2422, "step": 5195 }, { "epoch": 2.445176470588235, "grad_norm": 1.0066287530965836, "learning_rate": 0.00023919307134477572, "loss": 0.9258, "step": 5196 }, { "epoch": 2.4456470588235293, "grad_norm": 0.6962106952795757, "learning_rate": 0.00023915782296459007, "loss": 0.6875, "step": 5197 }, { "epoch": 2.4461176470588235, "grad_norm": 0.5507851711546004, "learning_rate": 0.00023912256696988904, "loss": 0.2266, "step": 5198 }, { "epoch": 2.4465882352941177, "grad_norm": 0.7107471012111839, "learning_rate": 0.00023908730336368364, "loss": 0.4434, "step": 5199 }, { "epoch": 2.447058823529412, "grad_norm": 1.2077752851736763, "learning_rate": 0.00023905203214898558, "loss": 1.2812, "step": 5200 }, { "epoch": 2.4475294117647057, "grad_norm": 1.1086384066609305, "learning_rate": 0.00023901675332880715, "loss": 1.1953, "step": 5201 }, { "epoch": 2.448, "grad_norm": 0.9368375731764853, "learning_rate": 0.00023898146690616134, "loss": 0.8008, "step": 5202 }, { "epoch": 2.448470588235294, "grad_norm": 1.190725745036527, "learning_rate": 0.0002389461728840618, "loss": 0.9805, "step": 5203 }, { "epoch": 2.4489411764705884, "grad_norm": 0.6003149978092757, "learning_rate": 0.00023891087126552292, "loss": 0.5273, "step": 5204 }, { "epoch": 2.449411764705882, "grad_norm": 1.3841935872711908, "learning_rate": 0.00023887556205355945, "loss": 0.832, "step": 5205 }, { "epoch": 2.4498823529411764, "grad_norm": 0.9094981491347666, "learning_rate": 0.00023884024525118702, "loss": 0.7344, "step": 5206 }, { "epoch": 2.4503529411764706, "grad_norm": 1.1714977669430153, "learning_rate": 0.00023880492086142193, "loss": 1.375, "step": 5207 }, { "epoch": 2.450823529411765, "grad_norm": 0.9332847718460745, "learning_rate": 0.00023876958888728098, "loss": 0.8203, "step": 5208 }, { "epoch": 2.451294117647059, "grad_norm": 0.9854705383865546, "learning_rate": 0.00023873424933178178, "loss": 1.0703, "step": 5209 }, { "epoch": 2.451764705882353, "grad_norm": 0.9373474814843002, "learning_rate": 0.00023869890219794238, "loss": 0.9219, "step": 5210 }, { "epoch": 2.452235294117647, "grad_norm": 0.8054544919929949, "learning_rate": 0.0002386635474887817, "loss": 0.3203, "step": 5211 }, { "epoch": 2.4527058823529413, "grad_norm": 0.7766630025577869, "learning_rate": 0.0002386281852073192, "loss": 0.8555, "step": 5212 }, { "epoch": 2.453176470588235, "grad_norm": 0.6782212861926913, "learning_rate": 0.00023859281535657496, "loss": 0.2773, "step": 5213 }, { "epoch": 2.4536470588235293, "grad_norm": 1.1313290916821093, "learning_rate": 0.00023855743793956967, "loss": 0.8555, "step": 5214 }, { "epoch": 2.4541176470588235, "grad_norm": 0.9217886915888603, "learning_rate": 0.0002385220529593249, "loss": 0.8477, "step": 5215 }, { "epoch": 2.4545882352941177, "grad_norm": 0.8469218464495647, "learning_rate": 0.00023848666041886258, "loss": 0.7461, "step": 5216 }, { "epoch": 2.455058823529412, "grad_norm": 0.9527737788028972, "learning_rate": 0.00023845126032120539, "loss": 1.5938, "step": 5217 }, { "epoch": 2.4555294117647057, "grad_norm": 0.7298451104194991, "learning_rate": 0.00023841585266937677, "loss": 0.4316, "step": 5218 }, { "epoch": 2.456, "grad_norm": 0.47109213869087896, "learning_rate": 0.00023838043746640065, "loss": 0.1533, "step": 5219 }, { "epoch": 2.456470588235294, "grad_norm": 0.8829584393866916, "learning_rate": 0.00023834501471530165, "loss": 0.7461, "step": 5220 }, { "epoch": 2.4569411764705884, "grad_norm": 0.4664184409620717, "learning_rate": 0.0002383095844191051, "loss": 0.2236, "step": 5221 }, { "epoch": 2.457411764705882, "grad_norm": 0.6998167607700777, "learning_rate": 0.0002382741465808368, "loss": 0.4805, "step": 5222 }, { "epoch": 2.4578823529411764, "grad_norm": 0.6370750049243685, "learning_rate": 0.0002382387012035235, "loss": 0.5352, "step": 5223 }, { "epoch": 2.4583529411764706, "grad_norm": 1.194420065825109, "learning_rate": 0.0002382032482901922, "loss": 1.4766, "step": 5224 }, { "epoch": 2.458823529411765, "grad_norm": 1.4440532228731768, "learning_rate": 0.00023816778784387094, "loss": 1.7734, "step": 5225 }, { "epoch": 2.4592941176470586, "grad_norm": 0.8863124761921989, "learning_rate": 0.00023813231986758815, "loss": 0.6641, "step": 5226 }, { "epoch": 2.459764705882353, "grad_norm": 0.423460364781996, "learning_rate": 0.00023809684436437294, "loss": 0.1309, "step": 5227 }, { "epoch": 2.460235294117647, "grad_norm": 0.7580009238081192, "learning_rate": 0.00023806136133725506, "loss": 0.5078, "step": 5228 }, { "epoch": 2.4607058823529413, "grad_norm": 0.6194383337779283, "learning_rate": 0.00023802587078926505, "loss": 0.2275, "step": 5229 }, { "epoch": 2.461176470588235, "grad_norm": 0.6857499347196937, "learning_rate": 0.00023799037272343387, "loss": 0.4121, "step": 5230 }, { "epoch": 2.4616470588235293, "grad_norm": 0.5328941162373143, "learning_rate": 0.00023795486714279327, "loss": 0.2422, "step": 5231 }, { "epoch": 2.4621176470588235, "grad_norm": 0.5135856266343501, "learning_rate": 0.0002379193540503756, "loss": 0.2158, "step": 5232 }, { "epoch": 2.4625882352941177, "grad_norm": 1.2145896154440103, "learning_rate": 0.00023788383344921384, "loss": 1.2812, "step": 5233 }, { "epoch": 2.463058823529412, "grad_norm": 0.9790935296594244, "learning_rate": 0.0002378483053423417, "loss": 1.2188, "step": 5234 }, { "epoch": 2.4635294117647057, "grad_norm": 0.9264383508297478, "learning_rate": 0.00023781276973279335, "loss": 0.8984, "step": 5235 }, { "epoch": 2.464, "grad_norm": 1.192822442617013, "learning_rate": 0.0002377772266236037, "loss": 1.0703, "step": 5236 }, { "epoch": 2.464470588235294, "grad_norm": 1.0212274669835022, "learning_rate": 0.0002377416760178084, "loss": 1.1484, "step": 5237 }, { "epoch": 2.4649411764705884, "grad_norm": 0.8324800398789358, "learning_rate": 0.00023770611791844364, "loss": 0.6602, "step": 5238 }, { "epoch": 2.465411764705882, "grad_norm": 0.9539627627730436, "learning_rate": 0.00023767055232854615, "loss": 1.4062, "step": 5239 }, { "epoch": 2.4658823529411764, "grad_norm": 0.9156659544034766, "learning_rate": 0.00023763497925115347, "loss": 0.3848, "step": 5240 }, { "epoch": 2.4663529411764706, "grad_norm": 1.0161082383829076, "learning_rate": 0.00023759939868930373, "loss": 1.1094, "step": 5241 }, { "epoch": 2.466823529411765, "grad_norm": 0.9575575589266142, "learning_rate": 0.00023756381064603564, "loss": 0.9141, "step": 5242 }, { "epoch": 2.4672941176470586, "grad_norm": 1.0565568112528072, "learning_rate": 0.0002375282151243887, "loss": 0.8516, "step": 5243 }, { "epoch": 2.467764705882353, "grad_norm": 0.946412794426698, "learning_rate": 0.00023749261212740278, "loss": 0.9023, "step": 5244 }, { "epoch": 2.468235294117647, "grad_norm": 0.681643886965831, "learning_rate": 0.00023745700165811868, "loss": 0.3008, "step": 5245 }, { "epoch": 2.4687058823529413, "grad_norm": 1.2247533747230637, "learning_rate": 0.00023742138371957765, "loss": 1.0, "step": 5246 }, { "epoch": 2.469176470588235, "grad_norm": 0.5760000272821425, "learning_rate": 0.00023738575831482164, "loss": 0.5625, "step": 5247 }, { "epoch": 2.4696470588235293, "grad_norm": 1.0958862855356941, "learning_rate": 0.0002373501254468933, "loss": 1.3047, "step": 5248 }, { "epoch": 2.4701176470588235, "grad_norm": 1.0700630985138735, "learning_rate": 0.00023731448511883575, "loss": 1.1406, "step": 5249 }, { "epoch": 2.4705882352941178, "grad_norm": 0.8886833642524284, "learning_rate": 0.00023727883733369292, "loss": 0.5039, "step": 5250 }, { "epoch": 2.471058823529412, "grad_norm": 0.7706147671517214, "learning_rate": 0.00023724318209450925, "loss": 0.5039, "step": 5251 }, { "epoch": 2.4715294117647058, "grad_norm": 0.914089071517524, "learning_rate": 0.00023720751940432992, "loss": 1.0859, "step": 5252 }, { "epoch": 2.472, "grad_norm": 0.73931811921891, "learning_rate": 0.0002371718492662007, "loss": 0.5508, "step": 5253 }, { "epoch": 2.472470588235294, "grad_norm": 1.1274767129625005, "learning_rate": 0.00023713617168316797, "loss": 1.2109, "step": 5254 }, { "epoch": 2.4729411764705884, "grad_norm": 0.7543242987063605, "learning_rate": 0.00023710048665827872, "loss": 1.0234, "step": 5255 }, { "epoch": 2.473411764705882, "grad_norm": 0.7532653048834494, "learning_rate": 0.00023706479419458075, "loss": 0.3203, "step": 5256 }, { "epoch": 2.4738823529411764, "grad_norm": 0.8911476859645353, "learning_rate": 0.00023702909429512228, "loss": 0.9727, "step": 5257 }, { "epoch": 2.4743529411764706, "grad_norm": 0.9353169384428712, "learning_rate": 0.00023699338696295228, "loss": 1.1484, "step": 5258 }, { "epoch": 2.474823529411765, "grad_norm": 0.9616890089815651, "learning_rate": 0.00023695767220112033, "loss": 1.0859, "step": 5259 }, { "epoch": 2.4752941176470586, "grad_norm": 1.00996730081516, "learning_rate": 0.0002369219500126767, "loss": 1.0234, "step": 5260 }, { "epoch": 2.475764705882353, "grad_norm": 1.420439668318261, "learning_rate": 0.00023688622040067217, "loss": 1.9844, "step": 5261 }, { "epoch": 2.476235294117647, "grad_norm": 0.9347829924816321, "learning_rate": 0.0002368504833681582, "loss": 1.2578, "step": 5262 }, { "epoch": 2.4767058823529413, "grad_norm": 0.7322480690680024, "learning_rate": 0.00023681473891818699, "loss": 0.4648, "step": 5263 }, { "epoch": 2.477176470588235, "grad_norm": 0.5895218414752545, "learning_rate": 0.00023677898705381123, "loss": 0.4766, "step": 5264 }, { "epoch": 2.4776470588235293, "grad_norm": 1.1486428777695052, "learning_rate": 0.00023674322777808432, "loss": 1.6406, "step": 5265 }, { "epoch": 2.4781176470588235, "grad_norm": 0.8175523033035033, "learning_rate": 0.0002367074610940603, "loss": 0.6992, "step": 5266 }, { "epoch": 2.4785882352941178, "grad_norm": 0.7141831439198932, "learning_rate": 0.00023667168700479383, "loss": 0.4121, "step": 5267 }, { "epoch": 2.479058823529412, "grad_norm": 0.940421653506081, "learning_rate": 0.0002366359055133401, "loss": 1.2734, "step": 5268 }, { "epoch": 2.4795294117647058, "grad_norm": 0.8691819478588284, "learning_rate": 0.00023660011662275516, "loss": 0.8125, "step": 5269 }, { "epoch": 2.48, "grad_norm": 0.9584958824698061, "learning_rate": 0.00023656432033609544, "loss": 0.8555, "step": 5270 }, { "epoch": 2.480470588235294, "grad_norm": 1.0760135303025489, "learning_rate": 0.0002365285166564182, "loss": 0.6992, "step": 5271 }, { "epoch": 2.4809411764705884, "grad_norm": 0.8206211703587447, "learning_rate": 0.00023649270558678123, "loss": 0.8398, "step": 5272 }, { "epoch": 2.481411764705882, "grad_norm": 0.6395334370986511, "learning_rate": 0.00023645688713024295, "loss": 0.4551, "step": 5273 }, { "epoch": 2.4818823529411764, "grad_norm": 0.9482863514355002, "learning_rate": 0.0002364210612898624, "loss": 1.0781, "step": 5274 }, { "epoch": 2.4823529411764707, "grad_norm": 0.8596075469356124, "learning_rate": 0.00023638522806869935, "loss": 0.7109, "step": 5275 }, { "epoch": 2.482823529411765, "grad_norm": 0.8026534162895409, "learning_rate": 0.0002363493874698141, "loss": 0.6367, "step": 5276 }, { "epoch": 2.4832941176470587, "grad_norm": 0.7852373885926203, "learning_rate": 0.0002363135394962676, "loss": 0.4121, "step": 5277 }, { "epoch": 2.483764705882353, "grad_norm": 1.3569095265050104, "learning_rate": 0.00023627768415112154, "loss": 1.0156, "step": 5278 }, { "epoch": 2.484235294117647, "grad_norm": 0.9353282072009352, "learning_rate": 0.00023624182143743797, "loss": 1.3594, "step": 5279 }, { "epoch": 2.4847058823529413, "grad_norm": 0.7629750190934674, "learning_rate": 0.00023620595135827985, "loss": 0.5781, "step": 5280 }, { "epoch": 2.485176470588235, "grad_norm": 0.840794036348513, "learning_rate": 0.0002361700739167107, "loss": 0.4434, "step": 5281 }, { "epoch": 2.4856470588235293, "grad_norm": 1.137213773702047, "learning_rate": 0.00023613418911579454, "loss": 1.2109, "step": 5282 }, { "epoch": 2.4861176470588235, "grad_norm": 0.6174669805370843, "learning_rate": 0.00023609829695859618, "loss": 0.7773, "step": 5283 }, { "epoch": 2.4865882352941178, "grad_norm": 0.6169019039156374, "learning_rate": 0.00023606239744818094, "loss": 0.3555, "step": 5284 }, { "epoch": 2.487058823529412, "grad_norm": 0.5298595152140316, "learning_rate": 0.0002360264905876148, "loss": 0.2158, "step": 5285 }, { "epoch": 2.4875294117647058, "grad_norm": 0.8060402828407235, "learning_rate": 0.0002359905763799644, "loss": 0.4902, "step": 5286 }, { "epoch": 2.488, "grad_norm": 0.6599438100290221, "learning_rate": 0.00023595465482829697, "loss": 0.4023, "step": 5287 }, { "epoch": 2.488470588235294, "grad_norm": 0.8314994176133895, "learning_rate": 0.0002359187259356805, "loss": 1.2188, "step": 5288 }, { "epoch": 2.488941176470588, "grad_norm": 0.8399598823874302, "learning_rate": 0.00023588278970518333, "loss": 0.7109, "step": 5289 }, { "epoch": 2.489411764705882, "grad_norm": 1.05263839725922, "learning_rate": 0.00023584684613987472, "loss": 1.4609, "step": 5290 }, { "epoch": 2.4898823529411764, "grad_norm": 0.7484480767615528, "learning_rate": 0.00023581089524282432, "loss": 0.9375, "step": 5291 }, { "epoch": 2.4903529411764707, "grad_norm": 0.5523081037410278, "learning_rate": 0.0002357749370171026, "loss": 0.5859, "step": 5292 }, { "epoch": 2.490823529411765, "grad_norm": 1.2244964103186347, "learning_rate": 0.00023573897146578053, "loss": 0.7969, "step": 5293 }, { "epoch": 2.4912941176470587, "grad_norm": 0.6894718335532353, "learning_rate": 0.00023570299859192976, "loss": 0.4824, "step": 5294 }, { "epoch": 2.491764705882353, "grad_norm": 0.9816141562533195, "learning_rate": 0.00023566701839862253, "loss": 0.7852, "step": 5295 }, { "epoch": 2.492235294117647, "grad_norm": 0.977776113093807, "learning_rate": 0.00023563103088893172, "loss": 0.9102, "step": 5296 }, { "epoch": 2.4927058823529413, "grad_norm": 1.3599732401468998, "learning_rate": 0.00023559503606593088, "loss": 1.2891, "step": 5297 }, { "epoch": 2.493176470588235, "grad_norm": 0.6774066340097618, "learning_rate": 0.0002355590339326941, "loss": 0.3887, "step": 5298 }, { "epoch": 2.4936470588235293, "grad_norm": 1.0884866225699341, "learning_rate": 0.00023552302449229612, "loss": 1.2734, "step": 5299 }, { "epoch": 2.4941176470588236, "grad_norm": 0.9184792547551967, "learning_rate": 0.00023548700774781242, "loss": 0.8906, "step": 5300 }, { "epoch": 2.4945882352941178, "grad_norm": 1.0943117294578892, "learning_rate": 0.00023545098370231893, "loss": 1.4688, "step": 5301 }, { "epoch": 2.495058823529412, "grad_norm": 1.0224032412937047, "learning_rate": 0.00023541495235889227, "loss": 1.2891, "step": 5302 }, { "epoch": 2.4955294117647058, "grad_norm": 0.6891598909443293, "learning_rate": 0.0002353789137206098, "loss": 0.7578, "step": 5303 }, { "epoch": 2.496, "grad_norm": 0.6682794428213494, "learning_rate": 0.00023534286779054923, "loss": 0.4414, "step": 5304 }, { "epoch": 2.496470588235294, "grad_norm": 1.1134533465330911, "learning_rate": 0.00023530681457178916, "loss": 1.0234, "step": 5305 }, { "epoch": 2.496941176470588, "grad_norm": 0.8940880990458157, "learning_rate": 0.00023527075406740876, "loss": 0.7891, "step": 5306 }, { "epoch": 2.497411764705882, "grad_norm": 1.1045693015933702, "learning_rate": 0.00023523468628048764, "loss": 1.2344, "step": 5307 }, { "epoch": 2.4978823529411764, "grad_norm": 0.8646756161173231, "learning_rate": 0.0002351986112141063, "loss": 0.7461, "step": 5308 }, { "epoch": 2.4983529411764707, "grad_norm": 0.8246209166123377, "learning_rate": 0.00023516252887134558, "loss": 1.0312, "step": 5309 }, { "epoch": 2.498823529411765, "grad_norm": 0.5061365193519302, "learning_rate": 0.00023512643925528725, "loss": 0.1494, "step": 5310 }, { "epoch": 2.4992941176470587, "grad_norm": 1.0592741686163294, "learning_rate": 0.00023509034236901348, "loss": 0.6523, "step": 5311 }, { "epoch": 2.499764705882353, "grad_norm": 1.391205895443806, "learning_rate": 0.00023505423821560705, "loss": 1.8594, "step": 5312 }, { "epoch": 2.500235294117647, "grad_norm": 0.6662655514357666, "learning_rate": 0.00023501812679815152, "loss": 0.3086, "step": 5313 }, { "epoch": 2.5007058823529413, "grad_norm": 0.7719766442719794, "learning_rate": 0.00023498200811973092, "loss": 0.4043, "step": 5314 }, { "epoch": 2.501176470588235, "grad_norm": 0.53431975878053, "learning_rate": 0.00023494588218343003, "loss": 0.208, "step": 5315 }, { "epoch": 2.5016470588235293, "grad_norm": 0.8270270996247342, "learning_rate": 0.0002349097489923341, "loss": 0.4414, "step": 5316 }, { "epoch": 2.5021176470588236, "grad_norm": 0.9520798790085485, "learning_rate": 0.00023487360854952915, "loss": 0.625, "step": 5317 }, { "epoch": 2.502588235294118, "grad_norm": 0.9290779274437383, "learning_rate": 0.00023483746085810176, "loss": 0.9961, "step": 5318 }, { "epoch": 2.503058823529412, "grad_norm": 1.0548393954970114, "learning_rate": 0.00023480130592113904, "loss": 0.8242, "step": 5319 }, { "epoch": 2.503529411764706, "grad_norm": 0.8871982810041074, "learning_rate": 0.00023476514374172887, "loss": 0.8672, "step": 5320 }, { "epoch": 2.504, "grad_norm": 0.7785547583857061, "learning_rate": 0.0002347289743229596, "loss": 0.498, "step": 5321 }, { "epoch": 2.5044705882352942, "grad_norm": 0.6400861650339028, "learning_rate": 0.00023469279766792037, "loss": 0.707, "step": 5322 }, { "epoch": 2.504941176470588, "grad_norm": 0.9566298757813017, "learning_rate": 0.0002346566137797008, "loss": 1.0547, "step": 5323 }, { "epoch": 2.5054117647058822, "grad_norm": 0.9298645426311576, "learning_rate": 0.0002346204226613912, "loss": 0.668, "step": 5324 }, { "epoch": 2.5058823529411764, "grad_norm": 0.7551746557826768, "learning_rate": 0.0002345842243160824, "loss": 0.5508, "step": 5325 }, { "epoch": 2.5063529411764707, "grad_norm": 1.0888806112660878, "learning_rate": 0.00023454801874686596, "loss": 1.1875, "step": 5326 }, { "epoch": 2.506823529411765, "grad_norm": 0.7994184375964988, "learning_rate": 0.00023451180595683409, "loss": 0.6289, "step": 5327 }, { "epoch": 2.5072941176470587, "grad_norm": 0.7040139947808769, "learning_rate": 0.0002344755859490794, "loss": 0.4961, "step": 5328 }, { "epoch": 2.507764705882353, "grad_norm": 1.2329084406735193, "learning_rate": 0.00023443935872669534, "loss": 1.5156, "step": 5329 }, { "epoch": 2.508235294117647, "grad_norm": 0.9134287769656011, "learning_rate": 0.00023440312429277583, "loss": 0.9609, "step": 5330 }, { "epoch": 2.5087058823529413, "grad_norm": 0.9808375315951625, "learning_rate": 0.00023436688265041555, "loss": 0.7266, "step": 5331 }, { "epoch": 2.509176470588235, "grad_norm": 0.8631221045577676, "learning_rate": 0.0002343306338027097, "loss": 0.8984, "step": 5332 }, { "epoch": 2.5096470588235293, "grad_norm": 0.786336720533533, "learning_rate": 0.00023429437775275406, "loss": 1.1875, "step": 5333 }, { "epoch": 2.5101176470588236, "grad_norm": 1.0504816178048915, "learning_rate": 0.00023425811450364508, "loss": 1.4688, "step": 5334 }, { "epoch": 2.510588235294118, "grad_norm": 1.3828949805336108, "learning_rate": 0.00023422184405847985, "loss": 1.0, "step": 5335 }, { "epoch": 2.511058823529412, "grad_norm": 0.8415960714288944, "learning_rate": 0.00023418556642035602, "loss": 0.498, "step": 5336 }, { "epoch": 2.511529411764706, "grad_norm": 1.2369897731190354, "learning_rate": 0.0002341492815923719, "loss": 1.8672, "step": 5337 }, { "epoch": 2.512, "grad_norm": 0.8309970359473333, "learning_rate": 0.00023411298957762636, "loss": 1.1797, "step": 5338 }, { "epoch": 2.5124705882352942, "grad_norm": 1.0781031296206909, "learning_rate": 0.00023407669037921902, "loss": 1.3281, "step": 5339 }, { "epoch": 2.512941176470588, "grad_norm": 1.077978698862514, "learning_rate": 0.00023404038400024984, "loss": 1.3203, "step": 5340 }, { "epoch": 2.5134117647058822, "grad_norm": 0.9686222173134853, "learning_rate": 0.00023400407044381965, "loss": 1.1094, "step": 5341 }, { "epoch": 2.5138823529411765, "grad_norm": 0.8286624319384918, "learning_rate": 0.00023396774971302985, "loss": 0.5078, "step": 5342 }, { "epoch": 2.5143529411764707, "grad_norm": 0.8205624927564392, "learning_rate": 0.00023393142181098235, "loss": 0.5, "step": 5343 }, { "epoch": 2.514823529411765, "grad_norm": 1.2487224973037598, "learning_rate": 0.00023389508674077973, "loss": 1.4609, "step": 5344 }, { "epoch": 2.5152941176470587, "grad_norm": 1.0178419320269214, "learning_rate": 0.00023385874450552522, "loss": 1.125, "step": 5345 }, { "epoch": 2.515764705882353, "grad_norm": 1.810127116664347, "learning_rate": 0.00023382239510832262, "loss": 2.2969, "step": 5346 }, { "epoch": 2.516235294117647, "grad_norm": 0.8488926648085815, "learning_rate": 0.0002337860385522763, "loss": 0.9219, "step": 5347 }, { "epoch": 2.516705882352941, "grad_norm": 1.2101768997928009, "learning_rate": 0.00023374967484049132, "loss": 1.1172, "step": 5348 }, { "epoch": 2.517176470588235, "grad_norm": 0.6454043532531911, "learning_rate": 0.00023371330397607335, "loss": 0.6719, "step": 5349 }, { "epoch": 2.5176470588235293, "grad_norm": 1.1691022044995574, "learning_rate": 0.00023367692596212858, "loss": 1.6328, "step": 5350 }, { "epoch": 2.5181176470588236, "grad_norm": 0.6423504402182661, "learning_rate": 0.0002336405408017639, "loss": 0.2734, "step": 5351 }, { "epoch": 2.518588235294118, "grad_norm": 1.1680019553624879, "learning_rate": 0.0002336041484980868, "loss": 0.9648, "step": 5352 }, { "epoch": 2.519058823529412, "grad_norm": 1.2429958475886798, "learning_rate": 0.0002335677490542053, "loss": 0.4492, "step": 5353 }, { "epoch": 2.519529411764706, "grad_norm": 0.5884591906262742, "learning_rate": 0.0002335313424732282, "loss": 0.6602, "step": 5354 }, { "epoch": 2.52, "grad_norm": 0.7069937718082245, "learning_rate": 0.0002334949287582647, "loss": 0.8359, "step": 5355 }, { "epoch": 2.5204705882352942, "grad_norm": 0.7915484126080554, "learning_rate": 0.00023345850791242477, "loss": 0.5664, "step": 5356 }, { "epoch": 2.520941176470588, "grad_norm": 1.0183293471221986, "learning_rate": 0.00023342207993881883, "loss": 1.4219, "step": 5357 }, { "epoch": 2.5214117647058822, "grad_norm": 0.5788511848185349, "learning_rate": 0.0002333856448405581, "loss": 0.7031, "step": 5358 }, { "epoch": 2.5218823529411765, "grad_norm": 0.9713463628101517, "learning_rate": 0.00023334920262075437, "loss": 1.1797, "step": 5359 }, { "epoch": 2.5223529411764707, "grad_norm": 0.7944826402123205, "learning_rate": 0.00023331275328251983, "loss": 1.3125, "step": 5360 }, { "epoch": 2.522823529411765, "grad_norm": 0.8034712950308877, "learning_rate": 0.00023327629682896758, "loss": 0.5586, "step": 5361 }, { "epoch": 2.5232941176470587, "grad_norm": 0.7978837559571246, "learning_rate": 0.00023323983326321106, "loss": 1.1328, "step": 5362 }, { "epoch": 2.523764705882353, "grad_norm": 0.2734134626356121, "learning_rate": 0.0002332033625883645, "loss": 0.0309, "step": 5363 }, { "epoch": 2.524235294117647, "grad_norm": 1.159223291272834, "learning_rate": 0.00023316688480754267, "loss": 1.4297, "step": 5364 }, { "epoch": 2.524705882352941, "grad_norm": 0.5273609351803327, "learning_rate": 0.00023313039992386094, "loss": 0.1992, "step": 5365 }, { "epoch": 2.525176470588235, "grad_norm": 0.6528804796305463, "learning_rate": 0.00023309390794043532, "loss": 0.4453, "step": 5366 }, { "epoch": 2.5256470588235294, "grad_norm": 1.3544531789693226, "learning_rate": 0.00023305740886038235, "loss": 2.0625, "step": 5367 }, { "epoch": 2.5261176470588236, "grad_norm": 0.4954261907768377, "learning_rate": 0.00023302090268681932, "loss": 0.2949, "step": 5368 }, { "epoch": 2.526588235294118, "grad_norm": 0.662934029371449, "learning_rate": 0.00023298438942286395, "loss": 0.3184, "step": 5369 }, { "epoch": 2.527058823529412, "grad_norm": 0.8165662619863275, "learning_rate": 0.00023294786907163473, "loss": 0.5234, "step": 5370 }, { "epoch": 2.527529411764706, "grad_norm": 0.7111893459332335, "learning_rate": 0.0002329113416362506, "loss": 0.4668, "step": 5371 }, { "epoch": 2.528, "grad_norm": 1.0434278002387978, "learning_rate": 0.00023287480711983125, "loss": 0.4492, "step": 5372 }, { "epoch": 2.5284705882352942, "grad_norm": 1.0059995256039123, "learning_rate": 0.00023283826552549682, "loss": 1.1562, "step": 5373 }, { "epoch": 2.528941176470588, "grad_norm": 0.9721474059443517, "learning_rate": 0.00023280171685636822, "loss": 1.2266, "step": 5374 }, { "epoch": 2.5294117647058822, "grad_norm": 0.31795685319065753, "learning_rate": 0.0002327651611155669, "loss": 0.0845, "step": 5375 }, { "epoch": 2.5298823529411765, "grad_norm": 0.7524227928246248, "learning_rate": 0.0002327285983062148, "loss": 0.6094, "step": 5376 }, { "epoch": 2.5303529411764707, "grad_norm": 0.9451845020040134, "learning_rate": 0.00023269202843143462, "loss": 0.8555, "step": 5377 }, { "epoch": 2.530823529411765, "grad_norm": 0.7159636544497104, "learning_rate": 0.00023265545149434968, "loss": 0.4238, "step": 5378 }, { "epoch": 2.5312941176470587, "grad_norm": 1.3403054253882944, "learning_rate": 0.0002326188674980837, "loss": 1.6953, "step": 5379 }, { "epoch": 2.531764705882353, "grad_norm": 0.6289645014093221, "learning_rate": 0.0002325822764457612, "loss": 0.4004, "step": 5380 }, { "epoch": 2.532235294117647, "grad_norm": 1.2779865837533688, "learning_rate": 0.0002325456783405072, "loss": 1.7578, "step": 5381 }, { "epoch": 2.532705882352941, "grad_norm": 0.7914809531605537, "learning_rate": 0.00023250907318544744, "loss": 0.4141, "step": 5382 }, { "epoch": 2.533176470588235, "grad_norm": 0.8565110092980367, "learning_rate": 0.0002324724609837081, "loss": 0.6133, "step": 5383 }, { "epoch": 2.5336470588235294, "grad_norm": 1.3454733220189614, "learning_rate": 0.0002324358417384161, "loss": 1.875, "step": 5384 }, { "epoch": 2.5341176470588236, "grad_norm": 1.271288716470642, "learning_rate": 0.00023239921545269884, "loss": 1.8984, "step": 5385 }, { "epoch": 2.534588235294118, "grad_norm": 0.9306966994306379, "learning_rate": 0.00023236258212968445, "loss": 0.668, "step": 5386 }, { "epoch": 2.535058823529412, "grad_norm": 0.7820914201088167, "learning_rate": 0.00023232594177250148, "loss": 0.3965, "step": 5387 }, { "epoch": 2.535529411764706, "grad_norm": 1.0967170733691192, "learning_rate": 0.00023228929438427932, "loss": 0.6055, "step": 5388 }, { "epoch": 2.536, "grad_norm": 0.9511497135170587, "learning_rate": 0.0002322526399681478, "loss": 1.2734, "step": 5389 }, { "epoch": 2.5364705882352943, "grad_norm": 0.23232624650358452, "learning_rate": 0.0002322159785272374, "loss": 0.0322, "step": 5390 }, { "epoch": 2.536941176470588, "grad_norm": 1.0838834548583778, "learning_rate": 0.00023217931006467912, "loss": 1.0078, "step": 5391 }, { "epoch": 2.5374117647058823, "grad_norm": 0.9745521525723304, "learning_rate": 0.00023214263458360467, "loss": 0.8164, "step": 5392 }, { "epoch": 2.5378823529411765, "grad_norm": 1.011003324862869, "learning_rate": 0.00023210595208714635, "loss": 1.5156, "step": 5393 }, { "epoch": 2.5383529411764707, "grad_norm": 0.9203881052883093, "learning_rate": 0.00023206926257843702, "loss": 0.6875, "step": 5394 }, { "epoch": 2.538823529411765, "grad_norm": 0.6590802161285604, "learning_rate": 0.00023203256606061008, "loss": 0.3711, "step": 5395 }, { "epoch": 2.5392941176470587, "grad_norm": 0.5725562491224294, "learning_rate": 0.00023199586253679964, "loss": 0.8867, "step": 5396 }, { "epoch": 2.539764705882353, "grad_norm": 0.9394045230177428, "learning_rate": 0.00023195915201014035, "loss": 0.9766, "step": 5397 }, { "epoch": 2.540235294117647, "grad_norm": 1.0825964538994686, "learning_rate": 0.00023192243448376745, "loss": 1.2422, "step": 5398 }, { "epoch": 2.540705882352941, "grad_norm": 0.8128790124877474, "learning_rate": 0.00023188570996081687, "loss": 0.7695, "step": 5399 }, { "epoch": 2.541176470588235, "grad_norm": 0.8629447036258628, "learning_rate": 0.00023184897844442495, "loss": 1.4297, "step": 5400 }, { "epoch": 2.5416470588235294, "grad_norm": 1.2192901239441916, "learning_rate": 0.00023181223993772886, "loss": 1.2109, "step": 5401 }, { "epoch": 2.5421176470588236, "grad_norm": 0.9461062847877837, "learning_rate": 0.00023177549444386618, "loss": 0.6836, "step": 5402 }, { "epoch": 2.542588235294118, "grad_norm": 0.7051664729112239, "learning_rate": 0.00023173874196597515, "loss": 0.3652, "step": 5403 }, { "epoch": 2.543058823529412, "grad_norm": 1.0278814234733262, "learning_rate": 0.00023170198250719464, "loss": 1.4766, "step": 5404 }, { "epoch": 2.543529411764706, "grad_norm": 1.0181523116629771, "learning_rate": 0.00023166521607066409, "loss": 1.4219, "step": 5405 }, { "epoch": 2.544, "grad_norm": 0.8969647845479297, "learning_rate": 0.00023162844265952348, "loss": 0.5234, "step": 5406 }, { "epoch": 2.5444705882352943, "grad_norm": 0.952626708401414, "learning_rate": 0.0002315916622769135, "loss": 0.9141, "step": 5407 }, { "epoch": 2.544941176470588, "grad_norm": 1.1511032649280821, "learning_rate": 0.00023155487492597536, "loss": 0.7383, "step": 5408 }, { "epoch": 2.5454117647058823, "grad_norm": 1.0851269762819546, "learning_rate": 0.0002315180806098509, "loss": 1.3438, "step": 5409 }, { "epoch": 2.5458823529411765, "grad_norm": 1.1371772900038648, "learning_rate": 0.00023148127933168245, "loss": 1.5781, "step": 5410 }, { "epoch": 2.5463529411764707, "grad_norm": 0.8257304016071914, "learning_rate": 0.00023144447109461314, "loss": 0.291, "step": 5411 }, { "epoch": 2.546823529411765, "grad_norm": 0.9719876659461969, "learning_rate": 0.0002314076559017865, "loss": 1.1719, "step": 5412 }, { "epoch": 2.5472941176470587, "grad_norm": 0.9483511910276156, "learning_rate": 0.00023137083375634676, "loss": 1.4297, "step": 5413 }, { "epoch": 2.547764705882353, "grad_norm": 0.9862121610665903, "learning_rate": 0.00023133400466143867, "loss": 0.7773, "step": 5414 }, { "epoch": 2.548235294117647, "grad_norm": 0.9102015345231099, "learning_rate": 0.00023129716862020766, "loss": 0.7305, "step": 5415 }, { "epoch": 2.548705882352941, "grad_norm": 0.8463498956085982, "learning_rate": 0.0002312603256357997, "loss": 0.5234, "step": 5416 }, { "epoch": 2.549176470588235, "grad_norm": 1.233625520645093, "learning_rate": 0.00023122347571136127, "loss": 1.5391, "step": 5417 }, { "epoch": 2.5496470588235294, "grad_norm": 0.7671466331201571, "learning_rate": 0.00023118661885003969, "loss": 0.5586, "step": 5418 }, { "epoch": 2.5501176470588236, "grad_norm": 0.9323182105976184, "learning_rate": 0.00023114975505498264, "loss": 0.9648, "step": 5419 }, { "epoch": 2.550588235294118, "grad_norm": 1.0974920278318203, "learning_rate": 0.00023111288432933847, "loss": 1.3125, "step": 5420 }, { "epoch": 2.5510588235294116, "grad_norm": 0.6979930171983102, "learning_rate": 0.0002310760066762561, "loss": 0.418, "step": 5421 }, { "epoch": 2.551529411764706, "grad_norm": 1.156117684753964, "learning_rate": 0.00023103912209888512, "loss": 0.9961, "step": 5422 }, { "epoch": 2.552, "grad_norm": 0.9923753149734339, "learning_rate": 0.00023100223060037558, "loss": 1.2188, "step": 5423 }, { "epoch": 2.5524705882352943, "grad_norm": 1.0992371805409509, "learning_rate": 0.00023096533218387827, "loss": 1.1641, "step": 5424 }, { "epoch": 2.552941176470588, "grad_norm": 1.24310260140291, "learning_rate": 0.00023092842685254442, "loss": 1.1484, "step": 5425 }, { "epoch": 2.5534117647058823, "grad_norm": 0.7715670138047127, "learning_rate": 0.000230891514609526, "loss": 1.2188, "step": 5426 }, { "epoch": 2.5538823529411765, "grad_norm": 0.9118017627799281, "learning_rate": 0.00023085459545797542, "loss": 1.1016, "step": 5427 }, { "epoch": 2.5543529411764707, "grad_norm": 0.7965427019756588, "learning_rate": 0.00023081766940104585, "loss": 0.416, "step": 5428 }, { "epoch": 2.554823529411765, "grad_norm": 0.8572820458417668, "learning_rate": 0.00023078073644189085, "loss": 0.5234, "step": 5429 }, { "epoch": 2.5552941176470587, "grad_norm": 0.8203940238682939, "learning_rate": 0.00023074379658366478, "loss": 0.5156, "step": 5430 }, { "epoch": 2.555764705882353, "grad_norm": 1.0439949396851795, "learning_rate": 0.0002307068498295224, "loss": 1.1719, "step": 5431 }, { "epoch": 2.556235294117647, "grad_norm": 0.9944030841392381, "learning_rate": 0.00023066989618261919, "loss": 0.8047, "step": 5432 }, { "epoch": 2.556705882352941, "grad_norm": 0.8848496361558887, "learning_rate": 0.00023063293564611114, "loss": 0.7383, "step": 5433 }, { "epoch": 2.557176470588235, "grad_norm": 1.1696781347029406, "learning_rate": 0.00023059596822315491, "loss": 0.9062, "step": 5434 }, { "epoch": 2.5576470588235294, "grad_norm": 0.7537901244693078, "learning_rate": 0.00023055899391690766, "loss": 0.8945, "step": 5435 }, { "epoch": 2.5581176470588236, "grad_norm": 0.8523807734347817, "learning_rate": 0.00023052201273052723, "loss": 0.7969, "step": 5436 }, { "epoch": 2.558588235294118, "grad_norm": 0.8597912323333718, "learning_rate": 0.00023048502466717192, "loss": 0.7305, "step": 5437 }, { "epoch": 2.5590588235294116, "grad_norm": 0.8733522843955173, "learning_rate": 0.00023044802973000073, "loss": 0.9727, "step": 5438 }, { "epoch": 2.559529411764706, "grad_norm": 0.5006431699399937, "learning_rate": 0.00023041102792217323, "loss": 0.1895, "step": 5439 }, { "epoch": 2.56, "grad_norm": 0.4855807626799689, "learning_rate": 0.00023037401924684946, "loss": 0.1914, "step": 5440 }, { "epoch": 2.5604705882352943, "grad_norm": 0.9470501937662757, "learning_rate": 0.00023033700370719026, "loss": 1.2266, "step": 5441 }, { "epoch": 2.560941176470588, "grad_norm": 0.8266562200648545, "learning_rate": 0.00023029998130635692, "loss": 1.0, "step": 5442 }, { "epoch": 2.5614117647058823, "grad_norm": 0.31591868554473335, "learning_rate": 0.00023026295204751126, "loss": 0.0645, "step": 5443 }, { "epoch": 2.5618823529411765, "grad_norm": 1.2142076883585382, "learning_rate": 0.00023022591593381582, "loss": 0.9883, "step": 5444 }, { "epoch": 2.5623529411764707, "grad_norm": 0.8876784534958565, "learning_rate": 0.0002301888729684337, "loss": 0.7422, "step": 5445 }, { "epoch": 2.562823529411765, "grad_norm": 1.3951101639034773, "learning_rate": 0.00023015182315452843, "loss": 1.3125, "step": 5446 }, { "epoch": 2.5632941176470587, "grad_norm": 0.6704428304264962, "learning_rate": 0.00023011476649526438, "loss": 0.252, "step": 5447 }, { "epoch": 2.563764705882353, "grad_norm": 0.5532306977640389, "learning_rate": 0.0002300777029938063, "loss": 0.2354, "step": 5448 }, { "epoch": 2.564235294117647, "grad_norm": 0.9719307725150944, "learning_rate": 0.0002300406326533196, "loss": 0.7344, "step": 5449 }, { "epoch": 2.564705882352941, "grad_norm": 0.6407855952488114, "learning_rate": 0.00023000355547697027, "loss": 0.2871, "step": 5450 }, { "epoch": 2.565176470588235, "grad_norm": 0.8584858952829261, "learning_rate": 0.00022996647146792488, "loss": 0.6641, "step": 5451 }, { "epoch": 2.5656470588235294, "grad_norm": 0.6777819743598813, "learning_rate": 0.0002299293806293506, "loss": 0.2988, "step": 5452 }, { "epoch": 2.5661176470588236, "grad_norm": 1.1876601898054266, "learning_rate": 0.00022989228296441515, "loss": 1.1406, "step": 5453 }, { "epoch": 2.566588235294118, "grad_norm": 1.1903143194062598, "learning_rate": 0.00022985517847628686, "loss": 0.8125, "step": 5454 }, { "epoch": 2.5670588235294116, "grad_norm": 1.0400063323757833, "learning_rate": 0.0002298180671681347, "loss": 1.2891, "step": 5455 }, { "epoch": 2.567529411764706, "grad_norm": 0.5301285482709583, "learning_rate": 0.00022978094904312803, "loss": 0.4961, "step": 5456 }, { "epoch": 2.568, "grad_norm": 0.9097603061137811, "learning_rate": 0.000229743824104437, "loss": 0.5, "step": 5457 }, { "epoch": 2.5684705882352943, "grad_norm": 1.0230198280338927, "learning_rate": 0.00022970669235523224, "loss": 1.4375, "step": 5458 }, { "epoch": 2.568941176470588, "grad_norm": 0.6824700642870365, "learning_rate": 0.00022966955379868502, "loss": 0.2754, "step": 5459 }, { "epoch": 2.5694117647058823, "grad_norm": 1.0139364448461181, "learning_rate": 0.00022963240843796714, "loss": 0.7812, "step": 5460 }, { "epoch": 2.5698823529411765, "grad_norm": 1.2655855596685208, "learning_rate": 0.000229595256276251, "loss": 1.75, "step": 5461 }, { "epoch": 2.5703529411764707, "grad_norm": 0.6792072883949668, "learning_rate": 0.00022955809731670955, "loss": 0.6328, "step": 5462 }, { "epoch": 2.570823529411765, "grad_norm": 0.8351901597599412, "learning_rate": 0.00022952093156251635, "loss": 1.2422, "step": 5463 }, { "epoch": 2.5712941176470587, "grad_norm": 0.763667218630852, "learning_rate": 0.0002294837590168455, "loss": 0.6211, "step": 5464 }, { "epoch": 2.571764705882353, "grad_norm": 1.0903194598193402, "learning_rate": 0.00022944657968287184, "loss": 1.5234, "step": 5465 }, { "epoch": 2.572235294117647, "grad_norm": 0.7187750756539527, "learning_rate": 0.00022940939356377052, "loss": 0.2734, "step": 5466 }, { "epoch": 2.572705882352941, "grad_norm": 0.8475589183255805, "learning_rate": 0.00022937220066271752, "loss": 0.9727, "step": 5467 }, { "epoch": 2.573176470588235, "grad_norm": 1.352220018780092, "learning_rate": 0.0002293350009828893, "loss": 1.6172, "step": 5468 }, { "epoch": 2.5736470588235294, "grad_norm": 1.4833710347844034, "learning_rate": 0.00022929779452746283, "loss": 0.7305, "step": 5469 }, { "epoch": 2.5741176470588236, "grad_norm": 0.6347234548840298, "learning_rate": 0.0002292605812996157, "loss": 0.2793, "step": 5470 }, { "epoch": 2.574588235294118, "grad_norm": 0.6370639485070113, "learning_rate": 0.00022922336130252622, "loss": 0.6562, "step": 5471 }, { "epoch": 2.5750588235294116, "grad_norm": 1.031406470730728, "learning_rate": 0.00022918613453937314, "loss": 1.6328, "step": 5472 }, { "epoch": 2.575529411764706, "grad_norm": 1.1021397976850822, "learning_rate": 0.00022914890101333573, "loss": 1.5391, "step": 5473 }, { "epoch": 2.576, "grad_norm": 1.4679368109141568, "learning_rate": 0.00022911166072759386, "loss": 1.8516, "step": 5474 }, { "epoch": 2.576470588235294, "grad_norm": 1.2611027414673892, "learning_rate": 0.00022907441368532822, "loss": 1.2578, "step": 5475 }, { "epoch": 2.576941176470588, "grad_norm": 0.9352805942603035, "learning_rate": 0.00022903715988971973, "loss": 0.8359, "step": 5476 }, { "epoch": 2.5774117647058823, "grad_norm": 0.7530588047631959, "learning_rate": 0.0002289998993439502, "loss": 0.6953, "step": 5477 }, { "epoch": 2.5778823529411765, "grad_norm": 0.7926497503171681, "learning_rate": 0.00022896263205120167, "loss": 0.6367, "step": 5478 }, { "epoch": 2.5783529411764707, "grad_norm": 0.7991883913303502, "learning_rate": 0.0002289253580146571, "loss": 0.4004, "step": 5479 }, { "epoch": 2.578823529411765, "grad_norm": 0.5037769266529687, "learning_rate": 0.00022888807723749988, "loss": 0.168, "step": 5480 }, { "epoch": 2.5792941176470587, "grad_norm": 1.1074512884326253, "learning_rate": 0.0002288507897229138, "loss": 1.5078, "step": 5481 }, { "epoch": 2.579764705882353, "grad_norm": 0.5235299875228067, "learning_rate": 0.00022881349547408363, "loss": 0.1992, "step": 5482 }, { "epoch": 2.580235294117647, "grad_norm": 1.0555748589208116, "learning_rate": 0.00022877619449419436, "loss": 1.0234, "step": 5483 }, { "epoch": 2.580705882352941, "grad_norm": 0.7587357132423742, "learning_rate": 0.00022873888678643164, "loss": 0.4375, "step": 5484 }, { "epoch": 2.581176470588235, "grad_norm": 0.9596720253075588, "learning_rate": 0.0002287015723539818, "loss": 1.0469, "step": 5485 }, { "epoch": 2.5816470588235294, "grad_norm": 0.7706656191378292, "learning_rate": 0.00022866425120003165, "loss": 0.498, "step": 5486 }, { "epoch": 2.5821176470588236, "grad_norm": 0.6240427948334446, "learning_rate": 0.0002286269233277686, "loss": 0.2949, "step": 5487 }, { "epoch": 2.582588235294118, "grad_norm": 0.9927911187375814, "learning_rate": 0.0002285895887403806, "loss": 0.9453, "step": 5488 }, { "epoch": 2.5830588235294116, "grad_norm": 0.7391881366715007, "learning_rate": 0.00022855224744105625, "loss": 0.6484, "step": 5489 }, { "epoch": 2.583529411764706, "grad_norm": 1.62153140049841, "learning_rate": 0.00022851489943298472, "loss": 1.4844, "step": 5490 }, { "epoch": 2.584, "grad_norm": 0.7572741276421805, "learning_rate": 0.0002284775447193556, "loss": 0.832, "step": 5491 }, { "epoch": 2.584470588235294, "grad_norm": 1.2222614442305808, "learning_rate": 0.00022844018330335928, "loss": 1.6797, "step": 5492 }, { "epoch": 2.584941176470588, "grad_norm": 1.2850220697766224, "learning_rate": 0.00022840281518818657, "loss": 1.2031, "step": 5493 }, { "epoch": 2.5854117647058823, "grad_norm": 0.47279906282690265, "learning_rate": 0.00022836544037702883, "loss": 0.1191, "step": 5494 }, { "epoch": 2.5858823529411765, "grad_norm": 0.9115593817006611, "learning_rate": 0.00022832805887307815, "loss": 0.8672, "step": 5495 }, { "epoch": 2.5863529411764707, "grad_norm": 1.0354321551780057, "learning_rate": 0.00022829067067952702, "loss": 0.8359, "step": 5496 }, { "epoch": 2.586823529411765, "grad_norm": 0.6847108215100969, "learning_rate": 0.00022825327579956863, "loss": 0.4414, "step": 5497 }, { "epoch": 2.5872941176470587, "grad_norm": 0.8982201458238828, "learning_rate": 0.0002282158742363966, "loss": 0.9219, "step": 5498 }, { "epoch": 2.587764705882353, "grad_norm": 0.5922264904401354, "learning_rate": 0.00022817846599320528, "loss": 0.3789, "step": 5499 }, { "epoch": 2.588235294117647, "grad_norm": 0.766305582650897, "learning_rate": 0.00022814105107318952, "loss": 0.3633, "step": 5500 }, { "epoch": 2.588705882352941, "grad_norm": 0.9959478378484281, "learning_rate": 0.00022810362947954475, "loss": 0.7891, "step": 5501 }, { "epoch": 2.589176470588235, "grad_norm": 0.9032146460189086, "learning_rate": 0.00022806620121546684, "loss": 1.1875, "step": 5502 }, { "epoch": 2.5896470588235294, "grad_norm": 1.1681219698823515, "learning_rate": 0.0002280287662841525, "loss": 1.5312, "step": 5503 }, { "epoch": 2.5901176470588236, "grad_norm": 0.7904873370853397, "learning_rate": 0.00022799132468879878, "loss": 0.5508, "step": 5504 }, { "epoch": 2.590588235294118, "grad_norm": 0.9432613708610341, "learning_rate": 0.00022795387643260336, "loss": 1.2969, "step": 5505 }, { "epoch": 2.5910588235294116, "grad_norm": 0.8698965766450525, "learning_rate": 0.00022791642151876455, "loss": 0.918, "step": 5506 }, { "epoch": 2.591529411764706, "grad_norm": 0.7748102913234091, "learning_rate": 0.00022787895995048118, "loss": 0.7344, "step": 5507 }, { "epoch": 2.592, "grad_norm": 0.840414515505405, "learning_rate": 0.00022784149173095262, "loss": 0.6562, "step": 5508 }, { "epoch": 2.592470588235294, "grad_norm": 0.9446747450780508, "learning_rate": 0.00022780401686337885, "loss": 0.5547, "step": 5509 }, { "epoch": 2.592941176470588, "grad_norm": 1.1183140482918195, "learning_rate": 0.00022776653535096044, "loss": 1.0156, "step": 5510 }, { "epoch": 2.5934117647058823, "grad_norm": 0.5823657312760979, "learning_rate": 0.00022772904719689844, "loss": 0.207, "step": 5511 }, { "epoch": 2.5938823529411765, "grad_norm": 0.6679359328940848, "learning_rate": 0.00022769155240439455, "loss": 1.0156, "step": 5512 }, { "epoch": 2.5943529411764708, "grad_norm": 0.769975244043238, "learning_rate": 0.00022765405097665103, "loss": 0.5938, "step": 5513 }, { "epoch": 2.594823529411765, "grad_norm": 0.7677935477401782, "learning_rate": 0.00022761654291687065, "loss": 0.4648, "step": 5514 }, { "epoch": 2.5952941176470588, "grad_norm": 1.0804349725748525, "learning_rate": 0.00022757902822825683, "loss": 1.1562, "step": 5515 }, { "epoch": 2.595764705882353, "grad_norm": 1.0833614940009733, "learning_rate": 0.00022754150691401345, "loss": 1.0234, "step": 5516 }, { "epoch": 2.596235294117647, "grad_norm": 0.5508611974724555, "learning_rate": 0.00022750397897734504, "loss": 0.3496, "step": 5517 }, { "epoch": 2.596705882352941, "grad_norm": 0.7293792084676745, "learning_rate": 0.00022746644442145665, "loss": 0.377, "step": 5518 }, { "epoch": 2.597176470588235, "grad_norm": 0.8594634413096227, "learning_rate": 0.00022742890324955396, "loss": 0.9219, "step": 5519 }, { "epoch": 2.5976470588235294, "grad_norm": 0.9992996300752692, "learning_rate": 0.00022739135546484315, "loss": 0.8047, "step": 5520 }, { "epoch": 2.5981176470588236, "grad_norm": 0.7555039307674657, "learning_rate": 0.00022735380107053094, "loss": 1.0312, "step": 5521 }, { "epoch": 2.598588235294118, "grad_norm": 0.9924212664073427, "learning_rate": 0.00022731624006982473, "loss": 1.0312, "step": 5522 }, { "epoch": 2.5990588235294116, "grad_norm": 0.5820458223833374, "learning_rate": 0.0002272786724659324, "loss": 0.3281, "step": 5523 }, { "epoch": 2.599529411764706, "grad_norm": 0.9387753679136576, "learning_rate": 0.00022724109826206235, "loss": 0.6328, "step": 5524 }, { "epoch": 2.6, "grad_norm": 0.728241423432107, "learning_rate": 0.00022720351746142366, "loss": 0.5586, "step": 5525 }, { "epoch": 2.600470588235294, "grad_norm": 0.7526323565475065, "learning_rate": 0.00022716593006722594, "loss": 0.4023, "step": 5526 }, { "epoch": 2.600941176470588, "grad_norm": 0.8708488745747166, "learning_rate": 0.0002271283360826793, "loss": 1.3203, "step": 5527 }, { "epoch": 2.6014117647058823, "grad_norm": 1.3311282214076008, "learning_rate": 0.0002270907355109944, "loss": 1.3672, "step": 5528 }, { "epoch": 2.6018823529411765, "grad_norm": 0.90719190656975, "learning_rate": 0.00022705312835538257, "loss": 0.8633, "step": 5529 }, { "epoch": 2.6023529411764708, "grad_norm": 1.0837216174889832, "learning_rate": 0.00022701551461905563, "loss": 1.375, "step": 5530 }, { "epoch": 2.602823529411765, "grad_norm": 0.7954423179264166, "learning_rate": 0.00022697789430522602, "loss": 1.1094, "step": 5531 }, { "epoch": 2.6032941176470588, "grad_norm": 0.8833660551091608, "learning_rate": 0.00022694026741710657, "loss": 1.2344, "step": 5532 }, { "epoch": 2.603764705882353, "grad_norm": 0.8467547583550308, "learning_rate": 0.00022690263395791095, "loss": 0.7383, "step": 5533 }, { "epoch": 2.604235294117647, "grad_norm": 0.9757145329967754, "learning_rate": 0.00022686499393085322, "loss": 0.9102, "step": 5534 }, { "epoch": 2.604705882352941, "grad_norm": 1.3063249974088567, "learning_rate": 0.00022682734733914792, "loss": 2.0, "step": 5535 }, { "epoch": 2.605176470588235, "grad_norm": 0.9017110276131339, "learning_rate": 0.00022678969418601038, "loss": 1.1562, "step": 5536 }, { "epoch": 2.6056470588235294, "grad_norm": 1.1796433139564038, "learning_rate": 0.0002267520344746563, "loss": 1.1641, "step": 5537 }, { "epoch": 2.6061176470588236, "grad_norm": 1.2377816518953333, "learning_rate": 0.000226714368208302, "loss": 1.8672, "step": 5538 }, { "epoch": 2.606588235294118, "grad_norm": 0.7882297027433046, "learning_rate": 0.00022667669539016435, "loss": 1.125, "step": 5539 }, { "epoch": 2.6070588235294117, "grad_norm": 1.544007677605226, "learning_rate": 0.00022663901602346082, "loss": 1.7656, "step": 5540 }, { "epoch": 2.607529411764706, "grad_norm": 0.8894368891783535, "learning_rate": 0.00022660133011140942, "loss": 1.1797, "step": 5541 }, { "epoch": 2.608, "grad_norm": 0.7762960455172859, "learning_rate": 0.00022656363765722872, "loss": 1.0703, "step": 5542 }, { "epoch": 2.608470588235294, "grad_norm": 0.5707242677658305, "learning_rate": 0.00022652593866413775, "loss": 0.1934, "step": 5543 }, { "epoch": 2.608941176470588, "grad_norm": 0.7070457881377418, "learning_rate": 0.00022648823313535635, "loss": 0.4375, "step": 5544 }, { "epoch": 2.6094117647058823, "grad_norm": 1.0800582774878094, "learning_rate": 0.00022645052107410463, "loss": 1.1484, "step": 5545 }, { "epoch": 2.6098823529411765, "grad_norm": 0.595590021548518, "learning_rate": 0.0002264128024836034, "loss": 0.4375, "step": 5546 }, { "epoch": 2.6103529411764708, "grad_norm": 0.8224843630774684, "learning_rate": 0.00022637507736707408, "loss": 0.6836, "step": 5547 }, { "epoch": 2.6108235294117645, "grad_norm": 0.8780675556880486, "learning_rate": 0.00022633734572773852, "loss": 0.4824, "step": 5548 }, { "epoch": 2.6112941176470588, "grad_norm": 1.0505033011987477, "learning_rate": 0.00022629960756881916, "loss": 1.5234, "step": 5549 }, { "epoch": 2.611764705882353, "grad_norm": 1.1487913999767496, "learning_rate": 0.00022626186289353913, "loss": 1.0078, "step": 5550 }, { "epoch": 2.612235294117647, "grad_norm": 0.8816139261196649, "learning_rate": 0.00022622411170512193, "loss": 0.6953, "step": 5551 }, { "epoch": 2.612705882352941, "grad_norm": 0.5718432981659678, "learning_rate": 0.0002261863540067917, "loss": 0.252, "step": 5552 }, { "epoch": 2.613176470588235, "grad_norm": 0.9972653178941217, "learning_rate": 0.00022614858980177314, "loss": 0.9062, "step": 5553 }, { "epoch": 2.6136470588235294, "grad_norm": 0.6761811039343142, "learning_rate": 0.00022611081909329153, "loss": 0.416, "step": 5554 }, { "epoch": 2.6141176470588237, "grad_norm": 0.9073835374169409, "learning_rate": 0.00022607304188457264, "loss": 0.8867, "step": 5555 }, { "epoch": 2.614588235294118, "grad_norm": 1.0129164986244632, "learning_rate": 0.0002260352581788428, "loss": 1.0938, "step": 5556 }, { "epoch": 2.6150588235294117, "grad_norm": 1.175075586451455, "learning_rate": 0.000225997467979329, "loss": 0.9961, "step": 5557 }, { "epoch": 2.615529411764706, "grad_norm": 1.235046576830232, "learning_rate": 0.00022595967128925866, "loss": 1.0312, "step": 5558 }, { "epoch": 2.616, "grad_norm": 0.8319583955653784, "learning_rate": 0.0002259218681118598, "loss": 0.9297, "step": 5559 }, { "epoch": 2.616470588235294, "grad_norm": 1.3705711722832452, "learning_rate": 0.000225884058450361, "loss": 0.9844, "step": 5560 }, { "epoch": 2.616941176470588, "grad_norm": 1.1187855658834687, "learning_rate": 0.00022584624230799144, "loss": 1.1484, "step": 5561 }, { "epoch": 2.6174117647058823, "grad_norm": 1.1710428990552366, "learning_rate": 0.00022580841968798072, "loss": 1.4062, "step": 5562 }, { "epoch": 2.6178823529411765, "grad_norm": 0.95765068851368, "learning_rate": 0.00022577059059355914, "loss": 0.8477, "step": 5563 }, { "epoch": 2.6183529411764708, "grad_norm": 0.5277768493589863, "learning_rate": 0.00022573275502795747, "loss": 0.5156, "step": 5564 }, { "epoch": 2.6188235294117646, "grad_norm": 0.7299992074924272, "learning_rate": 0.00022569491299440704, "loss": 0.7695, "step": 5565 }, { "epoch": 2.6192941176470588, "grad_norm": 0.8055031570822289, "learning_rate": 0.00022565706449613978, "loss": 0.416, "step": 5566 }, { "epoch": 2.619764705882353, "grad_norm": 0.9787317551410705, "learning_rate": 0.0002256192095363881, "loss": 0.7773, "step": 5567 }, { "epoch": 2.620235294117647, "grad_norm": 0.7076543890089566, "learning_rate": 0.00022558134811838504, "loss": 0.4199, "step": 5568 }, { "epoch": 2.620705882352941, "grad_norm": 0.6784313484201188, "learning_rate": 0.00022554348024536413, "loss": 0.4785, "step": 5569 }, { "epoch": 2.621176470588235, "grad_norm": 1.1599347792216397, "learning_rate": 0.00022550560592055942, "loss": 1.7344, "step": 5570 }, { "epoch": 2.6216470588235294, "grad_norm": 1.0251184865114662, "learning_rate": 0.00022546772514720566, "loss": 0.9297, "step": 5571 }, { "epoch": 2.6221176470588237, "grad_norm": 0.6524974783082937, "learning_rate": 0.00022542983792853797, "loss": 0.4043, "step": 5572 }, { "epoch": 2.622588235294118, "grad_norm": 0.9084312132228678, "learning_rate": 0.00022539194426779216, "loss": 0.7891, "step": 5573 }, { "epoch": 2.6230588235294117, "grad_norm": 1.033349472025906, "learning_rate": 0.00022535404416820453, "loss": 1.1484, "step": 5574 }, { "epoch": 2.623529411764706, "grad_norm": 1.2164980675312222, "learning_rate": 0.0002253161376330119, "loss": 1.1719, "step": 5575 }, { "epoch": 2.624, "grad_norm": 0.5037387838092872, "learning_rate": 0.00022527822466545164, "loss": 0.1846, "step": 5576 }, { "epoch": 2.624470588235294, "grad_norm": 1.04624819003529, "learning_rate": 0.00022524030526876183, "loss": 1.6172, "step": 5577 }, { "epoch": 2.624941176470588, "grad_norm": 0.7955359671874102, "learning_rate": 0.0002252023794461809, "loss": 1.2578, "step": 5578 }, { "epoch": 2.6254117647058823, "grad_norm": 0.8757273568757928, "learning_rate": 0.00022516444720094785, "loss": 0.707, "step": 5579 }, { "epoch": 2.6258823529411766, "grad_norm": 0.8885137663928736, "learning_rate": 0.00022512650853630234, "loss": 0.5938, "step": 5580 }, { "epoch": 2.626352941176471, "grad_norm": 1.046226668524503, "learning_rate": 0.00022508856345548455, "loss": 1.2734, "step": 5581 }, { "epoch": 2.6268235294117646, "grad_norm": 0.6376581896428903, "learning_rate": 0.00022505061196173513, "loss": 0.4395, "step": 5582 }, { "epoch": 2.627294117647059, "grad_norm": 0.6414680716857862, "learning_rate": 0.00022501265405829526, "loss": 0.6094, "step": 5583 }, { "epoch": 2.627764705882353, "grad_norm": 0.6366729864526937, "learning_rate": 0.00022497468974840687, "loss": 0.3594, "step": 5584 }, { "epoch": 2.6282352941176472, "grad_norm": 0.8961746077344886, "learning_rate": 0.00022493671903531218, "loss": 0.5977, "step": 5585 }, { "epoch": 2.628705882352941, "grad_norm": 0.8666291030498131, "learning_rate": 0.00022489874192225414, "loss": 0.5117, "step": 5586 }, { "epoch": 2.6291764705882352, "grad_norm": 0.9784746278634183, "learning_rate": 0.00022486075841247617, "loss": 1.3203, "step": 5587 }, { "epoch": 2.6296470588235294, "grad_norm": 0.6109354378322548, "learning_rate": 0.0002248227685092222, "loss": 0.2715, "step": 5588 }, { "epoch": 2.6301176470588237, "grad_norm": 0.9472094349198725, "learning_rate": 0.00022478477221573687, "loss": 0.9766, "step": 5589 }, { "epoch": 2.630588235294118, "grad_norm": 0.7249279006547098, "learning_rate": 0.00022474676953526508, "loss": 0.5586, "step": 5590 }, { "epoch": 2.6310588235294117, "grad_norm": 1.1300156188875412, "learning_rate": 0.00022470876047105262, "loss": 1.1484, "step": 5591 }, { "epoch": 2.631529411764706, "grad_norm": 0.8380854546623194, "learning_rate": 0.00022467074502634558, "loss": 1.3125, "step": 5592 }, { "epoch": 2.632, "grad_norm": 1.3346696701285647, "learning_rate": 0.00022463272320439063, "loss": 1.5156, "step": 5593 }, { "epoch": 2.632470588235294, "grad_norm": 1.54294182513783, "learning_rate": 0.00022459469500843502, "loss": 1.5156, "step": 5594 }, { "epoch": 2.632941176470588, "grad_norm": 0.76313031218321, "learning_rate": 0.0002245566604417266, "loss": 0.4961, "step": 5595 }, { "epoch": 2.6334117647058823, "grad_norm": 0.9926518440657933, "learning_rate": 0.0002245186195075137, "loss": 1.0078, "step": 5596 }, { "epoch": 2.6338823529411766, "grad_norm": 0.49888192199238934, "learning_rate": 0.0002244805722090451, "loss": 0.3398, "step": 5597 }, { "epoch": 2.634352941176471, "grad_norm": 1.352696133508456, "learning_rate": 0.0002244425185495704, "loss": 0.918, "step": 5598 }, { "epoch": 2.6348235294117646, "grad_norm": 0.8376786581491003, "learning_rate": 0.00022440445853233945, "loss": 0.7188, "step": 5599 }, { "epoch": 2.635294117647059, "grad_norm": 1.4757040874238938, "learning_rate": 0.00022436639216060275, "loss": 1.9297, "step": 5600 }, { "epoch": 2.635764705882353, "grad_norm": 0.9784456181527912, "learning_rate": 0.00022432831943761146, "loss": 1.1328, "step": 5601 }, { "epoch": 2.636235294117647, "grad_norm": 0.9435955012729209, "learning_rate": 0.00022429024036661708, "loss": 0.9609, "step": 5602 }, { "epoch": 2.636705882352941, "grad_norm": 0.8099011048288718, "learning_rate": 0.00022425215495087177, "loss": 1.0625, "step": 5603 }, { "epoch": 2.6371764705882352, "grad_norm": 1.1401028345306652, "learning_rate": 0.00022421406319362823, "loss": 1.4531, "step": 5604 }, { "epoch": 2.6376470588235295, "grad_norm": 0.7058595511343901, "learning_rate": 0.00022417596509813968, "loss": 0.7578, "step": 5605 }, { "epoch": 2.6381176470588237, "grad_norm": 6.214607702344594, "learning_rate": 0.0002241378606676599, "loss": 3.0781, "step": 5606 }, { "epoch": 2.638588235294118, "grad_norm": 0.8889599987589776, "learning_rate": 0.00022409974990544312, "loss": 0.5195, "step": 5607 }, { "epoch": 2.6390588235294117, "grad_norm": 0.7215623331395703, "learning_rate": 0.0002240616328147443, "loss": 0.4238, "step": 5608 }, { "epoch": 2.639529411764706, "grad_norm": 0.7400054879795824, "learning_rate": 0.00022402350939881873, "loss": 0.9688, "step": 5609 }, { "epoch": 2.64, "grad_norm": 0.6824190754376186, "learning_rate": 0.00022398537966092236, "loss": 0.3574, "step": 5610 }, { "epoch": 2.640470588235294, "grad_norm": 0.998351733922392, "learning_rate": 0.00022394724360431167, "loss": 1.0938, "step": 5611 }, { "epoch": 2.640941176470588, "grad_norm": 1.0421932773796025, "learning_rate": 0.00022390910123224373, "loss": 0.875, "step": 5612 }, { "epoch": 2.6414117647058823, "grad_norm": 1.0954769356949199, "learning_rate": 0.00022387095254797597, "loss": 1.4609, "step": 5613 }, { "epoch": 2.6418823529411766, "grad_norm": 0.8869478825449375, "learning_rate": 0.00022383279755476651, "loss": 0.8164, "step": 5614 }, { "epoch": 2.642352941176471, "grad_norm": 1.4249657568082652, "learning_rate": 0.00022379463625587406, "loss": 1.0391, "step": 5615 }, { "epoch": 2.6428235294117646, "grad_norm": 0.989349157062106, "learning_rate": 0.00022375646865455766, "loss": 1.0156, "step": 5616 }, { "epoch": 2.643294117647059, "grad_norm": 0.9585575563501952, "learning_rate": 0.00022371829475407713, "loss": 0.8242, "step": 5617 }, { "epoch": 2.643764705882353, "grad_norm": 1.050856121872048, "learning_rate": 0.00022368011455769251, "loss": 1.0234, "step": 5618 }, { "epoch": 2.644235294117647, "grad_norm": 0.6236740782391716, "learning_rate": 0.00022364192806866481, "loss": 0.2598, "step": 5619 }, { "epoch": 2.644705882352941, "grad_norm": 0.910596058123462, "learning_rate": 0.00022360373529025523, "loss": 1.0859, "step": 5620 }, { "epoch": 2.6451764705882352, "grad_norm": 0.43935492554550215, "learning_rate": 0.00022356553622572564, "loss": 0.2012, "step": 5621 }, { "epoch": 2.6456470588235295, "grad_norm": 1.1195355587879694, "learning_rate": 0.00022352733087833838, "loss": 1.6875, "step": 5622 }, { "epoch": 2.6461176470588237, "grad_norm": 0.3958755604743378, "learning_rate": 0.00022348911925135646, "loss": 0.0413, "step": 5623 }, { "epoch": 2.646588235294118, "grad_norm": 0.8943758037503099, "learning_rate": 0.00022345090134804325, "loss": 1.0156, "step": 5624 }, { "epoch": 2.6470588235294117, "grad_norm": 0.7248045535854246, "learning_rate": 0.0002234126771716628, "loss": 0.2715, "step": 5625 }, { "epoch": 2.647529411764706, "grad_norm": 0.7795905642381888, "learning_rate": 0.00022337444672547967, "loss": 0.875, "step": 5626 }, { "epoch": 2.648, "grad_norm": 0.7616873158234608, "learning_rate": 0.00022333621001275884, "loss": 0.9375, "step": 5627 }, { "epoch": 2.648470588235294, "grad_norm": 0.9067212695977679, "learning_rate": 0.00022329796703676597, "loss": 0.9609, "step": 5628 }, { "epoch": 2.648941176470588, "grad_norm": 0.9026203486163124, "learning_rate": 0.00022325971780076721, "loss": 0.6602, "step": 5629 }, { "epoch": 2.6494117647058824, "grad_norm": 1.5705784854209164, "learning_rate": 0.00022322146230802924, "loss": 1.375, "step": 5630 }, { "epoch": 2.6498823529411766, "grad_norm": 0.9177350324682947, "learning_rate": 0.0002231832005618192, "loss": 0.9258, "step": 5631 }, { "epoch": 2.650352941176471, "grad_norm": 0.8732027087845632, "learning_rate": 0.00022314493256540485, "loss": 0.6367, "step": 5632 }, { "epoch": 2.6508235294117646, "grad_norm": 0.6148917482743418, "learning_rate": 0.0002231066583220545, "loss": 0.5625, "step": 5633 }, { "epoch": 2.651294117647059, "grad_norm": 0.9913380862754058, "learning_rate": 0.00022306837783503698, "loss": 1.4297, "step": 5634 }, { "epoch": 2.651764705882353, "grad_norm": 0.9589201001541817, "learning_rate": 0.00022303009110762154, "loss": 1.25, "step": 5635 }, { "epoch": 2.652235294117647, "grad_norm": 0.8843295765464737, "learning_rate": 0.00022299179814307818, "loss": 0.9023, "step": 5636 }, { "epoch": 2.652705882352941, "grad_norm": 0.7956914850626353, "learning_rate": 0.0002229534989446772, "loss": 0.6367, "step": 5637 }, { "epoch": 2.6531764705882352, "grad_norm": 0.614558266325728, "learning_rate": 0.00022291519351568955, "loss": 0.3262, "step": 5638 }, { "epoch": 2.6536470588235295, "grad_norm": 1.2577715550097843, "learning_rate": 0.00022287688185938673, "loss": 1.2812, "step": 5639 }, { "epoch": 2.6541176470588237, "grad_norm": 0.982169269570468, "learning_rate": 0.00022283856397904077, "loss": 1.0703, "step": 5640 }, { "epoch": 2.654588235294118, "grad_norm": 0.807621281801289, "learning_rate": 0.00022280023987792416, "loss": 0.9883, "step": 5641 }, { "epoch": 2.6550588235294117, "grad_norm": 0.8957058529844548, "learning_rate": 0.00022276190955930998, "loss": 0.416, "step": 5642 }, { "epoch": 2.655529411764706, "grad_norm": 1.1031798697009634, "learning_rate": 0.00022272357302647187, "loss": 1.7969, "step": 5643 }, { "epoch": 2.656, "grad_norm": 0.7765667560739664, "learning_rate": 0.00022268523028268385, "loss": 0.6836, "step": 5644 }, { "epoch": 2.656470588235294, "grad_norm": 0.9844326926494363, "learning_rate": 0.00022264688133122069, "loss": 1.0156, "step": 5645 }, { "epoch": 2.656941176470588, "grad_norm": 0.7693386397606725, "learning_rate": 0.00022260852617535752, "loss": 0.7773, "step": 5646 }, { "epoch": 2.6574117647058824, "grad_norm": 1.8097771667605274, "learning_rate": 0.00022257016481837006, "loss": 1.0547, "step": 5647 }, { "epoch": 2.6578823529411766, "grad_norm": 0.8131751770224195, "learning_rate": 0.00022253179726353458, "loss": 0.8945, "step": 5648 }, { "epoch": 2.658352941176471, "grad_norm": 0.3216927424488076, "learning_rate": 0.00022249342351412783, "loss": 0.105, "step": 5649 }, { "epoch": 2.6588235294117646, "grad_norm": 0.3738424769753959, "learning_rate": 0.00022245504357342716, "loss": 0.4023, "step": 5650 }, { "epoch": 2.659294117647059, "grad_norm": 0.5767361197708384, "learning_rate": 0.00022241665744471037, "loss": 0.1963, "step": 5651 }, { "epoch": 2.659764705882353, "grad_norm": 0.7575866157233793, "learning_rate": 0.00022237826513125584, "loss": 0.5273, "step": 5652 }, { "epoch": 2.660235294117647, "grad_norm": 1.2041894373641806, "learning_rate": 0.00022233986663634241, "loss": 1.2812, "step": 5653 }, { "epoch": 2.660705882352941, "grad_norm": 0.7128950965979172, "learning_rate": 0.0002223014619632496, "loss": 0.4609, "step": 5654 }, { "epoch": 2.6611764705882353, "grad_norm": 0.8624343868311428, "learning_rate": 0.00022226305111525726, "loss": 0.6641, "step": 5655 }, { "epoch": 2.6616470588235295, "grad_norm": 1.4501315660272254, "learning_rate": 0.00022222463409564595, "loss": 2.0625, "step": 5656 }, { "epoch": 2.6621176470588237, "grad_norm": 0.8913030009446437, "learning_rate": 0.0002221862109076966, "loss": 1.3281, "step": 5657 }, { "epoch": 2.662588235294118, "grad_norm": 1.3747713328186517, "learning_rate": 0.00022214778155469077, "loss": 1.2969, "step": 5658 }, { "epoch": 2.6630588235294117, "grad_norm": 0.7343271359861292, "learning_rate": 0.0002221093460399105, "loss": 1.1172, "step": 5659 }, { "epoch": 2.663529411764706, "grad_norm": 0.9177996867427402, "learning_rate": 0.0002220709043666384, "loss": 1.3984, "step": 5660 }, { "epoch": 2.664, "grad_norm": 1.0930591298910877, "learning_rate": 0.0002220324565381576, "loss": 0.9062, "step": 5661 }, { "epoch": 2.664470588235294, "grad_norm": 1.2767295071013682, "learning_rate": 0.0002219940025577516, "loss": 1.8906, "step": 5662 }, { "epoch": 2.664941176470588, "grad_norm": 0.42949677873591247, "learning_rate": 0.00022195554242870473, "loss": 0.1709, "step": 5663 }, { "epoch": 2.6654117647058824, "grad_norm": 0.9921229425874011, "learning_rate": 0.0002219170761543016, "loss": 1.0312, "step": 5664 }, { "epoch": 2.6658823529411766, "grad_norm": 1.2860894642540686, "learning_rate": 0.00022187860373782737, "loss": 1.3203, "step": 5665 }, { "epoch": 2.666352941176471, "grad_norm": 0.9204475065958366, "learning_rate": 0.00022184012518256785, "loss": 0.832, "step": 5666 }, { "epoch": 2.6668235294117646, "grad_norm": 0.8160566138284007, "learning_rate": 0.00022180164049180926, "loss": 0.8125, "step": 5667 }, { "epoch": 2.667294117647059, "grad_norm": 1.1067470852703798, "learning_rate": 0.00022176314966883837, "loss": 1.5703, "step": 5668 }, { "epoch": 2.667764705882353, "grad_norm": 0.6863529512516142, "learning_rate": 0.0002217246527169425, "loss": 0.6836, "step": 5669 }, { "epoch": 2.668235294117647, "grad_norm": 1.1820748211599788, "learning_rate": 0.00022168614963940952, "loss": 1.2266, "step": 5670 }, { "epoch": 2.668705882352941, "grad_norm": 1.0053449208182583, "learning_rate": 0.00022164764043952772, "loss": 1.0078, "step": 5671 }, { "epoch": 2.6691764705882353, "grad_norm": 0.9064542983034427, "learning_rate": 0.00022160912512058598, "loss": 0.8164, "step": 5672 }, { "epoch": 2.6696470588235295, "grad_norm": 1.03507133373647, "learning_rate": 0.0002215706036858737, "loss": 1.0781, "step": 5673 }, { "epoch": 2.6701176470588237, "grad_norm": 1.1535303441164606, "learning_rate": 0.00022153207613868087, "loss": 1.7812, "step": 5674 }, { "epoch": 2.6705882352941175, "grad_norm": 0.7800684142648299, "learning_rate": 0.00022149354248229784, "loss": 0.4453, "step": 5675 }, { "epoch": 2.6710588235294117, "grad_norm": 0.5742952515353849, "learning_rate": 0.00022145500272001558, "loss": 0.3594, "step": 5676 }, { "epoch": 2.671529411764706, "grad_norm": 1.14072222217073, "learning_rate": 0.00022141645685512564, "loss": 0.6641, "step": 5677 }, { "epoch": 2.672, "grad_norm": 0.7738079794599014, "learning_rate": 0.00022137790489091999, "loss": 0.7383, "step": 5678 }, { "epoch": 2.672470588235294, "grad_norm": 0.5642581480076669, "learning_rate": 0.00022133934683069115, "loss": 0.3965, "step": 5679 }, { "epoch": 2.672941176470588, "grad_norm": 2.9709561537540963, "learning_rate": 0.00022130078267773218, "loss": 1.2266, "step": 5680 }, { "epoch": 2.6734117647058824, "grad_norm": 0.5078085187763846, "learning_rate": 0.00022126221243533664, "loss": 0.2002, "step": 5681 }, { "epoch": 2.6738823529411766, "grad_norm": 0.7393338984693986, "learning_rate": 0.00022122363610679863, "loss": 0.5078, "step": 5682 }, { "epoch": 2.674352941176471, "grad_norm": 0.6347617836345532, "learning_rate": 0.0002211850536954127, "loss": 0.3086, "step": 5683 }, { "epoch": 2.6748235294117646, "grad_norm": 0.986803765590389, "learning_rate": 0.00022114646520447406, "loss": 1.4297, "step": 5684 }, { "epoch": 2.675294117647059, "grad_norm": 0.8851170524957758, "learning_rate": 0.00022110787063727832, "loss": 0.8828, "step": 5685 }, { "epoch": 2.675764705882353, "grad_norm": 0.45357333199994304, "learning_rate": 0.00022106926999712165, "loss": 0.1357, "step": 5686 }, { "epoch": 2.676235294117647, "grad_norm": 0.6397190175878118, "learning_rate": 0.0002210306632873007, "loss": 0.3828, "step": 5687 }, { "epoch": 2.676705882352941, "grad_norm": 1.3531596866153381, "learning_rate": 0.00022099205051111272, "loss": 1.0625, "step": 5688 }, { "epoch": 2.6771764705882353, "grad_norm": 0.8387445958183902, "learning_rate": 0.0002209534316718554, "loss": 0.9922, "step": 5689 }, { "epoch": 2.6776470588235295, "grad_norm": 0.8940095737242908, "learning_rate": 0.000220914806772827, "loss": 1.0, "step": 5690 }, { "epoch": 2.6781176470588237, "grad_norm": 1.5384100339660176, "learning_rate": 0.0002208761758173263, "loss": 2.0312, "step": 5691 }, { "epoch": 2.6785882352941175, "grad_norm": 1.043287966475457, "learning_rate": 0.0002208375388086525, "loss": 1.2266, "step": 5692 }, { "epoch": 2.6790588235294117, "grad_norm": 0.38516366690587817, "learning_rate": 0.00022079889575010543, "loss": 0.1572, "step": 5693 }, { "epoch": 2.679529411764706, "grad_norm": 0.8472244454654653, "learning_rate": 0.0002207602466449854, "loss": 1.3047, "step": 5694 }, { "epoch": 2.68, "grad_norm": 0.703056974241518, "learning_rate": 0.00022072159149659327, "loss": 1.0078, "step": 5695 }, { "epoch": 2.680470588235294, "grad_norm": 0.3331023873974773, "learning_rate": 0.00022068293030823026, "loss": 0.0593, "step": 5696 }, { "epoch": 2.680941176470588, "grad_norm": 1.0091518345539954, "learning_rate": 0.00022064426308319838, "loss": 0.9883, "step": 5697 }, { "epoch": 2.6814117647058824, "grad_norm": 0.8201057429810705, "learning_rate": 0.0002206055898247999, "loss": 0.5352, "step": 5698 }, { "epoch": 2.6818823529411766, "grad_norm": 1.077749118837852, "learning_rate": 0.0002205669105363377, "loss": 1.5547, "step": 5699 }, { "epoch": 2.682352941176471, "grad_norm": 0.6966735942608424, "learning_rate": 0.00022052822522111522, "loss": 0.2393, "step": 5700 }, { "epoch": 2.6828235294117646, "grad_norm": 0.6115933114833412, "learning_rate": 0.0002204895338824364, "loss": 0.2393, "step": 5701 }, { "epoch": 2.683294117647059, "grad_norm": 0.817697298493514, "learning_rate": 0.0002204508365236056, "loss": 0.7734, "step": 5702 }, { "epoch": 2.683764705882353, "grad_norm": 0.9860604266089292, "learning_rate": 0.0002204121331479278, "loss": 1.0938, "step": 5703 }, { "epoch": 2.684235294117647, "grad_norm": 0.5622344499004959, "learning_rate": 0.0002203734237587085, "loss": 0.3652, "step": 5704 }, { "epoch": 2.684705882352941, "grad_norm": 1.1355428927729114, "learning_rate": 0.00022033470835925361, "loss": 1.4375, "step": 5705 }, { "epoch": 2.6851764705882353, "grad_norm": 1.0368910813851469, "learning_rate": 0.00022029598695286963, "loss": 1.0859, "step": 5706 }, { "epoch": 2.6856470588235295, "grad_norm": 0.7818519124171226, "learning_rate": 0.00022025725954286356, "loss": 0.4902, "step": 5707 }, { "epoch": 2.6861176470588237, "grad_norm": 1.1791252416545994, "learning_rate": 0.00022021852613254297, "loss": 0.7422, "step": 5708 }, { "epoch": 2.6865882352941175, "grad_norm": 0.8524531658964632, "learning_rate": 0.00022017978672521577, "loss": 0.6094, "step": 5709 }, { "epoch": 2.6870588235294117, "grad_norm": 0.40763338338729377, "learning_rate": 0.00022014104132419055, "loss": 0.0918, "step": 5710 }, { "epoch": 2.687529411764706, "grad_norm": 0.6442201285952587, "learning_rate": 0.0002201022899327764, "loss": 0.3242, "step": 5711 }, { "epoch": 2.6879999999999997, "grad_norm": 0.9271548428342212, "learning_rate": 0.00022006353255428281, "loss": 0.8477, "step": 5712 }, { "epoch": 2.688470588235294, "grad_norm": 1.1271548824781108, "learning_rate": 0.00022002476919201987, "loss": 1.4453, "step": 5713 }, { "epoch": 2.688941176470588, "grad_norm": 0.8358422579277304, "learning_rate": 0.00021998599984929818, "loss": 0.9531, "step": 5714 }, { "epoch": 2.6894117647058824, "grad_norm": 1.38763601247154, "learning_rate": 0.00021994722452942886, "loss": 2.0938, "step": 5715 }, { "epoch": 2.6898823529411766, "grad_norm": 0.7946391356413129, "learning_rate": 0.00021990844323572346, "loss": 0.4688, "step": 5716 }, { "epoch": 2.690352941176471, "grad_norm": 1.1058486197538333, "learning_rate": 0.0002198696559714941, "loss": 1.4375, "step": 5717 }, { "epoch": 2.6908235294117646, "grad_norm": 1.3170832509119588, "learning_rate": 0.0002198308627400534, "loss": 1.8516, "step": 5718 }, { "epoch": 2.691294117647059, "grad_norm": 0.8243493873187261, "learning_rate": 0.00021979206354471453, "loss": 0.5547, "step": 5719 }, { "epoch": 2.691764705882353, "grad_norm": 0.9827022385178271, "learning_rate": 0.00021975325838879107, "loss": 0.793, "step": 5720 }, { "epoch": 2.692235294117647, "grad_norm": 1.0524084475881785, "learning_rate": 0.0002197144472755972, "loss": 0.9531, "step": 5721 }, { "epoch": 2.692705882352941, "grad_norm": 1.0672049574533045, "learning_rate": 0.00021967563020844762, "loss": 1.875, "step": 5722 }, { "epoch": 2.6931764705882353, "grad_norm": 1.1175579854535491, "learning_rate": 0.00021963680719065746, "loss": 0.8594, "step": 5723 }, { "epoch": 2.6936470588235295, "grad_norm": 0.6802127055577498, "learning_rate": 0.0002195979782255424, "loss": 0.293, "step": 5724 }, { "epoch": 2.6941176470588237, "grad_norm": 1.0082320347445948, "learning_rate": 0.0002195591433164186, "loss": 1.2344, "step": 5725 }, { "epoch": 2.6945882352941175, "grad_norm": 0.8262048840108077, "learning_rate": 0.00021952030246660283, "loss": 0.543, "step": 5726 }, { "epoch": 2.6950588235294117, "grad_norm": 0.6205786728457076, "learning_rate": 0.00021948145567941217, "loss": 0.291, "step": 5727 }, { "epoch": 2.695529411764706, "grad_norm": 0.6533006790267482, "learning_rate": 0.0002194426029581644, "loss": 0.3438, "step": 5728 }, { "epoch": 2.6959999999999997, "grad_norm": 1.0093304540407277, "learning_rate": 0.00021940374430617777, "loss": 1.4844, "step": 5729 }, { "epoch": 2.696470588235294, "grad_norm": 0.8185912470002326, "learning_rate": 0.0002193648797267709, "loss": 1.1094, "step": 5730 }, { "epoch": 2.696941176470588, "grad_norm": 0.6834317726411774, "learning_rate": 0.0002193260092232631, "loss": 0.3828, "step": 5731 }, { "epoch": 2.6974117647058824, "grad_norm": 0.9873391614341921, "learning_rate": 0.00021928713279897408, "loss": 1.7578, "step": 5732 }, { "epoch": 2.6978823529411766, "grad_norm": 0.8938925914488172, "learning_rate": 0.00021924825045722404, "loss": 1.0703, "step": 5733 }, { "epoch": 2.698352941176471, "grad_norm": 0.9241076182791764, "learning_rate": 0.00021920936220133373, "loss": 0.9688, "step": 5734 }, { "epoch": 2.6988235294117646, "grad_norm": 0.8305246097945812, "learning_rate": 0.0002191704680346245, "loss": 0.5781, "step": 5735 }, { "epoch": 2.699294117647059, "grad_norm": 0.7114414425092302, "learning_rate": 0.00021913156796041794, "loss": 0.3262, "step": 5736 }, { "epoch": 2.699764705882353, "grad_norm": 0.6858322012479063, "learning_rate": 0.0002190926619820364, "loss": 0.3867, "step": 5737 }, { "epoch": 2.700235294117647, "grad_norm": 1.443587592013023, "learning_rate": 0.00021905375010280266, "loss": 0.7188, "step": 5738 }, { "epoch": 2.700705882352941, "grad_norm": 1.1104642351554104, "learning_rate": 0.00021901483232603995, "loss": 1.3125, "step": 5739 }, { "epoch": 2.7011764705882353, "grad_norm": 1.2051934012948096, "learning_rate": 0.000218975908655072, "loss": 1.4609, "step": 5740 }, { "epoch": 2.7016470588235295, "grad_norm": 0.873691154458211, "learning_rate": 0.0002189369790932232, "loss": 0.6484, "step": 5741 }, { "epoch": 2.7021176470588237, "grad_norm": 0.89078930540028, "learning_rate": 0.0002188980436438182, "loss": 0.8906, "step": 5742 }, { "epoch": 2.7025882352941175, "grad_norm": 1.0047765205556856, "learning_rate": 0.0002188591023101823, "loss": 1.5547, "step": 5743 }, { "epoch": 2.7030588235294117, "grad_norm": 1.3659395688476854, "learning_rate": 0.00021882015509564132, "loss": 1.7109, "step": 5744 }, { "epoch": 2.703529411764706, "grad_norm": 0.9551976845580988, "learning_rate": 0.00021878120200352158, "loss": 0.9375, "step": 5745 }, { "epoch": 2.7039999999999997, "grad_norm": 0.9266722185126136, "learning_rate": 0.00021874224303714983, "loss": 0.8125, "step": 5746 }, { "epoch": 2.704470588235294, "grad_norm": 0.7803591282385873, "learning_rate": 0.00021870327819985331, "loss": 1.2891, "step": 5747 }, { "epoch": 2.704941176470588, "grad_norm": 1.189963067474547, "learning_rate": 0.00021866430749495987, "loss": 1.7266, "step": 5748 }, { "epoch": 2.7054117647058824, "grad_norm": 0.7609359799437548, "learning_rate": 0.00021862533092579774, "loss": 0.6914, "step": 5749 }, { "epoch": 2.7058823529411766, "grad_norm": 0.8224711609413206, "learning_rate": 0.00021858634849569576, "loss": 0.5078, "step": 5750 }, { "epoch": 2.706352941176471, "grad_norm": 0.9046727600142983, "learning_rate": 0.00021854736020798319, "loss": 1.0469, "step": 5751 }, { "epoch": 2.7068235294117646, "grad_norm": 1.140232540351936, "learning_rate": 0.00021850836606598987, "loss": 1.1641, "step": 5752 }, { "epoch": 2.707294117647059, "grad_norm": 0.960597155789195, "learning_rate": 0.00021846936607304603, "loss": 0.8281, "step": 5753 }, { "epoch": 2.707764705882353, "grad_norm": 0.6496684238012226, "learning_rate": 0.00021843036023248247, "loss": 0.625, "step": 5754 }, { "epoch": 2.708235294117647, "grad_norm": 0.7309035608771525, "learning_rate": 0.00021839134854763055, "loss": 0.4414, "step": 5755 }, { "epoch": 2.708705882352941, "grad_norm": 0.901124703930704, "learning_rate": 0.00021835233102182199, "loss": 0.6797, "step": 5756 }, { "epoch": 2.7091764705882353, "grad_norm": 0.973595783703364, "learning_rate": 0.00021831330765838903, "loss": 0.7891, "step": 5757 }, { "epoch": 2.7096470588235295, "grad_norm": 0.677285135193564, "learning_rate": 0.00021827427846066462, "loss": 0.5391, "step": 5758 }, { "epoch": 2.7101176470588237, "grad_norm": 1.036949468692561, "learning_rate": 0.0002182352434319819, "loss": 1.1016, "step": 5759 }, { "epoch": 2.7105882352941175, "grad_norm": 0.6307676036769917, "learning_rate": 0.0002181962025756747, "loss": 0.3887, "step": 5760 }, { "epoch": 2.7110588235294117, "grad_norm": 2.1609735960600025, "learning_rate": 0.00021815715589507728, "loss": 2.1719, "step": 5761 }, { "epoch": 2.711529411764706, "grad_norm": 0.6829387987540567, "learning_rate": 0.00021811810339352446, "loss": 0.3789, "step": 5762 }, { "epoch": 2.7119999999999997, "grad_norm": 1.3119490919344474, "learning_rate": 0.0002180790450743515, "loss": 1.5, "step": 5763 }, { "epoch": 2.712470588235294, "grad_norm": 1.2936773612788206, "learning_rate": 0.0002180399809408941, "loss": 1.3438, "step": 5764 }, { "epoch": 2.712941176470588, "grad_norm": 0.848493170599694, "learning_rate": 0.00021800091099648865, "loss": 0.5508, "step": 5765 }, { "epoch": 2.7134117647058824, "grad_norm": 0.9116118074752093, "learning_rate": 0.0002179618352444719, "loss": 0.6016, "step": 5766 }, { "epoch": 2.7138823529411766, "grad_norm": 0.6810527293029738, "learning_rate": 0.00021792275368818102, "loss": 0.3809, "step": 5767 }, { "epoch": 2.714352941176471, "grad_norm": 0.4890750821887223, "learning_rate": 0.00021788366633095375, "loss": 0.3164, "step": 5768 }, { "epoch": 2.7148235294117646, "grad_norm": 0.9063242024605195, "learning_rate": 0.00021784457317612852, "loss": 0.8828, "step": 5769 }, { "epoch": 2.715294117647059, "grad_norm": 0.9802972546565898, "learning_rate": 0.0002178054742270439, "loss": 0.6367, "step": 5770 }, { "epoch": 2.715764705882353, "grad_norm": 1.1421527741101392, "learning_rate": 0.0002177663694870392, "loss": 1.1328, "step": 5771 }, { "epoch": 2.716235294117647, "grad_norm": 1.2696144514418686, "learning_rate": 0.00021772725895945422, "loss": 1.0625, "step": 5772 }, { "epoch": 2.716705882352941, "grad_norm": 0.6625584015517804, "learning_rate": 0.0002176881426476291, "loss": 0.293, "step": 5773 }, { "epoch": 2.7171764705882353, "grad_norm": 1.4498774442348625, "learning_rate": 0.00021764902055490456, "loss": 1.2031, "step": 5774 }, { "epoch": 2.7176470588235295, "grad_norm": 0.8009519912832921, "learning_rate": 0.00021760989268462182, "loss": 0.5859, "step": 5775 }, { "epoch": 2.7181176470588237, "grad_norm": 0.8266911045997948, "learning_rate": 0.00021757075904012266, "loss": 0.8359, "step": 5776 }, { "epoch": 2.7185882352941175, "grad_norm": 1.0348236500302719, "learning_rate": 0.00021753161962474924, "loss": 1.9453, "step": 5777 }, { "epoch": 2.7190588235294118, "grad_norm": 0.9116816743149853, "learning_rate": 0.00021749247444184422, "loss": 1.1875, "step": 5778 }, { "epoch": 2.719529411764706, "grad_norm": 1.2841385855511036, "learning_rate": 0.00021745332349475092, "loss": 1.8516, "step": 5779 }, { "epoch": 2.7199999999999998, "grad_norm": 1.0606021306052877, "learning_rate": 0.0002174141667868129, "loss": 1.3984, "step": 5780 }, { "epoch": 2.720470588235294, "grad_norm": 0.8004712012207295, "learning_rate": 0.00021737500432137435, "loss": 1.0625, "step": 5781 }, { "epoch": 2.720941176470588, "grad_norm": 0.7022708474219609, "learning_rate": 0.00021733583610177997, "loss": 0.2832, "step": 5782 }, { "epoch": 2.7214117647058824, "grad_norm": 1.1148094778396462, "learning_rate": 0.00021729666213137494, "loss": 0.75, "step": 5783 }, { "epoch": 2.7218823529411766, "grad_norm": 0.8002844340657576, "learning_rate": 0.00021725748241350486, "loss": 0.7227, "step": 5784 }, { "epoch": 2.722352941176471, "grad_norm": 1.1148397708561895, "learning_rate": 0.00021721829695151585, "loss": 0.8398, "step": 5785 }, { "epoch": 2.7228235294117646, "grad_norm": 0.8012758047224374, "learning_rate": 0.00021717910574875458, "loss": 0.5664, "step": 5786 }, { "epoch": 2.723294117647059, "grad_norm": 0.9075614576003382, "learning_rate": 0.00021713990880856818, "loss": 1.2422, "step": 5787 }, { "epoch": 2.723764705882353, "grad_norm": 0.9458714149908599, "learning_rate": 0.0002171007061343043, "loss": 0.9258, "step": 5788 }, { "epoch": 2.724235294117647, "grad_norm": 0.7143418313987981, "learning_rate": 0.0002170614977293109, "loss": 0.7969, "step": 5789 }, { "epoch": 2.724705882352941, "grad_norm": 0.47936813768182296, "learning_rate": 0.00021702228359693672, "loss": 0.5156, "step": 5790 }, { "epoch": 2.7251764705882353, "grad_norm": 1.2079625363555315, "learning_rate": 0.00021698306374053075, "loss": 1.9922, "step": 5791 }, { "epoch": 2.7256470588235295, "grad_norm": 1.1153502028476718, "learning_rate": 0.00021694383816344258, "loss": 1.4453, "step": 5792 }, { "epoch": 2.7261176470588238, "grad_norm": 0.8263594363242105, "learning_rate": 0.0002169046068690223, "loss": 0.5039, "step": 5793 }, { "epoch": 2.7265882352941175, "grad_norm": 0.8479692568259589, "learning_rate": 0.00021686536986062043, "loss": 0.5977, "step": 5794 }, { "epoch": 2.7270588235294118, "grad_norm": 0.692009279528225, "learning_rate": 0.000216826127141588, "loss": 0.3652, "step": 5795 }, { "epoch": 2.727529411764706, "grad_norm": 0.7299330311822633, "learning_rate": 0.0002167868787152765, "loss": 0.4434, "step": 5796 }, { "epoch": 2.7279999999999998, "grad_norm": 1.217740461317879, "learning_rate": 0.000216747624585038, "loss": 1.4766, "step": 5797 }, { "epoch": 2.728470588235294, "grad_norm": 1.5507496519934012, "learning_rate": 0.00021670836475422498, "loss": 1.1172, "step": 5798 }, { "epoch": 2.728941176470588, "grad_norm": 0.9870325159488043, "learning_rate": 0.00021666909922619037, "loss": 1.4219, "step": 5799 }, { "epoch": 2.7294117647058824, "grad_norm": 1.571398198576584, "learning_rate": 0.0002166298280042877, "loss": 1.4297, "step": 5800 }, { "epoch": 2.7298823529411766, "grad_norm": 0.5519950371307543, "learning_rate": 0.00021659055109187088, "loss": 0.1865, "step": 5801 }, { "epoch": 2.7303529411764704, "grad_norm": 1.0511898613512156, "learning_rate": 0.0002165512684922944, "loss": 0.6289, "step": 5802 }, { "epoch": 2.7308235294117646, "grad_norm": 0.6489240231915745, "learning_rate": 0.0002165119802089131, "loss": 0.377, "step": 5803 }, { "epoch": 2.731294117647059, "grad_norm": 1.5796749855061143, "learning_rate": 0.00021647268624508254, "loss": 1.3984, "step": 5804 }, { "epoch": 2.731764705882353, "grad_norm": 0.6703550577966164, "learning_rate": 0.00021643338660415844, "loss": 0.4199, "step": 5805 }, { "epoch": 2.732235294117647, "grad_norm": 0.7418649572032036, "learning_rate": 0.0002163940812894973, "loss": 0.5742, "step": 5806 }, { "epoch": 2.732705882352941, "grad_norm": 1.4211336128042422, "learning_rate": 0.00021635477030445597, "loss": 1.5781, "step": 5807 }, { "epoch": 2.7331764705882353, "grad_norm": 0.7976109779810332, "learning_rate": 0.0002163154536523918, "loss": 0.332, "step": 5808 }, { "epoch": 2.7336470588235295, "grad_norm": 1.0974886898840364, "learning_rate": 0.0002162761313366626, "loss": 1.6562, "step": 5809 }, { "epoch": 2.7341176470588238, "grad_norm": 0.7165506650982462, "learning_rate": 0.0002162368033606267, "loss": 0.4004, "step": 5810 }, { "epoch": 2.7345882352941175, "grad_norm": 0.895407038102111, "learning_rate": 0.00021619746972764288, "loss": 0.8359, "step": 5811 }, { "epoch": 2.7350588235294118, "grad_norm": 0.8360482503911386, "learning_rate": 0.00021615813044107045, "loss": 0.7422, "step": 5812 }, { "epoch": 2.735529411764706, "grad_norm": 1.2169014032933207, "learning_rate": 0.00021611878550426922, "loss": 1.625, "step": 5813 }, { "epoch": 2.7359999999999998, "grad_norm": 1.2855963562380952, "learning_rate": 0.00021607943492059936, "loss": 1.4062, "step": 5814 }, { "epoch": 2.736470588235294, "grad_norm": 0.8688085172293217, "learning_rate": 0.00021604007869342168, "loss": 0.7305, "step": 5815 }, { "epoch": 2.736941176470588, "grad_norm": 0.9853239585509584, "learning_rate": 0.00021600071682609733, "loss": 0.9219, "step": 5816 }, { "epoch": 2.7374117647058824, "grad_norm": 0.90309583058985, "learning_rate": 0.00021596134932198803, "loss": 0.7695, "step": 5817 }, { "epoch": 2.7378823529411767, "grad_norm": 1.8977265231713638, "learning_rate": 0.00021592197618445596, "loss": 2.3125, "step": 5818 }, { "epoch": 2.7383529411764704, "grad_norm": 0.9227646471315083, "learning_rate": 0.00021588259741686373, "loss": 1.2734, "step": 5819 }, { "epoch": 2.7388235294117647, "grad_norm": 0.38510389485623575, "learning_rate": 0.00021584321302257464, "loss": 0.1079, "step": 5820 }, { "epoch": 2.739294117647059, "grad_norm": 1.0812147057698915, "learning_rate": 0.00021580382300495215, "loss": 1.3203, "step": 5821 }, { "epoch": 2.739764705882353, "grad_norm": 0.9417327764666981, "learning_rate": 0.00021576442736736032, "loss": 0.4316, "step": 5822 }, { "epoch": 2.740235294117647, "grad_norm": 1.244845710208417, "learning_rate": 0.00021572502611316393, "loss": 2.1562, "step": 5823 }, { "epoch": 2.740705882352941, "grad_norm": 0.8093243443840951, "learning_rate": 0.0002156856192457279, "loss": 0.7539, "step": 5824 }, { "epoch": 2.7411764705882353, "grad_norm": 1.3013277004270993, "learning_rate": 0.00021564620676841782, "loss": 1.0078, "step": 5825 }, { "epoch": 2.7416470588235295, "grad_norm": 0.6031834128878352, "learning_rate": 0.0002156067886845996, "loss": 0.6484, "step": 5826 }, { "epoch": 2.7421176470588238, "grad_norm": 1.12834685957125, "learning_rate": 0.00021556736499763992, "loss": 1.5, "step": 5827 }, { "epoch": 2.7425882352941175, "grad_norm": 0.8928696648285822, "learning_rate": 0.00021552793571090568, "loss": 0.7031, "step": 5828 }, { "epoch": 2.7430588235294118, "grad_norm": 0.8802429599968428, "learning_rate": 0.0002154885008277642, "loss": 0.6445, "step": 5829 }, { "epoch": 2.743529411764706, "grad_norm": 1.0397254107010712, "learning_rate": 0.0002154490603515836, "loss": 1.2891, "step": 5830 }, { "epoch": 2.7439999999999998, "grad_norm": 0.8668117172386341, "learning_rate": 0.0002154096142857323, "loss": 0.8555, "step": 5831 }, { "epoch": 2.744470588235294, "grad_norm": 0.6439014484657629, "learning_rate": 0.000215370162633579, "loss": 0.2871, "step": 5832 }, { "epoch": 2.744941176470588, "grad_norm": 1.0330590321025603, "learning_rate": 0.00021533070539849316, "loss": 0.9766, "step": 5833 }, { "epoch": 2.7454117647058824, "grad_norm": 0.31463254060459434, "learning_rate": 0.00021529124258384466, "loss": 0.0481, "step": 5834 }, { "epoch": 2.7458823529411767, "grad_norm": 0.9816549648918751, "learning_rate": 0.00021525177419300383, "loss": 1.4922, "step": 5835 }, { "epoch": 2.7463529411764704, "grad_norm": 0.8924513444896688, "learning_rate": 0.0002152123002293414, "loss": 0.8633, "step": 5836 }, { "epoch": 2.7468235294117647, "grad_norm": 0.5413486097271895, "learning_rate": 0.0002151728206962287, "loss": 0.498, "step": 5837 }, { "epoch": 2.747294117647059, "grad_norm": 1.4814270660071875, "learning_rate": 0.00021513333559703738, "loss": 1.1094, "step": 5838 }, { "epoch": 2.7477647058823527, "grad_norm": 1.272693558422664, "learning_rate": 0.00021509384493513978, "loss": 1.8594, "step": 5839 }, { "epoch": 2.748235294117647, "grad_norm": 0.44086006864600724, "learning_rate": 0.00021505434871390848, "loss": 0.1855, "step": 5840 }, { "epoch": 2.748705882352941, "grad_norm": 0.7822059980073082, "learning_rate": 0.00021501484693671677, "loss": 0.4512, "step": 5841 }, { "epoch": 2.7491764705882353, "grad_norm": 1.001407442822012, "learning_rate": 0.00021497533960693823, "loss": 0.7812, "step": 5842 }, { "epoch": 2.7496470588235296, "grad_norm": 0.6260462289979409, "learning_rate": 0.00021493582672794693, "loss": 0.2891, "step": 5843 }, { "epoch": 2.7501176470588238, "grad_norm": 1.0774142020734117, "learning_rate": 0.00021489630830311758, "loss": 1.1875, "step": 5844 }, { "epoch": 2.7505882352941176, "grad_norm": 0.8579137003827231, "learning_rate": 0.00021485678433582516, "loss": 0.498, "step": 5845 }, { "epoch": 2.751058823529412, "grad_norm": 1.3162296960766433, "learning_rate": 0.00021481725482944518, "loss": 1.625, "step": 5846 }, { "epoch": 2.751529411764706, "grad_norm": 0.8285924288688508, "learning_rate": 0.0002147777197873538, "loss": 0.5039, "step": 5847 }, { "epoch": 2.752, "grad_norm": 1.170619907010905, "learning_rate": 0.0002147381792129274, "loss": 1.5, "step": 5848 }, { "epoch": 2.752470588235294, "grad_norm": 1.1098656574164831, "learning_rate": 0.00021469863310954296, "loss": 1.4297, "step": 5849 }, { "epoch": 2.7529411764705882, "grad_norm": 0.8454820373250961, "learning_rate": 0.00021465908148057787, "loss": 0.4727, "step": 5850 }, { "epoch": 2.7534117647058824, "grad_norm": 1.2474524173673869, "learning_rate": 0.00021461952432941006, "loss": 2.125, "step": 5851 }, { "epoch": 2.7538823529411767, "grad_norm": 1.4993804995195783, "learning_rate": 0.00021457996165941798, "loss": 1.3672, "step": 5852 }, { "epoch": 2.7543529411764704, "grad_norm": 0.7756395861612344, "learning_rate": 0.00021454039347398033, "loss": 0.6992, "step": 5853 }, { "epoch": 2.7548235294117647, "grad_norm": 0.5451585578604625, "learning_rate": 0.00021450081977647658, "loss": 0.1855, "step": 5854 }, { "epoch": 2.755294117647059, "grad_norm": 0.7238938385798541, "learning_rate": 0.00021446124057028642, "loss": 0.4355, "step": 5855 }, { "epoch": 2.7557647058823527, "grad_norm": 0.8356378102205116, "learning_rate": 0.00021442165585879014, "loss": 0.6797, "step": 5856 }, { "epoch": 2.756235294117647, "grad_norm": 0.9006628092971942, "learning_rate": 0.00021438206564536845, "loss": 0.8906, "step": 5857 }, { "epoch": 2.756705882352941, "grad_norm": 0.682625346582807, "learning_rate": 0.00021434246993340256, "loss": 0.7617, "step": 5858 }, { "epoch": 2.7571764705882353, "grad_norm": 1.1903884359536963, "learning_rate": 0.00021430286872627417, "loss": 1.3203, "step": 5859 }, { "epoch": 2.7576470588235296, "grad_norm": 0.9456729882200775, "learning_rate": 0.00021426326202736532, "loss": 0.4219, "step": 5860 }, { "epoch": 2.758117647058824, "grad_norm": 0.45534756767079065, "learning_rate": 0.00021422364984005876, "loss": 0.2246, "step": 5861 }, { "epoch": 2.7585882352941176, "grad_norm": 0.7827291485045165, "learning_rate": 0.00021418403216773746, "loss": 0.5703, "step": 5862 }, { "epoch": 2.759058823529412, "grad_norm": 0.7916486643905701, "learning_rate": 0.000214144409013785, "loss": 0.6523, "step": 5863 }, { "epoch": 2.759529411764706, "grad_norm": 1.235016265306768, "learning_rate": 0.00021410478038158537, "loss": 0.9844, "step": 5864 }, { "epoch": 2.76, "grad_norm": 0.7922162763225182, "learning_rate": 0.00021406514627452306, "loss": 0.6289, "step": 5865 }, { "epoch": 2.760470588235294, "grad_norm": 0.9893265828463178, "learning_rate": 0.00021402550669598303, "loss": 0.7383, "step": 5866 }, { "epoch": 2.7609411764705882, "grad_norm": 1.2693470817439816, "learning_rate": 0.00021398586164935067, "loss": 1.6094, "step": 5867 }, { "epoch": 2.7614117647058825, "grad_norm": 1.374989257581171, "learning_rate": 0.0002139462111380119, "loss": 1.1484, "step": 5868 }, { "epoch": 2.7618823529411767, "grad_norm": 1.021531824237861, "learning_rate": 0.00021390655516535304, "loss": 1.2578, "step": 5869 }, { "epoch": 2.7623529411764705, "grad_norm": 1.1980766411319028, "learning_rate": 0.00021386689373476087, "loss": 1.3281, "step": 5870 }, { "epoch": 2.7628235294117647, "grad_norm": 0.9068712661877222, "learning_rate": 0.00021382722684962276, "loss": 0.5625, "step": 5871 }, { "epoch": 2.763294117647059, "grad_norm": 0.8238295437487375, "learning_rate": 0.00021378755451332637, "loss": 1.0312, "step": 5872 }, { "epoch": 2.7637647058823527, "grad_norm": 1.2624948385286248, "learning_rate": 0.00021374787672926, "loss": 1.6328, "step": 5873 }, { "epoch": 2.764235294117647, "grad_norm": 0.8407116831303528, "learning_rate": 0.00021370819350081222, "loss": 1.1172, "step": 5874 }, { "epoch": 2.764705882352941, "grad_norm": 0.8859996154972264, "learning_rate": 0.00021366850483137226, "loss": 0.7383, "step": 5875 }, { "epoch": 2.7651764705882353, "grad_norm": 0.87086619343701, "learning_rate": 0.00021362881072432965, "loss": 0.9844, "step": 5876 }, { "epoch": 2.7656470588235296, "grad_norm": 1.0306952520141075, "learning_rate": 0.00021358911118307455, "loss": 0.7969, "step": 5877 }, { "epoch": 2.766117647058824, "grad_norm": 0.6844363960643277, "learning_rate": 0.00021354940621099745, "loss": 0.8359, "step": 5878 }, { "epoch": 2.7665882352941176, "grad_norm": 0.504500887568576, "learning_rate": 0.00021350969581148934, "loss": 0.2256, "step": 5879 }, { "epoch": 2.767058823529412, "grad_norm": 0.9394200681121068, "learning_rate": 0.00021346997998794172, "loss": 0.7422, "step": 5880 }, { "epoch": 2.767529411764706, "grad_norm": 2.376295876501953, "learning_rate": 0.00021343025874374645, "loss": 2.125, "step": 5881 }, { "epoch": 2.768, "grad_norm": 0.8665773679615507, "learning_rate": 0.000213390532082296, "loss": 0.7383, "step": 5882 }, { "epoch": 2.768470588235294, "grad_norm": 0.7534280619046043, "learning_rate": 0.00021335080000698317, "loss": 0.4707, "step": 5883 }, { "epoch": 2.7689411764705882, "grad_norm": 0.6412669556556522, "learning_rate": 0.00021331106252120126, "loss": 0.3633, "step": 5884 }, { "epoch": 2.7694117647058825, "grad_norm": 1.0275301015541864, "learning_rate": 0.00021327131962834407, "loss": 1.3125, "step": 5885 }, { "epoch": 2.7698823529411767, "grad_norm": 0.5854014777476444, "learning_rate": 0.00021323157133180587, "loss": 0.6719, "step": 5886 }, { "epoch": 2.7703529411764705, "grad_norm": 0.7907947924085632, "learning_rate": 0.0002131918176349813, "loss": 0.3809, "step": 5887 }, { "epoch": 2.7708235294117647, "grad_norm": 0.4483276047263087, "learning_rate": 0.0002131520585412656, "loss": 0.2324, "step": 5888 }, { "epoch": 2.771294117647059, "grad_norm": 0.8331822632019874, "learning_rate": 0.0002131122940540543, "loss": 0.5312, "step": 5889 }, { "epoch": 2.7717647058823527, "grad_norm": 0.7091973788175007, "learning_rate": 0.00021307252417674356, "loss": 0.3359, "step": 5890 }, { "epoch": 2.772235294117647, "grad_norm": 0.972715350787507, "learning_rate": 0.0002130327489127298, "loss": 1.2812, "step": 5891 }, { "epoch": 2.772705882352941, "grad_norm": 1.0413310124181232, "learning_rate": 0.0002129929682654102, "loss": 1.4922, "step": 5892 }, { "epoch": 2.7731764705882354, "grad_norm": 0.9039176856657312, "learning_rate": 0.0002129531822381821, "loss": 0.5469, "step": 5893 }, { "epoch": 2.7736470588235296, "grad_norm": 0.9805447651107901, "learning_rate": 0.00021291339083444344, "loss": 1.2188, "step": 5894 }, { "epoch": 2.774117647058824, "grad_norm": 1.2664756510591662, "learning_rate": 0.00021287359405759265, "loss": 1.2891, "step": 5895 }, { "epoch": 2.7745882352941176, "grad_norm": 1.0697480676188533, "learning_rate": 0.0002128337919110285, "loss": 1.5859, "step": 5896 }, { "epoch": 2.775058823529412, "grad_norm": 0.7523277003126211, "learning_rate": 0.00021279398439815034, "loss": 0.4062, "step": 5897 }, { "epoch": 2.775529411764706, "grad_norm": 0.6793092742943437, "learning_rate": 0.00021275417152235787, "loss": 0.4766, "step": 5898 }, { "epoch": 2.776, "grad_norm": 0.9361538291520402, "learning_rate": 0.00021271435328705138, "loss": 0.9609, "step": 5899 }, { "epoch": 2.776470588235294, "grad_norm": 0.896875534014107, "learning_rate": 0.00021267452969563153, "loss": 0.9414, "step": 5900 }, { "epoch": 2.7769411764705882, "grad_norm": 0.756972443682883, "learning_rate": 0.00021263470075149937, "loss": 0.4863, "step": 5901 }, { "epoch": 2.7774117647058825, "grad_norm": 0.5020465309081507, "learning_rate": 0.00021259486645805662, "loss": 0.2539, "step": 5902 }, { "epoch": 2.7778823529411767, "grad_norm": 1.021559477602974, "learning_rate": 0.00021255502681870518, "loss": 1.125, "step": 5903 }, { "epoch": 2.7783529411764705, "grad_norm": 0.578470101748657, "learning_rate": 0.00021251518183684767, "loss": 0.3242, "step": 5904 }, { "epoch": 2.7788235294117647, "grad_norm": 0.6867057183433357, "learning_rate": 0.00021247533151588693, "loss": 0.3555, "step": 5905 }, { "epoch": 2.779294117647059, "grad_norm": 0.5285623068471542, "learning_rate": 0.0002124354758592265, "loss": 0.1299, "step": 5906 }, { "epoch": 2.7797647058823527, "grad_norm": 1.4471880672701924, "learning_rate": 0.00021239561487027017, "loss": 1.2031, "step": 5907 }, { "epoch": 2.780235294117647, "grad_norm": 1.251042136028203, "learning_rate": 0.00021235574855242226, "loss": 1.4062, "step": 5908 }, { "epoch": 2.780705882352941, "grad_norm": 0.7609547414170904, "learning_rate": 0.00021231587690908758, "loss": 0.4707, "step": 5909 }, { "epoch": 2.7811764705882354, "grad_norm": 0.7859801559955616, "learning_rate": 0.00021227599994367137, "loss": 0.6133, "step": 5910 }, { "epoch": 2.7816470588235296, "grad_norm": 0.7992343067476747, "learning_rate": 0.0002122361176595793, "loss": 0.6172, "step": 5911 }, { "epoch": 2.782117647058824, "grad_norm": 1.3667611042420464, "learning_rate": 0.0002121962300602175, "loss": 1.7344, "step": 5912 }, { "epoch": 2.7825882352941176, "grad_norm": 0.6961561993570582, "learning_rate": 0.00021215633714899264, "loss": 0.5977, "step": 5913 }, { "epoch": 2.783058823529412, "grad_norm": 0.7529358609069068, "learning_rate": 0.00021211643892931167, "loss": 0.4023, "step": 5914 }, { "epoch": 2.783529411764706, "grad_norm": 0.9272082008797969, "learning_rate": 0.00021207653540458209, "loss": 0.7422, "step": 5915 }, { "epoch": 2.784, "grad_norm": 2.0280477765634486, "learning_rate": 0.00021203662657821198, "loss": 1.3438, "step": 5916 }, { "epoch": 2.784470588235294, "grad_norm": 1.171332829181307, "learning_rate": 0.00021199671245360964, "loss": 1.3125, "step": 5917 }, { "epoch": 2.7849411764705883, "grad_norm": 0.904376715049806, "learning_rate": 0.0002119567930341839, "loss": 0.6562, "step": 5918 }, { "epoch": 2.7854117647058825, "grad_norm": 1.2332225273825475, "learning_rate": 0.00021191686832334423, "loss": 1.7656, "step": 5919 }, { "epoch": 2.7858823529411767, "grad_norm": 0.8818512172658429, "learning_rate": 0.0002118769383245003, "loss": 1.3906, "step": 5920 }, { "epoch": 2.7863529411764705, "grad_norm": 0.42300223339054893, "learning_rate": 0.0002118370030410623, "loss": 0.1738, "step": 5921 }, { "epoch": 2.7868235294117647, "grad_norm": 0.7793368971430571, "learning_rate": 0.0002117970624764409, "loss": 0.4492, "step": 5922 }, { "epoch": 2.787294117647059, "grad_norm": 0.6117797298881789, "learning_rate": 0.00021175711663404728, "loss": 0.1699, "step": 5923 }, { "epoch": 2.7877647058823527, "grad_norm": 0.4738958856113477, "learning_rate": 0.000211717165517293, "loss": 0.207, "step": 5924 }, { "epoch": 2.788235294117647, "grad_norm": 0.9400469438375478, "learning_rate": 0.00021167720912959004, "loss": 0.7109, "step": 5925 }, { "epoch": 2.788705882352941, "grad_norm": 0.9850706551104537, "learning_rate": 0.00021163724747435091, "loss": 0.7188, "step": 5926 }, { "epoch": 2.7891764705882354, "grad_norm": 1.0589606059624295, "learning_rate": 0.00021159728055498856, "loss": 0.8203, "step": 5927 }, { "epoch": 2.7896470588235296, "grad_norm": 0.8092836260661953, "learning_rate": 0.00021155730837491625, "loss": 0.9648, "step": 5928 }, { "epoch": 2.7901176470588234, "grad_norm": 0.972036535300292, "learning_rate": 0.0002115173309375479, "loss": 1.1094, "step": 5929 }, { "epoch": 2.7905882352941176, "grad_norm": 1.0634978120558591, "learning_rate": 0.00021147734824629772, "loss": 0.9609, "step": 5930 }, { "epoch": 2.791058823529412, "grad_norm": 0.9607554864323737, "learning_rate": 0.0002114373603045805, "loss": 0.9102, "step": 5931 }, { "epoch": 2.791529411764706, "grad_norm": 0.8217782800073246, "learning_rate": 0.00021139736711581132, "loss": 0.8945, "step": 5932 }, { "epoch": 2.792, "grad_norm": 0.6810899939904488, "learning_rate": 0.00021135736868340587, "loss": 0.2617, "step": 5933 }, { "epoch": 2.792470588235294, "grad_norm": 0.7042628354486763, "learning_rate": 0.00021131736501078018, "loss": 0.4941, "step": 5934 }, { "epoch": 2.7929411764705883, "grad_norm": 0.2907525818169788, "learning_rate": 0.0002112773561013507, "loss": 0.0413, "step": 5935 }, { "epoch": 2.7934117647058825, "grad_norm": 2.530839016119866, "learning_rate": 0.0002112373419585345, "loss": 3.4844, "step": 5936 }, { "epoch": 2.7938823529411767, "grad_norm": 0.6870414411142972, "learning_rate": 0.00021119732258574893, "loss": 0.4316, "step": 5937 }, { "epoch": 2.7943529411764705, "grad_norm": 0.7263425353243401, "learning_rate": 0.0002111572979864118, "loss": 0.3809, "step": 5938 }, { "epoch": 2.7948235294117647, "grad_norm": 1.2913617755334366, "learning_rate": 0.00021111726816394147, "loss": 2.125, "step": 5939 }, { "epoch": 2.795294117647059, "grad_norm": 1.0963549503771006, "learning_rate": 0.00021107723312175665, "loss": 1.3516, "step": 5940 }, { "epoch": 2.7957647058823527, "grad_norm": 0.9214107357188764, "learning_rate": 0.00021103719286327658, "loss": 0.5234, "step": 5941 }, { "epoch": 2.796235294117647, "grad_norm": 0.908989590511627, "learning_rate": 0.0002109971473919208, "loss": 0.416, "step": 5942 }, { "epoch": 2.796705882352941, "grad_norm": 0.7932597234150471, "learning_rate": 0.00021095709671110942, "loss": 0.8594, "step": 5943 }, { "epoch": 2.7971764705882354, "grad_norm": 0.9775949790272824, "learning_rate": 0.00021091704082426308, "loss": 0.9961, "step": 5944 }, { "epoch": 2.7976470588235296, "grad_norm": 1.3851827467276023, "learning_rate": 0.0002108769797348026, "loss": 2.0156, "step": 5945 }, { "epoch": 2.7981176470588234, "grad_norm": 1.0749451638306557, "learning_rate": 0.00021083691344614947, "loss": 0.7695, "step": 5946 }, { "epoch": 2.7985882352941176, "grad_norm": 0.8567682936564471, "learning_rate": 0.00021079684196172554, "loss": 1.3594, "step": 5947 }, { "epoch": 2.799058823529412, "grad_norm": 1.105508310979029, "learning_rate": 0.00021075676528495307, "loss": 1.2266, "step": 5948 }, { "epoch": 2.799529411764706, "grad_norm": 0.6184939349260022, "learning_rate": 0.00021071668341925482, "loss": 0.6797, "step": 5949 }, { "epoch": 2.8, "grad_norm": 1.1907191574024165, "learning_rate": 0.00021067659636805403, "loss": 1.0469, "step": 5950 }, { "epoch": 2.800470588235294, "grad_norm": 0.38683110540009885, "learning_rate": 0.00021063650413477428, "loss": 0.083, "step": 5951 }, { "epoch": 2.8009411764705883, "grad_norm": 1.1735940732905237, "learning_rate": 0.00021059640672283964, "loss": 1.2734, "step": 5952 }, { "epoch": 2.8014117647058825, "grad_norm": 0.627319067358784, "learning_rate": 0.0002105563041356747, "loss": 0.3477, "step": 5953 }, { "epoch": 2.8018823529411767, "grad_norm": 0.91435181713921, "learning_rate": 0.00021051619637670431, "loss": 1.0156, "step": 5954 }, { "epoch": 2.8023529411764705, "grad_norm": 1.1912983587585046, "learning_rate": 0.00021047608344935395, "loss": 1.6719, "step": 5955 }, { "epoch": 2.8028235294117647, "grad_norm": 0.7224517034347654, "learning_rate": 0.0002104359653570494, "loss": 0.7969, "step": 5956 }, { "epoch": 2.803294117647059, "grad_norm": 0.8909799361207662, "learning_rate": 0.00021039584210321699, "loss": 1.0625, "step": 5957 }, { "epoch": 2.8037647058823527, "grad_norm": 0.5610682296469675, "learning_rate": 0.00021035571369128341, "loss": 0.5508, "step": 5958 }, { "epoch": 2.804235294117647, "grad_norm": 1.0094852063511983, "learning_rate": 0.00021031558012467582, "loss": 1.0703, "step": 5959 }, { "epoch": 2.804705882352941, "grad_norm": 1.0118188029794575, "learning_rate": 0.00021027544140682188, "loss": 0.9609, "step": 5960 }, { "epoch": 2.8051764705882354, "grad_norm": 1.0528984232442071, "learning_rate": 0.00021023529754114956, "loss": 1.1406, "step": 5961 }, { "epoch": 2.8056470588235296, "grad_norm": 0.8216448663049646, "learning_rate": 0.0002101951485310874, "loss": 0.8203, "step": 5962 }, { "epoch": 2.8061176470588234, "grad_norm": 0.9265100988964832, "learning_rate": 0.0002101549943800643, "loss": 1.4219, "step": 5963 }, { "epoch": 2.8065882352941176, "grad_norm": 0.999178083804766, "learning_rate": 0.00021011483509150963, "loss": 1.2812, "step": 5964 }, { "epoch": 2.807058823529412, "grad_norm": 0.6077128919775189, "learning_rate": 0.00021007467066885314, "loss": 0.5391, "step": 5965 }, { "epoch": 2.8075294117647056, "grad_norm": 0.6566726110996205, "learning_rate": 0.00021003450111552514, "loss": 0.375, "step": 5966 }, { "epoch": 2.808, "grad_norm": 1.1234605435938512, "learning_rate": 0.00020999432643495626, "loss": 1.5156, "step": 5967 }, { "epoch": 2.808470588235294, "grad_norm": 0.8634368141076794, "learning_rate": 0.00020995414663057768, "loss": 0.7383, "step": 5968 }, { "epoch": 2.8089411764705883, "grad_norm": 0.7900429938309325, "learning_rate": 0.00020991396170582088, "loss": 0.5078, "step": 5969 }, { "epoch": 2.8094117647058825, "grad_norm": 0.941661125824496, "learning_rate": 0.00020987377166411785, "loss": 1.0078, "step": 5970 }, { "epoch": 2.8098823529411767, "grad_norm": 0.7850643953535998, "learning_rate": 0.00020983357650890106, "loss": 0.8359, "step": 5971 }, { "epoch": 2.8103529411764705, "grad_norm": 1.0033767460052618, "learning_rate": 0.00020979337624360336, "loss": 0.7148, "step": 5972 }, { "epoch": 2.8108235294117647, "grad_norm": 0.737145180310122, "learning_rate": 0.000209753170871658, "loss": 0.7266, "step": 5973 }, { "epoch": 2.811294117647059, "grad_norm": 0.7475394268309432, "learning_rate": 0.00020971296039649882, "loss": 0.3203, "step": 5974 }, { "epoch": 2.8117647058823527, "grad_norm": 1.0031579389033114, "learning_rate": 0.00020967274482155992, "loss": 1.0, "step": 5975 }, { "epoch": 2.812235294117647, "grad_norm": 0.7718319601681508, "learning_rate": 0.00020963252415027588, "loss": 0.7461, "step": 5976 }, { "epoch": 2.812705882352941, "grad_norm": 0.8770533579906422, "learning_rate": 0.0002095922983860818, "loss": 0.5234, "step": 5977 }, { "epoch": 2.8131764705882354, "grad_norm": 0.8208463109415172, "learning_rate": 0.00020955206753241315, "loss": 0.9258, "step": 5978 }, { "epoch": 2.8136470588235296, "grad_norm": 0.828180019933176, "learning_rate": 0.00020951183159270586, "loss": 0.5859, "step": 5979 }, { "epoch": 2.8141176470588234, "grad_norm": 0.8594145283599371, "learning_rate": 0.0002094715905703962, "loss": 0.9375, "step": 5980 }, { "epoch": 2.8145882352941176, "grad_norm": 0.8773717907840707, "learning_rate": 0.00020943134446892102, "loss": 0.9922, "step": 5981 }, { "epoch": 2.815058823529412, "grad_norm": 1.0433009138342655, "learning_rate": 0.00020939109329171752, "loss": 1.1875, "step": 5982 }, { "epoch": 2.8155294117647056, "grad_norm": 0.6599695975736669, "learning_rate": 0.00020935083704222331, "loss": 0.252, "step": 5983 }, { "epoch": 2.816, "grad_norm": 0.8747066129986882, "learning_rate": 0.00020931057572387655, "loss": 1.0625, "step": 5984 }, { "epoch": 2.816470588235294, "grad_norm": 0.7909486876159811, "learning_rate": 0.00020927030934011568, "loss": 0.416, "step": 5985 }, { "epoch": 2.8169411764705883, "grad_norm": 0.9703111088335213, "learning_rate": 0.00020923003789437967, "loss": 1.2109, "step": 5986 }, { "epoch": 2.8174117647058825, "grad_norm": 1.005072518396254, "learning_rate": 0.00020918976139010784, "loss": 1.0234, "step": 5987 }, { "epoch": 2.8178823529411767, "grad_norm": 1.1029709957912914, "learning_rate": 0.00020914947983074016, "loss": 0.5859, "step": 5988 }, { "epoch": 2.8183529411764705, "grad_norm": 1.1226272697377793, "learning_rate": 0.00020910919321971675, "loss": 1.3047, "step": 5989 }, { "epoch": 2.8188235294117647, "grad_norm": 0.9724087350918814, "learning_rate": 0.00020906890156047826, "loss": 0.6562, "step": 5990 }, { "epoch": 2.819294117647059, "grad_norm": 0.8674248869233605, "learning_rate": 0.0002090286048564659, "loss": 0.8828, "step": 5991 }, { "epoch": 2.8197647058823527, "grad_norm": 0.9690119469434696, "learning_rate": 0.00020898830311112118, "loss": 0.7539, "step": 5992 }, { "epoch": 2.820235294117647, "grad_norm": 0.8099214670385239, "learning_rate": 0.00020894799632788596, "loss": 0.4375, "step": 5993 }, { "epoch": 2.820705882352941, "grad_norm": 0.8094850941158376, "learning_rate": 0.0002089076845102028, "loss": 0.5547, "step": 5994 }, { "epoch": 2.8211764705882354, "grad_norm": 0.5616805824877793, "learning_rate": 0.00020886736766151438, "loss": 0.2393, "step": 5995 }, { "epoch": 2.8216470588235296, "grad_norm": 0.9545289909919745, "learning_rate": 0.00020882704578526406, "loss": 0.625, "step": 5996 }, { "epoch": 2.8221176470588234, "grad_norm": 1.1688890719805183, "learning_rate": 0.0002087867188848954, "loss": 1.3047, "step": 5997 }, { "epoch": 2.8225882352941176, "grad_norm": 0.4668674213144363, "learning_rate": 0.0002087463869638527, "loss": 0.1504, "step": 5998 }, { "epoch": 2.823058823529412, "grad_norm": 0.794837834116743, "learning_rate": 0.00020870605002558038, "loss": 0.6875, "step": 5999 }, { "epoch": 2.8235294117647056, "grad_norm": 0.5642763654369711, "learning_rate": 0.00020866570807352337, "loss": 0.5234, "step": 6000 }, { "epoch": 2.824, "grad_norm": 1.2660565526706806, "learning_rate": 0.0002086253611111272, "loss": 1.6172, "step": 6001 }, { "epoch": 2.824470588235294, "grad_norm": 0.5471327939379343, "learning_rate": 0.00020858500914183763, "loss": 0.2793, "step": 6002 }, { "epoch": 2.8249411764705883, "grad_norm": 0.8879058740205614, "learning_rate": 0.0002085446521691009, "loss": 0.9219, "step": 6003 }, { "epoch": 2.8254117647058825, "grad_norm": 0.8937659376222966, "learning_rate": 0.00020850429019636372, "loss": 0.8594, "step": 6004 }, { "epoch": 2.8258823529411767, "grad_norm": 0.6907932996243478, "learning_rate": 0.00020846392322707323, "loss": 0.4102, "step": 6005 }, { "epoch": 2.8263529411764705, "grad_norm": 1.2454757317193215, "learning_rate": 0.00020842355126467693, "loss": 1.0781, "step": 6006 }, { "epoch": 2.8268235294117647, "grad_norm": 0.9033443124143, "learning_rate": 0.00020838317431262274, "loss": 0.9883, "step": 6007 }, { "epoch": 2.827294117647059, "grad_norm": 0.9839646861676291, "learning_rate": 0.0002083427923743591, "loss": 0.6953, "step": 6008 }, { "epoch": 2.8277647058823527, "grad_norm": 0.7440963265168244, "learning_rate": 0.00020830240545333489, "loss": 0.377, "step": 6009 }, { "epoch": 2.828235294117647, "grad_norm": 1.4910970277070248, "learning_rate": 0.00020826201355299918, "loss": 1.5469, "step": 6010 }, { "epoch": 2.828705882352941, "grad_norm": 1.0767510420104158, "learning_rate": 0.0002082216166768018, "loss": 1.1172, "step": 6011 }, { "epoch": 2.8291764705882354, "grad_norm": 0.9617340366498897, "learning_rate": 0.00020818121482819277, "loss": 0.7422, "step": 6012 }, { "epoch": 2.8296470588235296, "grad_norm": 0.5622520464519425, "learning_rate": 0.00020814080801062263, "loss": 0.5117, "step": 6013 }, { "epoch": 2.8301176470588234, "grad_norm": 0.9155983571217716, "learning_rate": 0.00020810039622754228, "loss": 0.8203, "step": 6014 }, { "epoch": 2.8305882352941176, "grad_norm": 1.5998338060754078, "learning_rate": 0.00020805997948240314, "loss": 1.9219, "step": 6015 }, { "epoch": 2.831058823529412, "grad_norm": 0.9476705715193297, "learning_rate": 0.00020801955777865694, "loss": 1.3672, "step": 6016 }, { "epoch": 2.8315294117647056, "grad_norm": 0.8306164329537369, "learning_rate": 0.00020797913111975593, "loss": 0.5195, "step": 6017 }, { "epoch": 2.832, "grad_norm": 0.925363839912463, "learning_rate": 0.00020793869950915276, "loss": 1.1797, "step": 6018 }, { "epoch": 2.832470588235294, "grad_norm": 0.9860411863661829, "learning_rate": 0.00020789826295030046, "loss": 1.0312, "step": 6019 }, { "epoch": 2.8329411764705883, "grad_norm": 0.98150899445512, "learning_rate": 0.0002078578214466525, "loss": 1.3359, "step": 6020 }, { "epoch": 2.8334117647058825, "grad_norm": 1.0285184655794024, "learning_rate": 0.00020781737500166279, "loss": 1.1562, "step": 6021 }, { "epoch": 2.8338823529411767, "grad_norm": 0.3891781723776986, "learning_rate": 0.00020777692361878569, "loss": 0.1348, "step": 6022 }, { "epoch": 2.8343529411764705, "grad_norm": 1.0736309469668404, "learning_rate": 0.0002077364673014759, "loss": 0.9219, "step": 6023 }, { "epoch": 2.8348235294117647, "grad_norm": 1.3460821211989498, "learning_rate": 0.00020769600605318857, "loss": 2.3125, "step": 6024 }, { "epoch": 2.835294117647059, "grad_norm": 1.3035580534556581, "learning_rate": 0.0002076555398773794, "loss": 1.3438, "step": 6025 }, { "epoch": 2.8357647058823527, "grad_norm": 0.9884204604824484, "learning_rate": 0.00020761506877750428, "loss": 0.8906, "step": 6026 }, { "epoch": 2.836235294117647, "grad_norm": 1.066402210334321, "learning_rate": 0.00020757459275701966, "loss": 0.9258, "step": 6027 }, { "epoch": 2.836705882352941, "grad_norm": 1.1698737198185671, "learning_rate": 0.0002075341118193825, "loss": 1.2812, "step": 6028 }, { "epoch": 2.8371764705882354, "grad_norm": 0.8959949062410071, "learning_rate": 0.00020749362596804994, "loss": 1.5859, "step": 6029 }, { "epoch": 2.8376470588235296, "grad_norm": 0.3753401345992222, "learning_rate": 0.00020745313520647971, "loss": 0.0776, "step": 6030 }, { "epoch": 2.8381176470588234, "grad_norm": 0.8876371152253059, "learning_rate": 0.00020741263953812992, "loss": 0.6172, "step": 6031 }, { "epoch": 2.8385882352941176, "grad_norm": 0.7389615169970151, "learning_rate": 0.0002073721389664591, "loss": 0.4336, "step": 6032 }, { "epoch": 2.839058823529412, "grad_norm": 0.7196182465152804, "learning_rate": 0.00020733163349492625, "loss": 0.334, "step": 6033 }, { "epoch": 2.8395294117647056, "grad_norm": 0.6898347441266894, "learning_rate": 0.00020729112312699066, "loss": 0.4336, "step": 6034 }, { "epoch": 2.84, "grad_norm": 1.1093778446219933, "learning_rate": 0.00020725060786611212, "loss": 1.1641, "step": 6035 }, { "epoch": 2.840470588235294, "grad_norm": 1.1061748730780514, "learning_rate": 0.0002072100877157509, "loss": 1.0938, "step": 6036 }, { "epoch": 2.8409411764705883, "grad_norm": 0.8943713948973683, "learning_rate": 0.00020716956267936759, "loss": 1.1406, "step": 6037 }, { "epoch": 2.8414117647058825, "grad_norm": 0.5304540589226873, "learning_rate": 0.0002071290327604231, "loss": 0.1738, "step": 6038 }, { "epoch": 2.8418823529411767, "grad_norm": 1.0721721776522708, "learning_rate": 0.00020708849796237914, "loss": 1.0469, "step": 6039 }, { "epoch": 2.8423529411764705, "grad_norm": 1.8795141286413612, "learning_rate": 0.00020704795828869736, "loss": 0.7109, "step": 6040 }, { "epoch": 2.8428235294117647, "grad_norm": 1.324459771839715, "learning_rate": 0.00020700741374284014, "loss": 2.2969, "step": 6041 }, { "epoch": 2.843294117647059, "grad_norm": 1.0570025805778365, "learning_rate": 0.00020696686432827017, "loss": 1.0547, "step": 6042 }, { "epoch": 2.8437647058823528, "grad_norm": 1.064152599249459, "learning_rate": 0.00020692631004845055, "loss": 1.1875, "step": 6043 }, { "epoch": 2.844235294117647, "grad_norm": 1.0026580599189334, "learning_rate": 0.00020688575090684484, "loss": 0.8789, "step": 6044 }, { "epoch": 2.844705882352941, "grad_norm": 0.9424816561123592, "learning_rate": 0.00020684518690691696, "loss": 1.0781, "step": 6045 }, { "epoch": 2.8451764705882354, "grad_norm": 0.8958846207137394, "learning_rate": 0.00020680461805213133, "loss": 1.0156, "step": 6046 }, { "epoch": 2.8456470588235296, "grad_norm": 0.8347989903327065, "learning_rate": 0.00020676404434595265, "loss": 1.0547, "step": 6047 }, { "epoch": 2.8461176470588234, "grad_norm": 1.111179552128968, "learning_rate": 0.00020672346579184617, "loss": 1.5859, "step": 6048 }, { "epoch": 2.8465882352941176, "grad_norm": 1.0163703937866964, "learning_rate": 0.00020668288239327753, "loss": 1.1875, "step": 6049 }, { "epoch": 2.847058823529412, "grad_norm": 1.1282394350916758, "learning_rate": 0.00020664229415371266, "loss": 0.8945, "step": 6050 }, { "epoch": 2.8475294117647056, "grad_norm": 0.7743959582592039, "learning_rate": 0.00020660170107661803, "loss": 0.75, "step": 6051 }, { "epoch": 2.848, "grad_norm": 0.9794680896233614, "learning_rate": 0.00020656110316546054, "loss": 0.793, "step": 6052 }, { "epoch": 2.848470588235294, "grad_norm": 0.9369465790829412, "learning_rate": 0.00020652050042370743, "loss": 0.7891, "step": 6053 }, { "epoch": 2.8489411764705883, "grad_norm": 0.8725170773941542, "learning_rate": 0.0002064798928548263, "loss": 0.5273, "step": 6054 }, { "epoch": 2.8494117647058825, "grad_norm": 0.9273112061350105, "learning_rate": 0.00020643928046228525, "loss": 0.7031, "step": 6055 }, { "epoch": 2.8498823529411763, "grad_norm": 1.0571738987401804, "learning_rate": 0.00020639866324955285, "loss": 1.0703, "step": 6056 }, { "epoch": 2.8503529411764705, "grad_norm": 0.5760367010642101, "learning_rate": 0.00020635804122009802, "loss": 0.3047, "step": 6057 }, { "epoch": 2.8508235294117648, "grad_norm": 0.9433287563568721, "learning_rate": 0.00020631741437738998, "loss": 0.8555, "step": 6058 }, { "epoch": 2.851294117647059, "grad_norm": 0.9876267609993756, "learning_rate": 0.00020627678272489854, "loss": 1.4219, "step": 6059 }, { "epoch": 2.8517647058823528, "grad_norm": 0.9414336944208304, "learning_rate": 0.00020623614626609385, "loss": 0.8555, "step": 6060 }, { "epoch": 2.852235294117647, "grad_norm": 0.6357521840739637, "learning_rate": 0.0002061955050044464, "loss": 0.3184, "step": 6061 }, { "epoch": 2.852705882352941, "grad_norm": 0.7953740783223564, "learning_rate": 0.00020615485894342716, "loss": 1.0781, "step": 6062 }, { "epoch": 2.8531764705882354, "grad_norm": 0.7779715742570088, "learning_rate": 0.00020611420808650758, "loss": 0.5508, "step": 6063 }, { "epoch": 2.8536470588235296, "grad_norm": 1.0210064580320675, "learning_rate": 0.00020607355243715936, "loss": 1.2031, "step": 6064 }, { "epoch": 2.8541176470588234, "grad_norm": 0.94099635727199, "learning_rate": 0.00020603289199885473, "loss": 0.4238, "step": 6065 }, { "epoch": 2.8545882352941176, "grad_norm": 0.619058959655097, "learning_rate": 0.00020599222677506628, "loss": 0.3262, "step": 6066 }, { "epoch": 2.855058823529412, "grad_norm": 0.8754782162737196, "learning_rate": 0.00020595155676926704, "loss": 0.6953, "step": 6067 }, { "epoch": 2.8555294117647056, "grad_norm": 0.6492686839636564, "learning_rate": 0.0002059108819849304, "loss": 0.2988, "step": 6068 }, { "epoch": 2.856, "grad_norm": 1.0243533801635982, "learning_rate": 0.0002058702024255302, "loss": 1.3906, "step": 6069 }, { "epoch": 2.856470588235294, "grad_norm": 0.6934675262675948, "learning_rate": 0.00020582951809454067, "loss": 0.2812, "step": 6070 }, { "epoch": 2.8569411764705883, "grad_norm": 0.48605636818285214, "learning_rate": 0.00020578882899543644, "loss": 0.1455, "step": 6071 }, { "epoch": 2.8574117647058825, "grad_norm": 0.7135618885886549, "learning_rate": 0.0002057481351316926, "loss": 0.3301, "step": 6072 }, { "epoch": 2.8578823529411763, "grad_norm": 1.3132866813070827, "learning_rate": 0.00020570743650678454, "loss": 1.375, "step": 6073 }, { "epoch": 2.8583529411764705, "grad_norm": 0.9706200100243588, "learning_rate": 0.00020566673312418817, "loss": 0.7734, "step": 6074 }, { "epoch": 2.8588235294117648, "grad_norm": 0.5509060638786714, "learning_rate": 0.00020562602498737975, "loss": 0.168, "step": 6075 }, { "epoch": 2.859294117647059, "grad_norm": 1.0670714518079307, "learning_rate": 0.00020558531209983593, "loss": 0.9062, "step": 6076 }, { "epoch": 2.8597647058823528, "grad_norm": 0.6694125366678091, "learning_rate": 0.00020554459446503386, "loss": 0.3125, "step": 6077 }, { "epoch": 2.860235294117647, "grad_norm": 0.8050745874531532, "learning_rate": 0.00020550387208645093, "loss": 0.707, "step": 6078 }, { "epoch": 2.860705882352941, "grad_norm": 0.7640208908920303, "learning_rate": 0.0002054631449675651, "loss": 0.7148, "step": 6079 }, { "epoch": 2.8611764705882354, "grad_norm": 1.3581242805665457, "learning_rate": 0.00020542241311185467, "loss": 1.75, "step": 6080 }, { "epoch": 2.8616470588235297, "grad_norm": 0.9205752984735959, "learning_rate": 0.0002053816765227983, "loss": 0.8594, "step": 6081 }, { "epoch": 2.8621176470588234, "grad_norm": 0.5814809632086263, "learning_rate": 0.0002053409352038751, "loss": 0.2598, "step": 6082 }, { "epoch": 2.8625882352941177, "grad_norm": 0.8480528498703617, "learning_rate": 0.00020530018915856457, "loss": 0.668, "step": 6083 }, { "epoch": 2.863058823529412, "grad_norm": 0.7415432317615487, "learning_rate": 0.00020525943839034668, "loss": 0.6562, "step": 6084 }, { "epoch": 2.8635294117647057, "grad_norm": 0.7092292049614408, "learning_rate": 0.0002052186829027017, "loss": 0.4023, "step": 6085 }, { "epoch": 2.864, "grad_norm": 1.0741899676018338, "learning_rate": 0.00020517792269911033, "loss": 1.5234, "step": 6086 }, { "epoch": 2.864470588235294, "grad_norm": 0.6789186321765658, "learning_rate": 0.0002051371577830537, "loss": 0.3984, "step": 6087 }, { "epoch": 2.8649411764705883, "grad_norm": 0.6305309348123542, "learning_rate": 0.00020509638815801342, "loss": 0.2656, "step": 6088 }, { "epoch": 2.8654117647058825, "grad_norm": 0.626930170790992, "learning_rate": 0.00020505561382747127, "loss": 0.3789, "step": 6089 }, { "epoch": 2.8658823529411763, "grad_norm": 0.9626257611199674, "learning_rate": 0.00020501483479490968, "loss": 1.3047, "step": 6090 }, { "epoch": 2.8663529411764705, "grad_norm": 0.7077368384783972, "learning_rate": 0.00020497405106381138, "loss": 0.7461, "step": 6091 }, { "epoch": 2.8668235294117648, "grad_norm": 0.9929848239335343, "learning_rate": 0.0002049332626376594, "loss": 1.2031, "step": 6092 }, { "epoch": 2.8672941176470585, "grad_norm": 1.3771809032561715, "learning_rate": 0.00020489246951993742, "loss": 2.0469, "step": 6093 }, { "epoch": 2.8677647058823528, "grad_norm": 1.067542115885805, "learning_rate": 0.00020485167171412926, "loss": 1.5391, "step": 6094 }, { "epoch": 2.868235294117647, "grad_norm": 1.1080368140710097, "learning_rate": 0.00020481086922371935, "loss": 0.7812, "step": 6095 }, { "epoch": 2.868705882352941, "grad_norm": 0.781373303176325, "learning_rate": 0.00020477006205219228, "loss": 0.4316, "step": 6096 }, { "epoch": 2.8691764705882354, "grad_norm": 0.8852045062402865, "learning_rate": 0.0002047292502030333, "loss": 0.9062, "step": 6097 }, { "epoch": 2.8696470588235297, "grad_norm": 0.8292332381505685, "learning_rate": 0.00020468843367972786, "loss": 0.7148, "step": 6098 }, { "epoch": 2.8701176470588234, "grad_norm": 0.8281268882631333, "learning_rate": 0.00020464761248576203, "loss": 0.5586, "step": 6099 }, { "epoch": 2.8705882352941177, "grad_norm": 0.6704400937202685, "learning_rate": 0.00020460678662462194, "loss": 0.3301, "step": 6100 }, { "epoch": 2.871058823529412, "grad_norm": 1.0099630385103295, "learning_rate": 0.0002045659560997945, "loss": 1.0, "step": 6101 }, { "epoch": 2.8715294117647057, "grad_norm": 0.7975404904592927, "learning_rate": 0.00020452512091476678, "loss": 0.6328, "step": 6102 }, { "epoch": 2.872, "grad_norm": 1.044716609482021, "learning_rate": 0.0002044842810730262, "loss": 1.2344, "step": 6103 }, { "epoch": 2.872470588235294, "grad_norm": 0.9375230163818925, "learning_rate": 0.00020444343657806087, "loss": 1.0859, "step": 6104 }, { "epoch": 2.8729411764705883, "grad_norm": 1.122345870265151, "learning_rate": 0.000204402587433359, "loss": 1.2656, "step": 6105 }, { "epoch": 2.8734117647058826, "grad_norm": 1.9071495715409448, "learning_rate": 0.0002043617336424093, "loss": 0.6562, "step": 6106 }, { "epoch": 2.8738823529411763, "grad_norm": 1.0962880275223716, "learning_rate": 0.00020432087520870086, "loss": 0.5703, "step": 6107 }, { "epoch": 2.8743529411764706, "grad_norm": 0.7616543318085182, "learning_rate": 0.0002042800121357233, "loss": 0.5469, "step": 6108 }, { "epoch": 2.8748235294117648, "grad_norm": 1.4241767087131436, "learning_rate": 0.00020423914442696642, "loss": 1.5469, "step": 6109 }, { "epoch": 2.8752941176470586, "grad_norm": 0.7191425302437056, "learning_rate": 0.00020419827208592058, "loss": 0.3203, "step": 6110 }, { "epoch": 2.875764705882353, "grad_norm": 1.1497281802930608, "learning_rate": 0.00020415739511607653, "loss": 1.0547, "step": 6111 }, { "epoch": 2.876235294117647, "grad_norm": 0.6293328381417588, "learning_rate": 0.0002041165135209252, "loss": 0.3477, "step": 6112 }, { "epoch": 2.8767058823529412, "grad_norm": 0.8996455010396927, "learning_rate": 0.0002040756273039582, "loss": 0.7461, "step": 6113 }, { "epoch": 2.8771764705882354, "grad_norm": 0.9117413927640878, "learning_rate": 0.0002040347364686674, "loss": 0.6914, "step": 6114 }, { "epoch": 2.8776470588235297, "grad_norm": 0.7394493512515616, "learning_rate": 0.00020399384101854506, "loss": 0.3691, "step": 6115 }, { "epoch": 2.8781176470588234, "grad_norm": 0.7709008156298472, "learning_rate": 0.00020395294095708387, "loss": 0.4727, "step": 6116 }, { "epoch": 2.8785882352941177, "grad_norm": 1.3500973781955332, "learning_rate": 0.00020391203628777687, "loss": 1.5156, "step": 6117 }, { "epoch": 2.879058823529412, "grad_norm": 0.9613744976059235, "learning_rate": 0.00020387112701411754, "loss": 0.6758, "step": 6118 }, { "epoch": 2.8795294117647057, "grad_norm": 1.3631416158332503, "learning_rate": 0.00020383021313959972, "loss": 1.5547, "step": 6119 }, { "epoch": 2.88, "grad_norm": 0.6893714273460196, "learning_rate": 0.00020378929466771767, "loss": 0.2852, "step": 6120 }, { "epoch": 2.880470588235294, "grad_norm": 0.9106348516766533, "learning_rate": 0.000203748371601966, "loss": 0.5273, "step": 6121 }, { "epoch": 2.8809411764705883, "grad_norm": 1.3002920579239987, "learning_rate": 0.0002037074439458398, "loss": 1.2734, "step": 6122 }, { "epoch": 2.8814117647058826, "grad_norm": 0.9561211318254909, "learning_rate": 0.0002036665117028344, "loss": 1.4297, "step": 6123 }, { "epoch": 2.8818823529411763, "grad_norm": 0.9829723091626696, "learning_rate": 0.0002036255748764457, "loss": 0.625, "step": 6124 }, { "epoch": 2.8823529411764706, "grad_norm": 0.8980126204206028, "learning_rate": 0.00020358463347016988, "loss": 0.7734, "step": 6125 }, { "epoch": 2.882823529411765, "grad_norm": 0.4696764478315362, "learning_rate": 0.00020354368748750357, "loss": 0.1543, "step": 6126 }, { "epoch": 2.8832941176470586, "grad_norm": 1.1489210680407898, "learning_rate": 0.00020350273693194363, "loss": 1.1406, "step": 6127 }, { "epoch": 2.883764705882353, "grad_norm": 1.1366726014149044, "learning_rate": 0.00020346178180698758, "loss": 1.1875, "step": 6128 }, { "epoch": 2.884235294117647, "grad_norm": 0.8368900035841156, "learning_rate": 0.00020342082211613312, "loss": 0.8828, "step": 6129 }, { "epoch": 2.8847058823529412, "grad_norm": 1.326934569916608, "learning_rate": 0.00020337985786287842, "loss": 2.1562, "step": 6130 }, { "epoch": 2.8851764705882355, "grad_norm": 1.0138722574669767, "learning_rate": 0.00020333888905072208, "loss": 1.0078, "step": 6131 }, { "epoch": 2.8856470588235297, "grad_norm": 0.7029548872659631, "learning_rate": 0.00020329791568316294, "loss": 0.7227, "step": 6132 }, { "epoch": 2.8861176470588235, "grad_norm": 1.8951568136947794, "learning_rate": 0.00020325693776370037, "loss": 1.8828, "step": 6133 }, { "epoch": 2.8865882352941177, "grad_norm": 0.9220019808513075, "learning_rate": 0.0002032159552958341, "loss": 1.0703, "step": 6134 }, { "epoch": 2.887058823529412, "grad_norm": 0.9564385028972974, "learning_rate": 0.00020317496828306427, "loss": 1.7969, "step": 6135 }, { "epoch": 2.8875294117647057, "grad_norm": 1.0938454422557602, "learning_rate": 0.00020313397672889132, "loss": 1.0078, "step": 6136 }, { "epoch": 2.888, "grad_norm": 1.157343055723956, "learning_rate": 0.00020309298063681607, "loss": 0.7656, "step": 6137 }, { "epoch": 2.888470588235294, "grad_norm": 0.9820492013694111, "learning_rate": 0.00020305198001033988, "loss": 1.4453, "step": 6138 }, { "epoch": 2.8889411764705883, "grad_norm": 0.5989836318918592, "learning_rate": 0.00020301097485296443, "loss": 0.5391, "step": 6139 }, { "epoch": 2.8894117647058826, "grad_norm": 1.2178504012133433, "learning_rate": 0.00020296996516819163, "loss": 1.4844, "step": 6140 }, { "epoch": 2.8898823529411763, "grad_norm": 1.2304090945109398, "learning_rate": 0.00020292895095952402, "loss": 0.7852, "step": 6141 }, { "epoch": 2.8903529411764706, "grad_norm": 0.6665681605127298, "learning_rate": 0.0002028879322304644, "loss": 0.2773, "step": 6142 }, { "epoch": 2.890823529411765, "grad_norm": 0.8296636790428902, "learning_rate": 0.00020284690898451593, "loss": 0.9766, "step": 6143 }, { "epoch": 2.8912941176470586, "grad_norm": 0.840436006389754, "learning_rate": 0.00020280588122518223, "loss": 0.5273, "step": 6144 }, { "epoch": 2.891764705882353, "grad_norm": 0.8768685175662975, "learning_rate": 0.00020276484895596726, "loss": 1.0469, "step": 6145 }, { "epoch": 2.892235294117647, "grad_norm": 0.5869427045226486, "learning_rate": 0.00020272381218037542, "loss": 0.1797, "step": 6146 }, { "epoch": 2.8927058823529412, "grad_norm": 0.8817314063633339, "learning_rate": 0.00020268277090191133, "loss": 1.2891, "step": 6147 }, { "epoch": 2.8931764705882355, "grad_norm": 0.6732917539116642, "learning_rate": 0.00020264172512408026, "loss": 0.4473, "step": 6148 }, { "epoch": 2.8936470588235297, "grad_norm": 0.4878636952299275, "learning_rate": 0.00020260067485038767, "loss": 0.2109, "step": 6149 }, { "epoch": 2.8941176470588235, "grad_norm": 1.379255258362574, "learning_rate": 0.0002025596200843394, "loss": 1.0938, "step": 6150 }, { "epoch": 2.8945882352941177, "grad_norm": 0.9701430418037901, "learning_rate": 0.00020251856082944178, "loss": 0.7461, "step": 6151 }, { "epoch": 2.895058823529412, "grad_norm": 1.1090347575416244, "learning_rate": 0.00020247749708920146, "loss": 1.1016, "step": 6152 }, { "epoch": 2.8955294117647057, "grad_norm": 0.9739861049641051, "learning_rate": 0.00020243642886712553, "loss": 1.25, "step": 6153 }, { "epoch": 2.896, "grad_norm": 1.3331137137421178, "learning_rate": 0.0002023953561667213, "loss": 1.3672, "step": 6154 }, { "epoch": 2.896470588235294, "grad_norm": 0.712177528323347, "learning_rate": 0.0002023542789914967, "loss": 0.2949, "step": 6155 }, { "epoch": 2.8969411764705884, "grad_norm": 0.8206685122646065, "learning_rate": 0.0002023131973449599, "loss": 0.8555, "step": 6156 }, { "epoch": 2.8974117647058826, "grad_norm": 0.6055664007812595, "learning_rate": 0.00020227211123061942, "loss": 0.3242, "step": 6157 }, { "epoch": 2.8978823529411764, "grad_norm": 1.0879549259174075, "learning_rate": 0.00020223102065198426, "loss": 1.0078, "step": 6158 }, { "epoch": 2.8983529411764706, "grad_norm": 1.2787462507233036, "learning_rate": 0.00020218992561256375, "loss": 1.2109, "step": 6159 }, { "epoch": 2.898823529411765, "grad_norm": 0.9461271813032948, "learning_rate": 0.0002021488261158676, "loss": 0.8828, "step": 6160 }, { "epoch": 2.8992941176470586, "grad_norm": 0.847233173952044, "learning_rate": 0.0002021077221654059, "loss": 0.582, "step": 6161 }, { "epoch": 2.899764705882353, "grad_norm": 0.9381133580263012, "learning_rate": 0.00020206661376468912, "loss": 1.4453, "step": 6162 }, { "epoch": 2.900235294117647, "grad_norm": 0.5427511990938138, "learning_rate": 0.00020202550091722818, "loss": 0.2441, "step": 6163 }, { "epoch": 2.9007058823529412, "grad_norm": 0.5927147767124615, "learning_rate": 0.00020198438362653426, "loss": 0.5469, "step": 6164 }, { "epoch": 2.9011764705882355, "grad_norm": 1.4329816853566408, "learning_rate": 0.00020194326189611897, "loss": 1.4844, "step": 6165 }, { "epoch": 2.9016470588235292, "grad_norm": 0.9002660212486876, "learning_rate": 0.00020190213572949435, "loss": 1.1016, "step": 6166 }, { "epoch": 2.9021176470588235, "grad_norm": 0.3942501524360411, "learning_rate": 0.0002018610051301728, "loss": 0.0815, "step": 6167 }, { "epoch": 2.9025882352941177, "grad_norm": 1.2240899295048133, "learning_rate": 0.00020181987010166699, "loss": 1.3281, "step": 6168 }, { "epoch": 2.903058823529412, "grad_norm": 1.1798236441503493, "learning_rate": 0.0002017787306474901, "loss": 1.4531, "step": 6169 }, { "epoch": 2.9035294117647057, "grad_norm": 1.1860895008555585, "learning_rate": 0.00020173758677115568, "loss": 0.9766, "step": 6170 }, { "epoch": 2.904, "grad_norm": 0.816241567647383, "learning_rate": 0.00020169643847617753, "loss": 1.1484, "step": 6171 }, { "epoch": 2.904470588235294, "grad_norm": 0.6457975575690112, "learning_rate": 0.00020165528576607, "loss": 0.3066, "step": 6172 }, { "epoch": 2.9049411764705884, "grad_norm": 0.8861372742560427, "learning_rate": 0.0002016141286443477, "loss": 1.0312, "step": 6173 }, { "epoch": 2.9054117647058826, "grad_norm": 0.5444050639591069, "learning_rate": 0.0002015729671145256, "loss": 0.2852, "step": 6174 }, { "epoch": 2.9058823529411764, "grad_norm": 1.082698084778778, "learning_rate": 0.0002015318011801192, "loss": 0.9375, "step": 6175 }, { "epoch": 2.9063529411764706, "grad_norm": 1.0819294073909915, "learning_rate": 0.0002014906308446442, "loss": 1.4609, "step": 6176 }, { "epoch": 2.906823529411765, "grad_norm": 0.8211689742494424, "learning_rate": 0.0002014494561116168, "loss": 1.1172, "step": 6177 }, { "epoch": 2.9072941176470586, "grad_norm": 0.8781685615672066, "learning_rate": 0.00020140827698455338, "loss": 0.6914, "step": 6178 }, { "epoch": 2.907764705882353, "grad_norm": 0.7335862025612797, "learning_rate": 0.00020136709346697107, "loss": 0.4824, "step": 6179 }, { "epoch": 2.908235294117647, "grad_norm": 1.1803564464833904, "learning_rate": 0.00020132590556238696, "loss": 1.3672, "step": 6180 }, { "epoch": 2.9087058823529413, "grad_norm": 1.0164315643500166, "learning_rate": 0.00020128471327431877, "loss": 1.6875, "step": 6181 }, { "epoch": 2.9091764705882355, "grad_norm": 1.1671064652012562, "learning_rate": 0.0002012435166062845, "loss": 1.2344, "step": 6182 }, { "epoch": 2.9096470588235293, "grad_norm": 1.1877074934717684, "learning_rate": 0.00020120231556180262, "loss": 0.9375, "step": 6183 }, { "epoch": 2.9101176470588235, "grad_norm": 0.8406022851311393, "learning_rate": 0.00020116111014439178, "loss": 0.9375, "step": 6184 }, { "epoch": 2.9105882352941177, "grad_norm": 0.8962761789444165, "learning_rate": 0.00020111990035757125, "loss": 1.1484, "step": 6185 }, { "epoch": 2.911058823529412, "grad_norm": 0.6262807267427749, "learning_rate": 0.00020107868620486047, "loss": 0.3242, "step": 6186 }, { "epoch": 2.9115294117647057, "grad_norm": 0.8960302795894068, "learning_rate": 0.0002010374676897793, "loss": 0.8125, "step": 6187 }, { "epoch": 2.912, "grad_norm": 1.1928804321770727, "learning_rate": 0.00020099624481584807, "loss": 1.7031, "step": 6188 }, { "epoch": 2.912470588235294, "grad_norm": 1.385362948347221, "learning_rate": 0.0002009550175865874, "loss": 1.3672, "step": 6189 }, { "epoch": 2.9129411764705884, "grad_norm": 0.3421356999561352, "learning_rate": 0.00020091378600551834, "loss": 0.1914, "step": 6190 }, { "epoch": 2.9134117647058826, "grad_norm": 0.9774208625066715, "learning_rate": 0.00020087255007616222, "loss": 0.7773, "step": 6191 }, { "epoch": 2.9138823529411764, "grad_norm": 0.9439509577250635, "learning_rate": 0.00020083130980204074, "loss": 0.7969, "step": 6192 }, { "epoch": 2.9143529411764706, "grad_norm": 1.2537459347795572, "learning_rate": 0.0002007900651866761, "loss": 1.3203, "step": 6193 }, { "epoch": 2.914823529411765, "grad_norm": 0.9218427834939814, "learning_rate": 0.00020074881623359083, "loss": 0.5703, "step": 6194 }, { "epoch": 2.9152941176470586, "grad_norm": 1.0145493189420416, "learning_rate": 0.00020070756294630765, "loss": 0.9453, "step": 6195 }, { "epoch": 2.915764705882353, "grad_norm": 1.2656758328537396, "learning_rate": 0.0002006663053283499, "loss": 1.2344, "step": 6196 }, { "epoch": 2.916235294117647, "grad_norm": 0.8930957126816321, "learning_rate": 0.00020062504338324116, "loss": 0.6992, "step": 6197 }, { "epoch": 2.9167058823529413, "grad_norm": 0.44287134889530216, "learning_rate": 0.0002005837771145054, "loss": 0.2041, "step": 6198 }, { "epoch": 2.9171764705882355, "grad_norm": 0.6079509139364586, "learning_rate": 0.00020054250652566696, "loss": 0.9375, "step": 6199 }, { "epoch": 2.9176470588235293, "grad_norm": 1.71189593140269, "learning_rate": 0.0002005012316202506, "loss": 1.3047, "step": 6200 }, { "epoch": 2.9181176470588235, "grad_norm": 0.8518779409476236, "learning_rate": 0.00020045995240178133, "loss": 0.5586, "step": 6201 }, { "epoch": 2.9185882352941177, "grad_norm": 0.6989258176559475, "learning_rate": 0.00020041866887378457, "loss": 0.3672, "step": 6202 }, { "epoch": 2.919058823529412, "grad_norm": 0.970322929286691, "learning_rate": 0.0002003773810397862, "loss": 1.1484, "step": 6203 }, { "epoch": 2.9195294117647057, "grad_norm": 1.0923833068603834, "learning_rate": 0.00020033608890331246, "loss": 1.1016, "step": 6204 }, { "epoch": 2.92, "grad_norm": 0.8338926182049914, "learning_rate": 0.00020029479246788973, "loss": 0.4062, "step": 6205 }, { "epoch": 2.920470588235294, "grad_norm": 0.7920034466679259, "learning_rate": 0.00020025349173704508, "loss": 0.6211, "step": 6206 }, { "epoch": 2.9209411764705884, "grad_norm": 1.114210211630089, "learning_rate": 0.00020021218671430573, "loss": 1.2734, "step": 6207 }, { "epoch": 2.9214117647058826, "grad_norm": 0.999647141867926, "learning_rate": 0.00020017087740319936, "loss": 1.1406, "step": 6208 }, { "epoch": 2.9218823529411764, "grad_norm": 0.8757484706469177, "learning_rate": 0.00020012956380725393, "loss": 0.8203, "step": 6209 }, { "epoch": 2.9223529411764706, "grad_norm": 1.1312004184255426, "learning_rate": 0.00020008824592999795, "loss": 1.2656, "step": 6210 }, { "epoch": 2.922823529411765, "grad_norm": 0.5226322314108683, "learning_rate": 0.00020004692377496004, "loss": 0.2598, "step": 6211 }, { "epoch": 2.9232941176470586, "grad_norm": 0.6933625175212432, "learning_rate": 0.00020000559734566933, "loss": 0.4629, "step": 6212 }, { "epoch": 2.923764705882353, "grad_norm": 0.6370607906444563, "learning_rate": 0.00019996426664565538, "loss": 0.2676, "step": 6213 }, { "epoch": 2.924235294117647, "grad_norm": 0.807932368673752, "learning_rate": 0.000199922931678448, "loss": 0.8555, "step": 6214 }, { "epoch": 2.9247058823529413, "grad_norm": 0.745792139139325, "learning_rate": 0.00019988159244757736, "loss": 0.5195, "step": 6215 }, { "epoch": 2.9251764705882355, "grad_norm": 0.4524013717362496, "learning_rate": 0.000199840248956574, "loss": 0.3105, "step": 6216 }, { "epoch": 2.9256470588235293, "grad_norm": 0.7119844435777909, "learning_rate": 0.000199798901208969, "loss": 0.4492, "step": 6217 }, { "epoch": 2.9261176470588235, "grad_norm": 1.697286196071266, "learning_rate": 0.0001997575492082935, "loss": 0.9961, "step": 6218 }, { "epoch": 2.9265882352941177, "grad_norm": 0.6062994500011583, "learning_rate": 0.0001997161929580793, "loss": 0.3711, "step": 6219 }, { "epoch": 2.9270588235294115, "grad_norm": 0.7883306159701253, "learning_rate": 0.00019967483246185834, "loss": 0.5898, "step": 6220 }, { "epoch": 2.9275294117647057, "grad_norm": 0.7565743128392055, "learning_rate": 0.00019963346772316307, "loss": 0.8477, "step": 6221 }, { "epoch": 2.928, "grad_norm": 0.5355256989264482, "learning_rate": 0.00019959209874552612, "loss": 0.293, "step": 6222 }, { "epoch": 2.928470588235294, "grad_norm": 0.7585713943165939, "learning_rate": 0.00019955072553248075, "loss": 0.457, "step": 6223 }, { "epoch": 2.9289411764705884, "grad_norm": 0.7162304146212974, "learning_rate": 0.0001995093480875604, "loss": 1.2734, "step": 6224 }, { "epoch": 2.9294117647058826, "grad_norm": 0.646830289863709, "learning_rate": 0.00019946796641429884, "loss": 0.7969, "step": 6225 }, { "epoch": 2.9298823529411764, "grad_norm": 0.8431851639906992, "learning_rate": 0.0001994265805162303, "loss": 0.9727, "step": 6226 }, { "epoch": 2.9303529411764706, "grad_norm": 0.8644805941579494, "learning_rate": 0.00019938519039688937, "loss": 0.6836, "step": 6227 }, { "epoch": 2.930823529411765, "grad_norm": 0.8288945476796481, "learning_rate": 0.00019934379605981097, "loss": 0.9648, "step": 6228 }, { "epoch": 2.9312941176470586, "grad_norm": 0.6485356218945831, "learning_rate": 0.0001993023975085303, "loss": 0.5586, "step": 6229 }, { "epoch": 2.931764705882353, "grad_norm": 1.062158167550415, "learning_rate": 0.00019926099474658305, "loss": 0.9023, "step": 6230 }, { "epoch": 2.932235294117647, "grad_norm": 0.9074371390871346, "learning_rate": 0.0001992195877775053, "loss": 1.0312, "step": 6231 }, { "epoch": 2.9327058823529413, "grad_norm": 0.5139090238784386, "learning_rate": 0.00019917817660483327, "loss": 0.3086, "step": 6232 }, { "epoch": 2.9331764705882355, "grad_norm": 0.8874580989203726, "learning_rate": 0.00019913676123210373, "loss": 0.6172, "step": 6233 }, { "epoch": 2.9336470588235293, "grad_norm": 1.0856572754380125, "learning_rate": 0.00019909534166285382, "loss": 1.0312, "step": 6234 }, { "epoch": 2.9341176470588235, "grad_norm": 0.8116526773110564, "learning_rate": 0.00019905391790062092, "loss": 0.8125, "step": 6235 }, { "epoch": 2.9345882352941177, "grad_norm": 1.0189958048822962, "learning_rate": 0.00019901248994894276, "loss": 1.1953, "step": 6236 }, { "epoch": 2.9350588235294115, "grad_norm": 1.3558378776930122, "learning_rate": 0.0001989710578113576, "loss": 1.3203, "step": 6237 }, { "epoch": 2.9355294117647057, "grad_norm": 0.940197257870853, "learning_rate": 0.00019892962149140392, "loss": 0.7383, "step": 6238 }, { "epoch": 2.936, "grad_norm": 0.7392412649288211, "learning_rate": 0.0001988881809926205, "loss": 0.291, "step": 6239 }, { "epoch": 2.936470588235294, "grad_norm": 0.9495855603530524, "learning_rate": 0.00019884673631854672, "loss": 1.6875, "step": 6240 }, { "epoch": 2.9369411764705884, "grad_norm": 0.5905283259325864, "learning_rate": 0.000198805287472722, "loss": 0.2334, "step": 6241 }, { "epoch": 2.9374117647058826, "grad_norm": 0.48883358985235026, "learning_rate": 0.0001987638344586864, "loss": 0.1582, "step": 6242 }, { "epoch": 2.9378823529411764, "grad_norm": 0.8291747322526214, "learning_rate": 0.00019872237727998004, "loss": 0.543, "step": 6243 }, { "epoch": 2.9383529411764706, "grad_norm": 0.9223327855780465, "learning_rate": 0.0001986809159401438, "loss": 0.832, "step": 6244 }, { "epoch": 2.938823529411765, "grad_norm": 0.72839766825176, "learning_rate": 0.00019863945044271846, "loss": 0.4023, "step": 6245 }, { "epoch": 2.9392941176470586, "grad_norm": 1.540626864847864, "learning_rate": 0.00019859798079124554, "loss": 0.6797, "step": 6246 }, { "epoch": 2.939764705882353, "grad_norm": 1.6801850031376118, "learning_rate": 0.00019855650698926666, "loss": 2.0938, "step": 6247 }, { "epoch": 2.940235294117647, "grad_norm": 0.8474508434251669, "learning_rate": 0.00019851502904032396, "loss": 0.5898, "step": 6248 }, { "epoch": 2.9407058823529413, "grad_norm": 0.7747022316873518, "learning_rate": 0.00019847354694795978, "loss": 0.4023, "step": 6249 }, { "epoch": 2.9411764705882355, "grad_norm": 0.9074518353937916, "learning_rate": 0.00019843206071571692, "loss": 0.8945, "step": 6250 }, { "epoch": 2.9416470588235293, "grad_norm": 0.5600044003127369, "learning_rate": 0.00019839057034713858, "loss": 0.2852, "step": 6251 }, { "epoch": 2.9421176470588235, "grad_norm": 0.702087165694097, "learning_rate": 0.00019834907584576812, "loss": 0.3887, "step": 6252 }, { "epoch": 2.9425882352941177, "grad_norm": 0.9911363207511541, "learning_rate": 0.00019830757721514943, "loss": 0.7266, "step": 6253 }, { "epoch": 2.9430588235294115, "grad_norm": 0.9624930701914157, "learning_rate": 0.00019826607445882667, "loss": 0.7383, "step": 6254 }, { "epoch": 2.9435294117647057, "grad_norm": 1.0214482622699055, "learning_rate": 0.00019822456758034445, "loss": 0.7617, "step": 6255 }, { "epoch": 2.944, "grad_norm": 0.9411773868378706, "learning_rate": 0.0001981830565832476, "loss": 0.957, "step": 6256 }, { "epoch": 2.944470588235294, "grad_norm": 1.262487894754331, "learning_rate": 0.0001981415414710814, "loss": 1.8594, "step": 6257 }, { "epoch": 2.9449411764705884, "grad_norm": 0.9338848463778947, "learning_rate": 0.00019810002224739142, "loss": 1.0391, "step": 6258 }, { "epoch": 2.9454117647058826, "grad_norm": 0.7433538339074064, "learning_rate": 0.00019805849891572363, "loss": 0.5234, "step": 6259 }, { "epoch": 2.9458823529411764, "grad_norm": 0.9287219288001145, "learning_rate": 0.00019801697147962428, "loss": 1.1328, "step": 6260 }, { "epoch": 2.9463529411764706, "grad_norm": 0.6933262604530085, "learning_rate": 0.00019797543994264004, "loss": 0.2988, "step": 6261 }, { "epoch": 2.946823529411765, "grad_norm": 0.8055612615362706, "learning_rate": 0.00019793390430831794, "loss": 1.1562, "step": 6262 }, { "epoch": 2.9472941176470586, "grad_norm": 0.7417549440262968, "learning_rate": 0.00019789236458020527, "loss": 0.3496, "step": 6263 }, { "epoch": 2.947764705882353, "grad_norm": 0.5191994504498557, "learning_rate": 0.00019785082076184977, "loss": 0.2402, "step": 6264 }, { "epoch": 2.948235294117647, "grad_norm": 0.6191583124251568, "learning_rate": 0.00019780927285679945, "loss": 0.457, "step": 6265 }, { "epoch": 2.9487058823529413, "grad_norm": 1.3117053693738296, "learning_rate": 0.00019776772086860277, "loss": 1.9766, "step": 6266 }, { "epoch": 2.9491764705882355, "grad_norm": 0.9115010171457427, "learning_rate": 0.0001977261648008084, "loss": 0.7266, "step": 6267 }, { "epoch": 2.9496470588235293, "grad_norm": 1.1294019407991167, "learning_rate": 0.00019768460465696545, "loss": 0.9805, "step": 6268 }, { "epoch": 2.9501176470588235, "grad_norm": 1.0103144721104267, "learning_rate": 0.0001976430404406234, "loss": 1.5547, "step": 6269 }, { "epoch": 2.9505882352941177, "grad_norm": 1.0648401162048784, "learning_rate": 0.00019760147215533197, "loss": 0.875, "step": 6270 }, { "epoch": 2.9510588235294115, "grad_norm": 0.6113698656196529, "learning_rate": 0.00019755989980464137, "loss": 0.1934, "step": 6271 }, { "epoch": 2.9515294117647057, "grad_norm": 0.3828674337027282, "learning_rate": 0.00019751832339210204, "loss": 0.1318, "step": 6272 }, { "epoch": 2.952, "grad_norm": 1.3914823267745968, "learning_rate": 0.00019747674292126477, "loss": 0.9648, "step": 6273 }, { "epoch": 2.952470588235294, "grad_norm": 0.5221394271909039, "learning_rate": 0.00019743515839568083, "loss": 0.2412, "step": 6274 }, { "epoch": 2.9529411764705884, "grad_norm": 0.282082147446472, "learning_rate": 0.00019739356981890167, "loss": 0.0913, "step": 6275 }, { "epoch": 2.9534117647058826, "grad_norm": 0.7022054269922826, "learning_rate": 0.00019735197719447916, "loss": 0.4102, "step": 6276 }, { "epoch": 2.9538823529411764, "grad_norm": 1.6051964601938338, "learning_rate": 0.00019731038052596558, "loss": 1.875, "step": 6277 }, { "epoch": 2.9543529411764706, "grad_norm": 1.954909796583651, "learning_rate": 0.00019726877981691347, "loss": 1.2891, "step": 6278 }, { "epoch": 2.954823529411765, "grad_norm": 0.8249387328027037, "learning_rate": 0.0001972271750708757, "loss": 0.75, "step": 6279 }, { "epoch": 2.9552941176470586, "grad_norm": 0.43451920895681295, "learning_rate": 0.0001971855662914055, "loss": 0.1299, "step": 6280 }, { "epoch": 2.955764705882353, "grad_norm": 2.6929841187075327, "learning_rate": 0.00019714395348205649, "loss": 2.5625, "step": 6281 }, { "epoch": 2.956235294117647, "grad_norm": 0.9940543275582331, "learning_rate": 0.00019710233664638264, "loss": 0.9258, "step": 6282 }, { "epoch": 2.9567058823529413, "grad_norm": 0.8867157649199796, "learning_rate": 0.00019706071578793819, "loss": 0.8711, "step": 6283 }, { "epoch": 2.9571764705882355, "grad_norm": 0.7648051344959753, "learning_rate": 0.00019701909091027778, "loss": 0.4824, "step": 6284 }, { "epoch": 2.9576470588235293, "grad_norm": 0.8549218199971144, "learning_rate": 0.0001969774620169564, "loss": 0.8477, "step": 6285 }, { "epoch": 2.9581176470588235, "grad_norm": 1.407762288819515, "learning_rate": 0.00019693582911152935, "loss": 1.2109, "step": 6286 }, { "epoch": 2.9585882352941177, "grad_norm": 0.7547536938115537, "learning_rate": 0.0001968941921975522, "loss": 0.7031, "step": 6287 }, { "epoch": 2.9590588235294115, "grad_norm": 0.8792021788640465, "learning_rate": 0.00019685255127858107, "loss": 0.6719, "step": 6288 }, { "epoch": 2.9595294117647057, "grad_norm": 0.6994314221248136, "learning_rate": 0.00019681090635817228, "loss": 0.4062, "step": 6289 }, { "epoch": 2.96, "grad_norm": 1.4666038247790245, "learning_rate": 0.00019676925743988244, "loss": 1.2578, "step": 6290 }, { "epoch": 2.960470588235294, "grad_norm": 0.7316145080497605, "learning_rate": 0.00019672760452726863, "loss": 0.6523, "step": 6291 }, { "epoch": 2.9609411764705884, "grad_norm": 0.9297658733173996, "learning_rate": 0.0001966859476238882, "loss": 0.8633, "step": 6292 }, { "epoch": 2.961411764705882, "grad_norm": 0.8746507454925544, "learning_rate": 0.00019664428673329883, "loss": 1.5625, "step": 6293 }, { "epoch": 2.9618823529411764, "grad_norm": 0.6029160435008966, "learning_rate": 0.00019660262185905856, "loss": 0.3086, "step": 6294 }, { "epoch": 2.9623529411764706, "grad_norm": 0.8403549729536742, "learning_rate": 0.0001965609530047259, "loss": 0.6406, "step": 6295 }, { "epoch": 2.962823529411765, "grad_norm": 0.7997911220427305, "learning_rate": 0.0001965192801738594, "loss": 0.6875, "step": 6296 }, { "epoch": 2.9632941176470586, "grad_norm": 0.5673796636840629, "learning_rate": 0.0001964776033700182, "loss": 0.2578, "step": 6297 }, { "epoch": 2.963764705882353, "grad_norm": 1.1952818778152172, "learning_rate": 0.00019643592259676164, "loss": 1.2578, "step": 6298 }, { "epoch": 2.964235294117647, "grad_norm": 1.0326967505031595, "learning_rate": 0.0001963942378576496, "loss": 0.7656, "step": 6299 }, { "epoch": 2.9647058823529413, "grad_norm": 1.1278652798142792, "learning_rate": 0.0001963525491562421, "loss": 1.2656, "step": 6300 }, { "epoch": 2.9651764705882355, "grad_norm": 0.7262353080350001, "learning_rate": 0.0001963108564960995, "loss": 0.3281, "step": 6301 }, { "epoch": 2.9656470588235293, "grad_norm": 0.6283408216309463, "learning_rate": 0.00019626915988078263, "loss": 0.4375, "step": 6302 }, { "epoch": 2.9661176470588235, "grad_norm": 0.658520737670331, "learning_rate": 0.00019622745931385254, "loss": 0.5547, "step": 6303 }, { "epoch": 2.9665882352941177, "grad_norm": 0.7958232548289187, "learning_rate": 0.00019618575479887066, "loss": 0.8516, "step": 6304 }, { "epoch": 2.9670588235294115, "grad_norm": 0.553211546291057, "learning_rate": 0.00019614404633939886, "loss": 0.5195, "step": 6305 }, { "epoch": 2.9675294117647057, "grad_norm": 0.9860508110102151, "learning_rate": 0.0001961023339389991, "loss": 1.1406, "step": 6306 }, { "epoch": 2.968, "grad_norm": 0.7212971975877194, "learning_rate": 0.00019606061760123392, "loss": 0.8086, "step": 6307 }, { "epoch": 2.968470588235294, "grad_norm": 0.8185830267201923, "learning_rate": 0.00019601889732966604, "loss": 0.4805, "step": 6308 }, { "epoch": 2.9689411764705884, "grad_norm": 0.7563122514153954, "learning_rate": 0.0001959771731278586, "loss": 0.8906, "step": 6309 }, { "epoch": 2.969411764705882, "grad_norm": 0.9775646120195847, "learning_rate": 0.0001959354449993751, "loss": 1.3672, "step": 6310 }, { "epoch": 2.9698823529411764, "grad_norm": 0.9478616356824464, "learning_rate": 0.00019589371294777925, "loss": 1.7734, "step": 6311 }, { "epoch": 2.9703529411764706, "grad_norm": 0.7451335905786758, "learning_rate": 0.00019585197697663526, "loss": 0.8516, "step": 6312 }, { "epoch": 2.970823529411765, "grad_norm": 0.743345206438643, "learning_rate": 0.0001958102370895075, "loss": 0.4355, "step": 6313 }, { "epoch": 2.9712941176470586, "grad_norm": 0.7231623192980157, "learning_rate": 0.00019576849328996075, "loss": 0.5117, "step": 6314 }, { "epoch": 2.971764705882353, "grad_norm": 1.0444370503734144, "learning_rate": 0.00019572674558156024, "loss": 0.8047, "step": 6315 }, { "epoch": 2.972235294117647, "grad_norm": 0.8221211611551998, "learning_rate": 0.00019568499396787134, "loss": 0.8477, "step": 6316 }, { "epoch": 2.9727058823529413, "grad_norm": 0.6589449723446545, "learning_rate": 0.0001956432384524599, "loss": 0.2773, "step": 6317 }, { "epoch": 2.9731764705882355, "grad_norm": 0.7094103323469158, "learning_rate": 0.000195601479038892, "loss": 0.4336, "step": 6318 }, { "epoch": 2.9736470588235293, "grad_norm": 0.25983378674136126, "learning_rate": 0.0001955597157307341, "loss": 0.0142, "step": 6319 }, { "epoch": 2.9741176470588235, "grad_norm": 0.7287959658041551, "learning_rate": 0.000195517948531553, "loss": 0.4746, "step": 6320 }, { "epoch": 2.9745882352941178, "grad_norm": 0.7723331575909074, "learning_rate": 0.00019547617744491587, "loss": 0.7734, "step": 6321 }, { "epoch": 2.9750588235294115, "grad_norm": 0.9992897294541093, "learning_rate": 0.00019543440247439008, "loss": 0.875, "step": 6322 }, { "epoch": 2.9755294117647058, "grad_norm": 0.9163380193028441, "learning_rate": 0.00019539262362354353, "loss": 1.1641, "step": 6323 }, { "epoch": 2.976, "grad_norm": 0.5115185484094166, "learning_rate": 0.00019535084089594422, "loss": 0.168, "step": 6324 }, { "epoch": 2.976470588235294, "grad_norm": 1.0890396429114109, "learning_rate": 0.00019530905429516066, "loss": 1.3984, "step": 6325 }, { "epoch": 2.9769411764705884, "grad_norm": 0.8641657696388874, "learning_rate": 0.00019526726382476165, "loss": 0.8438, "step": 6326 }, { "epoch": 2.977411764705882, "grad_norm": 0.7265378270365085, "learning_rate": 0.00019522546948831627, "loss": 0.8047, "step": 6327 }, { "epoch": 2.9778823529411764, "grad_norm": 0.600267161143349, "learning_rate": 0.00019518367128939392, "loss": 0.2402, "step": 6328 }, { "epoch": 2.9783529411764706, "grad_norm": 0.9244869370238451, "learning_rate": 0.0001951418692315645, "loss": 0.9062, "step": 6329 }, { "epoch": 2.978823529411765, "grad_norm": 1.3243574854180207, "learning_rate": 0.00019510006331839794, "loss": 1.9141, "step": 6330 }, { "epoch": 2.9792941176470586, "grad_norm": 1.020609834171367, "learning_rate": 0.00019505825355346483, "loss": 0.957, "step": 6331 }, { "epoch": 2.979764705882353, "grad_norm": 0.9110233310035762, "learning_rate": 0.0001950164399403358, "loss": 0.582, "step": 6332 }, { "epoch": 2.980235294117647, "grad_norm": 0.7618431538438044, "learning_rate": 0.00019497462248258202, "loss": 1.1875, "step": 6333 }, { "epoch": 2.9807058823529413, "grad_norm": 0.7654632043717944, "learning_rate": 0.0001949328011837749, "loss": 0.4609, "step": 6334 }, { "epoch": 2.9811764705882355, "grad_norm": 0.8212896308967009, "learning_rate": 0.00019489097604748613, "loss": 0.4512, "step": 6335 }, { "epoch": 2.9816470588235293, "grad_norm": 0.48130466658986304, "learning_rate": 0.00019484914707728785, "loss": 0.0815, "step": 6336 }, { "epoch": 2.9821176470588235, "grad_norm": 0.9250452739333986, "learning_rate": 0.00019480731427675238, "loss": 0.4863, "step": 6337 }, { "epoch": 2.9825882352941178, "grad_norm": 0.743517632056366, "learning_rate": 0.0001947654776494525, "loss": 0.668, "step": 6338 }, { "epoch": 2.9830588235294115, "grad_norm": 1.063550858480488, "learning_rate": 0.0001947236371989613, "loss": 0.7969, "step": 6339 }, { "epoch": 2.9835294117647058, "grad_norm": 0.8265267270426937, "learning_rate": 0.0001946817929288521, "loss": 1.0312, "step": 6340 }, { "epoch": 2.984, "grad_norm": 0.9444435334798749, "learning_rate": 0.0001946399448426986, "loss": 1.0391, "step": 6341 }, { "epoch": 2.984470588235294, "grad_norm": 0.9356461188074662, "learning_rate": 0.00019459809294407486, "loss": 1.0703, "step": 6342 }, { "epoch": 2.9849411764705884, "grad_norm": 0.9302873757013288, "learning_rate": 0.00019455623723655521, "loss": 0.8594, "step": 6343 }, { "epoch": 2.985411764705882, "grad_norm": 1.5279173273558506, "learning_rate": 0.00019451437772371438, "loss": 1.3516, "step": 6344 }, { "epoch": 2.9858823529411764, "grad_norm": 0.8862054338444273, "learning_rate": 0.00019447251440912736, "loss": 0.9922, "step": 6345 }, { "epoch": 2.9863529411764707, "grad_norm": 1.0772723283190815, "learning_rate": 0.00019443064729636947, "loss": 1.2891, "step": 6346 }, { "epoch": 2.9868235294117644, "grad_norm": 0.7956454700307982, "learning_rate": 0.00019438877638901635, "loss": 0.7461, "step": 6347 }, { "epoch": 2.9872941176470587, "grad_norm": 0.669586515655322, "learning_rate": 0.00019434690169064402, "loss": 0.293, "step": 6348 }, { "epoch": 2.987764705882353, "grad_norm": 1.0017092942395596, "learning_rate": 0.00019430502320482871, "loss": 0.5977, "step": 6349 }, { "epoch": 2.988235294117647, "grad_norm": 1.2223074158288423, "learning_rate": 0.00019426314093514717, "loss": 1.8359, "step": 6350 }, { "epoch": 2.9887058823529413, "grad_norm": 0.8935782018017323, "learning_rate": 0.0001942212548851763, "loss": 0.4766, "step": 6351 }, { "epoch": 2.9891764705882355, "grad_norm": 0.6910420251230842, "learning_rate": 0.00019417936505849328, "loss": 0.9414, "step": 6352 }, { "epoch": 2.9896470588235293, "grad_norm": 1.2184985582602363, "learning_rate": 0.00019413747145867583, "loss": 1.2344, "step": 6353 }, { "epoch": 2.9901176470588235, "grad_norm": 0.8576133416927126, "learning_rate": 0.00019409557408930184, "loss": 0.6211, "step": 6354 }, { "epoch": 2.9905882352941178, "grad_norm": 0.7951848335611387, "learning_rate": 0.0001940536729539495, "loss": 0.4316, "step": 6355 }, { "epoch": 2.9910588235294115, "grad_norm": 0.9617676451795949, "learning_rate": 0.00019401176805619746, "loss": 1.1172, "step": 6356 }, { "epoch": 2.9915294117647058, "grad_norm": 0.9646878068832389, "learning_rate": 0.0001939698593996245, "loss": 1.0625, "step": 6357 }, { "epoch": 2.992, "grad_norm": 0.9868002931191411, "learning_rate": 0.00019392794698780988, "loss": 1.4922, "step": 6358 }, { "epoch": 2.992470588235294, "grad_norm": 0.2479804957347555, "learning_rate": 0.00019388603082433315, "loss": 0.0732, "step": 6359 }, { "epoch": 2.9929411764705884, "grad_norm": 0.6896415401016672, "learning_rate": 0.00019384411091277412, "loss": 0.6367, "step": 6360 }, { "epoch": 2.993411764705882, "grad_norm": 0.3637193003830581, "learning_rate": 0.00019380218725671298, "loss": 0.103, "step": 6361 }, { "epoch": 2.9938823529411764, "grad_norm": 1.0343558875416996, "learning_rate": 0.00019376025985973013, "loss": 1.1953, "step": 6362 }, { "epoch": 2.9943529411764707, "grad_norm": 0.8516223871254014, "learning_rate": 0.00019371832872540652, "loss": 0.9219, "step": 6363 }, { "epoch": 2.9948235294117644, "grad_norm": 0.7254646841039442, "learning_rate": 0.00019367639385732316, "loss": 0.3047, "step": 6364 }, { "epoch": 2.9952941176470587, "grad_norm": 1.1844581221513593, "learning_rate": 0.00019363445525906152, "loss": 1.1484, "step": 6365 }, { "epoch": 2.995764705882353, "grad_norm": 1.2713838421912922, "learning_rate": 0.00019359251293420336, "loss": 1.5, "step": 6366 }, { "epoch": 2.996235294117647, "grad_norm": 1.538161398677476, "learning_rate": 0.00019355056688633086, "loss": 0.9961, "step": 6367 }, { "epoch": 2.9967058823529413, "grad_norm": 0.5107063718476736, "learning_rate": 0.0001935086171190263, "loss": 0.4043, "step": 6368 }, { "epoch": 2.9971764705882356, "grad_norm": 1.5305372115582878, "learning_rate": 0.00019346666363587238, "loss": 1.3438, "step": 6369 }, { "epoch": 2.9976470588235293, "grad_norm": 1.2093039238131558, "learning_rate": 0.00019342470644045223, "loss": 1.2734, "step": 6370 }, { "epoch": 2.9981176470588236, "grad_norm": 0.8118761061623283, "learning_rate": 0.00019338274553634913, "loss": 1.0547, "step": 6371 }, { "epoch": 2.9985882352941178, "grad_norm": 0.8367391677162748, "learning_rate": 0.00019334078092714683, "loss": 0.6445, "step": 6372 }, { "epoch": 2.9990588235294116, "grad_norm": 0.7371268209703118, "learning_rate": 0.00019329881261642918, "loss": 0.8125, "step": 6373 }, { "epoch": 2.9995294117647058, "grad_norm": 1.02956376165449, "learning_rate": 0.00019325684060778056, "loss": 1.1641, "step": 6374 }, { "epoch": 3.0, "grad_norm": 1.1916545214818965, "learning_rate": 0.00019321486490478563, "loss": 1.1094, "step": 6375 }, { "epoch": 3.0, "eval_loss": 0.8723981380462646, "eval_runtime": 36.6599, "eval_samples_per_second": 20.458, "eval_steps_per_second": 10.229, "step": 6375 }, { "epoch": 3.0004705882352942, "grad_norm": 0.866820827229502, "learning_rate": 0.0001931728855110292, "loss": 0.9492, "step": 6376 }, { "epoch": 3.0009411764705884, "grad_norm": 0.7365311482158337, "learning_rate": 0.00019313090243009665, "loss": 0.9414, "step": 6377 }, { "epoch": 3.0014117647058822, "grad_norm": 0.619659916038612, "learning_rate": 0.00019308891566557345, "loss": 0.1914, "step": 6378 }, { "epoch": 3.0018823529411764, "grad_norm": 0.8741398504132446, "learning_rate": 0.0001930469252210455, "loss": 0.6328, "step": 6379 }, { "epoch": 3.0023529411764707, "grad_norm": 0.5130135825229437, "learning_rate": 0.000193004931100099, "loss": 0.1807, "step": 6380 }, { "epoch": 3.002823529411765, "grad_norm": 0.9509736071447067, "learning_rate": 0.00019296293330632047, "loss": 1.3203, "step": 6381 }, { "epoch": 3.0032941176470587, "grad_norm": 0.4546039212879703, "learning_rate": 0.0001929209318432967, "loss": 0.1602, "step": 6382 }, { "epoch": 3.003764705882353, "grad_norm": 0.9614492591157758, "learning_rate": 0.00019287892671461475, "loss": 1.6016, "step": 6383 }, { "epoch": 3.004235294117647, "grad_norm": 0.6007613315805156, "learning_rate": 0.0001928369179238622, "loss": 0.2197, "step": 6384 }, { "epoch": 3.0047058823529413, "grad_norm": 0.848665476043217, "learning_rate": 0.00019279490547462674, "loss": 0.7812, "step": 6385 }, { "epoch": 3.005176470588235, "grad_norm": 1.1814508148581615, "learning_rate": 0.0001927528893704964, "loss": 1.0391, "step": 6386 }, { "epoch": 3.0056470588235293, "grad_norm": 1.2878720604596319, "learning_rate": 0.00019271086961505964, "loss": 0.7383, "step": 6387 }, { "epoch": 3.0061176470588236, "grad_norm": 0.8070820826178312, "learning_rate": 0.0001926688462119051, "loss": 0.6094, "step": 6388 }, { "epoch": 3.006588235294118, "grad_norm": 0.5037919035360751, "learning_rate": 0.00019262681916462187, "loss": 0.1826, "step": 6389 }, { "epoch": 3.0070588235294116, "grad_norm": 0.9427307909298784, "learning_rate": 0.00019258478847679914, "loss": 1.0156, "step": 6390 }, { "epoch": 3.007529411764706, "grad_norm": 0.8855009892379687, "learning_rate": 0.0001925427541520266, "loss": 1.0156, "step": 6391 }, { "epoch": 3.008, "grad_norm": 0.9107009319726408, "learning_rate": 0.00019250071619389422, "loss": 0.4551, "step": 6392 }, { "epoch": 3.0084705882352942, "grad_norm": 0.8595440752693208, "learning_rate": 0.00019245867460599217, "loss": 0.8125, "step": 6393 }, { "epoch": 3.0089411764705885, "grad_norm": 0.648665555881263, "learning_rate": 0.00019241662939191106, "loss": 0.291, "step": 6394 }, { "epoch": 3.0094117647058822, "grad_norm": 0.7768418474964063, "learning_rate": 0.00019237458055524178, "loss": 0.6172, "step": 6395 }, { "epoch": 3.0098823529411765, "grad_norm": 0.8219607412251744, "learning_rate": 0.00019233252809957546, "loss": 0.7891, "step": 6396 }, { "epoch": 3.0103529411764707, "grad_norm": 0.8167598027059696, "learning_rate": 0.00019229047202850357, "loss": 1.2188, "step": 6397 }, { "epoch": 3.010823529411765, "grad_norm": 1.5006091065944458, "learning_rate": 0.00019224841234561797, "loss": 1.4219, "step": 6398 }, { "epoch": 3.0112941176470587, "grad_norm": 1.0386653946310358, "learning_rate": 0.00019220634905451074, "loss": 0.9609, "step": 6399 }, { "epoch": 3.011764705882353, "grad_norm": 0.5926138863062438, "learning_rate": 0.00019216428215877425, "loss": 0.3008, "step": 6400 }, { "epoch": 3.012235294117647, "grad_norm": 0.8319119453505535, "learning_rate": 0.00019212221166200124, "loss": 0.4395, "step": 6401 }, { "epoch": 3.0127058823529413, "grad_norm": 0.8924732819212803, "learning_rate": 0.00019208013756778476, "loss": 0.6836, "step": 6402 }, { "epoch": 3.013176470588235, "grad_norm": 0.9139830301318583, "learning_rate": 0.0001920380598797181, "loss": 0.9141, "step": 6403 }, { "epoch": 3.0136470588235293, "grad_norm": 1.0942600044775346, "learning_rate": 0.000191995978601395, "loss": 1.75, "step": 6404 }, { "epoch": 3.0141176470588236, "grad_norm": 1.2533496206246655, "learning_rate": 0.0001919538937364093, "loss": 0.8867, "step": 6405 }, { "epoch": 3.014588235294118, "grad_norm": 1.242266772379572, "learning_rate": 0.00019191180528835527, "loss": 2.0312, "step": 6406 }, { "epoch": 3.0150588235294116, "grad_norm": 1.1301460787426056, "learning_rate": 0.00019186971326082752, "loss": 1.1016, "step": 6407 }, { "epoch": 3.015529411764706, "grad_norm": 0.617829846796623, "learning_rate": 0.0001918276176574209, "loss": 0.2373, "step": 6408 }, { "epoch": 3.016, "grad_norm": 0.7909594818506126, "learning_rate": 0.00019178551848173054, "loss": 0.4688, "step": 6409 }, { "epoch": 3.0164705882352942, "grad_norm": 1.046368509212463, "learning_rate": 0.00019174341573735196, "loss": 0.4473, "step": 6410 }, { "epoch": 3.016941176470588, "grad_norm": 0.9358902186313574, "learning_rate": 0.0001917013094278809, "loss": 1.3047, "step": 6411 }, { "epoch": 3.0174117647058822, "grad_norm": 1.434305478210088, "learning_rate": 0.0001916591995569135, "loss": 1.6172, "step": 6412 }, { "epoch": 3.0178823529411765, "grad_norm": 1.1433306835667953, "learning_rate": 0.00019161708612804615, "loss": 1.4844, "step": 6413 }, { "epoch": 3.0183529411764707, "grad_norm": 0.7000247623446028, "learning_rate": 0.00019157496914487545, "loss": 0.3594, "step": 6414 }, { "epoch": 3.018823529411765, "grad_norm": 0.8293062452877302, "learning_rate": 0.00019153284861099848, "loss": 0.8125, "step": 6415 }, { "epoch": 3.0192941176470587, "grad_norm": 1.096476931542395, "learning_rate": 0.00019149072453001254, "loss": 1.2266, "step": 6416 }, { "epoch": 3.019764705882353, "grad_norm": 0.931804606602948, "learning_rate": 0.00019144859690551518, "loss": 1.0234, "step": 6417 }, { "epoch": 3.020235294117647, "grad_norm": 0.7640052681780769, "learning_rate": 0.00019140646574110438, "loss": 0.7461, "step": 6418 }, { "epoch": 3.0207058823529414, "grad_norm": 1.045124256723912, "learning_rate": 0.0001913643310403783, "loss": 0.498, "step": 6419 }, { "epoch": 3.021176470588235, "grad_norm": 1.1648251651137984, "learning_rate": 0.00019132219280693544, "loss": 0.7539, "step": 6420 }, { "epoch": 3.0216470588235294, "grad_norm": 0.8722914857417708, "learning_rate": 0.00019128005104437462, "loss": 0.793, "step": 6421 }, { "epoch": 3.0221176470588236, "grad_norm": 1.288552356952127, "learning_rate": 0.00019123790575629503, "loss": 1.6016, "step": 6422 }, { "epoch": 3.022588235294118, "grad_norm": 0.6994863789056295, "learning_rate": 0.00019119575694629594, "loss": 0.3965, "step": 6423 }, { "epoch": 3.0230588235294116, "grad_norm": 0.777642166167168, "learning_rate": 0.00019115360461797716, "loss": 0.4707, "step": 6424 }, { "epoch": 3.023529411764706, "grad_norm": 0.7496850815961983, "learning_rate": 0.00019111144877493873, "loss": 0.3652, "step": 6425 }, { "epoch": 3.024, "grad_norm": 0.9970293454689487, "learning_rate": 0.0001910692894207809, "loss": 0.6562, "step": 6426 }, { "epoch": 3.0244705882352942, "grad_norm": 0.4692136660672241, "learning_rate": 0.00019102712655910426, "loss": 0.2012, "step": 6427 }, { "epoch": 3.024941176470588, "grad_norm": 1.6226892482859347, "learning_rate": 0.00019098496019350985, "loss": 1.3906, "step": 6428 }, { "epoch": 3.0254117647058822, "grad_norm": 0.6337843764444453, "learning_rate": 0.0001909427903275988, "loss": 0.1875, "step": 6429 }, { "epoch": 3.0258823529411765, "grad_norm": 1.0997179484807735, "learning_rate": 0.00019090061696497258, "loss": 0.8203, "step": 6430 }, { "epoch": 3.0263529411764707, "grad_norm": 1.163035942563219, "learning_rate": 0.0001908584401092331, "loss": 0.9883, "step": 6431 }, { "epoch": 3.026823529411765, "grad_norm": 1.2284428887421488, "learning_rate": 0.0001908162597639824, "loss": 1.5391, "step": 6432 }, { "epoch": 3.0272941176470587, "grad_norm": 1.0589104050052307, "learning_rate": 0.00019077407593282294, "loss": 1.0312, "step": 6433 }, { "epoch": 3.027764705882353, "grad_norm": 1.0568111011049124, "learning_rate": 0.00019073188861935735, "loss": 0.5547, "step": 6434 }, { "epoch": 3.028235294117647, "grad_norm": 1.0478584874799959, "learning_rate": 0.00019068969782718874, "loss": 1.4453, "step": 6435 }, { "epoch": 3.0287058823529414, "grad_norm": 0.9980317466147722, "learning_rate": 0.00019064750355992036, "loss": 0.6094, "step": 6436 }, { "epoch": 3.029176470588235, "grad_norm": 0.9263309176834527, "learning_rate": 0.00019060530582115578, "loss": 0.5039, "step": 6437 }, { "epoch": 3.0296470588235294, "grad_norm": 0.8226249660579402, "learning_rate": 0.00019056310461449885, "loss": 0.5078, "step": 6438 }, { "epoch": 3.0301176470588236, "grad_norm": 0.975605583413939, "learning_rate": 0.00019052089994355394, "loss": 0.7109, "step": 6439 }, { "epoch": 3.030588235294118, "grad_norm": 0.8982215896362229, "learning_rate": 0.00019047869181192535, "loss": 0.4844, "step": 6440 }, { "epoch": 3.0310588235294116, "grad_norm": 0.4867795850671066, "learning_rate": 0.00019043648022321793, "loss": 0.2637, "step": 6441 }, { "epoch": 3.031529411764706, "grad_norm": 0.8677708286608353, "learning_rate": 0.00019039426518103678, "loss": 0.5195, "step": 6442 }, { "epoch": 3.032, "grad_norm": 1.1659733447867104, "learning_rate": 0.00019035204668898725, "loss": 0.9102, "step": 6443 }, { "epoch": 3.0324705882352943, "grad_norm": 0.7376223759507649, "learning_rate": 0.00019030982475067498, "loss": 0.7773, "step": 6444 }, { "epoch": 3.032941176470588, "grad_norm": 0.5164590961602744, "learning_rate": 0.00019026759936970598, "loss": 0.2031, "step": 6445 }, { "epoch": 3.0334117647058823, "grad_norm": 0.6704391257741054, "learning_rate": 0.0001902253705496865, "loss": 0.3398, "step": 6446 }, { "epoch": 3.0338823529411765, "grad_norm": 1.1976285716480972, "learning_rate": 0.00019018313829422307, "loss": 1.5, "step": 6447 }, { "epoch": 3.0343529411764707, "grad_norm": 0.5656810587728188, "learning_rate": 0.0001901409026069225, "loss": 0.249, "step": 6448 }, { "epoch": 3.034823529411765, "grad_norm": 0.49827638694598736, "learning_rate": 0.000190098663491392, "loss": 0.5156, "step": 6449 }, { "epoch": 3.0352941176470587, "grad_norm": 0.9711925744913699, "learning_rate": 0.00019005642095123895, "loss": 0.6367, "step": 6450 }, { "epoch": 3.035764705882353, "grad_norm": 1.0671374014375494, "learning_rate": 0.00019001417499007103, "loss": 0.8164, "step": 6451 }, { "epoch": 3.036235294117647, "grad_norm": 0.8759975721677118, "learning_rate": 0.0001899719256114963, "loss": 0.918, "step": 6452 }, { "epoch": 3.0367058823529414, "grad_norm": 0.8997595782833402, "learning_rate": 0.00018992967281912312, "loss": 0.5195, "step": 6453 }, { "epoch": 3.037176470588235, "grad_norm": 0.7044642718933308, "learning_rate": 0.00018988741661655996, "loss": 0.332, "step": 6454 }, { "epoch": 3.0376470588235294, "grad_norm": 1.2254082133187432, "learning_rate": 0.00018984515700741579, "loss": 1.0156, "step": 6455 }, { "epoch": 3.0381176470588236, "grad_norm": 0.7810017829209822, "learning_rate": 0.0001898028939952998, "loss": 0.4297, "step": 6456 }, { "epoch": 3.038588235294118, "grad_norm": 0.4967958519843177, "learning_rate": 0.00018976062758382144, "loss": 0.1699, "step": 6457 }, { "epoch": 3.0390588235294116, "grad_norm": 0.5003914528637243, "learning_rate": 0.0001897183577765904, "loss": 0.1631, "step": 6458 }, { "epoch": 3.039529411764706, "grad_norm": 0.6568319317885023, "learning_rate": 0.00018967608457721684, "loss": 0.7773, "step": 6459 }, { "epoch": 3.04, "grad_norm": 1.5489190337678103, "learning_rate": 0.00018963380798931102, "loss": 1.9062, "step": 6460 }, { "epoch": 3.0404705882352943, "grad_norm": 1.1537868697918072, "learning_rate": 0.00018959152801648364, "loss": 1.3047, "step": 6461 }, { "epoch": 3.040941176470588, "grad_norm": 0.42379892389509305, "learning_rate": 0.00018954924466234554, "loss": 0.3145, "step": 6462 }, { "epoch": 3.0414117647058823, "grad_norm": 1.0939397638439703, "learning_rate": 0.00018950695793050796, "loss": 1.1016, "step": 6463 }, { "epoch": 3.0418823529411765, "grad_norm": 1.1681726884187917, "learning_rate": 0.0001894646678245824, "loss": 1.0859, "step": 6464 }, { "epoch": 3.0423529411764707, "grad_norm": 0.946560238032588, "learning_rate": 0.00018942237434818062, "loss": 0.8047, "step": 6465 }, { "epoch": 3.0428235294117645, "grad_norm": 0.967156049016425, "learning_rate": 0.00018938007750491468, "loss": 0.7695, "step": 6466 }, { "epoch": 3.0432941176470587, "grad_norm": 1.1057808590153613, "learning_rate": 0.00018933777729839701, "loss": 0.793, "step": 6467 }, { "epoch": 3.043764705882353, "grad_norm": 0.857065685630132, "learning_rate": 0.0001892954737322402, "loss": 1.0625, "step": 6468 }, { "epoch": 3.044235294117647, "grad_norm": 0.761405085419839, "learning_rate": 0.00018925316681005718, "loss": 0.3926, "step": 6469 }, { "epoch": 3.0447058823529414, "grad_norm": 0.9545772149332783, "learning_rate": 0.0001892108565354612, "loss": 1.0, "step": 6470 }, { "epoch": 3.045176470588235, "grad_norm": 0.9499923094077463, "learning_rate": 0.0001891685429120657, "loss": 0.9766, "step": 6471 }, { "epoch": 3.0456470588235294, "grad_norm": 0.9769240945925024, "learning_rate": 0.00018912622594348454, "loss": 1.2344, "step": 6472 }, { "epoch": 3.0461176470588236, "grad_norm": 1.2406484227318804, "learning_rate": 0.00018908390563333175, "loss": 1.0469, "step": 6473 }, { "epoch": 3.046588235294118, "grad_norm": 0.9587314286248019, "learning_rate": 0.00018904158198522177, "loss": 0.75, "step": 6474 }, { "epoch": 3.0470588235294116, "grad_norm": 0.8522366171766707, "learning_rate": 0.00018899925500276912, "loss": 0.7578, "step": 6475 }, { "epoch": 3.047529411764706, "grad_norm": 1.0036744172559648, "learning_rate": 0.00018895692468958884, "loss": 0.793, "step": 6476 }, { "epoch": 3.048, "grad_norm": 0.7930266837146128, "learning_rate": 0.00018891459104929608, "loss": 0.5312, "step": 6477 }, { "epoch": 3.0484705882352943, "grad_norm": 0.6468407517668071, "learning_rate": 0.0001888722540855064, "loss": 0.3164, "step": 6478 }, { "epoch": 3.048941176470588, "grad_norm": 0.9034246530416107, "learning_rate": 0.00018882991380183556, "loss": 1.1719, "step": 6479 }, { "epoch": 3.0494117647058823, "grad_norm": 1.0833409703359074, "learning_rate": 0.0001887875702018996, "loss": 0.8867, "step": 6480 }, { "epoch": 3.0498823529411765, "grad_norm": 1.2021268550764515, "learning_rate": 0.00018874522328931492, "loss": 1.0547, "step": 6481 }, { "epoch": 3.0503529411764707, "grad_norm": 0.8434171616707177, "learning_rate": 0.0001887028730676981, "loss": 1.1953, "step": 6482 }, { "epoch": 3.0508235294117645, "grad_norm": 1.1337403820081065, "learning_rate": 0.00018866051954066608, "loss": 1.1875, "step": 6483 }, { "epoch": 3.0512941176470587, "grad_norm": 1.118007678223176, "learning_rate": 0.00018861816271183613, "loss": 1.4219, "step": 6484 }, { "epoch": 3.051764705882353, "grad_norm": 1.0479387348184621, "learning_rate": 0.00018857580258482557, "loss": 0.8906, "step": 6485 }, { "epoch": 3.052235294117647, "grad_norm": 0.948540865121124, "learning_rate": 0.0001885334391632523, "loss": 0.7031, "step": 6486 }, { "epoch": 3.0527058823529414, "grad_norm": 1.2989004823282932, "learning_rate": 0.00018849107245073432, "loss": 1.5938, "step": 6487 }, { "epoch": 3.053176470588235, "grad_norm": 0.6783865934600866, "learning_rate": 0.00018844870245088996, "loss": 0.3281, "step": 6488 }, { "epoch": 3.0536470588235294, "grad_norm": 0.8728262631560687, "learning_rate": 0.00018840632916733777, "loss": 0.9297, "step": 6489 }, { "epoch": 3.0541176470588236, "grad_norm": 0.9300416675650435, "learning_rate": 0.00018836395260369676, "loss": 0.6758, "step": 6490 }, { "epoch": 3.054588235294118, "grad_norm": 1.4046609474936165, "learning_rate": 0.000188321572763586, "loss": 1.5234, "step": 6491 }, { "epoch": 3.0550588235294116, "grad_norm": 0.8974104696286804, "learning_rate": 0.00018827918965062492, "loss": 0.7188, "step": 6492 }, { "epoch": 3.055529411764706, "grad_norm": 0.5900170724469875, "learning_rate": 0.00018823680326843334, "loss": 0.166, "step": 6493 }, { "epoch": 3.056, "grad_norm": 0.6578457272970598, "learning_rate": 0.00018819441362063122, "loss": 0.418, "step": 6494 }, { "epoch": 3.0564705882352943, "grad_norm": 1.050727160759921, "learning_rate": 0.00018815202071083878, "loss": 1.1875, "step": 6495 }, { "epoch": 3.056941176470588, "grad_norm": 1.1769728331340055, "learning_rate": 0.00018810962454267666, "loss": 1.0625, "step": 6496 }, { "epoch": 3.0574117647058823, "grad_norm": 1.0478389088183053, "learning_rate": 0.00018806722511976567, "loss": 1.3906, "step": 6497 }, { "epoch": 3.0578823529411765, "grad_norm": 0.8268729508520437, "learning_rate": 0.00018802482244572693, "loss": 0.7969, "step": 6498 }, { "epoch": 3.0583529411764707, "grad_norm": 1.2166877869749755, "learning_rate": 0.00018798241652418182, "loss": 1.5234, "step": 6499 }, { "epoch": 3.0588235294117645, "grad_norm": 1.1083562673370235, "learning_rate": 0.00018794000735875208, "loss": 1.0703, "step": 6500 }, { "epoch": 3.0592941176470587, "grad_norm": 0.6425580767772966, "learning_rate": 0.00018789759495305963, "loss": 0.2422, "step": 6501 }, { "epoch": 3.059764705882353, "grad_norm": 0.8321866084713321, "learning_rate": 0.00018785517931072668, "loss": 0.5117, "step": 6502 }, { "epoch": 3.060235294117647, "grad_norm": 0.9094362555091435, "learning_rate": 0.0001878127604353757, "loss": 0.7969, "step": 6503 }, { "epoch": 3.0607058823529414, "grad_norm": 0.6919149760714106, "learning_rate": 0.00018777033833062954, "loss": 0.2988, "step": 6504 }, { "epoch": 3.061176470588235, "grad_norm": 0.622032997738327, "learning_rate": 0.00018772791300011122, "loss": 0.3672, "step": 6505 }, { "epoch": 3.0616470588235294, "grad_norm": 1.3699692074672227, "learning_rate": 0.00018768548444744404, "loss": 1.7344, "step": 6506 }, { "epoch": 3.0621176470588236, "grad_norm": 0.9541972488479893, "learning_rate": 0.00018764305267625169, "loss": 0.6094, "step": 6507 }, { "epoch": 3.062588235294118, "grad_norm": 1.038597790602823, "learning_rate": 0.000187600617690158, "loss": 0.7344, "step": 6508 }, { "epoch": 3.0630588235294116, "grad_norm": 1.0327663721392597, "learning_rate": 0.00018755817949278706, "loss": 0.9062, "step": 6509 }, { "epoch": 3.063529411764706, "grad_norm": 1.2997173578247339, "learning_rate": 0.00018751573808776346, "loss": 1.9141, "step": 6510 }, { "epoch": 3.064, "grad_norm": 0.6338539080299074, "learning_rate": 0.00018747329347871178, "loss": 0.3574, "step": 6511 }, { "epoch": 3.0644705882352943, "grad_norm": 1.1771922362513474, "learning_rate": 0.00018743084566925702, "loss": 0.9141, "step": 6512 }, { "epoch": 3.064941176470588, "grad_norm": 1.3531872039346116, "learning_rate": 0.00018738839466302443, "loss": 0.2598, "step": 6513 }, { "epoch": 3.0654117647058823, "grad_norm": 1.2112483491941834, "learning_rate": 0.00018734594046363956, "loss": 1.9219, "step": 6514 }, { "epoch": 3.0658823529411765, "grad_norm": 1.1597900515264865, "learning_rate": 0.00018730348307472824, "loss": 1.1016, "step": 6515 }, { "epoch": 3.0663529411764707, "grad_norm": 0.7709455515323137, "learning_rate": 0.0001872610224999164, "loss": 0.7852, "step": 6516 }, { "epoch": 3.0668235294117645, "grad_norm": 0.7530952020978353, "learning_rate": 0.00018721855874283058, "loss": 0.3398, "step": 6517 }, { "epoch": 3.0672941176470587, "grad_norm": 0.8416309887670214, "learning_rate": 0.00018717609180709721, "loss": 1.1016, "step": 6518 }, { "epoch": 3.067764705882353, "grad_norm": 1.0816692802272265, "learning_rate": 0.00018713362169634325, "loss": 1.0234, "step": 6519 }, { "epoch": 3.068235294117647, "grad_norm": 0.6659736220540767, "learning_rate": 0.00018709114841419588, "loss": 0.8047, "step": 6520 }, { "epoch": 3.0687058823529414, "grad_norm": 0.7738795272735862, "learning_rate": 0.0001870486719642825, "loss": 0.4805, "step": 6521 }, { "epoch": 3.069176470588235, "grad_norm": 1.0048701653232217, "learning_rate": 0.00018700619235023085, "loss": 1.3516, "step": 6522 }, { "epoch": 3.0696470588235294, "grad_norm": 0.7645420001540283, "learning_rate": 0.00018696370957566882, "loss": 1.2266, "step": 6523 }, { "epoch": 3.0701176470588236, "grad_norm": 0.8679600231706317, "learning_rate": 0.0001869212236442247, "loss": 0.5312, "step": 6524 }, { "epoch": 3.070588235294118, "grad_norm": 0.9086032420212071, "learning_rate": 0.00018687873455952697, "loss": 0.5859, "step": 6525 }, { "epoch": 3.0710588235294116, "grad_norm": 0.9275797893959397, "learning_rate": 0.00018683624232520446, "loss": 0.8711, "step": 6526 }, { "epoch": 3.071529411764706, "grad_norm": 0.871985775878915, "learning_rate": 0.00018679374694488608, "loss": 1.2109, "step": 6527 }, { "epoch": 3.072, "grad_norm": 0.888928280059318, "learning_rate": 0.00018675124842220135, "loss": 0.9727, "step": 6528 }, { "epoch": 3.0724705882352943, "grad_norm": 0.5271821331861819, "learning_rate": 0.0001867087467607797, "loss": 0.3008, "step": 6529 }, { "epoch": 3.072941176470588, "grad_norm": 0.8380355064773234, "learning_rate": 0.00018666624196425098, "loss": 0.7891, "step": 6530 }, { "epoch": 3.0734117647058823, "grad_norm": 1.145486900830468, "learning_rate": 0.00018662373403624538, "loss": 0.918, "step": 6531 }, { "epoch": 3.0738823529411765, "grad_norm": 1.211077144919547, "learning_rate": 0.00018658122298039327, "loss": 1.3281, "step": 6532 }, { "epoch": 3.0743529411764707, "grad_norm": 0.9875163913968908, "learning_rate": 0.00018653870880032527, "loss": 1.0469, "step": 6533 }, { "epoch": 3.0748235294117645, "grad_norm": 0.7672438428341123, "learning_rate": 0.00018649619149967232, "loss": 1.0156, "step": 6534 }, { "epoch": 3.0752941176470587, "grad_norm": 0.8453423908871176, "learning_rate": 0.00018645367108206565, "loss": 0.3555, "step": 6535 }, { "epoch": 3.075764705882353, "grad_norm": 1.0423252113933015, "learning_rate": 0.00018641114755113658, "loss": 0.9141, "step": 6536 }, { "epoch": 3.076235294117647, "grad_norm": 0.628951629173593, "learning_rate": 0.00018636862091051697, "loss": 0.2969, "step": 6537 }, { "epoch": 3.0767058823529414, "grad_norm": 0.6106344111989399, "learning_rate": 0.00018632609116383874, "loss": 0.1953, "step": 6538 }, { "epoch": 3.077176470588235, "grad_norm": 0.8904971693989776, "learning_rate": 0.00018628355831473416, "loss": 0.8633, "step": 6539 }, { "epoch": 3.0776470588235294, "grad_norm": 0.9596355405986012, "learning_rate": 0.00018624102236683567, "loss": 1.2109, "step": 6540 }, { "epoch": 3.0781176470588236, "grad_norm": 1.0295099800390823, "learning_rate": 0.00018619848332377615, "loss": 1.1172, "step": 6541 }, { "epoch": 3.078588235294118, "grad_norm": 1.1725283336578125, "learning_rate": 0.00018615594118918862, "loss": 0.9414, "step": 6542 }, { "epoch": 3.0790588235294116, "grad_norm": 0.7932865665054804, "learning_rate": 0.00018611339596670632, "loss": 0.4004, "step": 6543 }, { "epoch": 3.079529411764706, "grad_norm": 0.8200969970371379, "learning_rate": 0.00018607084765996285, "loss": 0.9492, "step": 6544 }, { "epoch": 3.08, "grad_norm": 0.7921472664522747, "learning_rate": 0.00018602829627259213, "loss": 0.3496, "step": 6545 }, { "epoch": 3.0804705882352943, "grad_norm": 1.1085851587071265, "learning_rate": 0.00018598574180822818, "loss": 0.957, "step": 6546 }, { "epoch": 3.080941176470588, "grad_norm": 0.7139605059541888, "learning_rate": 0.00018594318427050537, "loss": 0.3535, "step": 6547 }, { "epoch": 3.0814117647058823, "grad_norm": 1.033849460635936, "learning_rate": 0.00018590062366305828, "loss": 0.8789, "step": 6548 }, { "epoch": 3.0818823529411765, "grad_norm": 0.9684610282196147, "learning_rate": 0.0001858580599895219, "loss": 0.7969, "step": 6549 }, { "epoch": 3.0823529411764707, "grad_norm": 0.8758136326227642, "learning_rate": 0.00018581549325353126, "loss": 0.9805, "step": 6550 }, { "epoch": 3.0828235294117645, "grad_norm": 0.8757335875981621, "learning_rate": 0.00018577292345872188, "loss": 0.4258, "step": 6551 }, { "epoch": 3.0832941176470587, "grad_norm": 1.0433191755385087, "learning_rate": 0.00018573035060872933, "loss": 0.8672, "step": 6552 }, { "epoch": 3.083764705882353, "grad_norm": 1.1471648809981452, "learning_rate": 0.0001856877747071896, "loss": 0.6992, "step": 6553 }, { "epoch": 3.084235294117647, "grad_norm": 1.0489804145573438, "learning_rate": 0.00018564519575773884, "loss": 1.2578, "step": 6554 }, { "epoch": 3.084705882352941, "grad_norm": 0.8731958404840344, "learning_rate": 0.00018560261376401356, "loss": 0.6484, "step": 6555 }, { "epoch": 3.085176470588235, "grad_norm": 0.9005138586815497, "learning_rate": 0.0001855600287296504, "loss": 0.7773, "step": 6556 }, { "epoch": 3.0856470588235294, "grad_norm": 0.7857514211700202, "learning_rate": 0.00018551744065828634, "loss": 0.6445, "step": 6557 }, { "epoch": 3.0861176470588236, "grad_norm": 1.358433865990449, "learning_rate": 0.0001854748495535587, "loss": 1.6328, "step": 6558 }, { "epoch": 3.086588235294118, "grad_norm": 1.0674813126081442, "learning_rate": 0.0001854322554191049, "loss": 0.9336, "step": 6559 }, { "epoch": 3.0870588235294116, "grad_norm": 0.9239598077097645, "learning_rate": 0.00018538965825856264, "loss": 0.6562, "step": 6560 }, { "epoch": 3.087529411764706, "grad_norm": 1.0611138026014268, "learning_rate": 0.00018534705807557003, "loss": 1.0625, "step": 6561 }, { "epoch": 3.088, "grad_norm": 0.47954086304889343, "learning_rate": 0.00018530445487376524, "loss": 0.165, "step": 6562 }, { "epoch": 3.0884705882352943, "grad_norm": 1.0966130632958018, "learning_rate": 0.0001852618486567869, "loss": 1.1953, "step": 6563 }, { "epoch": 3.088941176470588, "grad_norm": 0.9209816538400345, "learning_rate": 0.00018521923942827362, "loss": 0.5977, "step": 6564 }, { "epoch": 3.0894117647058823, "grad_norm": 0.6442463624527373, "learning_rate": 0.00018517662719186464, "loss": 0.1562, "step": 6565 }, { "epoch": 3.0898823529411765, "grad_norm": 1.3755016600296506, "learning_rate": 0.0001851340119511991, "loss": 0.9883, "step": 6566 }, { "epoch": 3.0903529411764707, "grad_norm": 0.36719587508944534, "learning_rate": 0.00018509139370991666, "loss": 0.0869, "step": 6567 }, { "epoch": 3.0908235294117645, "grad_norm": 0.9514254458493528, "learning_rate": 0.000185048772471657, "loss": 0.9219, "step": 6568 }, { "epoch": 3.0912941176470587, "grad_norm": 1.0370519778552922, "learning_rate": 0.00018500614824006035, "loss": 0.6133, "step": 6569 }, { "epoch": 3.091764705882353, "grad_norm": 1.5074599732785021, "learning_rate": 0.0001849635210187669, "loss": 1.5312, "step": 6570 }, { "epoch": 3.092235294117647, "grad_norm": 0.936434206096833, "learning_rate": 0.00018492089081141724, "loss": 1.0078, "step": 6571 }, { "epoch": 3.092705882352941, "grad_norm": 0.9088749941401348, "learning_rate": 0.00018487825762165223, "loss": 0.5312, "step": 6572 }, { "epoch": 3.093176470588235, "grad_norm": 1.0068420031712644, "learning_rate": 0.00018483562145311296, "loss": 0.6172, "step": 6573 }, { "epoch": 3.0936470588235294, "grad_norm": 0.6565843024939306, "learning_rate": 0.00018479298230944071, "loss": 0.1885, "step": 6574 }, { "epoch": 3.0941176470588236, "grad_norm": 1.1455209004617357, "learning_rate": 0.00018475034019427717, "loss": 1.0312, "step": 6575 }, { "epoch": 3.094588235294118, "grad_norm": 1.0280238822778347, "learning_rate": 0.00018470769511126409, "loss": 1.0078, "step": 6576 }, { "epoch": 3.0950588235294116, "grad_norm": 1.4476257815624938, "learning_rate": 0.00018466504706404362, "loss": 1.4297, "step": 6577 }, { "epoch": 3.095529411764706, "grad_norm": 1.5498752520987085, "learning_rate": 0.00018462239605625808, "loss": 0.8906, "step": 6578 }, { "epoch": 3.096, "grad_norm": 1.0778868174765914, "learning_rate": 0.0001845797420915501, "loss": 1.1406, "step": 6579 }, { "epoch": 3.0964705882352943, "grad_norm": 1.1839205728722273, "learning_rate": 0.00018453708517356256, "loss": 0.9375, "step": 6580 }, { "epoch": 3.096941176470588, "grad_norm": 1.1282008374540569, "learning_rate": 0.0001844944253059385, "loss": 1.2422, "step": 6581 }, { "epoch": 3.0974117647058823, "grad_norm": 0.9262024256291455, "learning_rate": 0.00018445176249232134, "loss": 1.1016, "step": 6582 }, { "epoch": 3.0978823529411765, "grad_norm": 1.039998009511066, "learning_rate": 0.0001844090967363547, "loss": 0.9258, "step": 6583 }, { "epoch": 3.0983529411764708, "grad_norm": 0.9537121216317082, "learning_rate": 0.0001843664280416824, "loss": 1.2734, "step": 6584 }, { "epoch": 3.0988235294117645, "grad_norm": 1.1723496866445355, "learning_rate": 0.0001843237564119486, "loss": 0.4922, "step": 6585 }, { "epoch": 3.0992941176470588, "grad_norm": 1.0731399193415416, "learning_rate": 0.00018428108185079763, "loss": 1.0469, "step": 6586 }, { "epoch": 3.099764705882353, "grad_norm": 0.9174528931164304, "learning_rate": 0.00018423840436187412, "loss": 0.8711, "step": 6587 }, { "epoch": 3.100235294117647, "grad_norm": 0.6628743810308633, "learning_rate": 0.0001841957239488229, "loss": 0.5742, "step": 6588 }, { "epoch": 3.100705882352941, "grad_norm": 1.0005187985413868, "learning_rate": 0.00018415304061528918, "loss": 1.6875, "step": 6589 }, { "epoch": 3.101176470588235, "grad_norm": 0.6720224661434426, "learning_rate": 0.00018411035436491827, "loss": 0.3301, "step": 6590 }, { "epoch": 3.1016470588235294, "grad_norm": 1.0337669739275703, "learning_rate": 0.0001840676652013558, "loss": 0.8359, "step": 6591 }, { "epoch": 3.1021176470588236, "grad_norm": 0.6169789075805041, "learning_rate": 0.00018402497312824758, "loss": 0.5078, "step": 6592 }, { "epoch": 3.1025882352941174, "grad_norm": 0.6959113341436489, "learning_rate": 0.0001839822781492398, "loss": 0.5312, "step": 6593 }, { "epoch": 3.1030588235294116, "grad_norm": 0.7427393275277175, "learning_rate": 0.00018393958026797878, "loss": 0.5156, "step": 6594 }, { "epoch": 3.103529411764706, "grad_norm": 0.917102952280621, "learning_rate": 0.0001838968794881111, "loss": 0.6172, "step": 6595 }, { "epoch": 3.104, "grad_norm": 0.3540126803411886, "learning_rate": 0.00018385417581328364, "loss": 0.1001, "step": 6596 }, { "epoch": 3.1044705882352943, "grad_norm": 1.0550336807466447, "learning_rate": 0.00018381146924714356, "loss": 0.9492, "step": 6597 }, { "epoch": 3.104941176470588, "grad_norm": 0.6004450211931813, "learning_rate": 0.00018376875979333812, "loss": 0.2256, "step": 6598 }, { "epoch": 3.1054117647058823, "grad_norm": 0.9889686905841897, "learning_rate": 0.000183726047455515, "loss": 1.6328, "step": 6599 }, { "epoch": 3.1058823529411765, "grad_norm": 1.1172451768897318, "learning_rate": 0.000183683332237322, "loss": 1.1328, "step": 6600 }, { "epoch": 3.1063529411764708, "grad_norm": 0.48856563961134025, "learning_rate": 0.00018364061414240717, "loss": 0.0923, "step": 6601 }, { "epoch": 3.1068235294117645, "grad_norm": 2.69569781736109, "learning_rate": 0.00018359789317441892, "loss": 1.0859, "step": 6602 }, { "epoch": 3.1072941176470588, "grad_norm": 0.7711986740726504, "learning_rate": 0.00018355516933700578, "loss": 0.4238, "step": 6603 }, { "epoch": 3.107764705882353, "grad_norm": 0.9731846916962309, "learning_rate": 0.0001835124426338166, "loss": 1.1641, "step": 6604 }, { "epoch": 3.108235294117647, "grad_norm": 0.46832898091048886, "learning_rate": 0.00018346971306850045, "loss": 0.3945, "step": 6605 }, { "epoch": 3.108705882352941, "grad_norm": 1.081854739352113, "learning_rate": 0.00018342698064470662, "loss": 0.8477, "step": 6606 }, { "epoch": 3.109176470588235, "grad_norm": 1.5423636906651408, "learning_rate": 0.0001833842453660847, "loss": 1.0312, "step": 6607 }, { "epoch": 3.1096470588235294, "grad_norm": 1.4143336540499851, "learning_rate": 0.00018334150723628442, "loss": 1.2656, "step": 6608 }, { "epoch": 3.1101176470588237, "grad_norm": 1.0912497718998617, "learning_rate": 0.00018329876625895592, "loss": 1.375, "step": 6609 }, { "epoch": 3.1105882352941174, "grad_norm": 1.0249791149382306, "learning_rate": 0.0001832560224377495, "loss": 0.9609, "step": 6610 }, { "epoch": 3.1110588235294117, "grad_norm": 0.631843784609759, "learning_rate": 0.0001832132757763156, "loss": 0.2168, "step": 6611 }, { "epoch": 3.111529411764706, "grad_norm": 0.6058825643325083, "learning_rate": 0.000183170526278305, "loss": 0.2559, "step": 6612 }, { "epoch": 3.112, "grad_norm": 1.401925072858435, "learning_rate": 0.0001831277739473688, "loss": 0.8438, "step": 6613 }, { "epoch": 3.1124705882352943, "grad_norm": 1.154465479124093, "learning_rate": 0.00018308501878715823, "loss": 1.9141, "step": 6614 }, { "epoch": 3.112941176470588, "grad_norm": 1.055935985553616, "learning_rate": 0.00018304226080132472, "loss": 0.6953, "step": 6615 }, { "epoch": 3.1134117647058823, "grad_norm": 0.8398788925689005, "learning_rate": 0.00018299949999352013, "loss": 0.5352, "step": 6616 }, { "epoch": 3.1138823529411765, "grad_norm": 0.7668193352615845, "learning_rate": 0.00018295673636739635, "loss": 0.4336, "step": 6617 }, { "epoch": 3.1143529411764708, "grad_norm": 0.7448843812339598, "learning_rate": 0.00018291396992660565, "loss": 0.6367, "step": 6618 }, { "epoch": 3.1148235294117645, "grad_norm": 0.8527083102116123, "learning_rate": 0.00018287120067480046, "loss": 0.4297, "step": 6619 }, { "epoch": 3.1152941176470588, "grad_norm": 0.9590505609726591, "learning_rate": 0.0001828284286156335, "loss": 0.9414, "step": 6620 }, { "epoch": 3.115764705882353, "grad_norm": 0.8310059741852247, "learning_rate": 0.0001827856537527577, "loss": 0.5781, "step": 6621 }, { "epoch": 3.116235294117647, "grad_norm": 1.3847890956120097, "learning_rate": 0.00018274287608982626, "loss": 1.7578, "step": 6622 }, { "epoch": 3.116705882352941, "grad_norm": 0.5589249607882996, "learning_rate": 0.00018270009563049264, "loss": 0.3027, "step": 6623 }, { "epoch": 3.117176470588235, "grad_norm": 0.7938616664548905, "learning_rate": 0.00018265731237841043, "loss": 0.3281, "step": 6624 }, { "epoch": 3.1176470588235294, "grad_norm": 0.8358545256697775, "learning_rate": 0.00018261452633723356, "loss": 0.6406, "step": 6625 }, { "epoch": 3.1181176470588237, "grad_norm": 1.4330629714834866, "learning_rate": 0.00018257173751061613, "loss": 0.9336, "step": 6626 }, { "epoch": 3.1185882352941174, "grad_norm": 0.8394981058647359, "learning_rate": 0.00018252894590221262, "loss": 1.1484, "step": 6627 }, { "epoch": 3.1190588235294117, "grad_norm": 1.2613137945786799, "learning_rate": 0.00018248615151567754, "loss": 1.3672, "step": 6628 }, { "epoch": 3.119529411764706, "grad_norm": 0.862519644117742, "learning_rate": 0.00018244335435466577, "loss": 0.6641, "step": 6629 }, { "epoch": 3.12, "grad_norm": 0.6187215041391465, "learning_rate": 0.00018240055442283242, "loss": 0.457, "step": 6630 }, { "epoch": 3.1204705882352943, "grad_norm": 0.899698358053465, "learning_rate": 0.00018235775172383277, "loss": 0.457, "step": 6631 }, { "epoch": 3.120941176470588, "grad_norm": 0.8968878984748208, "learning_rate": 0.00018231494626132243, "loss": 0.5469, "step": 6632 }, { "epoch": 3.1214117647058823, "grad_norm": 1.0132589255625661, "learning_rate": 0.00018227213803895717, "loss": 1.2422, "step": 6633 }, { "epoch": 3.1218823529411766, "grad_norm": 0.7436385105180233, "learning_rate": 0.00018222932706039307, "loss": 0.4043, "step": 6634 }, { "epoch": 3.1223529411764708, "grad_norm": 0.786009333226571, "learning_rate": 0.0001821865133292863, "loss": 1.1797, "step": 6635 }, { "epoch": 3.1228235294117646, "grad_norm": 0.9640924622441254, "learning_rate": 0.00018214369684929347, "loss": 1.3984, "step": 6636 }, { "epoch": 3.1232941176470588, "grad_norm": 0.5938183797524659, "learning_rate": 0.00018210087762407122, "loss": 0.208, "step": 6637 }, { "epoch": 3.123764705882353, "grad_norm": 0.9559973780846025, "learning_rate": 0.00018205805565727658, "loss": 0.9609, "step": 6638 }, { "epoch": 3.124235294117647, "grad_norm": 0.900977428087371, "learning_rate": 0.00018201523095256672, "loss": 0.4316, "step": 6639 }, { "epoch": 3.124705882352941, "grad_norm": 0.8642455194946572, "learning_rate": 0.00018197240351359913, "loss": 0.4102, "step": 6640 }, { "epoch": 3.1251764705882352, "grad_norm": 0.7390657410186743, "learning_rate": 0.0001819295733440315, "loss": 0.4062, "step": 6641 }, { "epoch": 3.1256470588235294, "grad_norm": 0.814019131290074, "learning_rate": 0.00018188674044752163, "loss": 0.4688, "step": 6642 }, { "epoch": 3.1261176470588237, "grad_norm": 0.9735070137566414, "learning_rate": 0.00018184390482772774, "loss": 1.0156, "step": 6643 }, { "epoch": 3.1265882352941174, "grad_norm": 0.8224644429766867, "learning_rate": 0.0001818010664883082, "loss": 0.5586, "step": 6644 }, { "epoch": 3.1270588235294117, "grad_norm": 0.97362094673867, "learning_rate": 0.0001817582254329216, "loss": 0.9023, "step": 6645 }, { "epoch": 3.127529411764706, "grad_norm": 1.0482116548905873, "learning_rate": 0.0001817153816652267, "loss": 0.9688, "step": 6646 }, { "epoch": 3.128, "grad_norm": 1.0276893757813117, "learning_rate": 0.0001816725351888827, "loss": 0.7891, "step": 6647 }, { "epoch": 3.1284705882352943, "grad_norm": 0.7809309233661873, "learning_rate": 0.00018162968600754888, "loss": 1.4609, "step": 6648 }, { "epoch": 3.128941176470588, "grad_norm": 1.0567863633160883, "learning_rate": 0.00018158683412488466, "loss": 0.9102, "step": 6649 }, { "epoch": 3.1294117647058823, "grad_norm": 1.044593561396743, "learning_rate": 0.00018154397954454993, "loss": 1.3359, "step": 6650 }, { "epoch": 3.1298823529411766, "grad_norm": 1.147519312091916, "learning_rate": 0.0001815011222702046, "loss": 0.6953, "step": 6651 }, { "epoch": 3.130352941176471, "grad_norm": 1.1297179267809796, "learning_rate": 0.00018145826230550892, "loss": 1.2578, "step": 6652 }, { "epoch": 3.1308235294117646, "grad_norm": 1.748521274905261, "learning_rate": 0.00018141539965412334, "loss": 1.9375, "step": 6653 }, { "epoch": 3.131294117647059, "grad_norm": 0.682861416087631, "learning_rate": 0.00018137253431970853, "loss": 0.3965, "step": 6654 }, { "epoch": 3.131764705882353, "grad_norm": 1.2748338256097698, "learning_rate": 0.0001813296663059254, "loss": 1.3438, "step": 6655 }, { "epoch": 3.1322352941176472, "grad_norm": 0.6810847413336647, "learning_rate": 0.00018128679561643508, "loss": 0.5234, "step": 6656 }, { "epoch": 3.132705882352941, "grad_norm": 1.8191941604173432, "learning_rate": 0.000181243922254899, "loss": 1.5078, "step": 6657 }, { "epoch": 3.1331764705882352, "grad_norm": 0.47421896079051634, "learning_rate": 0.0001812010462249787, "loss": 0.582, "step": 6658 }, { "epoch": 3.1336470588235295, "grad_norm": 0.903344875504946, "learning_rate": 0.00018115816753033601, "loss": 1.1016, "step": 6659 }, { "epoch": 3.1341176470588237, "grad_norm": 0.9168296045659776, "learning_rate": 0.00018111528617463296, "loss": 0.375, "step": 6660 }, { "epoch": 3.1345882352941175, "grad_norm": 0.8224739674712939, "learning_rate": 0.00018107240216153184, "loss": 0.5703, "step": 6661 }, { "epoch": 3.1350588235294117, "grad_norm": 0.8778740965611821, "learning_rate": 0.00018102951549469522, "loss": 1.1953, "step": 6662 }, { "epoch": 3.135529411764706, "grad_norm": 0.9674750049087271, "learning_rate": 0.00018098662617778573, "loss": 1.1875, "step": 6663 }, { "epoch": 3.136, "grad_norm": 1.3617666414161989, "learning_rate": 0.0001809437342144664, "loss": 1.3594, "step": 6664 }, { "epoch": 3.1364705882352943, "grad_norm": 0.4044474797175623, "learning_rate": 0.00018090083960840042, "loss": 0.0918, "step": 6665 }, { "epoch": 3.136941176470588, "grad_norm": 0.9763864734446784, "learning_rate": 0.0001808579423632511, "loss": 1.1172, "step": 6666 }, { "epoch": 3.1374117647058823, "grad_norm": 1.1721542280384176, "learning_rate": 0.0001808150424826822, "loss": 1.4141, "step": 6667 }, { "epoch": 3.1378823529411766, "grad_norm": 0.653719409112031, "learning_rate": 0.0001807721399703575, "loss": 0.3379, "step": 6668 }, { "epoch": 3.138352941176471, "grad_norm": 1.1736527519165492, "learning_rate": 0.0001807292348299411, "loss": 0.7773, "step": 6669 }, { "epoch": 3.1388235294117646, "grad_norm": 0.8612778033861062, "learning_rate": 0.00018068632706509733, "loss": 0.6172, "step": 6670 }, { "epoch": 3.139294117647059, "grad_norm": 0.8408759225341264, "learning_rate": 0.00018064341667949074, "loss": 0.75, "step": 6671 }, { "epoch": 3.139764705882353, "grad_norm": 0.9543877270732528, "learning_rate": 0.00018060050367678607, "loss": 1.0547, "step": 6672 }, { "epoch": 3.1402352941176472, "grad_norm": 1.1129904151661751, "learning_rate": 0.00018055758806064822, "loss": 0.6797, "step": 6673 }, { "epoch": 3.140705882352941, "grad_norm": 0.8007289039813672, "learning_rate": 0.00018051466983474252, "loss": 0.5859, "step": 6674 }, { "epoch": 3.1411764705882352, "grad_norm": 0.8896783762299839, "learning_rate": 0.00018047174900273435, "loss": 0.7852, "step": 6675 }, { "epoch": 3.1416470588235295, "grad_norm": 1.127766304110543, "learning_rate": 0.00018042882556828927, "loss": 1.2812, "step": 6676 }, { "epoch": 3.1421176470588237, "grad_norm": 0.5210329470864774, "learning_rate": 0.0001803858995350733, "loss": 0.103, "step": 6677 }, { "epoch": 3.1425882352941175, "grad_norm": 1.2620514426772629, "learning_rate": 0.00018034297090675246, "loss": 1.7109, "step": 6678 }, { "epoch": 3.1430588235294117, "grad_norm": 0.8825153336268385, "learning_rate": 0.00018030003968699308, "loss": 0.5469, "step": 6679 }, { "epoch": 3.143529411764706, "grad_norm": 0.8708871894821736, "learning_rate": 0.00018025710587946166, "loss": 0.9141, "step": 6680 }, { "epoch": 3.144, "grad_norm": 1.068520999769298, "learning_rate": 0.00018021416948782503, "loss": 0.9297, "step": 6681 }, { "epoch": 3.1444705882352944, "grad_norm": 0.9560157326536921, "learning_rate": 0.00018017123051575012, "loss": 0.5742, "step": 6682 }, { "epoch": 3.144941176470588, "grad_norm": 0.8392737031560973, "learning_rate": 0.0001801282889669041, "loss": 0.6641, "step": 6683 }, { "epoch": 3.1454117647058824, "grad_norm": 0.4347015510632458, "learning_rate": 0.00018008534484495445, "loss": 0.1289, "step": 6684 }, { "epoch": 3.1458823529411766, "grad_norm": 1.0046110115957874, "learning_rate": 0.0001800423981535688, "loss": 0.9453, "step": 6685 }, { "epoch": 3.1463529411764704, "grad_norm": 0.8982734775201339, "learning_rate": 0.000179999448896415, "loss": 1.2344, "step": 6686 }, { "epoch": 3.1468235294117646, "grad_norm": 0.6523916973028837, "learning_rate": 0.00017995649707716103, "loss": 0.3379, "step": 6687 }, { "epoch": 3.147294117647059, "grad_norm": 0.9698281165086797, "learning_rate": 0.00017991354269947533, "loss": 0.8438, "step": 6688 }, { "epoch": 3.147764705882353, "grad_norm": 0.6942588416640583, "learning_rate": 0.00017987058576702638, "loss": 0.332, "step": 6689 }, { "epoch": 3.1482352941176472, "grad_norm": 1.0521215777549544, "learning_rate": 0.00017982762628348286, "loss": 0.6836, "step": 6690 }, { "epoch": 3.148705882352941, "grad_norm": 0.834791866922409, "learning_rate": 0.00017978466425251375, "loss": 0.7305, "step": 6691 }, { "epoch": 3.1491764705882352, "grad_norm": 0.6754518101694249, "learning_rate": 0.00017974169967778825, "loss": 0.4199, "step": 6692 }, { "epoch": 3.1496470588235295, "grad_norm": 1.2240666834773195, "learning_rate": 0.00017969873256297572, "loss": 0.9531, "step": 6693 }, { "epoch": 3.1501176470588237, "grad_norm": 0.9371344552493922, "learning_rate": 0.0001796557629117457, "loss": 0.7383, "step": 6694 }, { "epoch": 3.1505882352941175, "grad_norm": 1.1112682589432974, "learning_rate": 0.00017961279072776812, "loss": 1.1797, "step": 6695 }, { "epoch": 3.1510588235294117, "grad_norm": 1.0083816352432702, "learning_rate": 0.00017956981601471294, "loss": 0.8906, "step": 6696 }, { "epoch": 3.151529411764706, "grad_norm": 1.1106552626398234, "learning_rate": 0.00017952683877625042, "loss": 1.5, "step": 6697 }, { "epoch": 3.152, "grad_norm": 0.4686023891341013, "learning_rate": 0.00017948385901605106, "loss": 0.1147, "step": 6698 }, { "epoch": 3.152470588235294, "grad_norm": 0.6429523271385863, "learning_rate": 0.0001794408767377855, "loss": 0.8125, "step": 6699 }, { "epoch": 3.152941176470588, "grad_norm": 0.9982719794804792, "learning_rate": 0.00017939789194512472, "loss": 0.6953, "step": 6700 }, { "epoch": 3.1534117647058824, "grad_norm": 1.0680160943825272, "learning_rate": 0.0001793549046417397, "loss": 0.9492, "step": 6701 }, { "epoch": 3.1538823529411766, "grad_norm": 1.2126038438314268, "learning_rate": 0.00017931191483130186, "loss": 1.3281, "step": 6702 }, { "epoch": 3.1543529411764704, "grad_norm": 1.0908316363444122, "learning_rate": 0.00017926892251748274, "loss": 0.9609, "step": 6703 }, { "epoch": 3.1548235294117646, "grad_norm": 1.385351505173085, "learning_rate": 0.00017922592770395405, "loss": 1.8594, "step": 6704 }, { "epoch": 3.155294117647059, "grad_norm": 0.995338215752106, "learning_rate": 0.00017918293039438781, "loss": 0.7031, "step": 6705 }, { "epoch": 3.155764705882353, "grad_norm": 0.8286962409220997, "learning_rate": 0.00017913993059245616, "loss": 0.4375, "step": 6706 }, { "epoch": 3.1562352941176472, "grad_norm": 0.6767287006403574, "learning_rate": 0.00017909692830183152, "loss": 0.6211, "step": 6707 }, { "epoch": 3.156705882352941, "grad_norm": 1.0056106372622862, "learning_rate": 0.00017905392352618647, "loss": 1.3047, "step": 6708 }, { "epoch": 3.1571764705882353, "grad_norm": 1.4727295814945178, "learning_rate": 0.00017901091626919383, "loss": 2.25, "step": 6709 }, { "epoch": 3.1576470588235295, "grad_norm": 0.6307784287940581, "learning_rate": 0.00017896790653452668, "loss": 0.3164, "step": 6710 }, { "epoch": 3.1581176470588237, "grad_norm": 0.9924236635055916, "learning_rate": 0.00017892489432585817, "loss": 1.1953, "step": 6711 }, { "epoch": 3.1585882352941175, "grad_norm": 0.8300999539114745, "learning_rate": 0.00017888187964686182, "loss": 0.4512, "step": 6712 }, { "epoch": 3.1590588235294117, "grad_norm": 1.0667807466163801, "learning_rate": 0.00017883886250121133, "loss": 1.3125, "step": 6713 }, { "epoch": 3.159529411764706, "grad_norm": 0.5008262323062479, "learning_rate": 0.00017879584289258048, "loss": 0.1465, "step": 6714 }, { "epoch": 3.16, "grad_norm": 0.8084834016976968, "learning_rate": 0.00017875282082464343, "loss": 0.6836, "step": 6715 }, { "epoch": 3.160470588235294, "grad_norm": 0.9137419515890198, "learning_rate": 0.00017870979630107446, "loss": 0.5898, "step": 6716 }, { "epoch": 3.160941176470588, "grad_norm": 1.0449514942918188, "learning_rate": 0.00017866676932554802, "loss": 0.9258, "step": 6717 }, { "epoch": 3.1614117647058824, "grad_norm": 0.9534506111436302, "learning_rate": 0.0001786237399017389, "loss": 0.4434, "step": 6718 }, { "epoch": 3.1618823529411766, "grad_norm": 0.9303293640658178, "learning_rate": 0.000178580708033322, "loss": 1.1797, "step": 6719 }, { "epoch": 3.1623529411764704, "grad_norm": 1.2264840246444821, "learning_rate": 0.00017853767372397243, "loss": 1.5625, "step": 6720 }, { "epoch": 3.1628235294117646, "grad_norm": 1.1211749750701765, "learning_rate": 0.00017849463697736554, "loss": 1.0391, "step": 6721 }, { "epoch": 3.163294117647059, "grad_norm": 0.6348517320435846, "learning_rate": 0.0001784515977971769, "loss": 0.3242, "step": 6722 }, { "epoch": 3.163764705882353, "grad_norm": 0.7268723933193182, "learning_rate": 0.00017840855618708227, "loss": 0.3828, "step": 6723 }, { "epoch": 3.1642352941176473, "grad_norm": 0.6264464691658328, "learning_rate": 0.00017836551215075758, "loss": 0.2773, "step": 6724 }, { "epoch": 3.164705882352941, "grad_norm": 0.8318999341085171, "learning_rate": 0.000178322465691879, "loss": 1.0469, "step": 6725 }, { "epoch": 3.1651764705882353, "grad_norm": 0.7762561311082666, "learning_rate": 0.00017827941681412298, "loss": 0.5703, "step": 6726 }, { "epoch": 3.1656470588235295, "grad_norm": 0.8870694510203153, "learning_rate": 0.00017823636552116603, "loss": 1.1484, "step": 6727 }, { "epoch": 3.1661176470588237, "grad_norm": 1.094095878253502, "learning_rate": 0.00017819331181668496, "loss": 1.3281, "step": 6728 }, { "epoch": 3.1665882352941175, "grad_norm": 1.1412960938066912, "learning_rate": 0.00017815025570435677, "loss": 1.3281, "step": 6729 }, { "epoch": 3.1670588235294117, "grad_norm": 1.1810414541960832, "learning_rate": 0.0001781071971878587, "loss": 1.3359, "step": 6730 }, { "epoch": 3.167529411764706, "grad_norm": 0.8542668641722835, "learning_rate": 0.00017806413627086807, "loss": 1.3125, "step": 6731 }, { "epoch": 3.168, "grad_norm": 0.74257612695271, "learning_rate": 0.0001780210729570626, "loss": 0.3555, "step": 6732 }, { "epoch": 3.168470588235294, "grad_norm": 0.9104544089165979, "learning_rate": 0.00017797800725012002, "loss": 1.0703, "step": 6733 }, { "epoch": 3.168941176470588, "grad_norm": 0.9963330712110081, "learning_rate": 0.00017793493915371842, "loss": 0.9219, "step": 6734 }, { "epoch": 3.1694117647058824, "grad_norm": 0.5277354810026398, "learning_rate": 0.00017789186867153594, "loss": 0.1924, "step": 6735 }, { "epoch": 3.1698823529411766, "grad_norm": 1.3036282165250759, "learning_rate": 0.00017784879580725113, "loss": 0.9688, "step": 6736 }, { "epoch": 3.1703529411764704, "grad_norm": 1.7635314631599308, "learning_rate": 0.00017780572056454256, "loss": 1.1172, "step": 6737 }, { "epoch": 3.1708235294117646, "grad_norm": 1.14288125715004, "learning_rate": 0.00017776264294708905, "loss": 0.8711, "step": 6738 }, { "epoch": 3.171294117647059, "grad_norm": 0.5026998539838653, "learning_rate": 0.00017771956295856964, "loss": 0.1147, "step": 6739 }, { "epoch": 3.171764705882353, "grad_norm": 1.246425731201086, "learning_rate": 0.0001776764806026636, "loss": 2.25, "step": 6740 }, { "epoch": 3.1722352941176473, "grad_norm": 0.9656665212638817, "learning_rate": 0.00017763339588305037, "loss": 0.7539, "step": 6741 }, { "epoch": 3.172705882352941, "grad_norm": 0.8197423163751779, "learning_rate": 0.00017759030880340956, "loss": 0.5039, "step": 6742 }, { "epoch": 3.1731764705882353, "grad_norm": 0.9828317362756549, "learning_rate": 0.00017754721936742108, "loss": 0.7578, "step": 6743 }, { "epoch": 3.1736470588235295, "grad_norm": 0.8588034376138921, "learning_rate": 0.00017750412757876492, "loss": 0.7461, "step": 6744 }, { "epoch": 3.1741176470588237, "grad_norm": 0.6074093881408374, "learning_rate": 0.00017746103344112136, "loss": 0.2393, "step": 6745 }, { "epoch": 3.1745882352941175, "grad_norm": 1.1492743561255203, "learning_rate": 0.00017741793695817093, "loss": 1.3594, "step": 6746 }, { "epoch": 3.1750588235294117, "grad_norm": 0.38332382043042534, "learning_rate": 0.00017737483813359413, "loss": 0.0762, "step": 6747 }, { "epoch": 3.175529411764706, "grad_norm": 1.0158813247990615, "learning_rate": 0.00017733173697107187, "loss": 0.6797, "step": 6748 }, { "epoch": 3.176, "grad_norm": 0.9082126429340818, "learning_rate": 0.00017728863347428524, "loss": 0.6172, "step": 6749 }, { "epoch": 3.176470588235294, "grad_norm": 0.5235469769569621, "learning_rate": 0.00017724552764691545, "loss": 0.3398, "step": 6750 }, { "epoch": 3.176941176470588, "grad_norm": 0.9773905932657521, "learning_rate": 0.000177202419492644, "loss": 0.6719, "step": 6751 }, { "epoch": 3.1774117647058824, "grad_norm": 0.7281073316532602, "learning_rate": 0.00017715930901515242, "loss": 0.2559, "step": 6752 }, { "epoch": 3.1778823529411766, "grad_norm": 0.7341465037512935, "learning_rate": 0.0001771161962181227, "loss": 1.0, "step": 6753 }, { "epoch": 3.1783529411764704, "grad_norm": 1.0022909585940878, "learning_rate": 0.0001770730811052368, "loss": 1.2891, "step": 6754 }, { "epoch": 3.1788235294117646, "grad_norm": 0.8420510501640359, "learning_rate": 0.00017702996368017693, "loss": 1.1484, "step": 6755 }, { "epoch": 3.179294117647059, "grad_norm": 0.77160770345113, "learning_rate": 0.00017698684394662563, "loss": 0.3848, "step": 6756 }, { "epoch": 3.179764705882353, "grad_norm": 0.5937812726924939, "learning_rate": 0.0001769437219082655, "loss": 0.1885, "step": 6757 }, { "epoch": 3.1802352941176473, "grad_norm": 0.8162522498661573, "learning_rate": 0.00017690059756877933, "loss": 0.3926, "step": 6758 }, { "epoch": 3.180705882352941, "grad_norm": 1.1923284551664761, "learning_rate": 0.00017685747093185017, "loss": 1.0156, "step": 6759 }, { "epoch": 3.1811764705882353, "grad_norm": 1.366165234154034, "learning_rate": 0.00017681434200116132, "loss": 0.7812, "step": 6760 }, { "epoch": 3.1816470588235295, "grad_norm": 1.2894473762456087, "learning_rate": 0.0001767712107803961, "loss": 1.4375, "step": 6761 }, { "epoch": 3.1821176470588237, "grad_norm": 0.8533430980230968, "learning_rate": 0.00017672807727323818, "loss": 0.75, "step": 6762 }, { "epoch": 3.1825882352941175, "grad_norm": 1.034294239768554, "learning_rate": 0.00017668494148337133, "loss": 0.7305, "step": 6763 }, { "epoch": 3.1830588235294117, "grad_norm": 0.7113489279886407, "learning_rate": 0.00017664180341447964, "loss": 0.3867, "step": 6764 }, { "epoch": 3.183529411764706, "grad_norm": 0.7289738835518353, "learning_rate": 0.0001765986630702472, "loss": 0.5117, "step": 6765 }, { "epoch": 3.184, "grad_norm": 1.088723019645059, "learning_rate": 0.00017655552045435846, "loss": 0.9805, "step": 6766 }, { "epoch": 3.184470588235294, "grad_norm": 0.5041908278104603, "learning_rate": 0.00017651237557049808, "loss": 0.1309, "step": 6767 }, { "epoch": 3.184941176470588, "grad_norm": 0.8990613221656903, "learning_rate": 0.00017646922842235076, "loss": 1.3125, "step": 6768 }, { "epoch": 3.1854117647058824, "grad_norm": 1.2868738039718113, "learning_rate": 0.00017642607901360147, "loss": 1.5391, "step": 6769 }, { "epoch": 3.1858823529411766, "grad_norm": 1.1629631405889727, "learning_rate": 0.00017638292734793547, "loss": 1.2734, "step": 6770 }, { "epoch": 3.1863529411764704, "grad_norm": 1.203307496417627, "learning_rate": 0.00017633977342903807, "loss": 0.9062, "step": 6771 }, { "epoch": 3.1868235294117646, "grad_norm": 0.3396221394473925, "learning_rate": 0.0001762966172605948, "loss": 0.0679, "step": 6772 }, { "epoch": 3.187294117647059, "grad_norm": 0.8227763188272392, "learning_rate": 0.00017625345884629141, "loss": 0.5039, "step": 6773 }, { "epoch": 3.187764705882353, "grad_norm": 0.4261855938891174, "learning_rate": 0.00017621029818981387, "loss": 0.0781, "step": 6774 }, { "epoch": 3.1882352941176473, "grad_norm": 0.7013231323433107, "learning_rate": 0.00017616713529484835, "loss": 0.4277, "step": 6775 }, { "epoch": 3.188705882352941, "grad_norm": 1.2270986583324064, "learning_rate": 0.00017612397016508104, "loss": 1.1094, "step": 6776 }, { "epoch": 3.1891764705882353, "grad_norm": 0.7388880374449013, "learning_rate": 0.0001760808028041986, "loss": 1.0625, "step": 6777 }, { "epoch": 3.1896470588235295, "grad_norm": 0.6055562795534953, "learning_rate": 0.0001760376332158877, "loss": 0.2109, "step": 6778 }, { "epoch": 3.1901176470588237, "grad_norm": 1.2128114178928466, "learning_rate": 0.00017599446140383512, "loss": 1.7578, "step": 6779 }, { "epoch": 3.1905882352941175, "grad_norm": 0.953251127374295, "learning_rate": 0.00017595128737172812, "loss": 1.0234, "step": 6780 }, { "epoch": 3.1910588235294117, "grad_norm": 1.9096119230687028, "learning_rate": 0.00017590811112325384, "loss": 0.9414, "step": 6781 }, { "epoch": 3.191529411764706, "grad_norm": 0.9997165640590238, "learning_rate": 0.00017586493266209978, "loss": 0.9141, "step": 6782 }, { "epoch": 3.192, "grad_norm": 0.9605700711428131, "learning_rate": 0.00017582175199195365, "loss": 0.6211, "step": 6783 }, { "epoch": 3.192470588235294, "grad_norm": 1.1598270745621435, "learning_rate": 0.00017577856911650324, "loss": 1.3906, "step": 6784 }, { "epoch": 3.192941176470588, "grad_norm": 1.0946660205382315, "learning_rate": 0.00017573538403943656, "loss": 1.125, "step": 6785 }, { "epoch": 3.1934117647058824, "grad_norm": 0.7838047322951178, "learning_rate": 0.00017569219676444187, "loss": 0.4297, "step": 6786 }, { "epoch": 3.1938823529411766, "grad_norm": 0.8319887485549939, "learning_rate": 0.00017564900729520751, "loss": 1.3828, "step": 6787 }, { "epoch": 3.1943529411764704, "grad_norm": 1.1542647644466029, "learning_rate": 0.0001756058156354222, "loss": 1.25, "step": 6788 }, { "epoch": 3.1948235294117646, "grad_norm": 0.823259950631156, "learning_rate": 0.00017556262178877459, "loss": 0.4102, "step": 6789 }, { "epoch": 3.195294117647059, "grad_norm": 1.5667336073258276, "learning_rate": 0.00017551942575895368, "loss": 0.9805, "step": 6790 }, { "epoch": 3.195764705882353, "grad_norm": 0.9364535986745323, "learning_rate": 0.00017547622754964868, "loss": 0.8359, "step": 6791 }, { "epoch": 3.1962352941176473, "grad_norm": 1.3327785715663023, "learning_rate": 0.0001754330271645489, "loss": 1.3359, "step": 6792 }, { "epoch": 3.196705882352941, "grad_norm": 0.5701563672239157, "learning_rate": 0.00017538982460734375, "loss": 0.5352, "step": 6793 }, { "epoch": 3.1971764705882353, "grad_norm": 1.3702012445004266, "learning_rate": 0.00017534661988172315, "loss": 0.8477, "step": 6794 }, { "epoch": 3.1976470588235295, "grad_norm": 1.2582523126061198, "learning_rate": 0.00017530341299137685, "loss": 1.3516, "step": 6795 }, { "epoch": 3.1981176470588237, "grad_norm": 1.0262503768855147, "learning_rate": 0.00017526020393999496, "loss": 1.4219, "step": 6796 }, { "epoch": 3.1985882352941175, "grad_norm": 0.9747744729182581, "learning_rate": 0.00017521699273126778, "loss": 0.9258, "step": 6797 }, { "epoch": 3.1990588235294117, "grad_norm": 0.8497449090288428, "learning_rate": 0.00017517377936888568, "loss": 0.6328, "step": 6798 }, { "epoch": 3.199529411764706, "grad_norm": 0.8394531281943044, "learning_rate": 0.0001751305638565394, "loss": 0.543, "step": 6799 }, { "epoch": 3.2, "grad_norm": 0.44261722668174885, "learning_rate": 0.00017508734619791966, "loss": 0.0825, "step": 6800 }, { "epoch": 3.200470588235294, "grad_norm": 1.0457787195477946, "learning_rate": 0.00017504412639671754, "loss": 0.6836, "step": 6801 }, { "epoch": 3.200941176470588, "grad_norm": 0.6965867689973966, "learning_rate": 0.00017500090445662415, "loss": 0.5156, "step": 6802 }, { "epoch": 3.2014117647058824, "grad_norm": 0.5144682430226261, "learning_rate": 0.0001749576803813309, "loss": 0.2617, "step": 6803 }, { "epoch": 3.2018823529411766, "grad_norm": 0.8218465069189599, "learning_rate": 0.00017491445417452934, "loss": 0.75, "step": 6804 }, { "epoch": 3.2023529411764704, "grad_norm": 0.7275643769522253, "learning_rate": 0.00017487122583991123, "loss": 0.2598, "step": 6805 }, { "epoch": 3.2028235294117646, "grad_norm": 1.0394938894964318, "learning_rate": 0.00017482799538116834, "loss": 0.9883, "step": 6806 }, { "epoch": 3.203294117647059, "grad_norm": 0.7909408172489423, "learning_rate": 0.00017478476280199294, "loss": 0.4355, "step": 6807 }, { "epoch": 3.203764705882353, "grad_norm": 1.1514878392198364, "learning_rate": 0.0001747415281060772, "loss": 1.1172, "step": 6808 }, { "epoch": 3.2042352941176473, "grad_norm": 1.0137923746996875, "learning_rate": 0.00017469829129711361, "loss": 0.4707, "step": 6809 }, { "epoch": 3.204705882352941, "grad_norm": 0.7103744371239065, "learning_rate": 0.00017465505237879476, "loss": 0.2227, "step": 6810 }, { "epoch": 3.2051764705882353, "grad_norm": 0.6708525489709528, "learning_rate": 0.00017461181135481355, "loss": 0.293, "step": 6811 }, { "epoch": 3.2056470588235295, "grad_norm": 1.1328452391833403, "learning_rate": 0.00017456856822886297, "loss": 0.8555, "step": 6812 }, { "epoch": 3.2061176470588233, "grad_norm": 0.736748037655369, "learning_rate": 0.00017452532300463609, "loss": 0.5312, "step": 6813 }, { "epoch": 3.2065882352941175, "grad_norm": 0.8923074160774761, "learning_rate": 0.00017448207568582634, "loss": 0.6523, "step": 6814 }, { "epoch": 3.2070588235294117, "grad_norm": 0.9000195272677995, "learning_rate": 0.00017443882627612726, "loss": 0.8047, "step": 6815 }, { "epoch": 3.207529411764706, "grad_norm": 0.9960080799574861, "learning_rate": 0.00017439557477923255, "loss": 0.5469, "step": 6816 }, { "epoch": 3.208, "grad_norm": 1.6725181219960483, "learning_rate": 0.00017435232119883604, "loss": 1.6328, "step": 6817 }, { "epoch": 3.208470588235294, "grad_norm": 0.6536969443778804, "learning_rate": 0.0001743090655386319, "loss": 0.3984, "step": 6818 }, { "epoch": 3.208941176470588, "grad_norm": 0.9200144650916544, "learning_rate": 0.00017426580780231435, "loss": 0.9141, "step": 6819 }, { "epoch": 3.2094117647058824, "grad_norm": 1.1249314092519602, "learning_rate": 0.0001742225479935777, "loss": 0.7852, "step": 6820 }, { "epoch": 3.2098823529411766, "grad_norm": 1.2522913067650785, "learning_rate": 0.00017417928611611675, "loss": 0.8398, "step": 6821 }, { "epoch": 3.2103529411764704, "grad_norm": 1.2581551090097507, "learning_rate": 0.0001741360221736261, "loss": 1.5781, "step": 6822 }, { "epoch": 3.2108235294117646, "grad_norm": 0.9647310846404518, "learning_rate": 0.0001740927561698008, "loss": 1.1641, "step": 6823 }, { "epoch": 3.211294117647059, "grad_norm": 0.9573093302062701, "learning_rate": 0.00017404948810833594, "loss": 1.0078, "step": 6824 }, { "epoch": 3.211764705882353, "grad_norm": 0.9944101573719338, "learning_rate": 0.0001740062179929268, "loss": 0.6836, "step": 6825 }, { "epoch": 3.212235294117647, "grad_norm": 0.794418367982676, "learning_rate": 0.00017396294582726896, "loss": 0.4375, "step": 6826 }, { "epoch": 3.212705882352941, "grad_norm": 1.1567715689302676, "learning_rate": 0.00017391967161505795, "loss": 1.1797, "step": 6827 }, { "epoch": 3.2131764705882353, "grad_norm": 0.33858573439520306, "learning_rate": 0.00017387639535998967, "loss": 0.0791, "step": 6828 }, { "epoch": 3.2136470588235295, "grad_norm": 0.7002322675765461, "learning_rate": 0.00017383311706576016, "loss": 0.6602, "step": 6829 }, { "epoch": 3.2141176470588233, "grad_norm": 0.8036658773633553, "learning_rate": 0.00017378983673606548, "loss": 0.8594, "step": 6830 }, { "epoch": 3.2145882352941175, "grad_norm": 0.6500355111158821, "learning_rate": 0.00017374655437460206, "loss": 0.3027, "step": 6831 }, { "epoch": 3.2150588235294117, "grad_norm": 0.7235017267266712, "learning_rate": 0.00017370326998506646, "loss": 0.6602, "step": 6832 }, { "epoch": 3.215529411764706, "grad_norm": 1.003973278383388, "learning_rate": 0.00017365998357115535, "loss": 0.8633, "step": 6833 }, { "epoch": 3.216, "grad_norm": 0.9742360658485354, "learning_rate": 0.00017361669513656558, "loss": 1.3828, "step": 6834 }, { "epoch": 3.216470588235294, "grad_norm": 1.0295360098639217, "learning_rate": 0.0001735734046849942, "loss": 1.2656, "step": 6835 }, { "epoch": 3.216941176470588, "grad_norm": 1.1667739456635087, "learning_rate": 0.0001735301122201384, "loss": 0.8867, "step": 6836 }, { "epoch": 3.2174117647058824, "grad_norm": 1.0452424060664545, "learning_rate": 0.00017348681774569563, "loss": 1.0078, "step": 6837 }, { "epoch": 3.2178823529411766, "grad_norm": 0.9552651122848526, "learning_rate": 0.00017344352126536343, "loss": 1.0625, "step": 6838 }, { "epoch": 3.2183529411764704, "grad_norm": 1.1093324785472065, "learning_rate": 0.00017340022278283953, "loss": 1.0469, "step": 6839 }, { "epoch": 3.2188235294117646, "grad_norm": 0.8586382276352903, "learning_rate": 0.00017335692230182183, "loss": 0.9414, "step": 6840 }, { "epoch": 3.219294117647059, "grad_norm": 1.064518333305457, "learning_rate": 0.00017331361982600836, "loss": 0.6289, "step": 6841 }, { "epoch": 3.219764705882353, "grad_norm": 1.4478653786111706, "learning_rate": 0.00017327031535909745, "loss": 1.4609, "step": 6842 }, { "epoch": 3.220235294117647, "grad_norm": 1.1564963489878122, "learning_rate": 0.00017322700890478744, "loss": 1.6953, "step": 6843 }, { "epoch": 3.220705882352941, "grad_norm": 1.3302515825789365, "learning_rate": 0.00017318370046677693, "loss": 1.4297, "step": 6844 }, { "epoch": 3.2211764705882353, "grad_norm": 1.013051152205503, "learning_rate": 0.00017314039004876473, "loss": 0.9023, "step": 6845 }, { "epoch": 3.2216470588235295, "grad_norm": 0.6399781024458497, "learning_rate": 0.0001730970776544497, "loss": 0.2334, "step": 6846 }, { "epoch": 3.2221176470588233, "grad_norm": 0.549261678358139, "learning_rate": 0.00017305376328753092, "loss": 0.4297, "step": 6847 }, { "epoch": 3.2225882352941175, "grad_norm": 1.1606748506832518, "learning_rate": 0.00017301044695170769, "loss": 0.9648, "step": 6848 }, { "epoch": 3.2230588235294118, "grad_norm": 1.0107310409700154, "learning_rate": 0.00017296712865067944, "loss": 0.6523, "step": 6849 }, { "epoch": 3.223529411764706, "grad_norm": 0.40914530123212534, "learning_rate": 0.00017292380838814577, "loss": 0.0928, "step": 6850 }, { "epoch": 3.224, "grad_norm": 0.8384299673906803, "learning_rate": 0.00017288048616780636, "loss": 0.4277, "step": 6851 }, { "epoch": 3.224470588235294, "grad_norm": 1.156076433509624, "learning_rate": 0.00017283716199336123, "loss": 1.0156, "step": 6852 }, { "epoch": 3.224941176470588, "grad_norm": 0.9756872463313953, "learning_rate": 0.0001727938358685105, "loss": 1.25, "step": 6853 }, { "epoch": 3.2254117647058824, "grad_norm": 0.9445334053299586, "learning_rate": 0.0001727505077969543, "loss": 0.8125, "step": 6854 }, { "epoch": 3.2258823529411766, "grad_norm": 0.954160982468654, "learning_rate": 0.00017270717778239316, "loss": 1.2812, "step": 6855 }, { "epoch": 3.2263529411764704, "grad_norm": 0.7357701618655285, "learning_rate": 0.0001726638458285277, "loss": 0.5859, "step": 6856 }, { "epoch": 3.2268235294117646, "grad_norm": 0.837472131884519, "learning_rate": 0.00017262051193905865, "loss": 0.3906, "step": 6857 }, { "epoch": 3.227294117647059, "grad_norm": 1.0867273538931708, "learning_rate": 0.00017257717611768685, "loss": 0.7852, "step": 6858 }, { "epoch": 3.227764705882353, "grad_norm": 1.2595019288955567, "learning_rate": 0.00017253383836811356, "loss": 1.4844, "step": 6859 }, { "epoch": 3.228235294117647, "grad_norm": 1.043273994383748, "learning_rate": 0.0001724904986940399, "loss": 0.957, "step": 6860 }, { "epoch": 3.228705882352941, "grad_norm": 1.0836833871711995, "learning_rate": 0.00017244715709916733, "loss": 1.0625, "step": 6861 }, { "epoch": 3.2291764705882353, "grad_norm": 1.100475427691386, "learning_rate": 0.00017240381358719746, "loss": 1.3906, "step": 6862 }, { "epoch": 3.2296470588235295, "grad_norm": 0.9843772443997063, "learning_rate": 0.000172360468161832, "loss": 0.7969, "step": 6863 }, { "epoch": 3.2301176470588233, "grad_norm": 0.8155436781262343, "learning_rate": 0.00017231712082677291, "loss": 0.543, "step": 6864 }, { "epoch": 3.2305882352941175, "grad_norm": 0.8278637022553116, "learning_rate": 0.0001722737715857222, "loss": 0.7305, "step": 6865 }, { "epoch": 3.2310588235294118, "grad_norm": 1.0586850970685244, "learning_rate": 0.00017223042044238219, "loss": 0.7891, "step": 6866 }, { "epoch": 3.231529411764706, "grad_norm": 0.8251111772868818, "learning_rate": 0.0001721870674004552, "loss": 0.5859, "step": 6867 }, { "epoch": 3.232, "grad_norm": 0.9771055284130218, "learning_rate": 0.00017214371246364384, "loss": 0.668, "step": 6868 }, { "epoch": 3.232470588235294, "grad_norm": 0.6841186391868761, "learning_rate": 0.00017210035563565082, "loss": 0.291, "step": 6869 }, { "epoch": 3.232941176470588, "grad_norm": 0.7856422148058992, "learning_rate": 0.00017205699692017902, "loss": 0.7812, "step": 6870 }, { "epoch": 3.2334117647058824, "grad_norm": 1.0545981768870436, "learning_rate": 0.0001720136363209315, "loss": 0.6602, "step": 6871 }, { "epoch": 3.2338823529411767, "grad_norm": 0.8745161402129825, "learning_rate": 0.0001719702738416115, "loss": 1.2109, "step": 6872 }, { "epoch": 3.2343529411764704, "grad_norm": 0.875119798960484, "learning_rate": 0.0001719269094859223, "loss": 0.4277, "step": 6873 }, { "epoch": 3.2348235294117647, "grad_norm": 0.5274101632145953, "learning_rate": 0.0001718835432575675, "loss": 0.3926, "step": 6874 }, { "epoch": 3.235294117647059, "grad_norm": 0.8185194653423474, "learning_rate": 0.00017184017516025076, "loss": 0.4199, "step": 6875 }, { "epoch": 3.235764705882353, "grad_norm": 0.8883253832786477, "learning_rate": 0.0001717968051976759, "loss": 0.8828, "step": 6876 }, { "epoch": 3.236235294117647, "grad_norm": 0.8422334389282053, "learning_rate": 0.00017175343337354698, "loss": 0.7812, "step": 6877 }, { "epoch": 3.236705882352941, "grad_norm": 1.243352322704315, "learning_rate": 0.00017171005969156817, "loss": 1.4531, "step": 6878 }, { "epoch": 3.2371764705882353, "grad_norm": 0.5535134688951131, "learning_rate": 0.00017166668415544376, "loss": 0.1982, "step": 6879 }, { "epoch": 3.2376470588235295, "grad_norm": 0.9517201350820008, "learning_rate": 0.0001716233067688783, "loss": 0.9688, "step": 6880 }, { "epoch": 3.2381176470588233, "grad_norm": 1.0375894906123322, "learning_rate": 0.00017157992753557633, "loss": 1.0234, "step": 6881 }, { "epoch": 3.2385882352941175, "grad_norm": 0.9389301758443239, "learning_rate": 0.0001715365464592427, "loss": 1.3438, "step": 6882 }, { "epoch": 3.2390588235294118, "grad_norm": 0.8325794402147355, "learning_rate": 0.00017149316354358237, "loss": 0.7266, "step": 6883 }, { "epoch": 3.239529411764706, "grad_norm": 1.1100297047304688, "learning_rate": 0.00017144977879230046, "loss": 0.7188, "step": 6884 }, { "epoch": 3.24, "grad_norm": 1.0707006278087334, "learning_rate": 0.00017140639220910222, "loss": 0.8945, "step": 6885 }, { "epoch": 3.240470588235294, "grad_norm": 0.6090122195090342, "learning_rate": 0.00017136300379769312, "loss": 0.2676, "step": 6886 }, { "epoch": 3.240941176470588, "grad_norm": 1.0899706159695168, "learning_rate": 0.00017131961356177869, "loss": 1.2969, "step": 6887 }, { "epoch": 3.2414117647058824, "grad_norm": 1.204197726090496, "learning_rate": 0.00017127622150506472, "loss": 1.6953, "step": 6888 }, { "epoch": 3.2418823529411767, "grad_norm": 1.0677608950332091, "learning_rate": 0.000171232827631257, "loss": 1.3594, "step": 6889 }, { "epoch": 3.2423529411764704, "grad_norm": 0.7358233457481309, "learning_rate": 0.00017118943194406169, "loss": 0.5742, "step": 6890 }, { "epoch": 3.2428235294117647, "grad_norm": 0.8435245604179689, "learning_rate": 0.00017114603444718496, "loss": 0.7109, "step": 6891 }, { "epoch": 3.243294117647059, "grad_norm": 1.182509163317527, "learning_rate": 0.00017110263514433317, "loss": 1.5938, "step": 6892 }, { "epoch": 3.243764705882353, "grad_norm": 1.2812708045991323, "learning_rate": 0.00017105923403921283, "loss": 1.3906, "step": 6893 }, { "epoch": 3.244235294117647, "grad_norm": 1.0299084206888347, "learning_rate": 0.00017101583113553063, "loss": 0.5352, "step": 6894 }, { "epoch": 3.244705882352941, "grad_norm": 0.6909003197854267, "learning_rate": 0.00017097242643699333, "loss": 0.7734, "step": 6895 }, { "epoch": 3.2451764705882353, "grad_norm": 1.0153534113617422, "learning_rate": 0.00017092901994730797, "loss": 0.9805, "step": 6896 }, { "epoch": 3.2456470588235296, "grad_norm": 0.9071805394796557, "learning_rate": 0.00017088561167018164, "loss": 0.6016, "step": 6897 }, { "epoch": 3.2461176470588233, "grad_norm": 1.058084063720317, "learning_rate": 0.0001708422016093216, "loss": 1.2891, "step": 6898 }, { "epoch": 3.2465882352941176, "grad_norm": 0.7594710791707051, "learning_rate": 0.00017079878976843533, "loss": 0.3223, "step": 6899 }, { "epoch": 3.2470588235294118, "grad_norm": 1.0361375577340362, "learning_rate": 0.00017075537615123042, "loss": 0.8711, "step": 6900 }, { "epoch": 3.247529411764706, "grad_norm": 0.7550371766727898, "learning_rate": 0.0001707119607614146, "loss": 0.3418, "step": 6901 }, { "epoch": 3.248, "grad_norm": 1.148815287424079, "learning_rate": 0.0001706685436026957, "loss": 1.1172, "step": 6902 }, { "epoch": 3.248470588235294, "grad_norm": 0.9400210919818842, "learning_rate": 0.0001706251246787818, "loss": 1.1562, "step": 6903 }, { "epoch": 3.248941176470588, "grad_norm": 0.9360280606762305, "learning_rate": 0.0001705817039933811, "loss": 0.457, "step": 6904 }, { "epoch": 3.2494117647058824, "grad_norm": 1.125391630579584, "learning_rate": 0.00017053828155020193, "loss": 1.1562, "step": 6905 }, { "epoch": 3.2498823529411767, "grad_norm": 1.2255548546196557, "learning_rate": 0.00017049485735295275, "loss": 1.2344, "step": 6906 }, { "epoch": 3.2503529411764704, "grad_norm": 0.5393385449846396, "learning_rate": 0.00017045143140534226, "loss": 0.1846, "step": 6907 }, { "epoch": 3.2508235294117647, "grad_norm": 0.8038599128217004, "learning_rate": 0.0001704080037110792, "loss": 0.7891, "step": 6908 }, { "epoch": 3.251294117647059, "grad_norm": 1.2431070551867223, "learning_rate": 0.00017036457427387248, "loss": 1.0156, "step": 6909 }, { "epoch": 3.251764705882353, "grad_norm": 0.7464750946581304, "learning_rate": 0.00017032114309743128, "loss": 0.3477, "step": 6910 }, { "epoch": 3.252235294117647, "grad_norm": 0.5190833255300612, "learning_rate": 0.0001702777101854648, "loss": 0.1387, "step": 6911 }, { "epoch": 3.252705882352941, "grad_norm": 1.1958015023849493, "learning_rate": 0.0001702342755416824, "loss": 1.1094, "step": 6912 }, { "epoch": 3.2531764705882353, "grad_norm": 1.1048319797581676, "learning_rate": 0.00017019083916979357, "loss": 1.125, "step": 6913 }, { "epoch": 3.2536470588235296, "grad_norm": 0.8622380036194555, "learning_rate": 0.0001701474010735081, "loss": 0.8672, "step": 6914 }, { "epoch": 3.2541176470588233, "grad_norm": 0.9568073302616743, "learning_rate": 0.00017010396125653572, "loss": 0.8867, "step": 6915 }, { "epoch": 3.2545882352941176, "grad_norm": 1.1767252825002126, "learning_rate": 0.00017006051972258645, "loss": 0.6953, "step": 6916 }, { "epoch": 3.255058823529412, "grad_norm": 0.7795969731605586, "learning_rate": 0.0001700170764753704, "loss": 0.5664, "step": 6917 }, { "epoch": 3.255529411764706, "grad_norm": 1.0075381984135892, "learning_rate": 0.00016997363151859785, "loss": 0.9883, "step": 6918 }, { "epoch": 3.2560000000000002, "grad_norm": 1.117244088439303, "learning_rate": 0.00016993018485597915, "loss": 0.8711, "step": 6919 }, { "epoch": 3.256470588235294, "grad_norm": 1.0993686001988834, "learning_rate": 0.00016988673649122492, "loss": 1.6406, "step": 6920 }, { "epoch": 3.2569411764705882, "grad_norm": 0.7790982055850005, "learning_rate": 0.00016984328642804587, "loss": 1.0156, "step": 6921 }, { "epoch": 3.2574117647058825, "grad_norm": 1.3150818880594424, "learning_rate": 0.00016979983467015284, "loss": 1.4219, "step": 6922 }, { "epoch": 3.2578823529411762, "grad_norm": 1.105581094713546, "learning_rate": 0.00016975638122125678, "loss": 0.9492, "step": 6923 }, { "epoch": 3.2583529411764705, "grad_norm": 1.0143403191076354, "learning_rate": 0.00016971292608506888, "loss": 0.9531, "step": 6924 }, { "epoch": 3.2588235294117647, "grad_norm": 0.7402982109133538, "learning_rate": 0.0001696694692653004, "loss": 0.2295, "step": 6925 }, { "epoch": 3.259294117647059, "grad_norm": 0.7158736318109848, "learning_rate": 0.00016962601076566275, "loss": 0.2754, "step": 6926 }, { "epoch": 3.259764705882353, "grad_norm": 0.8404380731456124, "learning_rate": 0.00016958255058986753, "loss": 0.6602, "step": 6927 }, { "epoch": 3.260235294117647, "grad_norm": 1.2329264179168629, "learning_rate": 0.00016953908874162645, "loss": 0.5312, "step": 6928 }, { "epoch": 3.260705882352941, "grad_norm": 0.7219424738285793, "learning_rate": 0.00016949562522465135, "loss": 0.3418, "step": 6929 }, { "epoch": 3.2611764705882353, "grad_norm": 0.5976090727110389, "learning_rate": 0.00016945216004265418, "loss": 0.4336, "step": 6930 }, { "epoch": 3.2616470588235296, "grad_norm": 0.972508166290649, "learning_rate": 0.0001694086931993472, "loss": 0.7109, "step": 6931 }, { "epoch": 3.2621176470588233, "grad_norm": 0.9116125253130651, "learning_rate": 0.0001693652246984426, "loss": 0.8125, "step": 6932 }, { "epoch": 3.2625882352941176, "grad_norm": 0.8310863047284526, "learning_rate": 0.0001693217545436528, "loss": 1.2344, "step": 6933 }, { "epoch": 3.263058823529412, "grad_norm": 0.612730309399416, "learning_rate": 0.0001692782827386904, "loss": 0.25, "step": 6934 }, { "epoch": 3.263529411764706, "grad_norm": 0.9431938003754389, "learning_rate": 0.00016923480928726813, "loss": 1.2812, "step": 6935 }, { "epoch": 3.2640000000000002, "grad_norm": 0.7253707626477262, "learning_rate": 0.00016919133419309875, "loss": 0.2891, "step": 6936 }, { "epoch": 3.264470588235294, "grad_norm": 0.8994484531583641, "learning_rate": 0.00016914785745989532, "loss": 1.0938, "step": 6937 }, { "epoch": 3.2649411764705882, "grad_norm": 0.6964746754513252, "learning_rate": 0.00016910437909137098, "loss": 0.3496, "step": 6938 }, { "epoch": 3.2654117647058825, "grad_norm": 1.0894529385522966, "learning_rate": 0.00016906089909123892, "loss": 1.1406, "step": 6939 }, { "epoch": 3.2658823529411762, "grad_norm": 0.9838768548595486, "learning_rate": 0.00016901741746321253, "loss": 0.8867, "step": 6940 }, { "epoch": 3.2663529411764705, "grad_norm": 0.44232065796394365, "learning_rate": 0.0001689739342110055, "loss": 0.1602, "step": 6941 }, { "epoch": 3.2668235294117647, "grad_norm": 1.0174673944387904, "learning_rate": 0.00016893044933833135, "loss": 1.5391, "step": 6942 }, { "epoch": 3.267294117647059, "grad_norm": 1.0255220223226469, "learning_rate": 0.00016888696284890402, "loss": 0.9492, "step": 6943 }, { "epoch": 3.267764705882353, "grad_norm": 0.6270283817875315, "learning_rate": 0.0001688434747464374, "loss": 0.1963, "step": 6944 }, { "epoch": 3.268235294117647, "grad_norm": 1.1049115435239034, "learning_rate": 0.00016879998503464561, "loss": 0.9766, "step": 6945 }, { "epoch": 3.268705882352941, "grad_norm": 0.6479401577162198, "learning_rate": 0.00016875649371724292, "loss": 0.3047, "step": 6946 }, { "epoch": 3.2691764705882354, "grad_norm": 0.6389395527189858, "learning_rate": 0.00016871300079794363, "loss": 0.6641, "step": 6947 }, { "epoch": 3.2696470588235296, "grad_norm": 0.7992181769871327, "learning_rate": 0.0001686695062804623, "loss": 0.7734, "step": 6948 }, { "epoch": 3.2701176470588234, "grad_norm": 0.9250653856574923, "learning_rate": 0.00016862601016851358, "loss": 0.5898, "step": 6949 }, { "epoch": 3.2705882352941176, "grad_norm": 1.2400697971052306, "learning_rate": 0.00016858251246581216, "loss": 1.125, "step": 6950 }, { "epoch": 3.271058823529412, "grad_norm": 0.9530027027408237, "learning_rate": 0.0001685390131760731, "loss": 0.9297, "step": 6951 }, { "epoch": 3.271529411764706, "grad_norm": 1.1152693156713234, "learning_rate": 0.00016849551230301135, "loss": 0.5469, "step": 6952 }, { "epoch": 3.2720000000000002, "grad_norm": 1.2651471913551178, "learning_rate": 0.00016845200985034215, "loss": 1.0703, "step": 6953 }, { "epoch": 3.272470588235294, "grad_norm": 1.1659919589445666, "learning_rate": 0.0001684085058217808, "loss": 1.2344, "step": 6954 }, { "epoch": 3.2729411764705882, "grad_norm": 1.7875623724995569, "learning_rate": 0.00016836500022104274, "loss": 1.4922, "step": 6955 }, { "epoch": 3.2734117647058825, "grad_norm": 0.6971029979781856, "learning_rate": 0.00016832149305184363, "loss": 0.2207, "step": 6956 }, { "epoch": 3.2738823529411762, "grad_norm": 0.9696901574048795, "learning_rate": 0.00016827798431789907, "loss": 0.7188, "step": 6957 }, { "epoch": 3.2743529411764705, "grad_norm": 2.0245385819391353, "learning_rate": 0.00016823447402292504, "loss": 0.8594, "step": 6958 }, { "epoch": 3.2748235294117647, "grad_norm": 1.2461164484857592, "learning_rate": 0.00016819096217063752, "loss": 0.9922, "step": 6959 }, { "epoch": 3.275294117647059, "grad_norm": 1.2194819043950031, "learning_rate": 0.00016814744876475252, "loss": 1.5078, "step": 6960 }, { "epoch": 3.275764705882353, "grad_norm": 0.9292288277173928, "learning_rate": 0.00016810393380898645, "loss": 0.9492, "step": 6961 }, { "epoch": 3.276235294117647, "grad_norm": 1.025531103555209, "learning_rate": 0.0001680604173070556, "loss": 1.0703, "step": 6962 }, { "epoch": 3.276705882352941, "grad_norm": 1.011204120712964, "learning_rate": 0.00016801689926267653, "loss": 0.6914, "step": 6963 }, { "epoch": 3.2771764705882354, "grad_norm": 0.78747287699161, "learning_rate": 0.00016797337967956592, "loss": 0.7852, "step": 6964 }, { "epoch": 3.2776470588235296, "grad_norm": 0.699577152155678, "learning_rate": 0.00016792985856144049, "loss": 0.6016, "step": 6965 }, { "epoch": 3.2781176470588234, "grad_norm": 0.748850413172106, "learning_rate": 0.00016788633591201723, "loss": 1.0, "step": 6966 }, { "epoch": 3.2785882352941176, "grad_norm": 1.9622623013798155, "learning_rate": 0.0001678428117350131, "loss": 0.625, "step": 6967 }, { "epoch": 3.279058823529412, "grad_norm": 1.1724829196305044, "learning_rate": 0.0001677992860341454, "loss": 1.0859, "step": 6968 }, { "epoch": 3.279529411764706, "grad_norm": 1.175787571585695, "learning_rate": 0.00016775575881313132, "loss": 1.375, "step": 6969 }, { "epoch": 3.2800000000000002, "grad_norm": 1.2736312491910775, "learning_rate": 0.00016771223007568836, "loss": 0.9766, "step": 6970 }, { "epoch": 3.280470588235294, "grad_norm": 1.0156591153745307, "learning_rate": 0.00016766869982553405, "loss": 1.5312, "step": 6971 }, { "epoch": 3.2809411764705882, "grad_norm": 0.911402172696875, "learning_rate": 0.0001676251680663861, "loss": 0.7227, "step": 6972 }, { "epoch": 3.2814117647058825, "grad_norm": 0.7873244270467685, "learning_rate": 0.0001675816348019624, "loss": 0.3691, "step": 6973 }, { "epoch": 3.2818823529411763, "grad_norm": 0.8438141555632428, "learning_rate": 0.0001675381000359808, "loss": 0.7344, "step": 6974 }, { "epoch": 3.2823529411764705, "grad_norm": 0.9402276685504326, "learning_rate": 0.0001674945637721595, "loss": 0.8555, "step": 6975 }, { "epoch": 3.2828235294117647, "grad_norm": 0.9380029997803432, "learning_rate": 0.00016745102601421663, "loss": 0.8125, "step": 6976 }, { "epoch": 3.283294117647059, "grad_norm": 0.7231711742652407, "learning_rate": 0.00016740748676587045, "loss": 0.6562, "step": 6977 }, { "epoch": 3.283764705882353, "grad_norm": 0.6547412310360622, "learning_rate": 0.0001673639460308396, "loss": 0.6289, "step": 6978 }, { "epoch": 3.284235294117647, "grad_norm": 0.9285987008784313, "learning_rate": 0.00016732040381284258, "loss": 0.6055, "step": 6979 }, { "epoch": 3.284705882352941, "grad_norm": 1.0562707739535862, "learning_rate": 0.00016727686011559813, "loss": 0.6602, "step": 6980 }, { "epoch": 3.2851764705882354, "grad_norm": 1.0958680173479798, "learning_rate": 0.00016723331494282502, "loss": 1.4844, "step": 6981 }, { "epoch": 3.2856470588235296, "grad_norm": 1.1710146801674928, "learning_rate": 0.00016718976829824234, "loss": 1.5234, "step": 6982 }, { "epoch": 3.2861176470588234, "grad_norm": 0.9741146915186746, "learning_rate": 0.00016714622018556915, "loss": 1.0391, "step": 6983 }, { "epoch": 3.2865882352941176, "grad_norm": 0.6292959878425286, "learning_rate": 0.0001671026706085246, "loss": 0.5078, "step": 6984 }, { "epoch": 3.287058823529412, "grad_norm": 0.8074561606105299, "learning_rate": 0.0001670591195708281, "loss": 0.3008, "step": 6985 }, { "epoch": 3.287529411764706, "grad_norm": 1.2377781585952086, "learning_rate": 0.00016701556707619914, "loss": 0.7383, "step": 6986 }, { "epoch": 3.288, "grad_norm": 1.2686720646102858, "learning_rate": 0.00016697201312835727, "loss": 1.0469, "step": 6987 }, { "epoch": 3.288470588235294, "grad_norm": 0.9238485888622975, "learning_rate": 0.00016692845773102222, "loss": 0.9883, "step": 6988 }, { "epoch": 3.2889411764705883, "grad_norm": 1.3919678878984318, "learning_rate": 0.0001668849008879139, "loss": 1.4062, "step": 6989 }, { "epoch": 3.2894117647058825, "grad_norm": 1.0759422885194245, "learning_rate": 0.00016684134260275217, "loss": 1.2188, "step": 6990 }, { "epoch": 3.2898823529411763, "grad_norm": 0.7833282278702478, "learning_rate": 0.0001667977828792572, "loss": 1.1484, "step": 6991 }, { "epoch": 3.2903529411764705, "grad_norm": 0.8791309058936435, "learning_rate": 0.00016675422172114917, "loss": 0.5625, "step": 6992 }, { "epoch": 3.2908235294117647, "grad_norm": 1.1434254910505959, "learning_rate": 0.00016671065913214845, "loss": 1.3203, "step": 6993 }, { "epoch": 3.291294117647059, "grad_norm": 0.6334806475336894, "learning_rate": 0.00016666709511597546, "loss": 0.5312, "step": 6994 }, { "epoch": 3.291764705882353, "grad_norm": 0.8397511677491991, "learning_rate": 0.0001666235296763508, "loss": 0.5, "step": 6995 }, { "epoch": 3.292235294117647, "grad_norm": 1.026373903062254, "learning_rate": 0.00016657996281699514, "loss": 0.7773, "step": 6996 }, { "epoch": 3.292705882352941, "grad_norm": 0.965515560059472, "learning_rate": 0.00016653639454162938, "loss": 0.8633, "step": 6997 }, { "epoch": 3.2931764705882354, "grad_norm": 0.5761640943658217, "learning_rate": 0.00016649282485397437, "loss": 0.21, "step": 6998 }, { "epoch": 3.2936470588235296, "grad_norm": 1.428104703228059, "learning_rate": 0.00016644925375775124, "loss": 0.9805, "step": 6999 }, { "epoch": 3.2941176470588234, "grad_norm": 0.723442002011222, "learning_rate": 0.00016640568125668117, "loss": 0.293, "step": 7000 }, { "epoch": 3.2945882352941176, "grad_norm": 1.0461806599088914, "learning_rate": 0.00016636210735448539, "loss": 0.7539, "step": 7001 }, { "epoch": 3.295058823529412, "grad_norm": 0.9186660670654947, "learning_rate": 0.00016631853205488546, "loss": 0.9648, "step": 7002 }, { "epoch": 3.295529411764706, "grad_norm": 1.0571097257908664, "learning_rate": 0.00016627495536160284, "loss": 1.1953, "step": 7003 }, { "epoch": 3.296, "grad_norm": 0.9172372869999549, "learning_rate": 0.0001662313772783592, "loss": 1.4531, "step": 7004 }, { "epoch": 3.296470588235294, "grad_norm": 0.9248819929172795, "learning_rate": 0.00016618779780887627, "loss": 1.1797, "step": 7005 }, { "epoch": 3.2969411764705883, "grad_norm": 0.6636538716496193, "learning_rate": 0.00016614421695687606, "loss": 0.1816, "step": 7006 }, { "epoch": 3.2974117647058825, "grad_norm": 0.947691755377726, "learning_rate": 0.00016610063472608053, "loss": 0.7227, "step": 7007 }, { "epoch": 3.2978823529411763, "grad_norm": 0.6335439368116482, "learning_rate": 0.0001660570511202118, "loss": 0.2949, "step": 7008 }, { "epoch": 3.2983529411764705, "grad_norm": 0.9589377878440362, "learning_rate": 0.00016601346614299214, "loss": 0.5195, "step": 7009 }, { "epoch": 3.2988235294117647, "grad_norm": 0.7078193766899186, "learning_rate": 0.00016596987979814395, "loss": 0.3574, "step": 7010 }, { "epoch": 3.299294117647059, "grad_norm": 0.6889899514610486, "learning_rate": 0.0001659262920893897, "loss": 0.2275, "step": 7011 }, { "epoch": 3.299764705882353, "grad_norm": 1.043937453847103, "learning_rate": 0.00016588270302045197, "loss": 1.4062, "step": 7012 }, { "epoch": 3.300235294117647, "grad_norm": 0.5418441204648201, "learning_rate": 0.00016583911259505352, "loss": 0.2617, "step": 7013 }, { "epoch": 3.300705882352941, "grad_norm": 0.7489137861886428, "learning_rate": 0.00016579552081691715, "loss": 0.3105, "step": 7014 }, { "epoch": 3.3011764705882354, "grad_norm": 0.8307098559109353, "learning_rate": 0.0001657519276897658, "loss": 1.0156, "step": 7015 }, { "epoch": 3.3016470588235296, "grad_norm": 0.9860117557037957, "learning_rate": 0.0001657083332173226, "loss": 0.75, "step": 7016 }, { "epoch": 3.3021176470588234, "grad_norm": 0.9911681143244896, "learning_rate": 0.0001656647374033107, "loss": 0.6055, "step": 7017 }, { "epoch": 3.3025882352941176, "grad_norm": 0.6301131695623722, "learning_rate": 0.00016562114025145332, "loss": 0.3477, "step": 7018 }, { "epoch": 3.303058823529412, "grad_norm": 1.0707162318453733, "learning_rate": 0.00016557754176547404, "loss": 1.0625, "step": 7019 }, { "epoch": 3.303529411764706, "grad_norm": 0.5069374547523919, "learning_rate": 0.0001655339419490962, "loss": 0.4785, "step": 7020 }, { "epoch": 3.304, "grad_norm": 1.3474752081012609, "learning_rate": 0.00016549034080604359, "loss": 1.7266, "step": 7021 }, { "epoch": 3.304470588235294, "grad_norm": 1.0347624050361637, "learning_rate": 0.00016544673834003985, "loss": 0.6836, "step": 7022 }, { "epoch": 3.3049411764705883, "grad_norm": 0.5062444610495632, "learning_rate": 0.00016540313455480895, "loss": 0.1738, "step": 7023 }, { "epoch": 3.3054117647058825, "grad_norm": 0.5769591846406443, "learning_rate": 0.00016535952945407473, "loss": 0.2949, "step": 7024 }, { "epoch": 3.3058823529411763, "grad_norm": 1.9176050821031965, "learning_rate": 0.00016531592304156137, "loss": 1.2656, "step": 7025 }, { "epoch": 3.3063529411764705, "grad_norm": 0.8316612772403391, "learning_rate": 0.0001652723153209931, "loss": 1.1953, "step": 7026 }, { "epoch": 3.3068235294117647, "grad_norm": 0.6765538894767344, "learning_rate": 0.0001652287062960942, "loss": 0.2383, "step": 7027 }, { "epoch": 3.307294117647059, "grad_norm": 0.9606048606778507, "learning_rate": 0.00016518509597058902, "loss": 0.4473, "step": 7028 }, { "epoch": 3.307764705882353, "grad_norm": 0.8966789433458089, "learning_rate": 0.0001651414843482022, "loss": 0.6445, "step": 7029 }, { "epoch": 3.308235294117647, "grad_norm": 0.9736426810440019, "learning_rate": 0.00016509787143265836, "loss": 1.1875, "step": 7030 }, { "epoch": 3.308705882352941, "grad_norm": 0.939915979975969, "learning_rate": 0.0001650542572276822, "loss": 1.0156, "step": 7031 }, { "epoch": 3.3091764705882354, "grad_norm": 0.6847496431825888, "learning_rate": 0.00016501064173699866, "loss": 0.2812, "step": 7032 }, { "epoch": 3.309647058823529, "grad_norm": 0.9518688509126152, "learning_rate": 0.0001649670249643327, "loss": 0.625, "step": 7033 }, { "epoch": 3.3101176470588234, "grad_norm": 1.172431840238526, "learning_rate": 0.0001649234069134094, "loss": 1.2891, "step": 7034 }, { "epoch": 3.3105882352941176, "grad_norm": 0.8482349486961778, "learning_rate": 0.00016487978758795397, "loss": 0.4336, "step": 7035 }, { "epoch": 3.311058823529412, "grad_norm": 0.968232825430674, "learning_rate": 0.00016483616699169163, "loss": 0.9844, "step": 7036 }, { "epoch": 3.311529411764706, "grad_norm": 1.0898327110298176, "learning_rate": 0.0001647925451283479, "loss": 1.25, "step": 7037 }, { "epoch": 3.312, "grad_norm": 0.9071717012397518, "learning_rate": 0.00016474892200164825, "loss": 0.6602, "step": 7038 }, { "epoch": 3.312470588235294, "grad_norm": 0.9947506699711928, "learning_rate": 0.0001647052976153183, "loss": 0.5859, "step": 7039 }, { "epoch": 3.3129411764705883, "grad_norm": 1.0019345922072453, "learning_rate": 0.00016466167197308384, "loss": 0.8008, "step": 7040 }, { "epoch": 3.3134117647058825, "grad_norm": 1.0727907272865371, "learning_rate": 0.00016461804507867067, "loss": 1.4141, "step": 7041 }, { "epoch": 3.3138823529411763, "grad_norm": 0.856474096691168, "learning_rate": 0.00016457441693580472, "loss": 0.5977, "step": 7042 }, { "epoch": 3.3143529411764705, "grad_norm": 1.0003623752988582, "learning_rate": 0.0001645307875482121, "loss": 0.8594, "step": 7043 }, { "epoch": 3.3148235294117647, "grad_norm": 1.0520026466480432, "learning_rate": 0.00016448715691961898, "loss": 1.1641, "step": 7044 }, { "epoch": 3.315294117647059, "grad_norm": 1.3239429300111072, "learning_rate": 0.00016444352505375155, "loss": 1.5078, "step": 7045 }, { "epoch": 3.315764705882353, "grad_norm": 0.7289656741172188, "learning_rate": 0.00016439989195433626, "loss": 0.1816, "step": 7046 }, { "epoch": 3.316235294117647, "grad_norm": 0.6449692272857303, "learning_rate": 0.00016435625762509955, "loss": 0.4277, "step": 7047 }, { "epoch": 3.316705882352941, "grad_norm": 0.7344832915771468, "learning_rate": 0.00016431262206976805, "loss": 0.3379, "step": 7048 }, { "epoch": 3.3171764705882354, "grad_norm": 1.3497210388962353, "learning_rate": 0.00016426898529206836, "loss": 1.6641, "step": 7049 }, { "epoch": 3.317647058823529, "grad_norm": 1.0201544725388028, "learning_rate": 0.00016422534729572738, "loss": 1.4609, "step": 7050 }, { "epoch": 3.3181176470588234, "grad_norm": 0.818737631698472, "learning_rate": 0.00016418170808447198, "loss": 0.4395, "step": 7051 }, { "epoch": 3.3185882352941176, "grad_norm": 0.6943277117441937, "learning_rate": 0.00016413806766202912, "loss": 0.2793, "step": 7052 }, { "epoch": 3.319058823529412, "grad_norm": 0.6040427837098735, "learning_rate": 0.00016409442603212594, "loss": 0.3223, "step": 7053 }, { "epoch": 3.319529411764706, "grad_norm": 0.9797761475527803, "learning_rate": 0.00016405078319848965, "loss": 1.4141, "step": 7054 }, { "epoch": 3.32, "grad_norm": 1.0253265937397593, "learning_rate": 0.0001640071391648476, "loss": 0.6523, "step": 7055 }, { "epoch": 3.320470588235294, "grad_norm": 0.6017878860656745, "learning_rate": 0.0001639634939349271, "loss": 0.2227, "step": 7056 }, { "epoch": 3.3209411764705883, "grad_norm": 0.7594419315352212, "learning_rate": 0.00016391984751245572, "loss": 0.6133, "step": 7057 }, { "epoch": 3.3214117647058825, "grad_norm": 0.32228073265491536, "learning_rate": 0.00016387619990116116, "loss": 0.4844, "step": 7058 }, { "epoch": 3.3218823529411763, "grad_norm": 1.0419553902736352, "learning_rate": 0.00016383255110477102, "loss": 1.2891, "step": 7059 }, { "epoch": 3.3223529411764705, "grad_norm": 1.036888181891881, "learning_rate": 0.00016378890112701312, "loss": 1.5078, "step": 7060 }, { "epoch": 3.3228235294117647, "grad_norm": 1.0276894746105874, "learning_rate": 0.0001637452499716155, "loss": 0.668, "step": 7061 }, { "epoch": 3.323294117647059, "grad_norm": 0.24514532544388484, "learning_rate": 0.0001637015976423061, "loss": 0.0189, "step": 7062 }, { "epoch": 3.323764705882353, "grad_norm": 0.7867923143318494, "learning_rate": 0.00016365794414281303, "loss": 0.6016, "step": 7063 }, { "epoch": 3.324235294117647, "grad_norm": 1.139503624556628, "learning_rate": 0.00016361428947686457, "loss": 1.6719, "step": 7064 }, { "epoch": 3.324705882352941, "grad_norm": 1.032365280813308, "learning_rate": 0.00016357063364818902, "loss": 1.2578, "step": 7065 }, { "epoch": 3.3251764705882354, "grad_norm": 0.44972337341862745, "learning_rate": 0.00016352697666051474, "loss": 0.1191, "step": 7066 }, { "epoch": 3.325647058823529, "grad_norm": 0.8696579247441925, "learning_rate": 0.0001634833185175703, "loss": 0.4316, "step": 7067 }, { "epoch": 3.3261176470588234, "grad_norm": 1.2054307852985242, "learning_rate": 0.00016343965922308438, "loss": 1.0391, "step": 7068 }, { "epoch": 3.3265882352941176, "grad_norm": 0.5073983279094443, "learning_rate": 0.00016339599878078563, "loss": 0.4316, "step": 7069 }, { "epoch": 3.327058823529412, "grad_norm": 1.2457383720056618, "learning_rate": 0.00016335233719440282, "loss": 1.0, "step": 7070 }, { "epoch": 3.327529411764706, "grad_norm": 0.6934153395330235, "learning_rate": 0.000163308674467665, "loss": 0.2695, "step": 7071 }, { "epoch": 3.328, "grad_norm": 0.9908844715050298, "learning_rate": 0.00016326501060430105, "loss": 1.1953, "step": 7072 }, { "epoch": 3.328470588235294, "grad_norm": 0.8601943127294472, "learning_rate": 0.00016322134560804017, "loss": 0.8008, "step": 7073 }, { "epoch": 3.3289411764705883, "grad_norm": 0.7554907472082867, "learning_rate": 0.00016317767948261148, "loss": 0.4062, "step": 7074 }, { "epoch": 3.3294117647058825, "grad_norm": 1.3111962181377876, "learning_rate": 0.00016313401223174434, "loss": 0.957, "step": 7075 }, { "epoch": 3.3298823529411763, "grad_norm": 1.0959870752639094, "learning_rate": 0.00016309034385916815, "loss": 1.6719, "step": 7076 }, { "epoch": 3.3303529411764705, "grad_norm": 1.0550385817132533, "learning_rate": 0.00016304667436861233, "loss": 1.3438, "step": 7077 }, { "epoch": 3.3308235294117647, "grad_norm": 0.9589775961315744, "learning_rate": 0.00016300300376380657, "loss": 1.2891, "step": 7078 }, { "epoch": 3.331294117647059, "grad_norm": 0.6076696628184782, "learning_rate": 0.00016295933204848052, "loss": 0.3184, "step": 7079 }, { "epoch": 3.331764705882353, "grad_norm": 1.03289915485475, "learning_rate": 0.00016291565922636392, "loss": 0.4629, "step": 7080 }, { "epoch": 3.332235294117647, "grad_norm": 0.7040478167774953, "learning_rate": 0.0001628719853011867, "loss": 0.3223, "step": 7081 }, { "epoch": 3.332705882352941, "grad_norm": 0.8314728518910846, "learning_rate": 0.00016282831027667881, "loss": 0.4043, "step": 7082 }, { "epoch": 3.3331764705882354, "grad_norm": 0.9702645473642078, "learning_rate": 0.00016278463415657027, "loss": 0.5078, "step": 7083 }, { "epoch": 3.333647058823529, "grad_norm": 1.3298728989336412, "learning_rate": 0.00016274095694459128, "loss": 1.3125, "step": 7084 }, { "epoch": 3.3341176470588234, "grad_norm": 0.6905901784348805, "learning_rate": 0.00016269727864447206, "loss": 0.5664, "step": 7085 }, { "epoch": 3.3345882352941176, "grad_norm": 1.5080964156240806, "learning_rate": 0.00016265359925994298, "loss": 1.5625, "step": 7086 }, { "epoch": 3.335058823529412, "grad_norm": 0.6878458358056033, "learning_rate": 0.00016260991879473441, "loss": 0.3379, "step": 7087 }, { "epoch": 3.335529411764706, "grad_norm": 0.8010218604748142, "learning_rate": 0.000162566237252577, "loss": 0.5156, "step": 7088 }, { "epoch": 3.336, "grad_norm": 0.8873158903801085, "learning_rate": 0.00016252255463720125, "loss": 0.5117, "step": 7089 }, { "epoch": 3.336470588235294, "grad_norm": 0.8349194383376077, "learning_rate": 0.0001624788709523379, "loss": 0.5195, "step": 7090 }, { "epoch": 3.3369411764705883, "grad_norm": 1.1464863332640476, "learning_rate": 0.00016243518620171778, "loss": 0.7109, "step": 7091 }, { "epoch": 3.3374117647058825, "grad_norm": 0.7290436712207837, "learning_rate": 0.00016239150038907177, "loss": 0.4043, "step": 7092 }, { "epoch": 3.3378823529411763, "grad_norm": 0.6891421183534381, "learning_rate": 0.00016234781351813079, "loss": 0.3418, "step": 7093 }, { "epoch": 3.3383529411764705, "grad_norm": 1.1653385435630907, "learning_rate": 0.00016230412559262606, "loss": 1.2656, "step": 7094 }, { "epoch": 3.3388235294117647, "grad_norm": 0.8401657700741443, "learning_rate": 0.00016226043661628862, "loss": 0.4297, "step": 7095 }, { "epoch": 3.339294117647059, "grad_norm": 0.9427568967984394, "learning_rate": 0.00016221674659284976, "loss": 0.5312, "step": 7096 }, { "epoch": 3.339764705882353, "grad_norm": 1.156151484610269, "learning_rate": 0.00016217305552604076, "loss": 0.8477, "step": 7097 }, { "epoch": 3.340235294117647, "grad_norm": 0.8071470825218077, "learning_rate": 0.00016212936341959312, "loss": 0.4805, "step": 7098 }, { "epoch": 3.340705882352941, "grad_norm": 0.9048227052285821, "learning_rate": 0.0001620856702772384, "loss": 1.0547, "step": 7099 }, { "epoch": 3.3411764705882354, "grad_norm": 1.1501981138538369, "learning_rate": 0.00016204197610270816, "loss": 2.0156, "step": 7100 }, { "epoch": 3.341647058823529, "grad_norm": 0.7987071724005032, "learning_rate": 0.00016199828089973406, "loss": 0.5156, "step": 7101 }, { "epoch": 3.3421176470588234, "grad_norm": 1.2287725044027415, "learning_rate": 0.00016195458467204793, "loss": 0.9336, "step": 7102 }, { "epoch": 3.3425882352941176, "grad_norm": 0.5156947973500247, "learning_rate": 0.00016191088742338166, "loss": 0.248, "step": 7103 }, { "epoch": 3.343058823529412, "grad_norm": 0.8909107362004659, "learning_rate": 0.00016186718915746714, "loss": 0.5938, "step": 7104 }, { "epoch": 3.343529411764706, "grad_norm": 0.9999312885848359, "learning_rate": 0.00016182348987803648, "loss": 1.1953, "step": 7105 }, { "epoch": 3.344, "grad_norm": 0.9897659949364518, "learning_rate": 0.0001617797895888218, "loss": 1.1484, "step": 7106 }, { "epoch": 3.344470588235294, "grad_norm": 0.7544787931270013, "learning_rate": 0.00016173608829355527, "loss": 0.4336, "step": 7107 }, { "epoch": 3.3449411764705883, "grad_norm": 1.2144475980860026, "learning_rate": 0.00016169238599596927, "loss": 0.6406, "step": 7108 }, { "epoch": 3.3454117647058825, "grad_norm": 1.1697501108825108, "learning_rate": 0.00016164868269979619, "loss": 1.1953, "step": 7109 }, { "epoch": 3.3458823529411763, "grad_norm": 1.101800501113885, "learning_rate": 0.0001616049784087684, "loss": 0.7617, "step": 7110 }, { "epoch": 3.3463529411764705, "grad_norm": 0.44841963198090773, "learning_rate": 0.00016156127312661855, "loss": 0.0981, "step": 7111 }, { "epoch": 3.3468235294117648, "grad_norm": 0.5635194564704816, "learning_rate": 0.0001615175668570793, "loss": 0.2129, "step": 7112 }, { "epoch": 3.347294117647059, "grad_norm": 0.8185389760909624, "learning_rate": 0.0001614738596038833, "loss": 0.75, "step": 7113 }, { "epoch": 3.3477647058823528, "grad_norm": 1.2009325839729306, "learning_rate": 0.00016143015137076345, "loss": 1.3203, "step": 7114 }, { "epoch": 3.348235294117647, "grad_norm": 0.8690301734098568, "learning_rate": 0.0001613864421614526, "loss": 0.5039, "step": 7115 }, { "epoch": 3.348705882352941, "grad_norm": 0.9587855483905849, "learning_rate": 0.00016134273197968374, "loss": 1.1953, "step": 7116 }, { "epoch": 3.3491764705882354, "grad_norm": 1.3045622407925104, "learning_rate": 0.0001612990208291899, "loss": 1.6719, "step": 7117 }, { "epoch": 3.349647058823529, "grad_norm": 0.7766944040025555, "learning_rate": 0.0001612553087137043, "loss": 0.3105, "step": 7118 }, { "epoch": 3.3501176470588234, "grad_norm": 0.9638576006259466, "learning_rate": 0.0001612115956369601, "loss": 0.9062, "step": 7119 }, { "epoch": 3.3505882352941176, "grad_norm": 1.200250339132728, "learning_rate": 0.00016116788160269065, "loss": 1.1719, "step": 7120 }, { "epoch": 3.351058823529412, "grad_norm": 1.0590253692387899, "learning_rate": 0.0001611241666146293, "loss": 1.3594, "step": 7121 }, { "epoch": 3.351529411764706, "grad_norm": 0.8030653365962799, "learning_rate": 0.00016108045067650962, "loss": 0.5312, "step": 7122 }, { "epoch": 3.352, "grad_norm": 1.1660826925402057, "learning_rate": 0.0001610367337920651, "loss": 2.0, "step": 7123 }, { "epoch": 3.352470588235294, "grad_norm": 0.7594396766028668, "learning_rate": 0.00016099301596502936, "loss": 0.4004, "step": 7124 }, { "epoch": 3.3529411764705883, "grad_norm": 1.0868442180811122, "learning_rate": 0.00016094929719913612, "loss": 1.0547, "step": 7125 }, { "epoch": 3.3534117647058825, "grad_norm": 1.0252071413749957, "learning_rate": 0.0001609055774981192, "loss": 1.3984, "step": 7126 }, { "epoch": 3.3538823529411763, "grad_norm": 0.9949940170938812, "learning_rate": 0.0001608618568657125, "loss": 0.9023, "step": 7127 }, { "epoch": 3.3543529411764705, "grad_norm": 1.0962949001757103, "learning_rate": 0.0001608181353056499, "loss": 0.6719, "step": 7128 }, { "epoch": 3.3548235294117648, "grad_norm": 0.8136037407455791, "learning_rate": 0.0001607744128216655, "loss": 0.5977, "step": 7129 }, { "epoch": 3.355294117647059, "grad_norm": 0.5906770698818049, "learning_rate": 0.00016073068941749343, "loss": 0.2969, "step": 7130 }, { "epoch": 3.3557647058823528, "grad_norm": 1.2415420281870284, "learning_rate": 0.0001606869650968678, "loss": 1.5703, "step": 7131 }, { "epoch": 3.356235294117647, "grad_norm": 0.8917248084619174, "learning_rate": 0.00016064323986352295, "loss": 0.5859, "step": 7132 }, { "epoch": 3.356705882352941, "grad_norm": 0.8568536070492044, "learning_rate": 0.00016059951372119323, "loss": 0.8281, "step": 7133 }, { "epoch": 3.3571764705882354, "grad_norm": 1.0089462456847782, "learning_rate": 0.000160555786673613, "loss": 0.6914, "step": 7134 }, { "epoch": 3.357647058823529, "grad_norm": 0.9291697982542565, "learning_rate": 0.00016051205872451686, "loss": 0.6133, "step": 7135 }, { "epoch": 3.3581176470588234, "grad_norm": 0.8454223002832478, "learning_rate": 0.00016046832987763935, "loss": 0.5, "step": 7136 }, { "epoch": 3.3585882352941177, "grad_norm": 1.3077077371457397, "learning_rate": 0.0001604246001367151, "loss": 1.0078, "step": 7137 }, { "epoch": 3.359058823529412, "grad_norm": 0.6873908372142564, "learning_rate": 0.00016038086950547885, "loss": 0.2451, "step": 7138 }, { "epoch": 3.359529411764706, "grad_norm": 0.8958214493831558, "learning_rate": 0.00016033713798766546, "loss": 1.2734, "step": 7139 }, { "epoch": 3.36, "grad_norm": 0.9152273126802098, "learning_rate": 0.00016029340558700977, "loss": 0.4863, "step": 7140 }, { "epoch": 3.360470588235294, "grad_norm": 1.3029320276516951, "learning_rate": 0.00016024967230724672, "loss": 1.2734, "step": 7141 }, { "epoch": 3.3609411764705883, "grad_norm": 1.1109167581668558, "learning_rate": 0.00016020593815211138, "loss": 1.3359, "step": 7142 }, { "epoch": 3.3614117647058825, "grad_norm": 0.5938516507387264, "learning_rate": 0.00016016220312533888, "loss": 0.2871, "step": 7143 }, { "epoch": 3.3618823529411763, "grad_norm": 0.9674416024931948, "learning_rate": 0.00016011846723066445, "loss": 0.6328, "step": 7144 }, { "epoch": 3.3623529411764705, "grad_norm": 0.65023583742118, "learning_rate": 0.0001600747304718232, "loss": 0.4883, "step": 7145 }, { "epoch": 3.3628235294117648, "grad_norm": 0.8498619308864994, "learning_rate": 0.00016003099285255058, "loss": 0.7109, "step": 7146 }, { "epoch": 3.363294117647059, "grad_norm": 0.828543911412513, "learning_rate": 0.000159987254376582, "loss": 0.9258, "step": 7147 }, { "epoch": 3.3637647058823528, "grad_norm": 0.9731408915932642, "learning_rate": 0.00015994351504765287, "loss": 0.9102, "step": 7148 }, { "epoch": 3.364235294117647, "grad_norm": 0.9683030922312893, "learning_rate": 0.00015989977486949883, "loss": 1.0391, "step": 7149 }, { "epoch": 3.364705882352941, "grad_norm": 0.8413031552422781, "learning_rate": 0.00015985603384585542, "loss": 0.8281, "step": 7150 }, { "epoch": 3.3651764705882354, "grad_norm": 0.9623894119953156, "learning_rate": 0.0001598122919804584, "loss": 0.8789, "step": 7151 }, { "epoch": 3.365647058823529, "grad_norm": 1.0454124209431483, "learning_rate": 0.0001597685492770435, "loss": 1.0, "step": 7152 }, { "epoch": 3.3661176470588234, "grad_norm": 0.8287411874298696, "learning_rate": 0.00015972480573934656, "loss": 0.2949, "step": 7153 }, { "epoch": 3.3665882352941177, "grad_norm": 0.6180407831589869, "learning_rate": 0.00015968106137110352, "loss": 0.543, "step": 7154 }, { "epoch": 3.367058823529412, "grad_norm": 0.9820266316228977, "learning_rate": 0.00015963731617605034, "loss": 1.1953, "step": 7155 }, { "epoch": 3.367529411764706, "grad_norm": 0.31493093126141863, "learning_rate": 0.00015959357015792313, "loss": 0.0515, "step": 7156 }, { "epoch": 3.368, "grad_norm": 1.240766939779663, "learning_rate": 0.00015954982332045794, "loss": 1.4062, "step": 7157 }, { "epoch": 3.368470588235294, "grad_norm": 0.8387751292797683, "learning_rate": 0.000159506075667391, "loss": 0.4883, "step": 7158 }, { "epoch": 3.3689411764705883, "grad_norm": 0.9391832800691327, "learning_rate": 0.0001594623272024586, "loss": 0.5469, "step": 7159 }, { "epoch": 3.369411764705882, "grad_norm": 1.0796925450489134, "learning_rate": 0.000159418577929397, "loss": 0.5312, "step": 7160 }, { "epoch": 3.3698823529411763, "grad_norm": 0.9416074326014002, "learning_rate": 0.0001593748278519427, "loss": 0.9492, "step": 7161 }, { "epoch": 3.3703529411764706, "grad_norm": 0.9806435377823258, "learning_rate": 0.00015933107697383206, "loss": 1.1328, "step": 7162 }, { "epoch": 3.3708235294117648, "grad_norm": 1.0819768998832908, "learning_rate": 0.0001592873252988017, "loss": 1.0391, "step": 7163 }, { "epoch": 3.371294117647059, "grad_norm": 0.6237278849986632, "learning_rate": 0.0001592435728305882, "loss": 0.4023, "step": 7164 }, { "epoch": 3.3717647058823528, "grad_norm": 0.2749173952238275, "learning_rate": 0.00015919981957292825, "loss": 0.0645, "step": 7165 }, { "epoch": 3.372235294117647, "grad_norm": 0.9598482482510792, "learning_rate": 0.00015915606552955856, "loss": 0.8047, "step": 7166 }, { "epoch": 3.372705882352941, "grad_norm": 0.9924457399886455, "learning_rate": 0.000159112310704216, "loss": 0.6758, "step": 7167 }, { "epoch": 3.3731764705882354, "grad_norm": 1.4517229127830582, "learning_rate": 0.00015906855510063744, "loss": 0.9805, "step": 7168 }, { "epoch": 3.373647058823529, "grad_norm": 1.1340189779149001, "learning_rate": 0.0001590247987225597, "loss": 1.0469, "step": 7169 }, { "epoch": 3.3741176470588234, "grad_norm": 1.1282367772963868, "learning_rate": 0.00015898104157371993, "loss": 0.9219, "step": 7170 }, { "epoch": 3.3745882352941177, "grad_norm": 0.8204460326651702, "learning_rate": 0.00015893728365785518, "loss": 0.6211, "step": 7171 }, { "epoch": 3.375058823529412, "grad_norm": 1.2890212247392525, "learning_rate": 0.00015889352497870253, "loss": 1.1719, "step": 7172 }, { "epoch": 3.375529411764706, "grad_norm": 1.076357281637492, "learning_rate": 0.00015884976553999925, "loss": 0.5664, "step": 7173 }, { "epoch": 3.376, "grad_norm": 0.422436727281153, "learning_rate": 0.0001588060053454826, "loss": 0.2041, "step": 7174 }, { "epoch": 3.376470588235294, "grad_norm": 1.1144577453452744, "learning_rate": 0.0001587622443988899, "loss": 1.1016, "step": 7175 }, { "epoch": 3.3769411764705883, "grad_norm": 0.7933289327200062, "learning_rate": 0.0001587184827039585, "loss": 0.9062, "step": 7176 }, { "epoch": 3.377411764705882, "grad_norm": 1.1272987732613728, "learning_rate": 0.00015867472026442596, "loss": 0.9023, "step": 7177 }, { "epoch": 3.3778823529411763, "grad_norm": 0.6657143777000355, "learning_rate": 0.00015863095708402972, "loss": 0.3125, "step": 7178 }, { "epoch": 3.3783529411764706, "grad_norm": 1.0682843042054901, "learning_rate": 0.00015858719316650744, "loss": 0.7109, "step": 7179 }, { "epoch": 3.378823529411765, "grad_norm": 0.9573950564329163, "learning_rate": 0.00015854342851559673, "loss": 0.5039, "step": 7180 }, { "epoch": 3.379294117647059, "grad_norm": 0.6102437735529944, "learning_rate": 0.00015849966313503537, "loss": 0.2129, "step": 7181 }, { "epoch": 3.379764705882353, "grad_norm": 1.4077507558429798, "learning_rate": 0.00015845589702856102, "loss": 2.2344, "step": 7182 }, { "epoch": 3.380235294117647, "grad_norm": 0.8888465607396719, "learning_rate": 0.0001584121301999116, "loss": 0.8164, "step": 7183 }, { "epoch": 3.3807058823529412, "grad_norm": 0.580925680226031, "learning_rate": 0.00015836836265282502, "loss": 0.4531, "step": 7184 }, { "epoch": 3.3811764705882354, "grad_norm": 1.1635353172713088, "learning_rate": 0.0001583245943910392, "loss": 0.8672, "step": 7185 }, { "epoch": 3.3816470588235292, "grad_norm": 0.9888793148094442, "learning_rate": 0.0001582808254182922, "loss": 0.9766, "step": 7186 }, { "epoch": 3.3821176470588235, "grad_norm": 3.0461128911680304, "learning_rate": 0.0001582370557383221, "loss": 1.4609, "step": 7187 }, { "epoch": 3.3825882352941177, "grad_norm": 0.5864144106980332, "learning_rate": 0.00015819328535486704, "loss": 0.209, "step": 7188 }, { "epoch": 3.383058823529412, "grad_norm": 1.1653938374532327, "learning_rate": 0.0001581495142716652, "loss": 1.1953, "step": 7189 }, { "epoch": 3.383529411764706, "grad_norm": 1.1660516516131132, "learning_rate": 0.00015810574249245488, "loss": 1.1094, "step": 7190 }, { "epoch": 3.384, "grad_norm": 4.116206623800933, "learning_rate": 0.00015806197002097438, "loss": 1.375, "step": 7191 }, { "epoch": 3.384470588235294, "grad_norm": 1.2922305392593558, "learning_rate": 0.00015801819686096213, "loss": 1.2109, "step": 7192 }, { "epoch": 3.3849411764705883, "grad_norm": 0.8241687652480698, "learning_rate": 0.00015797442301615652, "loss": 0.6094, "step": 7193 }, { "epoch": 3.385411764705882, "grad_norm": 1.0159338909537123, "learning_rate": 0.00015793064849029608, "loss": 0.668, "step": 7194 }, { "epoch": 3.3858823529411763, "grad_norm": 0.6256655370160992, "learning_rate": 0.00015788687328711933, "loss": 0.4531, "step": 7195 }, { "epoch": 3.3863529411764706, "grad_norm": 0.6364815048240303, "learning_rate": 0.00015784309741036492, "loss": 0.4395, "step": 7196 }, { "epoch": 3.386823529411765, "grad_norm": 0.8855451204393509, "learning_rate": 0.00015779932086377153, "loss": 0.8086, "step": 7197 }, { "epoch": 3.387294117647059, "grad_norm": 0.8696283486854226, "learning_rate": 0.00015775554365107791, "loss": 0.3574, "step": 7198 }, { "epoch": 3.387764705882353, "grad_norm": 1.07352944687809, "learning_rate": 0.0001577117657760228, "loss": 0.4961, "step": 7199 }, { "epoch": 3.388235294117647, "grad_norm": 0.6456564695152696, "learning_rate": 0.00015766798724234506, "loss": 0.4297, "step": 7200 }, { "epoch": 3.3887058823529412, "grad_norm": 0.6071533550462546, "learning_rate": 0.00015762420805378363, "loss": 0.1797, "step": 7201 }, { "epoch": 3.3891764705882355, "grad_norm": 1.0512150353863357, "learning_rate": 0.00015758042821407742, "loss": 1.0234, "step": 7202 }, { "epoch": 3.3896470588235292, "grad_norm": 0.7078164734120506, "learning_rate": 0.00015753664772696544, "loss": 1.0078, "step": 7203 }, { "epoch": 3.3901176470588235, "grad_norm": 0.5677043648784015, "learning_rate": 0.00015749286659618678, "loss": 0.1943, "step": 7204 }, { "epoch": 3.3905882352941177, "grad_norm": 0.618590886398587, "learning_rate": 0.00015744908482548059, "loss": 0.1602, "step": 7205 }, { "epoch": 3.391058823529412, "grad_norm": 0.7885689623862502, "learning_rate": 0.00015740530241858597, "loss": 0.4414, "step": 7206 }, { "epoch": 3.391529411764706, "grad_norm": 0.7120759210030844, "learning_rate": 0.00015736151937924224, "loss": 0.3789, "step": 7207 }, { "epoch": 3.392, "grad_norm": 0.9687867449190584, "learning_rate": 0.00015731773571118864, "loss": 0.918, "step": 7208 }, { "epoch": 3.392470588235294, "grad_norm": 0.8393127963516763, "learning_rate": 0.00015727395141816448, "loss": 0.5039, "step": 7209 }, { "epoch": 3.3929411764705883, "grad_norm": 0.9614760506151965, "learning_rate": 0.00015723016650390922, "loss": 0.9922, "step": 7210 }, { "epoch": 3.393411764705882, "grad_norm": 1.0808356338435121, "learning_rate": 0.0001571863809721623, "loss": 1.0156, "step": 7211 }, { "epoch": 3.3938823529411764, "grad_norm": 1.3095510568513746, "learning_rate": 0.00015714259482666317, "loss": 0.6875, "step": 7212 }, { "epoch": 3.3943529411764706, "grad_norm": 1.0921571205813074, "learning_rate": 0.00015709880807115138, "loss": 0.7148, "step": 7213 }, { "epoch": 3.394823529411765, "grad_norm": 0.7869941094756399, "learning_rate": 0.0001570550207093666, "loss": 0.8281, "step": 7214 }, { "epoch": 3.395294117647059, "grad_norm": 1.1392881438628353, "learning_rate": 0.00015701123274504845, "loss": 1.0312, "step": 7215 }, { "epoch": 3.395764705882353, "grad_norm": 0.8023991800364264, "learning_rate": 0.00015696744418193664, "loss": 0.3691, "step": 7216 }, { "epoch": 3.396235294117647, "grad_norm": 0.29687214125202005, "learning_rate": 0.00015692365502377088, "loss": 0.0288, "step": 7217 }, { "epoch": 3.3967058823529412, "grad_norm": 0.9864049390186559, "learning_rate": 0.00015687986527429103, "loss": 0.6523, "step": 7218 }, { "epoch": 3.3971764705882355, "grad_norm": 1.1402018871168007, "learning_rate": 0.00015683607493723699, "loss": 1.5, "step": 7219 }, { "epoch": 3.3976470588235292, "grad_norm": 1.203713907603914, "learning_rate": 0.00015679228401634855, "loss": 1.1719, "step": 7220 }, { "epoch": 3.3981176470588235, "grad_norm": 1.5129972951188655, "learning_rate": 0.0001567484925153658, "loss": 1.0234, "step": 7221 }, { "epoch": 3.3985882352941177, "grad_norm": 1.0723767805617412, "learning_rate": 0.00015670470043802867, "loss": 0.9023, "step": 7222 }, { "epoch": 3.399058823529412, "grad_norm": 1.0033133919181365, "learning_rate": 0.0001566609077880772, "loss": 1.6875, "step": 7223 }, { "epoch": 3.3995294117647057, "grad_norm": 0.7347476537846264, "learning_rate": 0.00015661711456925158, "loss": 0.3203, "step": 7224 }, { "epoch": 3.4, "grad_norm": 1.2069040333003358, "learning_rate": 0.0001565733207852919, "loss": 1.3125, "step": 7225 }, { "epoch": 3.400470588235294, "grad_norm": 0.6591569765431519, "learning_rate": 0.00015652952643993841, "loss": 0.5391, "step": 7226 }, { "epoch": 3.4009411764705884, "grad_norm": 0.948978507885629, "learning_rate": 0.00015648573153693128, "loss": 1.5234, "step": 7227 }, { "epoch": 3.401411764705882, "grad_norm": 0.9190056570685936, "learning_rate": 0.0001564419360800109, "loss": 0.8281, "step": 7228 }, { "epoch": 3.4018823529411764, "grad_norm": 0.6398158429675752, "learning_rate": 0.0001563981400729176, "loss": 0.7734, "step": 7229 }, { "epoch": 3.4023529411764706, "grad_norm": 0.6968423170106848, "learning_rate": 0.0001563543435193917, "loss": 0.5352, "step": 7230 }, { "epoch": 3.402823529411765, "grad_norm": 0.75574630818069, "learning_rate": 0.0001563105464231737, "loss": 0.2617, "step": 7231 }, { "epoch": 3.403294117647059, "grad_norm": 0.9285636194458948, "learning_rate": 0.0001562667487880041, "loss": 1.1094, "step": 7232 }, { "epoch": 3.403764705882353, "grad_norm": 1.0529899002318062, "learning_rate": 0.00015622295061762346, "loss": 1.6172, "step": 7233 }, { "epoch": 3.404235294117647, "grad_norm": 1.176819346499566, "learning_rate": 0.00015617915191577222, "loss": 0.8203, "step": 7234 }, { "epoch": 3.4047058823529412, "grad_norm": 0.4428065083902368, "learning_rate": 0.0001561353526861912, "loss": 0.1533, "step": 7235 }, { "epoch": 3.4051764705882355, "grad_norm": 0.8250445652485339, "learning_rate": 0.0001560915529326209, "loss": 0.5117, "step": 7236 }, { "epoch": 3.4056470588235292, "grad_norm": 1.0052637984872042, "learning_rate": 0.00015604775265880212, "loss": 1.3047, "step": 7237 }, { "epoch": 3.4061176470588235, "grad_norm": 0.8212099777548567, "learning_rate": 0.0001560039518684756, "loss": 0.4062, "step": 7238 }, { "epoch": 3.4065882352941177, "grad_norm": 1.0121038979514192, "learning_rate": 0.00015596015056538215, "loss": 1.3125, "step": 7239 }, { "epoch": 3.407058823529412, "grad_norm": 0.8070058934322897, "learning_rate": 0.0001559163487532626, "loss": 0.6523, "step": 7240 }, { "epoch": 3.4075294117647057, "grad_norm": 0.6460672717928989, "learning_rate": 0.00015587254643585784, "loss": 0.2871, "step": 7241 }, { "epoch": 3.408, "grad_norm": 0.6563530001340317, "learning_rate": 0.00015582874361690885, "loss": 0.2891, "step": 7242 }, { "epoch": 3.408470588235294, "grad_norm": 0.46225134899111964, "learning_rate": 0.00015578494030015652, "loss": 0.1328, "step": 7243 }, { "epoch": 3.4089411764705884, "grad_norm": 0.6073638255155083, "learning_rate": 0.00015574113648934192, "loss": 0.3047, "step": 7244 }, { "epoch": 3.409411764705882, "grad_norm": 0.4990821062411035, "learning_rate": 0.0001556973321882061, "loss": 0.1748, "step": 7245 }, { "epoch": 3.4098823529411764, "grad_norm": 1.1966134870106804, "learning_rate": 0.0001556535274004902, "loss": 0.9922, "step": 7246 }, { "epoch": 3.4103529411764706, "grad_norm": 0.9297241946998819, "learning_rate": 0.00015560972212993527, "loss": 1.9297, "step": 7247 }, { "epoch": 3.410823529411765, "grad_norm": 1.1052195990268356, "learning_rate": 0.0001555659163802826, "loss": 0.9766, "step": 7248 }, { "epoch": 3.411294117647059, "grad_norm": 1.0995542469679265, "learning_rate": 0.00015552211015527336, "loss": 0.918, "step": 7249 }, { "epoch": 3.411764705882353, "grad_norm": 0.9438862380394629, "learning_rate": 0.00015547830345864885, "loss": 0.8086, "step": 7250 }, { "epoch": 3.412235294117647, "grad_norm": 0.24834590814770258, "learning_rate": 0.00015543449629415028, "loss": 0.0278, "step": 7251 }, { "epoch": 3.4127058823529413, "grad_norm": 0.6809103512968134, "learning_rate": 0.00015539068866551916, "loss": 0.6758, "step": 7252 }, { "epoch": 3.4131764705882355, "grad_norm": 1.2395224060898922, "learning_rate": 0.00015534688057649673, "loss": 1.2422, "step": 7253 }, { "epoch": 3.4136470588235293, "grad_norm": 1.5724254948142933, "learning_rate": 0.0001553030720308245, "loss": 2.0, "step": 7254 }, { "epoch": 3.4141176470588235, "grad_norm": 0.7715708282135362, "learning_rate": 0.00015525926303224387, "loss": 0.3906, "step": 7255 }, { "epoch": 3.4145882352941177, "grad_norm": 0.8841266868663972, "learning_rate": 0.0001552154535844964, "loss": 0.9023, "step": 7256 }, { "epoch": 3.415058823529412, "grad_norm": 1.0110807590027266, "learning_rate": 0.00015517164369132364, "loss": 1.2266, "step": 7257 }, { "epoch": 3.4155294117647057, "grad_norm": 0.9042919134883013, "learning_rate": 0.0001551278333564671, "loss": 1.2344, "step": 7258 }, { "epoch": 3.416, "grad_norm": 0.8021036201791703, "learning_rate": 0.00015508402258366846, "loss": 0.8516, "step": 7259 }, { "epoch": 3.416470588235294, "grad_norm": 0.8697517939572506, "learning_rate": 0.00015504021137666938, "loss": 0.6562, "step": 7260 }, { "epoch": 3.4169411764705884, "grad_norm": 1.2018128580524543, "learning_rate": 0.00015499639973921148, "loss": 1.0703, "step": 7261 }, { "epoch": 3.417411764705882, "grad_norm": 1.2331018585038158, "learning_rate": 0.00015495258767503657, "loss": 1.7266, "step": 7262 }, { "epoch": 3.4178823529411764, "grad_norm": 1.235489876927632, "learning_rate": 0.0001549087751878864, "loss": 0.707, "step": 7263 }, { "epoch": 3.4183529411764706, "grad_norm": 1.0240663272831785, "learning_rate": 0.00015486496228150266, "loss": 1.2656, "step": 7264 }, { "epoch": 3.418823529411765, "grad_norm": 0.41102488583585683, "learning_rate": 0.00015482114895962735, "loss": 0.3438, "step": 7265 }, { "epoch": 3.419294117647059, "grad_norm": 0.8294011750539142, "learning_rate": 0.00015477733522600227, "loss": 1.3047, "step": 7266 }, { "epoch": 3.419764705882353, "grad_norm": 0.42749137154326994, "learning_rate": 0.00015473352108436934, "loss": 0.1621, "step": 7267 }, { "epoch": 3.420235294117647, "grad_norm": 1.0222663952021933, "learning_rate": 0.00015468970653847043, "loss": 0.7969, "step": 7268 }, { "epoch": 3.4207058823529413, "grad_norm": 0.8805314545434081, "learning_rate": 0.00015464589159204764, "loss": 0.9531, "step": 7269 }, { "epoch": 3.4211764705882355, "grad_norm": 0.9310523594191756, "learning_rate": 0.0001546020762488429, "loss": 0.6758, "step": 7270 }, { "epoch": 3.4216470588235293, "grad_norm": 0.7517355076243816, "learning_rate": 0.00015455826051259826, "loss": 0.332, "step": 7271 }, { "epoch": 3.4221176470588235, "grad_norm": 0.5353600427766064, "learning_rate": 0.00015451444438705584, "loss": 0.1396, "step": 7272 }, { "epoch": 3.4225882352941177, "grad_norm": 1.0445044992570602, "learning_rate": 0.00015447062787595772, "loss": 1.3203, "step": 7273 }, { "epoch": 3.423058823529412, "grad_norm": 1.0083257844917222, "learning_rate": 0.00015442681098304603, "loss": 0.8164, "step": 7274 }, { "epoch": 3.4235294117647057, "grad_norm": 1.1069587167085513, "learning_rate": 0.000154382993712063, "loss": 1.6016, "step": 7275 }, { "epoch": 3.424, "grad_norm": 0.9780358482438948, "learning_rate": 0.00015433917606675075, "loss": 0.9961, "step": 7276 }, { "epoch": 3.424470588235294, "grad_norm": 1.5567854110445707, "learning_rate": 0.00015429535805085162, "loss": 1.4688, "step": 7277 }, { "epoch": 3.4249411764705884, "grad_norm": 1.0889367256780556, "learning_rate": 0.00015425153966810783, "loss": 0.8125, "step": 7278 }, { "epoch": 3.425411764705882, "grad_norm": 1.0800481882221362, "learning_rate": 0.00015420772092226173, "loss": 1.0078, "step": 7279 }, { "epoch": 3.4258823529411764, "grad_norm": 1.9482767075583056, "learning_rate": 0.00015416390181705562, "loss": 1.5859, "step": 7280 }, { "epoch": 3.4263529411764706, "grad_norm": 0.6680810229629959, "learning_rate": 0.00015412008235623186, "loss": 0.2617, "step": 7281 }, { "epoch": 3.426823529411765, "grad_norm": 0.9066038366349838, "learning_rate": 0.00015407626254353282, "loss": 0.2539, "step": 7282 }, { "epoch": 3.427294117647059, "grad_norm": 1.0378696231453082, "learning_rate": 0.00015403244238270102, "loss": 0.9453, "step": 7283 }, { "epoch": 3.427764705882353, "grad_norm": 0.6741388559565012, "learning_rate": 0.00015398862187747888, "loss": 0.6016, "step": 7284 }, { "epoch": 3.428235294117647, "grad_norm": 1.1281877717851592, "learning_rate": 0.0001539448010316088, "loss": 0.7383, "step": 7285 }, { "epoch": 3.4287058823529413, "grad_norm": 0.9174373985628529, "learning_rate": 0.0001539009798488334, "loss": 0.7461, "step": 7286 }, { "epoch": 3.429176470588235, "grad_norm": 1.052216972974051, "learning_rate": 0.0001538571583328952, "loss": 1.3594, "step": 7287 }, { "epoch": 3.4296470588235293, "grad_norm": 1.0716102700005565, "learning_rate": 0.0001538133364875367, "loss": 1.0859, "step": 7288 }, { "epoch": 3.4301176470588235, "grad_norm": 1.1255297055454676, "learning_rate": 0.00015376951431650063, "loss": 1.0156, "step": 7289 }, { "epoch": 3.4305882352941177, "grad_norm": 0.8973495783487542, "learning_rate": 0.00015372569182352952, "loss": 1.0469, "step": 7290 }, { "epoch": 3.431058823529412, "grad_norm": 1.1195615309288285, "learning_rate": 0.00015368186901236605, "loss": 1.2109, "step": 7291 }, { "epoch": 3.4315294117647057, "grad_norm": 0.39694492817587756, "learning_rate": 0.00015363804588675292, "loss": 0.1016, "step": 7292 }, { "epoch": 3.432, "grad_norm": 0.9299382050194784, "learning_rate": 0.00015359422245043278, "loss": 1.1484, "step": 7293 }, { "epoch": 3.432470588235294, "grad_norm": 2.696652475980386, "learning_rate": 0.00015355039870714848, "loss": 1.0938, "step": 7294 }, { "epoch": 3.4329411764705884, "grad_norm": 0.6651046093934295, "learning_rate": 0.00015350657466064264, "loss": 0.4766, "step": 7295 }, { "epoch": 3.433411764705882, "grad_norm": 1.4036841445714703, "learning_rate": 0.00015346275031465818, "loss": 1.75, "step": 7296 }, { "epoch": 3.4338823529411764, "grad_norm": 1.103517412753304, "learning_rate": 0.00015341892567293785, "loss": 1.4766, "step": 7297 }, { "epoch": 3.4343529411764706, "grad_norm": 0.917520620458617, "learning_rate": 0.0001533751007392245, "loss": 0.6055, "step": 7298 }, { "epoch": 3.434823529411765, "grad_norm": 1.2453048384885934, "learning_rate": 0.00015333127551726096, "loss": 1.9062, "step": 7299 }, { "epoch": 3.435294117647059, "grad_norm": 1.1983027148439045, "learning_rate": 0.0001532874500107902, "loss": 0.9258, "step": 7300 }, { "epoch": 3.435764705882353, "grad_norm": 0.9864427074137242, "learning_rate": 0.00015324362422355507, "loss": 0.7422, "step": 7301 }, { "epoch": 3.436235294117647, "grad_norm": 0.2934049446182062, "learning_rate": 0.00015319979815929847, "loss": 0.027, "step": 7302 }, { "epoch": 3.4367058823529413, "grad_norm": 1.6325273267541618, "learning_rate": 0.00015315597182176348, "loss": 1.6328, "step": 7303 }, { "epoch": 3.437176470588235, "grad_norm": 0.8325180873536728, "learning_rate": 0.000153112145214693, "loss": 0.4531, "step": 7304 }, { "epoch": 3.4376470588235293, "grad_norm": 1.0895072935729635, "learning_rate": 0.00015306831834183008, "loss": 0.7578, "step": 7305 }, { "epoch": 3.4381176470588235, "grad_norm": 0.8624657099979257, "learning_rate": 0.00015302449120691765, "loss": 0.9414, "step": 7306 }, { "epoch": 3.4385882352941177, "grad_norm": 1.022975537046819, "learning_rate": 0.00015298066381369888, "loss": 0.8477, "step": 7307 }, { "epoch": 3.439058823529412, "grad_norm": 0.7905071849240456, "learning_rate": 0.0001529368361659168, "loss": 1.1094, "step": 7308 }, { "epoch": 3.4395294117647057, "grad_norm": 1.6374633233565505, "learning_rate": 0.0001528930082673145, "loss": 2.0625, "step": 7309 }, { "epoch": 3.44, "grad_norm": 1.5683768068363508, "learning_rate": 0.0001528491801216351, "loss": 1.7344, "step": 7310 }, { "epoch": 3.440470588235294, "grad_norm": 1.4866967877295343, "learning_rate": 0.00015280535173262176, "loss": 1.6797, "step": 7311 }, { "epoch": 3.4409411764705884, "grad_norm": 0.8660725164604287, "learning_rate": 0.0001527615231040176, "loss": 0.3398, "step": 7312 }, { "epoch": 3.441411764705882, "grad_norm": 0.8525271230732806, "learning_rate": 0.00015271769423956584, "loss": 1.0625, "step": 7313 }, { "epoch": 3.4418823529411764, "grad_norm": 0.9072214729582384, "learning_rate": 0.00015267386514300965, "loss": 0.7578, "step": 7314 }, { "epoch": 3.4423529411764706, "grad_norm": 1.9064517515629753, "learning_rate": 0.00015263003581809226, "loss": 1.8828, "step": 7315 }, { "epoch": 3.442823529411765, "grad_norm": 1.065870833168561, "learning_rate": 0.00015258620626855691, "loss": 1.0312, "step": 7316 }, { "epoch": 3.443294117647059, "grad_norm": 1.1167134988968739, "learning_rate": 0.00015254237649814687, "loss": 0.4668, "step": 7317 }, { "epoch": 3.443764705882353, "grad_norm": 1.1441808023994464, "learning_rate": 0.00015249854651060542, "loss": 1.0781, "step": 7318 }, { "epoch": 3.444235294117647, "grad_norm": 2.1312135034429485, "learning_rate": 0.00015245471630967583, "loss": 2.3438, "step": 7319 }, { "epoch": 3.4447058823529413, "grad_norm": 0.8994822692149803, "learning_rate": 0.00015241088589910138, "loss": 0.9609, "step": 7320 }, { "epoch": 3.445176470588235, "grad_norm": 1.1927735125543282, "learning_rate": 0.0001523670552826255, "loss": 1.1406, "step": 7321 }, { "epoch": 3.4456470588235293, "grad_norm": 0.3138301886607716, "learning_rate": 0.00015232322446399147, "loss": 0.0586, "step": 7322 }, { "epoch": 3.4461176470588235, "grad_norm": 0.5575992394767705, "learning_rate": 0.00015227939344694268, "loss": 0.1709, "step": 7323 }, { "epoch": 3.4465882352941177, "grad_norm": 0.9233385494380681, "learning_rate": 0.0001522355622352225, "loss": 0.5117, "step": 7324 }, { "epoch": 3.447058823529412, "grad_norm": 0.8355073540635222, "learning_rate": 0.00015219173083257437, "loss": 0.9219, "step": 7325 }, { "epoch": 3.4475294117647057, "grad_norm": 1.1110203400413916, "learning_rate": 0.00015214789924274168, "loss": 0.957, "step": 7326 }, { "epoch": 3.448, "grad_norm": 1.2678316086952488, "learning_rate": 0.00015210406746946783, "loss": 1.8438, "step": 7327 }, { "epoch": 3.448470588235294, "grad_norm": 1.0466977339242836, "learning_rate": 0.0001520602355164963, "loss": 1.0234, "step": 7328 }, { "epoch": 3.4489411764705884, "grad_norm": 0.9187770162342359, "learning_rate": 0.00015201640338757054, "loss": 0.5469, "step": 7329 }, { "epoch": 3.449411764705882, "grad_norm": 1.3356346917673747, "learning_rate": 0.00015197257108643406, "loss": 0.7617, "step": 7330 }, { "epoch": 3.4498823529411764, "grad_norm": 1.6263169532249597, "learning_rate": 0.00015192873861683036, "loss": 0.8594, "step": 7331 }, { "epoch": 3.4503529411764706, "grad_norm": 1.0036138597042688, "learning_rate": 0.00015188490598250288, "loss": 0.9922, "step": 7332 }, { "epoch": 3.450823529411765, "grad_norm": 0.8089596651633593, "learning_rate": 0.0001518410731871952, "loss": 0.3809, "step": 7333 }, { "epoch": 3.451294117647059, "grad_norm": 1.0123311464860827, "learning_rate": 0.00015179724023465084, "loss": 0.8516, "step": 7334 }, { "epoch": 3.451764705882353, "grad_norm": 1.1228302556683591, "learning_rate": 0.00015175340712861337, "loss": 0.9414, "step": 7335 }, { "epoch": 3.452235294117647, "grad_norm": 1.0687592578975318, "learning_rate": 0.00015170957387282627, "loss": 0.7422, "step": 7336 }, { "epoch": 3.4527058823529413, "grad_norm": 1.0146049799266907, "learning_rate": 0.00015166574047103322, "loss": 0.5781, "step": 7337 }, { "epoch": 3.453176470588235, "grad_norm": 0.9711707550268883, "learning_rate": 0.00015162190692697778, "loss": 1.1172, "step": 7338 }, { "epoch": 3.4536470588235293, "grad_norm": 1.1877004358432475, "learning_rate": 0.00015157807324440347, "loss": 0.5156, "step": 7339 }, { "epoch": 3.4541176470588235, "grad_norm": 1.2045356106480212, "learning_rate": 0.00015153423942705402, "loss": 1.2188, "step": 7340 }, { "epoch": 3.4545882352941177, "grad_norm": 1.0120591552142297, "learning_rate": 0.000151490405478673, "loss": 0.7461, "step": 7341 }, { "epoch": 3.455058823529412, "grad_norm": 1.0672653721917995, "learning_rate": 0.000151446571403004, "loss": 1.2812, "step": 7342 }, { "epoch": 3.4555294117647057, "grad_norm": 0.9153329250510431, "learning_rate": 0.0001514027372037907, "loss": 0.832, "step": 7343 }, { "epoch": 3.456, "grad_norm": 1.1343680964436667, "learning_rate": 0.0001513589028847768, "loss": 0.4629, "step": 7344 }, { "epoch": 3.456470588235294, "grad_norm": 1.1809828565437261, "learning_rate": 0.00015131506844970596, "loss": 1.1484, "step": 7345 }, { "epoch": 3.4569411764705884, "grad_norm": 0.3729953103400408, "learning_rate": 0.00015127123390232177, "loss": 0.0972, "step": 7346 }, { "epoch": 3.457411764705882, "grad_norm": 0.9564851975642696, "learning_rate": 0.00015122739924636796, "loss": 0.7344, "step": 7347 }, { "epoch": 3.4578823529411764, "grad_norm": 0.983335837263955, "learning_rate": 0.00015118356448558828, "loss": 0.5352, "step": 7348 }, { "epoch": 3.4583529411764706, "grad_norm": 0.8926104809979792, "learning_rate": 0.00015113972962372635, "loss": 0.7109, "step": 7349 }, { "epoch": 3.458823529411765, "grad_norm": 0.9996310601895841, "learning_rate": 0.00015109589466452594, "loss": 1.6094, "step": 7350 }, { "epoch": 3.4592941176470586, "grad_norm": 0.6251286333859339, "learning_rate": 0.00015105205961173074, "loss": 0.1963, "step": 7351 }, { "epoch": 3.459764705882353, "grad_norm": 0.5944677158657006, "learning_rate": 0.00015100822446908452, "loss": 0.3652, "step": 7352 }, { "epoch": 3.460235294117647, "grad_norm": 1.140940747272728, "learning_rate": 0.00015096438924033095, "loss": 0.8242, "step": 7353 }, { "epoch": 3.4607058823529413, "grad_norm": 0.9382565541374487, "learning_rate": 0.00015092055392921383, "loss": 0.7695, "step": 7354 }, { "epoch": 3.461176470588235, "grad_norm": 0.898609715946486, "learning_rate": 0.00015087671853947684, "loss": 0.457, "step": 7355 }, { "epoch": 3.4616470588235293, "grad_norm": 1.287876862540887, "learning_rate": 0.00015083288307486388, "loss": 1.0703, "step": 7356 }, { "epoch": 3.4621176470588235, "grad_norm": 1.0995733002937542, "learning_rate": 0.00015078904753911853, "loss": 0.8203, "step": 7357 }, { "epoch": 3.4625882352941177, "grad_norm": 0.934893681982369, "learning_rate": 0.0001507452119359847, "loss": 0.5156, "step": 7358 }, { "epoch": 3.463058823529412, "grad_norm": 0.8521277445743318, "learning_rate": 0.0001507013762692061, "loss": 0.8086, "step": 7359 }, { "epoch": 3.4635294117647057, "grad_norm": 1.0659306051382924, "learning_rate": 0.0001506575405425265, "loss": 1.2344, "step": 7360 }, { "epoch": 3.464, "grad_norm": 1.0007866282681632, "learning_rate": 0.00015061370475968975, "loss": 0.8281, "step": 7361 }, { "epoch": 3.464470588235294, "grad_norm": 0.9588116546701949, "learning_rate": 0.0001505698689244396, "loss": 0.6992, "step": 7362 }, { "epoch": 3.4649411764705884, "grad_norm": 0.36544224669074826, "learning_rate": 0.0001505260330405198, "loss": 0.2949, "step": 7363 }, { "epoch": 3.465411764705882, "grad_norm": 1.3188153766113315, "learning_rate": 0.00015048219711167422, "loss": 1.1719, "step": 7364 }, { "epoch": 3.4658823529411764, "grad_norm": 1.0436463162230112, "learning_rate": 0.00015043836114164667, "loss": 0.5859, "step": 7365 }, { "epoch": 3.4663529411764706, "grad_norm": 1.0225036498457771, "learning_rate": 0.00015039452513418094, "loss": 0.2285, "step": 7366 }, { "epoch": 3.466823529411765, "grad_norm": 0.8556211150997309, "learning_rate": 0.00015035068909302079, "loss": 0.9102, "step": 7367 }, { "epoch": 3.4672941176470586, "grad_norm": 0.8679016774619034, "learning_rate": 0.00015030685302191011, "loss": 1.2812, "step": 7368 }, { "epoch": 3.467764705882353, "grad_norm": 0.9213132812988257, "learning_rate": 0.00015026301692459265, "loss": 1.3906, "step": 7369 }, { "epoch": 3.468235294117647, "grad_norm": 0.9895781050624509, "learning_rate": 0.00015021918080481225, "loss": 0.9102, "step": 7370 }, { "epoch": 3.4687058823529413, "grad_norm": 1.2034700499990183, "learning_rate": 0.00015017534466631272, "loss": 1.1328, "step": 7371 }, { "epoch": 3.469176470588235, "grad_norm": 1.104443045673689, "learning_rate": 0.00015013150851283795, "loss": 1.5391, "step": 7372 }, { "epoch": 3.4696470588235293, "grad_norm": 0.9595530777660993, "learning_rate": 0.0001500876723481317, "loss": 0.5781, "step": 7373 }, { "epoch": 3.4701176470588235, "grad_norm": 0.6750727904704082, "learning_rate": 0.00015004383617593772, "loss": 0.4375, "step": 7374 }, { "epoch": 3.4705882352941178, "grad_norm": 0.9716368194106891, "learning_rate": 0.00015, "loss": 0.4785, "step": 7375 }, { "epoch": 3.471058823529412, "grad_norm": 1.5988679533384267, "learning_rate": 0.00014995616382406225, "loss": 1.3516, "step": 7376 }, { "epoch": 3.4715294117647058, "grad_norm": 0.8782899906107564, "learning_rate": 0.00014991232765186832, "loss": 0.4746, "step": 7377 }, { "epoch": 3.472, "grad_norm": 0.9009435176160814, "learning_rate": 0.00014986849148716205, "loss": 0.4219, "step": 7378 }, { "epoch": 3.472470588235294, "grad_norm": 0.6349997206609103, "learning_rate": 0.00014982465533368725, "loss": 0.1719, "step": 7379 }, { "epoch": 3.4729411764705884, "grad_norm": 0.6459327026098112, "learning_rate": 0.00014978081919518772, "loss": 0.2129, "step": 7380 }, { "epoch": 3.473411764705882, "grad_norm": 0.6874464128331782, "learning_rate": 0.00014973698307540735, "loss": 0.3379, "step": 7381 }, { "epoch": 3.4738823529411764, "grad_norm": 0.9231008903729245, "learning_rate": 0.00014969314697808991, "loss": 0.5117, "step": 7382 }, { "epoch": 3.4743529411764706, "grad_norm": 0.9423672846745189, "learning_rate": 0.00014964931090697922, "loss": 0.9883, "step": 7383 }, { "epoch": 3.474823529411765, "grad_norm": 0.897535351103932, "learning_rate": 0.00014960547486581906, "loss": 0.8516, "step": 7384 }, { "epoch": 3.4752941176470586, "grad_norm": 1.1068601277156338, "learning_rate": 0.0001495616388583533, "loss": 1.0312, "step": 7385 }, { "epoch": 3.475764705882353, "grad_norm": 0.7015997636897036, "learning_rate": 0.00014951780288832575, "loss": 0.8828, "step": 7386 }, { "epoch": 3.476235294117647, "grad_norm": 1.179099584301101, "learning_rate": 0.0001494739669594802, "loss": 0.9492, "step": 7387 }, { "epoch": 3.4767058823529413, "grad_norm": 1.0720872017162413, "learning_rate": 0.00014943013107556044, "loss": 0.793, "step": 7388 }, { "epoch": 3.477176470588235, "grad_norm": 0.8145344472833285, "learning_rate": 0.00014938629524031028, "loss": 0.5156, "step": 7389 }, { "epoch": 3.4776470588235293, "grad_norm": 0.6385766171010621, "learning_rate": 0.00014934245945747349, "loss": 0.4512, "step": 7390 }, { "epoch": 3.4781176470588235, "grad_norm": 1.067112967390156, "learning_rate": 0.00014929862373079392, "loss": 0.668, "step": 7391 }, { "epoch": 3.4785882352941178, "grad_norm": 1.1119138453091855, "learning_rate": 0.00014925478806401533, "loss": 1.0625, "step": 7392 }, { "epoch": 3.479058823529412, "grad_norm": 1.026906516167177, "learning_rate": 0.0001492109524608815, "loss": 1.0, "step": 7393 }, { "epoch": 3.4795294117647058, "grad_norm": 1.0845256203851101, "learning_rate": 0.00014916711692513618, "loss": 0.7969, "step": 7394 }, { "epoch": 3.48, "grad_norm": 0.5785697295665094, "learning_rate": 0.00014912328146052316, "loss": 0.1895, "step": 7395 }, { "epoch": 3.480470588235294, "grad_norm": 1.2149666480600925, "learning_rate": 0.00014907944607078623, "loss": 0.5234, "step": 7396 }, { "epoch": 3.4809411764705884, "grad_norm": 0.9633180966712864, "learning_rate": 0.00014903561075966908, "loss": 1.5781, "step": 7397 }, { "epoch": 3.481411764705882, "grad_norm": 0.7099506358260204, "learning_rate": 0.00014899177553091543, "loss": 0.291, "step": 7398 }, { "epoch": 3.4818823529411764, "grad_norm": 0.9658399800492969, "learning_rate": 0.0001489479403882692, "loss": 0.9922, "step": 7399 }, { "epoch": 3.4823529411764707, "grad_norm": 1.061185681002683, "learning_rate": 0.00014890410533547404, "loss": 0.5781, "step": 7400 }, { "epoch": 3.482823529411765, "grad_norm": 0.9305151598896385, "learning_rate": 0.00014886027037627362, "loss": 1.3047, "step": 7401 }, { "epoch": 3.4832941176470587, "grad_norm": 0.9218179768735633, "learning_rate": 0.0001488164355144117, "loss": 0.9219, "step": 7402 }, { "epoch": 3.483764705882353, "grad_norm": 0.9287311979692938, "learning_rate": 0.00014877260075363199, "loss": 0.8125, "step": 7403 }, { "epoch": 3.484235294117647, "grad_norm": 0.8959829398285716, "learning_rate": 0.00014872876609767823, "loss": 0.7656, "step": 7404 }, { "epoch": 3.4847058823529413, "grad_norm": 0.8832998665113012, "learning_rate": 0.00014868493155029402, "loss": 0.7188, "step": 7405 }, { "epoch": 3.485176470588235, "grad_norm": 1.0843157877885727, "learning_rate": 0.00014864109711522315, "loss": 1.5156, "step": 7406 }, { "epoch": 3.4856470588235293, "grad_norm": 1.026402146269482, "learning_rate": 0.00014859726279620926, "loss": 0.7305, "step": 7407 }, { "epoch": 3.4861176470588235, "grad_norm": 0.9561953261017146, "learning_rate": 0.00014855342859699597, "loss": 0.5547, "step": 7408 }, { "epoch": 3.4865882352941178, "grad_norm": 0.8011212251009112, "learning_rate": 0.000148509594521327, "loss": 0.957, "step": 7409 }, { "epoch": 3.487058823529412, "grad_norm": 1.0616737245959622, "learning_rate": 0.00014846576057294596, "loss": 1.1641, "step": 7410 }, { "epoch": 3.4875294117647058, "grad_norm": 1.0661901499697246, "learning_rate": 0.0001484219267555965, "loss": 0.5898, "step": 7411 }, { "epoch": 3.488, "grad_norm": 1.5616655156160588, "learning_rate": 0.0001483780930730222, "loss": 1.0391, "step": 7412 }, { "epoch": 3.488470588235294, "grad_norm": 0.9161158227064393, "learning_rate": 0.00014833425952896675, "loss": 0.5156, "step": 7413 }, { "epoch": 3.488941176470588, "grad_norm": 0.7453575756913546, "learning_rate": 0.00014829042612717373, "loss": 0.3906, "step": 7414 }, { "epoch": 3.489411764705882, "grad_norm": 0.8137384012180223, "learning_rate": 0.00014824659287138663, "loss": 0.334, "step": 7415 }, { "epoch": 3.4898823529411764, "grad_norm": 1.020329238282583, "learning_rate": 0.00014820275976534916, "loss": 0.7188, "step": 7416 }, { "epoch": 3.4903529411764707, "grad_norm": 0.8085556014277149, "learning_rate": 0.0001481589268128048, "loss": 0.5664, "step": 7417 }, { "epoch": 3.490823529411765, "grad_norm": 1.0720970732906858, "learning_rate": 0.0001481150940174971, "loss": 0.8789, "step": 7418 }, { "epoch": 3.4912941176470587, "grad_norm": 1.0719564537575528, "learning_rate": 0.00014807126138316964, "loss": 1.2422, "step": 7419 }, { "epoch": 3.491764705882353, "grad_norm": 1.2048420913333102, "learning_rate": 0.00014802742891356592, "loss": 0.793, "step": 7420 }, { "epoch": 3.492235294117647, "grad_norm": 1.1214243845807104, "learning_rate": 0.00014798359661242943, "loss": 1.1016, "step": 7421 }, { "epoch": 3.4927058823529413, "grad_norm": 0.8035604662907321, "learning_rate": 0.00014793976448350366, "loss": 0.8828, "step": 7422 }, { "epoch": 3.493176470588235, "grad_norm": 0.7169484718790774, "learning_rate": 0.00014789593253053217, "loss": 0.4062, "step": 7423 }, { "epoch": 3.4936470588235293, "grad_norm": 0.8349088916123351, "learning_rate": 0.00014785210075725835, "loss": 0.5859, "step": 7424 }, { "epoch": 3.4941176470588236, "grad_norm": 0.6515084118788579, "learning_rate": 0.0001478082691674256, "loss": 0.2285, "step": 7425 }, { "epoch": 3.4945882352941178, "grad_norm": 0.9232797616206819, "learning_rate": 0.00014776443776477746, "loss": 1.1953, "step": 7426 }, { "epoch": 3.495058823529412, "grad_norm": 1.258413647019514, "learning_rate": 0.00014772060655305732, "loss": 1.1172, "step": 7427 }, { "epoch": 3.4955294117647058, "grad_norm": 0.6461627162776183, "learning_rate": 0.0001476767755360085, "loss": 0.4238, "step": 7428 }, { "epoch": 3.496, "grad_norm": 0.44816053642284354, "learning_rate": 0.00014763294471737448, "loss": 0.1035, "step": 7429 }, { "epoch": 3.496470588235294, "grad_norm": 0.8534607692491781, "learning_rate": 0.00014758911410089862, "loss": 0.9102, "step": 7430 }, { "epoch": 3.496941176470588, "grad_norm": 1.148572848943152, "learning_rate": 0.0001475452836903242, "loss": 1.0703, "step": 7431 }, { "epoch": 3.497411764705882, "grad_norm": 1.0192490728426464, "learning_rate": 0.0001475014534893946, "loss": 0.9609, "step": 7432 }, { "epoch": 3.4978823529411764, "grad_norm": 1.1946087273183956, "learning_rate": 0.00014745762350185316, "loss": 1.5469, "step": 7433 }, { "epoch": 3.4983529411764707, "grad_norm": 1.156267830925703, "learning_rate": 0.0001474137937314431, "loss": 0.7344, "step": 7434 }, { "epoch": 3.498823529411765, "grad_norm": 1.027565701760355, "learning_rate": 0.00014736996418190774, "loss": 0.7188, "step": 7435 }, { "epoch": 3.4992941176470587, "grad_norm": 0.6180505140454842, "learning_rate": 0.00014732613485699035, "loss": 0.3281, "step": 7436 }, { "epoch": 3.499764705882353, "grad_norm": 1.29693189850442, "learning_rate": 0.0001472823057604342, "loss": 1.4062, "step": 7437 }, { "epoch": 3.500235294117647, "grad_norm": 0.9276216079673367, "learning_rate": 0.0001472384768959824, "loss": 1.4844, "step": 7438 }, { "epoch": 3.5007058823529413, "grad_norm": 1.3392822212425084, "learning_rate": 0.00014719464826737827, "loss": 1.4922, "step": 7439 }, { "epoch": 3.501176470588235, "grad_norm": 0.8952748938640314, "learning_rate": 0.0001471508198783649, "loss": 0.6719, "step": 7440 }, { "epoch": 3.5016470588235293, "grad_norm": 0.8978159749432334, "learning_rate": 0.0001471069917326855, "loss": 0.5391, "step": 7441 }, { "epoch": 3.5021176470588236, "grad_norm": 0.9581494208653958, "learning_rate": 0.0001470631638340832, "loss": 0.832, "step": 7442 }, { "epoch": 3.502588235294118, "grad_norm": 0.6663150108855361, "learning_rate": 0.00014701933618630107, "loss": 0.5195, "step": 7443 }, { "epoch": 3.503058823529412, "grad_norm": 1.1075580971055365, "learning_rate": 0.00014697550879308233, "loss": 1.3828, "step": 7444 }, { "epoch": 3.503529411764706, "grad_norm": 1.5232908343830156, "learning_rate": 0.00014693168165816995, "loss": 1.1094, "step": 7445 }, { "epoch": 3.504, "grad_norm": 0.5175884080181118, "learning_rate": 0.00014688785478530697, "loss": 0.4883, "step": 7446 }, { "epoch": 3.5044705882352942, "grad_norm": 1.0893923868961346, "learning_rate": 0.00014684402817823647, "loss": 0.6094, "step": 7447 }, { "epoch": 3.504941176470588, "grad_norm": 0.7570512275878253, "learning_rate": 0.00014680020184070148, "loss": 0.3184, "step": 7448 }, { "epoch": 3.5054117647058822, "grad_norm": 1.023750956304533, "learning_rate": 0.0001467563757764449, "loss": 0.6992, "step": 7449 }, { "epoch": 3.5058823529411764, "grad_norm": 1.1023289162802432, "learning_rate": 0.00014671254998920976, "loss": 0.8398, "step": 7450 }, { "epoch": 3.5063529411764707, "grad_norm": 0.6178890454473451, "learning_rate": 0.000146668724482739, "loss": 0.5273, "step": 7451 }, { "epoch": 3.506823529411765, "grad_norm": 0.5973036439916891, "learning_rate": 0.0001466248992607755, "loss": 0.2471, "step": 7452 }, { "epoch": 3.5072941176470587, "grad_norm": 0.7225721773724495, "learning_rate": 0.00014658107432706212, "loss": 0.6211, "step": 7453 }, { "epoch": 3.507764705882353, "grad_norm": 0.8299887875049373, "learning_rate": 0.0001465372496853418, "loss": 0.5781, "step": 7454 }, { "epoch": 3.508235294117647, "grad_norm": 0.966460984926077, "learning_rate": 0.00014649342533935733, "loss": 0.7656, "step": 7455 }, { "epoch": 3.5087058823529413, "grad_norm": 0.4630050361822918, "learning_rate": 0.00014644960129285152, "loss": 0.167, "step": 7456 }, { "epoch": 3.509176470588235, "grad_norm": 0.8254156611902753, "learning_rate": 0.0001464057775495672, "loss": 0.7148, "step": 7457 }, { "epoch": 3.5096470588235293, "grad_norm": 0.5079550451584125, "learning_rate": 0.0001463619541132471, "loss": 0.124, "step": 7458 }, { "epoch": 3.5101176470588236, "grad_norm": 0.9276083847817088, "learning_rate": 0.00014631813098763392, "loss": 1.5312, "step": 7459 }, { "epoch": 3.510588235294118, "grad_norm": 0.865014048913494, "learning_rate": 0.00014627430817647048, "loss": 0.4473, "step": 7460 }, { "epoch": 3.511058823529412, "grad_norm": 1.0477858905145112, "learning_rate": 0.00014623048568349937, "loss": 0.5273, "step": 7461 }, { "epoch": 3.511529411764706, "grad_norm": 1.1808095232671745, "learning_rate": 0.0001461866635124633, "loss": 0.8164, "step": 7462 }, { "epoch": 3.512, "grad_norm": 1.1816513013082706, "learning_rate": 0.0001461428416671048, "loss": 0.6484, "step": 7463 }, { "epoch": 3.5124705882352942, "grad_norm": 1.1798890799326793, "learning_rate": 0.0001460990201511666, "loss": 0.8906, "step": 7464 }, { "epoch": 3.512941176470588, "grad_norm": 0.8287501395373815, "learning_rate": 0.0001460551989683912, "loss": 0.416, "step": 7465 }, { "epoch": 3.5134117647058822, "grad_norm": 0.9845605847520414, "learning_rate": 0.00014601137812252112, "loss": 1.0859, "step": 7466 }, { "epoch": 3.5138823529411765, "grad_norm": 1.3742472853838852, "learning_rate": 0.00014596755761729895, "loss": 1.5625, "step": 7467 }, { "epoch": 3.5143529411764707, "grad_norm": 1.1022378036106457, "learning_rate": 0.00014592373745646715, "loss": 1.2109, "step": 7468 }, { "epoch": 3.514823529411765, "grad_norm": 0.7941164446887472, "learning_rate": 0.00014587991764376812, "loss": 0.2598, "step": 7469 }, { "epoch": 3.5152941176470587, "grad_norm": 1.702297848683833, "learning_rate": 0.00014583609818294438, "loss": 2.0625, "step": 7470 }, { "epoch": 3.515764705882353, "grad_norm": 0.8232495058777304, "learning_rate": 0.00014579227907773827, "loss": 0.9023, "step": 7471 }, { "epoch": 3.516235294117647, "grad_norm": 1.1032032015156306, "learning_rate": 0.00014574846033189217, "loss": 0.6133, "step": 7472 }, { "epoch": 3.516705882352941, "grad_norm": 0.9244808531616998, "learning_rate": 0.00014570464194914835, "loss": 0.6758, "step": 7473 }, { "epoch": 3.517176470588235, "grad_norm": 0.7213485262120435, "learning_rate": 0.00014566082393324925, "loss": 0.9141, "step": 7474 }, { "epoch": 3.5176470588235293, "grad_norm": 0.7190129805970288, "learning_rate": 0.00014561700628793706, "loss": 0.2734, "step": 7475 }, { "epoch": 3.5181176470588236, "grad_norm": 1.0470957816798423, "learning_rate": 0.000145573189016954, "loss": 0.4785, "step": 7476 }, { "epoch": 3.518588235294118, "grad_norm": 0.9360376049637057, "learning_rate": 0.0001455293721240423, "loss": 0.9492, "step": 7477 }, { "epoch": 3.519058823529412, "grad_norm": 0.9981872635057291, "learning_rate": 0.0001454855556129442, "loss": 0.6094, "step": 7478 }, { "epoch": 3.519529411764706, "grad_norm": 1.2367093772426518, "learning_rate": 0.00014544173948740174, "loss": 1.7344, "step": 7479 }, { "epoch": 3.52, "grad_norm": 1.599600776853381, "learning_rate": 0.00014539792375115713, "loss": 1.4453, "step": 7480 }, { "epoch": 3.5204705882352942, "grad_norm": 0.7859335342466699, "learning_rate": 0.0001453541084079524, "loss": 0.7852, "step": 7481 }, { "epoch": 3.520941176470588, "grad_norm": 1.1120362068098673, "learning_rate": 0.0001453102934615296, "loss": 1.0859, "step": 7482 }, { "epoch": 3.5214117647058822, "grad_norm": 0.48768469068801373, "learning_rate": 0.0001452664789156307, "loss": 0.1562, "step": 7483 }, { "epoch": 3.5218823529411765, "grad_norm": 0.9476316412511171, "learning_rate": 0.00014522266477399776, "loss": 0.7656, "step": 7484 }, { "epoch": 3.5223529411764707, "grad_norm": 1.10404105132535, "learning_rate": 0.00014517885104037268, "loss": 1.0703, "step": 7485 }, { "epoch": 3.522823529411765, "grad_norm": 0.8664536738205423, "learning_rate": 0.00014513503771849734, "loss": 1.4062, "step": 7486 }, { "epoch": 3.5232941176470587, "grad_norm": 0.8624846172877604, "learning_rate": 0.00014509122481211358, "loss": 1.1719, "step": 7487 }, { "epoch": 3.523764705882353, "grad_norm": 1.1346887961232999, "learning_rate": 0.0001450474123249634, "loss": 1.0547, "step": 7488 }, { "epoch": 3.524235294117647, "grad_norm": 1.0684136010097556, "learning_rate": 0.00014500360026078847, "loss": 1.2344, "step": 7489 }, { "epoch": 3.524705882352941, "grad_norm": 0.7071518545767361, "learning_rate": 0.00014495978862333063, "loss": 0.4238, "step": 7490 }, { "epoch": 3.525176470588235, "grad_norm": 0.6240828734625227, "learning_rate": 0.00014491597741633148, "loss": 0.2793, "step": 7491 }, { "epoch": 3.5256470588235294, "grad_norm": 0.9954367480464088, "learning_rate": 0.00014487216664353287, "loss": 1.0469, "step": 7492 }, { "epoch": 3.5261176470588236, "grad_norm": 1.3328317111820436, "learning_rate": 0.00014482835630867636, "loss": 1.8359, "step": 7493 }, { "epoch": 3.526588235294118, "grad_norm": 0.5266984987704814, "learning_rate": 0.00014478454641550354, "loss": 0.1533, "step": 7494 }, { "epoch": 3.527058823529412, "grad_norm": 0.5054377752880301, "learning_rate": 0.0001447407369677561, "loss": 0.3926, "step": 7495 }, { "epoch": 3.527529411764706, "grad_norm": 0.9592552733525188, "learning_rate": 0.0001446969279691755, "loss": 0.8086, "step": 7496 }, { "epoch": 3.528, "grad_norm": 1.0492636250765817, "learning_rate": 0.00014465311942350324, "loss": 1.4609, "step": 7497 }, { "epoch": 3.5284705882352942, "grad_norm": 0.8472783592295976, "learning_rate": 0.00014460931133448084, "loss": 1.3047, "step": 7498 }, { "epoch": 3.528941176470588, "grad_norm": 0.7777383812696682, "learning_rate": 0.0001445655037058497, "loss": 0.3145, "step": 7499 }, { "epoch": 3.5294117647058822, "grad_norm": 0.7575382780328886, "learning_rate": 0.00014452169654135115, "loss": 0.5742, "step": 7500 }, { "epoch": 3.5298823529411765, "grad_norm": 1.0268476730354374, "learning_rate": 0.0001444778898447266, "loss": 1.4297, "step": 7501 }, { "epoch": 3.5303529411764707, "grad_norm": 1.0715537793762222, "learning_rate": 0.00014443408361971739, "loss": 0.3516, "step": 7502 }, { "epoch": 3.530823529411765, "grad_norm": 1.4212422346972533, "learning_rate": 0.0001443902778700647, "loss": 1.75, "step": 7503 }, { "epoch": 3.5312941176470587, "grad_norm": 0.924116694705031, "learning_rate": 0.0001443464725995098, "loss": 0.7578, "step": 7504 }, { "epoch": 3.531764705882353, "grad_norm": 0.8007162702069924, "learning_rate": 0.0001443026678117939, "loss": 0.3867, "step": 7505 }, { "epoch": 3.532235294117647, "grad_norm": 0.9150980407115421, "learning_rate": 0.00014425886351065808, "loss": 0.832, "step": 7506 }, { "epoch": 3.532705882352941, "grad_norm": 1.1512408595730217, "learning_rate": 0.00014421505969984346, "loss": 0.9766, "step": 7507 }, { "epoch": 3.533176470588235, "grad_norm": 0.687165347301728, "learning_rate": 0.00014417125638309115, "loss": 0.5977, "step": 7508 }, { "epoch": 3.5336470588235294, "grad_norm": 1.0097025329780505, "learning_rate": 0.00014412745356414213, "loss": 0.8828, "step": 7509 }, { "epoch": 3.5341176470588236, "grad_norm": 0.9274923171412341, "learning_rate": 0.00014408365124673737, "loss": 0.7227, "step": 7510 }, { "epoch": 3.534588235294118, "grad_norm": 1.191648550615007, "learning_rate": 0.00014403984943461783, "loss": 0.7148, "step": 7511 }, { "epoch": 3.535058823529412, "grad_norm": 0.949105938317823, "learning_rate": 0.0001439960481315244, "loss": 0.957, "step": 7512 }, { "epoch": 3.535529411764706, "grad_norm": 1.2542755534594785, "learning_rate": 0.00014395224734119788, "loss": 1.2891, "step": 7513 }, { "epoch": 3.536, "grad_norm": 0.8435099863743468, "learning_rate": 0.00014390844706737907, "loss": 0.707, "step": 7514 }, { "epoch": 3.5364705882352943, "grad_norm": 0.6110307571157386, "learning_rate": 0.00014386464731380881, "loss": 0.1689, "step": 7515 }, { "epoch": 3.536941176470588, "grad_norm": 0.6608565559959894, "learning_rate": 0.00014382084808422775, "loss": 0.4102, "step": 7516 }, { "epoch": 3.5374117647058823, "grad_norm": 1.3671577454586696, "learning_rate": 0.00014377704938237654, "loss": 1.4219, "step": 7517 }, { "epoch": 3.5378823529411765, "grad_norm": 0.909491491519062, "learning_rate": 0.00014373325121199587, "loss": 0.8203, "step": 7518 }, { "epoch": 3.5383529411764707, "grad_norm": 1.7953974727486015, "learning_rate": 0.00014368945357682628, "loss": 1.1875, "step": 7519 }, { "epoch": 3.538823529411765, "grad_norm": 1.316702635337872, "learning_rate": 0.0001436456564806083, "loss": 1.0469, "step": 7520 }, { "epoch": 3.5392941176470587, "grad_norm": 0.7353931571035954, "learning_rate": 0.00014360185992708243, "loss": 0.4785, "step": 7521 }, { "epoch": 3.539764705882353, "grad_norm": 1.0158473312937142, "learning_rate": 0.00014355806391998913, "loss": 1.0703, "step": 7522 }, { "epoch": 3.540235294117647, "grad_norm": 0.6837425343724169, "learning_rate": 0.00014351426846306875, "loss": 0.6914, "step": 7523 }, { "epoch": 3.540705882352941, "grad_norm": 0.7203954182872885, "learning_rate": 0.00014347047356006161, "loss": 0.4785, "step": 7524 }, { "epoch": 3.541176470588235, "grad_norm": 1.1755884769857843, "learning_rate": 0.00014342667921470812, "loss": 1.3047, "step": 7525 }, { "epoch": 3.5416470588235294, "grad_norm": 0.7001224339607277, "learning_rate": 0.00014338288543074845, "loss": 0.3828, "step": 7526 }, { "epoch": 3.5421176470588236, "grad_norm": 0.7056811870572686, "learning_rate": 0.00014333909221192278, "loss": 0.4434, "step": 7527 }, { "epoch": 3.542588235294118, "grad_norm": 1.2243542415790358, "learning_rate": 0.00014329529956197135, "loss": 1.125, "step": 7528 }, { "epoch": 3.543058823529412, "grad_norm": 1.1784082902877007, "learning_rate": 0.00014325150748463423, "loss": 1.1797, "step": 7529 }, { "epoch": 3.543529411764706, "grad_norm": 0.5052718024926894, "learning_rate": 0.00014320771598365142, "loss": 0.2227, "step": 7530 }, { "epoch": 3.544, "grad_norm": 1.3413479972101334, "learning_rate": 0.00014316392506276302, "loss": 1.25, "step": 7531 }, { "epoch": 3.5444705882352943, "grad_norm": 1.2470243702440518, "learning_rate": 0.0001431201347257089, "loss": 1.3594, "step": 7532 }, { "epoch": 3.544941176470588, "grad_norm": 0.7071785078550403, "learning_rate": 0.00014307634497622907, "loss": 0.3516, "step": 7533 }, { "epoch": 3.5454117647058823, "grad_norm": 0.7424290207745049, "learning_rate": 0.00014303255581806336, "loss": 0.6797, "step": 7534 }, { "epoch": 3.5458823529411765, "grad_norm": 0.8848520660238793, "learning_rate": 0.0001429887672549515, "loss": 0.9102, "step": 7535 }, { "epoch": 3.5463529411764707, "grad_norm": 1.27790572590642, "learning_rate": 0.00014294497929063336, "loss": 1.2422, "step": 7536 }, { "epoch": 3.546823529411765, "grad_norm": 0.9198052587993618, "learning_rate": 0.0001429011919288486, "loss": 0.6094, "step": 7537 }, { "epoch": 3.5472941176470587, "grad_norm": 0.8870519618214755, "learning_rate": 0.0001428574051733368, "loss": 0.6602, "step": 7538 }, { "epoch": 3.547764705882353, "grad_norm": 0.9270555619174273, "learning_rate": 0.0001428136190278377, "loss": 0.707, "step": 7539 }, { "epoch": 3.548235294117647, "grad_norm": 1.2360257120843208, "learning_rate": 0.00014276983349609075, "loss": 0.9727, "step": 7540 }, { "epoch": 3.548705882352941, "grad_norm": 1.5015401014213903, "learning_rate": 0.0001427260485818355, "loss": 1.0703, "step": 7541 }, { "epoch": 3.549176470588235, "grad_norm": 0.7374486686712297, "learning_rate": 0.00014268226428881136, "loss": 0.6406, "step": 7542 }, { "epoch": 3.5496470588235294, "grad_norm": 1.1546462986888943, "learning_rate": 0.00014263848062075776, "loss": 1.1484, "step": 7543 }, { "epoch": 3.5501176470588236, "grad_norm": 0.7335057109099952, "learning_rate": 0.00014259469758141403, "loss": 0.707, "step": 7544 }, { "epoch": 3.550588235294118, "grad_norm": 1.7263183506519948, "learning_rate": 0.00014255091517451941, "loss": 2.1562, "step": 7545 }, { "epoch": 3.5510588235294116, "grad_norm": 1.077096141730709, "learning_rate": 0.0001425071334038132, "loss": 1.2266, "step": 7546 }, { "epoch": 3.551529411764706, "grad_norm": 0.9640576044471688, "learning_rate": 0.00014246335227303456, "loss": 0.5273, "step": 7547 }, { "epoch": 3.552, "grad_norm": 1.1495843829479016, "learning_rate": 0.00014241957178592258, "loss": 0.875, "step": 7548 }, { "epoch": 3.5524705882352943, "grad_norm": 0.9293264190141615, "learning_rate": 0.00014237579194621637, "loss": 1.0078, "step": 7549 }, { "epoch": 3.552941176470588, "grad_norm": 0.8554145872287385, "learning_rate": 0.00014233201275765494, "loss": 0.3203, "step": 7550 }, { "epoch": 3.5534117647058823, "grad_norm": 1.245865053829095, "learning_rate": 0.0001422882342239772, "loss": 1.2578, "step": 7551 }, { "epoch": 3.5538823529411765, "grad_norm": 1.0325967320061225, "learning_rate": 0.00014224445634892206, "loss": 1.1562, "step": 7552 }, { "epoch": 3.5543529411764707, "grad_norm": 0.574734051582859, "learning_rate": 0.00014220067913622844, "loss": 0.1602, "step": 7553 }, { "epoch": 3.554823529411765, "grad_norm": 1.3015203451709365, "learning_rate": 0.00014215690258963508, "loss": 0.9102, "step": 7554 }, { "epoch": 3.5552941176470587, "grad_norm": 1.0581903349306294, "learning_rate": 0.00014211312671288064, "loss": 0.875, "step": 7555 }, { "epoch": 3.555764705882353, "grad_norm": 0.982953315162556, "learning_rate": 0.00014206935150970392, "loss": 0.5703, "step": 7556 }, { "epoch": 3.556235294117647, "grad_norm": 0.6617130378328009, "learning_rate": 0.00014202557698384348, "loss": 0.2441, "step": 7557 }, { "epoch": 3.556705882352941, "grad_norm": 0.8886221637098826, "learning_rate": 0.00014198180313903784, "loss": 0.75, "step": 7558 }, { "epoch": 3.557176470588235, "grad_norm": 0.8930885858482371, "learning_rate": 0.0001419380299790256, "loss": 0.5898, "step": 7559 }, { "epoch": 3.5576470588235294, "grad_norm": 1.2468935744166534, "learning_rate": 0.00014189425750754512, "loss": 1.3516, "step": 7560 }, { "epoch": 3.5581176470588236, "grad_norm": 0.7312451400019055, "learning_rate": 0.00014185048572833478, "loss": 0.418, "step": 7561 }, { "epoch": 3.558588235294118, "grad_norm": 1.165612854584037, "learning_rate": 0.00014180671464513296, "loss": 1.125, "step": 7562 }, { "epoch": 3.5590588235294116, "grad_norm": 1.2585250198386022, "learning_rate": 0.0001417629442616779, "loss": 1.4297, "step": 7563 }, { "epoch": 3.559529411764706, "grad_norm": 0.5974806071893747, "learning_rate": 0.00014171917458170782, "loss": 0.6094, "step": 7564 }, { "epoch": 3.56, "grad_norm": 1.123357298841047, "learning_rate": 0.0001416754056089608, "loss": 0.6602, "step": 7565 }, { "epoch": 3.5604705882352943, "grad_norm": 0.7475954060638963, "learning_rate": 0.000141631637347175, "loss": 0.2988, "step": 7566 }, { "epoch": 3.560941176470588, "grad_norm": 0.8489590305390821, "learning_rate": 0.0001415878698000884, "loss": 0.8828, "step": 7567 }, { "epoch": 3.5614117647058823, "grad_norm": 1.222862422768492, "learning_rate": 0.00014154410297143898, "loss": 0.7461, "step": 7568 }, { "epoch": 3.5618823529411765, "grad_norm": 0.43672965757348187, "learning_rate": 0.00014150033686496469, "loss": 0.4473, "step": 7569 }, { "epoch": 3.5623529411764707, "grad_norm": 1.0142109679209248, "learning_rate": 0.0001414565714844033, "loss": 0.9609, "step": 7570 }, { "epoch": 3.562823529411765, "grad_norm": 1.4992530419614611, "learning_rate": 0.00014141280683349253, "loss": 2.1875, "step": 7571 }, { "epoch": 3.5632941176470587, "grad_norm": 1.075156529816739, "learning_rate": 0.00014136904291597028, "loss": 0.9844, "step": 7572 }, { "epoch": 3.563764705882353, "grad_norm": 1.1763974578920946, "learning_rate": 0.00014132527973557407, "loss": 0.7891, "step": 7573 }, { "epoch": 3.564235294117647, "grad_norm": 0.8554910384750025, "learning_rate": 0.00014128151729604152, "loss": 1.1094, "step": 7574 }, { "epoch": 3.564705882352941, "grad_norm": 1.4013407948484133, "learning_rate": 0.00014123775560111009, "loss": 0.9453, "step": 7575 }, { "epoch": 3.565176470588235, "grad_norm": 0.8891298618205153, "learning_rate": 0.00014119399465451738, "loss": 0.8945, "step": 7576 }, { "epoch": 3.5656470588235294, "grad_norm": 0.8932905838512197, "learning_rate": 0.0001411502344600007, "loss": 0.4004, "step": 7577 }, { "epoch": 3.5661176470588236, "grad_norm": 0.5690330153136799, "learning_rate": 0.00014110647502129742, "loss": 0.1982, "step": 7578 }, { "epoch": 3.566588235294118, "grad_norm": 1.180777134043347, "learning_rate": 0.00014106271634214477, "loss": 0.8789, "step": 7579 }, { "epoch": 3.5670588235294116, "grad_norm": 0.7351452556593586, "learning_rate": 0.00014101895842628001, "loss": 0.2988, "step": 7580 }, { "epoch": 3.567529411764706, "grad_norm": 0.7413551488189278, "learning_rate": 0.00014097520127744024, "loss": 0.3203, "step": 7581 }, { "epoch": 3.568, "grad_norm": 0.9877700627805057, "learning_rate": 0.00014093144489936259, "loss": 0.9883, "step": 7582 }, { "epoch": 3.5684705882352943, "grad_norm": 0.9229164354454158, "learning_rate": 0.00014088768929578397, "loss": 0.793, "step": 7583 }, { "epoch": 3.568941176470588, "grad_norm": 1.0239004476551632, "learning_rate": 0.00014084393447044138, "loss": 0.582, "step": 7584 }, { "epoch": 3.5694117647058823, "grad_norm": 0.849549511136911, "learning_rate": 0.00014080018042707173, "loss": 0.3164, "step": 7585 }, { "epoch": 3.5698823529411765, "grad_norm": 0.36317867899461553, "learning_rate": 0.00014075642716941176, "loss": 0.1396, "step": 7586 }, { "epoch": 3.5703529411764707, "grad_norm": 0.8863656641882923, "learning_rate": 0.0001407126747011983, "loss": 0.3691, "step": 7587 }, { "epoch": 3.570823529411765, "grad_norm": 1.162240367562745, "learning_rate": 0.00014066892302616794, "loss": 1.6016, "step": 7588 }, { "epoch": 3.5712941176470587, "grad_norm": 1.1784945910324485, "learning_rate": 0.0001406251721480573, "loss": 0.2217, "step": 7589 }, { "epoch": 3.571764705882353, "grad_norm": 1.133183860394275, "learning_rate": 0.000140581422070603, "loss": 1.3984, "step": 7590 }, { "epoch": 3.572235294117647, "grad_norm": 0.8732636153326596, "learning_rate": 0.0001405376727975414, "loss": 1.1016, "step": 7591 }, { "epoch": 3.572705882352941, "grad_norm": 0.6574576560574659, "learning_rate": 0.000140493924332609, "loss": 0.21, "step": 7592 }, { "epoch": 3.573176470588235, "grad_norm": 1.2419749208892208, "learning_rate": 0.00014045017667954203, "loss": 0.7617, "step": 7593 }, { "epoch": 3.5736470588235294, "grad_norm": 1.457482575037173, "learning_rate": 0.00014040642984207687, "loss": 1.0391, "step": 7594 }, { "epoch": 3.5741176470588236, "grad_norm": 0.9347720816831644, "learning_rate": 0.00014036268382394964, "loss": 0.6914, "step": 7595 }, { "epoch": 3.574588235294118, "grad_norm": 1.146257149770789, "learning_rate": 0.00014031893862889646, "loss": 0.9062, "step": 7596 }, { "epoch": 3.5750588235294116, "grad_norm": 1.0489385131413953, "learning_rate": 0.00014027519426065344, "loss": 1.1016, "step": 7597 }, { "epoch": 3.575529411764706, "grad_norm": 0.962532187572642, "learning_rate": 0.00014023145072295654, "loss": 0.8672, "step": 7598 }, { "epoch": 3.576, "grad_norm": 0.9670109984138355, "learning_rate": 0.0001401877080195416, "loss": 0.5156, "step": 7599 }, { "epoch": 3.576470588235294, "grad_norm": 0.7138416425560606, "learning_rate": 0.00014014396615414458, "loss": 0.3496, "step": 7600 }, { "epoch": 3.576941176470588, "grad_norm": 0.4548243534669854, "learning_rate": 0.0001401002251305012, "loss": 0.1045, "step": 7601 }, { "epoch": 3.5774117647058823, "grad_norm": 1.1338062771294997, "learning_rate": 0.0001400564849523471, "loss": 0.582, "step": 7602 }, { "epoch": 3.5778823529411765, "grad_norm": 1.1172524066714007, "learning_rate": 0.00014001274562341798, "loss": 1.2812, "step": 7603 }, { "epoch": 3.5783529411764707, "grad_norm": 1.396280894397082, "learning_rate": 0.0001399690071474494, "loss": 1.5, "step": 7604 }, { "epoch": 3.578823529411765, "grad_norm": 0.6598511199606525, "learning_rate": 0.0001399252695281768, "loss": 0.3477, "step": 7605 }, { "epoch": 3.5792941176470587, "grad_norm": 1.1875052467875635, "learning_rate": 0.00013988153276933555, "loss": 1.0156, "step": 7606 }, { "epoch": 3.579764705882353, "grad_norm": 1.2706390674540098, "learning_rate": 0.0001398377968746611, "loss": 1.3672, "step": 7607 }, { "epoch": 3.580235294117647, "grad_norm": 1.0015655621511617, "learning_rate": 0.0001397940618478886, "loss": 0.6719, "step": 7608 }, { "epoch": 3.580705882352941, "grad_norm": 0.7286389450059148, "learning_rate": 0.00013975032769275326, "loss": 0.6719, "step": 7609 }, { "epoch": 3.581176470588235, "grad_norm": 1.3270172381843415, "learning_rate": 0.00013970659441299026, "loss": 1.9219, "step": 7610 }, { "epoch": 3.5816470588235294, "grad_norm": 0.9965866833332447, "learning_rate": 0.00013966286201233457, "loss": 1.5469, "step": 7611 }, { "epoch": 3.5821176470588236, "grad_norm": 0.73103121381718, "learning_rate": 0.00013961913049452115, "loss": 0.3086, "step": 7612 }, { "epoch": 3.582588235294118, "grad_norm": 1.0339601976696118, "learning_rate": 0.00013957539986328492, "loss": 0.9023, "step": 7613 }, { "epoch": 3.5830588235294116, "grad_norm": 1.1654865185023138, "learning_rate": 0.00013953167012236068, "loss": 0.875, "step": 7614 }, { "epoch": 3.583529411764706, "grad_norm": 1.0650374442376287, "learning_rate": 0.00013948794127548316, "loss": 1.0781, "step": 7615 }, { "epoch": 3.584, "grad_norm": 0.7881770708879012, "learning_rate": 0.00013944421332638698, "loss": 0.4824, "step": 7616 }, { "epoch": 3.584470588235294, "grad_norm": 1.2018225260357167, "learning_rate": 0.0001394004862788068, "loss": 1.4219, "step": 7617 }, { "epoch": 3.584941176470588, "grad_norm": 0.7298554361312134, "learning_rate": 0.00013935676013647708, "loss": 0.375, "step": 7618 }, { "epoch": 3.5854117647058823, "grad_norm": 1.01440856990772, "learning_rate": 0.0001393130349031322, "loss": 1.0469, "step": 7619 }, { "epoch": 3.5858823529411765, "grad_norm": 0.66817862744067, "learning_rate": 0.00013926931058250657, "loss": 0.4043, "step": 7620 }, { "epoch": 3.5863529411764707, "grad_norm": 0.9698486609772101, "learning_rate": 0.00013922558717833445, "loss": 0.9688, "step": 7621 }, { "epoch": 3.586823529411765, "grad_norm": 1.2888608630568106, "learning_rate": 0.00013918186469435005, "loss": 0.8008, "step": 7622 }, { "epoch": 3.5872941176470587, "grad_norm": 1.0346267491467196, "learning_rate": 0.00013913814313428748, "loss": 1.0625, "step": 7623 }, { "epoch": 3.587764705882353, "grad_norm": 0.3719659219734529, "learning_rate": 0.00013909442250188074, "loss": 0.1089, "step": 7624 }, { "epoch": 3.588235294117647, "grad_norm": 0.7888498721165785, "learning_rate": 0.00013905070280086386, "loss": 0.3535, "step": 7625 }, { "epoch": 3.588705882352941, "grad_norm": 1.5602825697748233, "learning_rate": 0.00013900698403497062, "loss": 1.5156, "step": 7626 }, { "epoch": 3.589176470588235, "grad_norm": 1.4244589251015674, "learning_rate": 0.00013896326620793488, "loss": 0.6953, "step": 7627 }, { "epoch": 3.5896470588235294, "grad_norm": 0.7861234587145559, "learning_rate": 0.00013891954932349033, "loss": 0.3281, "step": 7628 }, { "epoch": 3.5901176470588236, "grad_norm": 1.0384886217636897, "learning_rate": 0.00013887583338537064, "loss": 0.7656, "step": 7629 }, { "epoch": 3.590588235294118, "grad_norm": 1.3050258726450676, "learning_rate": 0.0001388321183973093, "loss": 1.2656, "step": 7630 }, { "epoch": 3.5910588235294116, "grad_norm": 0.6496913157553779, "learning_rate": 0.0001387884043630399, "loss": 0.5508, "step": 7631 }, { "epoch": 3.591529411764706, "grad_norm": 1.047463412433807, "learning_rate": 0.0001387446912862957, "loss": 0.8594, "step": 7632 }, { "epoch": 3.592, "grad_norm": 0.8389021470105864, "learning_rate": 0.0001387009791708101, "loss": 1.0391, "step": 7633 }, { "epoch": 3.592470588235294, "grad_norm": 1.078170615450321, "learning_rate": 0.00013865726802031627, "loss": 0.8086, "step": 7634 }, { "epoch": 3.592941176470588, "grad_norm": 0.860999689273862, "learning_rate": 0.0001386135578385474, "loss": 0.5625, "step": 7635 }, { "epoch": 3.5934117647058823, "grad_norm": 1.1336763717032035, "learning_rate": 0.00013856984862923655, "loss": 1.0156, "step": 7636 }, { "epoch": 3.5938823529411765, "grad_norm": 1.0350251352899855, "learning_rate": 0.00013852614039611667, "loss": 1.0547, "step": 7637 }, { "epoch": 3.5943529411764708, "grad_norm": 1.036537939739563, "learning_rate": 0.0001384824331429207, "loss": 0.8711, "step": 7638 }, { "epoch": 3.594823529411765, "grad_norm": 1.2095692977385115, "learning_rate": 0.00013843872687338142, "loss": 0.7266, "step": 7639 }, { "epoch": 3.5952941176470588, "grad_norm": 0.3724390787986809, "learning_rate": 0.00013839502159123158, "loss": 0.1533, "step": 7640 }, { "epoch": 3.595764705882353, "grad_norm": 1.2106903043643558, "learning_rate": 0.00013835131730020382, "loss": 0.9062, "step": 7641 }, { "epoch": 3.596235294117647, "grad_norm": 0.8529403242222944, "learning_rate": 0.0001383076140040307, "loss": 0.3594, "step": 7642 }, { "epoch": 3.596705882352941, "grad_norm": 1.0211391648538408, "learning_rate": 0.0001382639117064447, "loss": 1.0781, "step": 7643 }, { "epoch": 3.597176470588235, "grad_norm": 1.182629349783044, "learning_rate": 0.0001382202104111782, "loss": 0.9219, "step": 7644 }, { "epoch": 3.5976470588235294, "grad_norm": 0.7857245566055852, "learning_rate": 0.0001381765101219635, "loss": 0.3281, "step": 7645 }, { "epoch": 3.5981176470588236, "grad_norm": 0.7306593200120165, "learning_rate": 0.00013813281084253286, "loss": 0.3125, "step": 7646 }, { "epoch": 3.598588235294118, "grad_norm": 1.0955143776515082, "learning_rate": 0.00013808911257661834, "loss": 0.9492, "step": 7647 }, { "epoch": 3.5990588235294116, "grad_norm": 1.0493121768724096, "learning_rate": 0.00013804541532795204, "loss": 0.5781, "step": 7648 }, { "epoch": 3.599529411764706, "grad_norm": 0.8126597646910685, "learning_rate": 0.00013800171910026594, "loss": 0.2949, "step": 7649 }, { "epoch": 3.6, "grad_norm": 1.4059379467657314, "learning_rate": 0.00013795802389729184, "loss": 1.3516, "step": 7650 }, { "epoch": 3.600470588235294, "grad_norm": 0.9871410807703561, "learning_rate": 0.0001379143297227616, "loss": 0.8242, "step": 7651 }, { "epoch": 3.600941176470588, "grad_norm": 1.0219541697369732, "learning_rate": 0.00013787063658040686, "loss": 1.1016, "step": 7652 }, { "epoch": 3.6014117647058823, "grad_norm": 0.8997926804831065, "learning_rate": 0.00013782694447395926, "loss": 0.5625, "step": 7653 }, { "epoch": 3.6018823529411765, "grad_norm": 0.654784489461791, "learning_rate": 0.0001377832534071503, "loss": 0.3184, "step": 7654 }, { "epoch": 3.6023529411764708, "grad_norm": 1.0180898295036211, "learning_rate": 0.0001377395633837114, "loss": 0.6914, "step": 7655 }, { "epoch": 3.602823529411765, "grad_norm": 1.00339862220104, "learning_rate": 0.000137695874407374, "loss": 1.2891, "step": 7656 }, { "epoch": 3.6032941176470588, "grad_norm": 1.020585900536795, "learning_rate": 0.0001376521864818692, "loss": 0.7305, "step": 7657 }, { "epoch": 3.603764705882353, "grad_norm": 1.1236092506058928, "learning_rate": 0.00013760849961092826, "loss": 1.0, "step": 7658 }, { "epoch": 3.604235294117647, "grad_norm": 0.31385452817804516, "learning_rate": 0.00013756481379828225, "loss": 0.0579, "step": 7659 }, { "epoch": 3.604705882352941, "grad_norm": 0.8673042535555724, "learning_rate": 0.00013752112904766208, "loss": 0.7305, "step": 7660 }, { "epoch": 3.605176470588235, "grad_norm": 1.1787828078435383, "learning_rate": 0.00013747744536279878, "loss": 1.6719, "step": 7661 }, { "epoch": 3.6056470588235294, "grad_norm": 1.0387152432988271, "learning_rate": 0.00013743376274742304, "loss": 0.6445, "step": 7662 }, { "epoch": 3.6061176470588236, "grad_norm": 0.9739112749971714, "learning_rate": 0.00013739008120526559, "loss": 0.9688, "step": 7663 }, { "epoch": 3.606588235294118, "grad_norm": 0.9888804194408317, "learning_rate": 0.00013734640074005702, "loss": 1.1719, "step": 7664 }, { "epoch": 3.6070588235294117, "grad_norm": 0.9870852480583562, "learning_rate": 0.0001373027213555279, "loss": 0.8398, "step": 7665 }, { "epoch": 3.607529411764706, "grad_norm": 0.45383216336655324, "learning_rate": 0.0001372590430554087, "loss": 0.2832, "step": 7666 }, { "epoch": 3.608, "grad_norm": 1.0330082444787767, "learning_rate": 0.0001372153658434297, "loss": 1.5625, "step": 7667 }, { "epoch": 3.608470588235294, "grad_norm": 1.0144663419214628, "learning_rate": 0.00013717168972332116, "loss": 0.8633, "step": 7668 }, { "epoch": 3.608941176470588, "grad_norm": 0.8539957689447394, "learning_rate": 0.00013712801469881324, "loss": 0.6172, "step": 7669 }, { "epoch": 3.6094117647058823, "grad_norm": 1.7363127619412668, "learning_rate": 0.00013708434077363603, "loss": 2.0625, "step": 7670 }, { "epoch": 3.6098823529411765, "grad_norm": 1.3803705463167737, "learning_rate": 0.00013704066795151946, "loss": 1.2812, "step": 7671 }, { "epoch": 3.6103529411764708, "grad_norm": 0.9270682632365749, "learning_rate": 0.00013699699623619338, "loss": 0.7852, "step": 7672 }, { "epoch": 3.6108235294117645, "grad_norm": 1.0246512319438354, "learning_rate": 0.00013695332563138762, "loss": 0.6094, "step": 7673 }, { "epoch": 3.6112941176470588, "grad_norm": 1.321968976102578, "learning_rate": 0.00013690965614083185, "loss": 0.9805, "step": 7674 }, { "epoch": 3.611764705882353, "grad_norm": 1.0468654839290885, "learning_rate": 0.00013686598776825563, "loss": 0.7812, "step": 7675 }, { "epoch": 3.612235294117647, "grad_norm": 1.2273361333079031, "learning_rate": 0.00013682232051738852, "loss": 0.9023, "step": 7676 }, { "epoch": 3.612705882352941, "grad_norm": 1.0376076892423252, "learning_rate": 0.00013677865439195986, "loss": 1.0391, "step": 7677 }, { "epoch": 3.613176470588235, "grad_norm": 1.328617509756435, "learning_rate": 0.00013673498939569892, "loss": 1.8672, "step": 7678 }, { "epoch": 3.6136470588235294, "grad_norm": 1.3232408199000554, "learning_rate": 0.000136691325532335, "loss": 1.6406, "step": 7679 }, { "epoch": 3.6141176470588237, "grad_norm": 0.9759437560424564, "learning_rate": 0.00013664766280559716, "loss": 0.6836, "step": 7680 }, { "epoch": 3.614588235294118, "grad_norm": 1.0454807214742772, "learning_rate": 0.00013660400121921434, "loss": 0.4199, "step": 7681 }, { "epoch": 3.6150588235294117, "grad_norm": 0.8418578853849935, "learning_rate": 0.00013656034077691562, "loss": 0.8867, "step": 7682 }, { "epoch": 3.615529411764706, "grad_norm": 1.2603739189517673, "learning_rate": 0.00013651668148242967, "loss": 1.7188, "step": 7683 }, { "epoch": 3.616, "grad_norm": 1.2007470967553109, "learning_rate": 0.00013647302333948528, "loss": 0.8984, "step": 7684 }, { "epoch": 3.616470588235294, "grad_norm": 1.2113018025146365, "learning_rate": 0.000136429366351811, "loss": 1.1016, "step": 7685 }, { "epoch": 3.616941176470588, "grad_norm": 1.328781126912191, "learning_rate": 0.00013638571052313543, "loss": 0.8047, "step": 7686 }, { "epoch": 3.6174117647058823, "grad_norm": 1.0168332355876237, "learning_rate": 0.00013634205585718698, "loss": 0.5469, "step": 7687 }, { "epoch": 3.6178823529411765, "grad_norm": 0.8135950313569, "learning_rate": 0.00013629840235769388, "loss": 1.25, "step": 7688 }, { "epoch": 3.6183529411764708, "grad_norm": 0.8815327993188126, "learning_rate": 0.0001362547500283845, "loss": 0.4453, "step": 7689 }, { "epoch": 3.6188235294117646, "grad_norm": 1.0012163039169069, "learning_rate": 0.00013621109887298685, "loss": 1.2734, "step": 7690 }, { "epoch": 3.6192941176470588, "grad_norm": 0.7621118964645581, "learning_rate": 0.00013616744889522898, "loss": 0.4492, "step": 7691 }, { "epoch": 3.619764705882353, "grad_norm": 0.9681741713261494, "learning_rate": 0.00013612380009883886, "loss": 0.8438, "step": 7692 }, { "epoch": 3.620235294117647, "grad_norm": 1.1991316912058523, "learning_rate": 0.00013608015248754425, "loss": 1.3047, "step": 7693 }, { "epoch": 3.620705882352941, "grad_norm": 0.928646262920153, "learning_rate": 0.00013603650606507293, "loss": 0.3926, "step": 7694 }, { "epoch": 3.621176470588235, "grad_norm": 0.90231100206483, "learning_rate": 0.0001359928608351524, "loss": 0.9023, "step": 7695 }, { "epoch": 3.6216470588235294, "grad_norm": 0.6371682363947458, "learning_rate": 0.00013594921680151032, "loss": 0.6953, "step": 7696 }, { "epoch": 3.6221176470588237, "grad_norm": 0.8107942731725754, "learning_rate": 0.00013590557396787406, "loss": 0.6133, "step": 7697 }, { "epoch": 3.622588235294118, "grad_norm": 0.9159307078668495, "learning_rate": 0.00013586193233797086, "loss": 0.5391, "step": 7698 }, { "epoch": 3.6230588235294117, "grad_norm": 0.6350507636836514, "learning_rate": 0.00013581829191552802, "loss": 0.4902, "step": 7699 }, { "epoch": 3.623529411764706, "grad_norm": 0.9026518880597639, "learning_rate": 0.00013577465270427262, "loss": 0.4492, "step": 7700 }, { "epoch": 3.624, "grad_norm": 0.5054968290032879, "learning_rate": 0.00013573101470793161, "loss": 0.2559, "step": 7701 }, { "epoch": 3.624470588235294, "grad_norm": 1.167492471560258, "learning_rate": 0.00013568737793023198, "loss": 0.8867, "step": 7702 }, { "epoch": 3.624941176470588, "grad_norm": 0.7485673671470348, "learning_rate": 0.00013564374237490047, "loss": 0.2539, "step": 7703 }, { "epoch": 3.6254117647058823, "grad_norm": 0.7744241928924899, "learning_rate": 0.0001356001080456638, "loss": 0.2656, "step": 7704 }, { "epoch": 3.6258823529411766, "grad_norm": 0.850468670787224, "learning_rate": 0.00013555647494624846, "loss": 0.5273, "step": 7705 }, { "epoch": 3.626352941176471, "grad_norm": 1.0612057287418761, "learning_rate": 0.00013551284308038105, "loss": 1.2266, "step": 7706 }, { "epoch": 3.6268235294117646, "grad_norm": 1.1590536189107625, "learning_rate": 0.00013546921245178792, "loss": 0.9805, "step": 7707 }, { "epoch": 3.627294117647059, "grad_norm": 0.9232074115126762, "learning_rate": 0.00013542558306419526, "loss": 0.5039, "step": 7708 }, { "epoch": 3.627764705882353, "grad_norm": 0.8673871988734797, "learning_rate": 0.00013538195492132928, "loss": 0.5938, "step": 7709 }, { "epoch": 3.6282352941176472, "grad_norm": 0.9566752420757584, "learning_rate": 0.00013533832802691613, "loss": 1.1016, "step": 7710 }, { "epoch": 3.628705882352941, "grad_norm": 1.0300490870575594, "learning_rate": 0.00013529470238468166, "loss": 1.3359, "step": 7711 }, { "epoch": 3.6291764705882352, "grad_norm": 0.7492673987789008, "learning_rate": 0.00013525107799835172, "loss": 0.2988, "step": 7712 }, { "epoch": 3.6296470588235294, "grad_norm": 0.7199618549155828, "learning_rate": 0.00013520745487165207, "loss": 0.2363, "step": 7713 }, { "epoch": 3.6301176470588237, "grad_norm": 0.8806658621388745, "learning_rate": 0.00013516383300830835, "loss": 0.8164, "step": 7714 }, { "epoch": 3.630588235294118, "grad_norm": 1.0254926558893611, "learning_rate": 0.00013512021241204604, "loss": 1.2344, "step": 7715 }, { "epoch": 3.6310588235294117, "grad_norm": 0.5443595229553345, "learning_rate": 0.00013507659308659053, "loss": 0.1973, "step": 7716 }, { "epoch": 3.631529411764706, "grad_norm": 1.445059681421171, "learning_rate": 0.00013503297503566727, "loss": 1.4453, "step": 7717 }, { "epoch": 3.632, "grad_norm": 0.8525373633686151, "learning_rate": 0.00013498935826300132, "loss": 0.6445, "step": 7718 }, { "epoch": 3.632470588235294, "grad_norm": 1.03227078992756, "learning_rate": 0.00013494574277231772, "loss": 0.6836, "step": 7719 }, { "epoch": 3.632941176470588, "grad_norm": 0.5194555771162769, "learning_rate": 0.00013490212856734162, "loss": 0.3184, "step": 7720 }, { "epoch": 3.6334117647058823, "grad_norm": 0.8802786948324607, "learning_rate": 0.00013485851565179778, "loss": 0.7695, "step": 7721 }, { "epoch": 3.6338823529411766, "grad_norm": 0.7165799696938548, "learning_rate": 0.00013481490402941098, "loss": 0.4609, "step": 7722 }, { "epoch": 3.634352941176471, "grad_norm": 0.7743523887574097, "learning_rate": 0.00013477129370390582, "loss": 0.4902, "step": 7723 }, { "epoch": 3.6348235294117646, "grad_norm": 0.7640076332611555, "learning_rate": 0.00013472768467900686, "loss": 1.2109, "step": 7724 }, { "epoch": 3.635294117647059, "grad_norm": 0.7925372702439955, "learning_rate": 0.0001346840769584386, "loss": 0.7305, "step": 7725 }, { "epoch": 3.635764705882353, "grad_norm": 1.0274648831991027, "learning_rate": 0.00013464047054592525, "loss": 0.9453, "step": 7726 }, { "epoch": 3.636235294117647, "grad_norm": 0.8954477989264203, "learning_rate": 0.00013459686544519108, "loss": 0.4707, "step": 7727 }, { "epoch": 3.636705882352941, "grad_norm": 0.8055220997341977, "learning_rate": 0.00013455326165996015, "loss": 0.5, "step": 7728 }, { "epoch": 3.6371764705882352, "grad_norm": 0.6475273776546261, "learning_rate": 0.0001345096591939564, "loss": 0.2324, "step": 7729 }, { "epoch": 3.6376470588235295, "grad_norm": 1.10221374393742, "learning_rate": 0.00013446605805090376, "loss": 1.0, "step": 7730 }, { "epoch": 3.6381176470588237, "grad_norm": 0.48253262243954537, "learning_rate": 0.000134422458234526, "loss": 0.1523, "step": 7731 }, { "epoch": 3.638588235294118, "grad_norm": 0.8449434235865037, "learning_rate": 0.00013437885974854665, "loss": 0.4199, "step": 7732 }, { "epoch": 3.6390588235294117, "grad_norm": 1.0334487526587102, "learning_rate": 0.00013433526259668931, "loss": 0.7109, "step": 7733 }, { "epoch": 3.639529411764706, "grad_norm": 0.8913736813057999, "learning_rate": 0.0001342916667826774, "loss": 0.4883, "step": 7734 }, { "epoch": 3.64, "grad_norm": 0.6018709342335832, "learning_rate": 0.0001342480723102342, "loss": 0.3984, "step": 7735 }, { "epoch": 3.640470588235294, "grad_norm": 0.830865214404191, "learning_rate": 0.00013420447918308285, "loss": 0.332, "step": 7736 }, { "epoch": 3.640941176470588, "grad_norm": 0.5629981324706219, "learning_rate": 0.00013416088740494649, "loss": 0.5508, "step": 7737 }, { "epoch": 3.6414117647058823, "grad_norm": 1.2234385781933836, "learning_rate": 0.00013411729697954803, "loss": 0.9688, "step": 7738 }, { "epoch": 3.6418823529411766, "grad_norm": 0.8059974294832009, "learning_rate": 0.00013407370791061027, "loss": 0.4199, "step": 7739 }, { "epoch": 3.642352941176471, "grad_norm": 0.7956928494292469, "learning_rate": 0.00013403012020185603, "loss": 0.4902, "step": 7740 }, { "epoch": 3.6428235294117646, "grad_norm": 0.6174975564773058, "learning_rate": 0.00013398653385700784, "loss": 0.6602, "step": 7741 }, { "epoch": 3.643294117647059, "grad_norm": 0.7411613336853632, "learning_rate": 0.00013394294887978818, "loss": 0.7578, "step": 7742 }, { "epoch": 3.643764705882353, "grad_norm": 1.012818585841034, "learning_rate": 0.0001338993652739195, "loss": 0.9375, "step": 7743 }, { "epoch": 3.644235294117647, "grad_norm": 0.8336199891751903, "learning_rate": 0.00013385578304312397, "loss": 0.4648, "step": 7744 }, { "epoch": 3.644705882352941, "grad_norm": 1.1026221777078882, "learning_rate": 0.00013381220219112373, "loss": 0.9297, "step": 7745 }, { "epoch": 3.6451764705882352, "grad_norm": 0.7085736410541413, "learning_rate": 0.00013376862272164084, "loss": 0.3438, "step": 7746 }, { "epoch": 3.6456470588235295, "grad_norm": 1.2224780672125048, "learning_rate": 0.00013372504463839718, "loss": 0.8906, "step": 7747 }, { "epoch": 3.6461176470588237, "grad_norm": 0.8692755851461997, "learning_rate": 0.00013368146794511457, "loss": 0.8633, "step": 7748 }, { "epoch": 3.646588235294118, "grad_norm": 0.7876752161028544, "learning_rate": 0.0001336378926455146, "loss": 0.5938, "step": 7749 }, { "epoch": 3.6470588235294117, "grad_norm": 0.36898666705312977, "learning_rate": 0.00013359431874331886, "loss": 0.0605, "step": 7750 }, { "epoch": 3.647529411764706, "grad_norm": 0.9573100448265035, "learning_rate": 0.0001335507462422488, "loss": 0.4863, "step": 7751 }, { "epoch": 3.648, "grad_norm": 1.7231316054538721, "learning_rate": 0.00013350717514602563, "loss": 1.2422, "step": 7752 }, { "epoch": 3.648470588235294, "grad_norm": 1.1730808402187836, "learning_rate": 0.00013346360545837062, "loss": 0.8555, "step": 7753 }, { "epoch": 3.648941176470588, "grad_norm": 0.4592856628435191, "learning_rate": 0.0001334200371830048, "loss": 0.1226, "step": 7754 }, { "epoch": 3.6494117647058824, "grad_norm": 0.8842185662515373, "learning_rate": 0.00013337647032364919, "loss": 0.9531, "step": 7755 }, { "epoch": 3.6498823529411766, "grad_norm": 0.624945099063881, "learning_rate": 0.0001333329048840245, "loss": 0.3125, "step": 7756 }, { "epoch": 3.650352941176471, "grad_norm": 0.7235353710678842, "learning_rate": 0.0001332893408678515, "loss": 0.7734, "step": 7757 }, { "epoch": 3.6508235294117646, "grad_norm": 0.966043830023376, "learning_rate": 0.00013324577827885078, "loss": 1.1016, "step": 7758 }, { "epoch": 3.651294117647059, "grad_norm": 0.9990164320875391, "learning_rate": 0.00013320221712074276, "loss": 1.4922, "step": 7759 }, { "epoch": 3.651764705882353, "grad_norm": 0.7952542558352178, "learning_rate": 0.00013315865739724778, "loss": 0.4863, "step": 7760 }, { "epoch": 3.652235294117647, "grad_norm": 0.7895427733563684, "learning_rate": 0.00013311509911208608, "loss": 0.9961, "step": 7761 }, { "epoch": 3.652705882352941, "grad_norm": 0.7797812318062766, "learning_rate": 0.00013307154226897773, "loss": 0.4355, "step": 7762 }, { "epoch": 3.6531764705882352, "grad_norm": 0.7840523348322253, "learning_rate": 0.0001330279868716427, "loss": 0.7109, "step": 7763 }, { "epoch": 3.6536470588235295, "grad_norm": 1.2603639383542213, "learning_rate": 0.00013298443292380084, "loss": 1.3906, "step": 7764 }, { "epoch": 3.6541176470588237, "grad_norm": 0.9434478367767806, "learning_rate": 0.00013294088042917188, "loss": 0.3379, "step": 7765 }, { "epoch": 3.654588235294118, "grad_norm": 1.2379784043895343, "learning_rate": 0.0001328973293914754, "loss": 1.4531, "step": 7766 }, { "epoch": 3.6550588235294117, "grad_norm": 0.8671492771981537, "learning_rate": 0.00013285377981443085, "loss": 0.6914, "step": 7767 }, { "epoch": 3.655529411764706, "grad_norm": 1.1507950384870247, "learning_rate": 0.00013281023170175766, "loss": 0.9727, "step": 7768 }, { "epoch": 3.656, "grad_norm": 1.2647395334252762, "learning_rate": 0.00013276668505717498, "loss": 1.4922, "step": 7769 }, { "epoch": 3.656470588235294, "grad_norm": 0.9318311422071559, "learning_rate": 0.00013272313988440187, "loss": 1.0703, "step": 7770 }, { "epoch": 3.656941176470588, "grad_norm": 0.9922434184294404, "learning_rate": 0.0001326795961871574, "loss": 1.0312, "step": 7771 }, { "epoch": 3.6574117647058824, "grad_norm": 0.669166206266777, "learning_rate": 0.00013263605396916038, "loss": 0.293, "step": 7772 }, { "epoch": 3.6578823529411766, "grad_norm": 0.8073264665474674, "learning_rate": 0.00013259251323412952, "loss": 0.7852, "step": 7773 }, { "epoch": 3.658352941176471, "grad_norm": 1.07938420461029, "learning_rate": 0.0001325489739857834, "loss": 1.3672, "step": 7774 }, { "epoch": 3.6588235294117646, "grad_norm": 1.508195660287553, "learning_rate": 0.0001325054362278405, "loss": 0.7539, "step": 7775 }, { "epoch": 3.659294117647059, "grad_norm": 0.8089752330432269, "learning_rate": 0.00013246189996401916, "loss": 0.8242, "step": 7776 }, { "epoch": 3.659764705882353, "grad_norm": 0.4412741569489591, "learning_rate": 0.00013241836519803756, "loss": 0.0605, "step": 7777 }, { "epoch": 3.660235294117647, "grad_norm": 1.1796859020466253, "learning_rate": 0.00013237483193361387, "loss": 0.957, "step": 7778 }, { "epoch": 3.660705882352941, "grad_norm": 0.6299346955816817, "learning_rate": 0.00013233130017446595, "loss": 0.3691, "step": 7779 }, { "epoch": 3.6611764705882353, "grad_norm": 1.1774582681296144, "learning_rate": 0.00013228776992431165, "loss": 1.0391, "step": 7780 }, { "epoch": 3.6616470588235295, "grad_norm": 0.9663369905190409, "learning_rate": 0.00013224424118686868, "loss": 0.625, "step": 7781 }, { "epoch": 3.6621176470588237, "grad_norm": 0.6667509014109342, "learning_rate": 0.00013220071396585463, "loss": 0.2812, "step": 7782 }, { "epoch": 3.662588235294118, "grad_norm": 1.0230997323855702, "learning_rate": 0.0001321571882649869, "loss": 1.4141, "step": 7783 }, { "epoch": 3.6630588235294117, "grad_norm": 0.8349147790781594, "learning_rate": 0.00013211366408798277, "loss": 0.4668, "step": 7784 }, { "epoch": 3.663529411764706, "grad_norm": 0.6874544091507189, "learning_rate": 0.00013207014143855951, "loss": 0.3926, "step": 7785 }, { "epoch": 3.664, "grad_norm": 0.9255850093796358, "learning_rate": 0.00013202662032043408, "loss": 0.5078, "step": 7786 }, { "epoch": 3.664470588235294, "grad_norm": 1.5089593305536795, "learning_rate": 0.00013198310073732344, "loss": 0.957, "step": 7787 }, { "epoch": 3.664941176470588, "grad_norm": 0.625722527344842, "learning_rate": 0.0001319395826929444, "loss": 0.1855, "step": 7788 }, { "epoch": 3.6654117647058824, "grad_norm": 0.9481520815212033, "learning_rate": 0.00013189606619101358, "loss": 0.9766, "step": 7789 }, { "epoch": 3.6658823529411766, "grad_norm": 1.0938766854619437, "learning_rate": 0.00013185255123524745, "loss": 0.75, "step": 7790 }, { "epoch": 3.666352941176471, "grad_norm": 2.2735219639960773, "learning_rate": 0.00013180903782936254, "loss": 1.6953, "step": 7791 }, { "epoch": 3.6668235294117646, "grad_norm": 1.028253585220834, "learning_rate": 0.000131765525977075, "loss": 0.9609, "step": 7792 }, { "epoch": 3.667294117647059, "grad_norm": 1.483611861232603, "learning_rate": 0.0001317220156821009, "loss": 2.0781, "step": 7793 }, { "epoch": 3.667764705882353, "grad_norm": 0.6720476555853796, "learning_rate": 0.00013167850694815642, "loss": 0.3047, "step": 7794 }, { "epoch": 3.668235294117647, "grad_norm": 1.2345272564619023, "learning_rate": 0.0001316349997789573, "loss": 1.1719, "step": 7795 }, { "epoch": 3.668705882352941, "grad_norm": 1.0826456634098807, "learning_rate": 0.00013159149417821923, "loss": 0.9844, "step": 7796 }, { "epoch": 3.6691764705882353, "grad_norm": 0.6317663478630998, "learning_rate": 0.00013154799014965785, "loss": 0.1377, "step": 7797 }, { "epoch": 3.6696470588235295, "grad_norm": 1.073692733161192, "learning_rate": 0.00013150448769698857, "loss": 1.4531, "step": 7798 }, { "epoch": 3.6701176470588237, "grad_norm": 0.972500637730694, "learning_rate": 0.00013146098682392685, "loss": 0.9609, "step": 7799 }, { "epoch": 3.6705882352941175, "grad_norm": 1.3296783438062685, "learning_rate": 0.0001314174875341878, "loss": 0.9922, "step": 7800 }, { "epoch": 3.6710588235294117, "grad_norm": 1.051714259831449, "learning_rate": 0.00013137398983148642, "loss": 1.0156, "step": 7801 }, { "epoch": 3.671529411764706, "grad_norm": 0.6430005377944018, "learning_rate": 0.00013133049371953767, "loss": 0.4199, "step": 7802 }, { "epoch": 3.672, "grad_norm": 0.7444681901706808, "learning_rate": 0.00013128699920205634, "loss": 0.3379, "step": 7803 }, { "epoch": 3.672470588235294, "grad_norm": 0.7654799245919045, "learning_rate": 0.00013124350628275708, "loss": 0.8281, "step": 7804 }, { "epoch": 3.672941176470588, "grad_norm": 0.9518853106076451, "learning_rate": 0.00013120001496535433, "loss": 0.6875, "step": 7805 }, { "epoch": 3.6734117647058824, "grad_norm": 0.8433090660238702, "learning_rate": 0.00013115652525356257, "loss": 0.4883, "step": 7806 }, { "epoch": 3.6738823529411766, "grad_norm": 0.9036008674934306, "learning_rate": 0.00013111303715109598, "loss": 0.8867, "step": 7807 }, { "epoch": 3.674352941176471, "grad_norm": 1.062569890714889, "learning_rate": 0.0001310695506616686, "loss": 1.0391, "step": 7808 }, { "epoch": 3.6748235294117646, "grad_norm": 0.7876412391936164, "learning_rate": 0.0001310260657889945, "loss": 0.8281, "step": 7809 }, { "epoch": 3.675294117647059, "grad_norm": 0.8455756352050243, "learning_rate": 0.00013098258253678744, "loss": 0.332, "step": 7810 }, { "epoch": 3.675764705882353, "grad_norm": 1.0684793062205113, "learning_rate": 0.00013093910090876108, "loss": 1.6328, "step": 7811 }, { "epoch": 3.676235294117647, "grad_norm": 0.9357795457572615, "learning_rate": 0.00013089562090862905, "loss": 0.8906, "step": 7812 }, { "epoch": 3.676705882352941, "grad_norm": 1.231872947368737, "learning_rate": 0.00013085214254010465, "loss": 0.832, "step": 7813 }, { "epoch": 3.6771764705882353, "grad_norm": 0.8563509596775017, "learning_rate": 0.00013080866580690126, "loss": 0.9258, "step": 7814 }, { "epoch": 3.6776470588235295, "grad_norm": 0.3608470047027866, "learning_rate": 0.00013076519071273187, "loss": 0.0291, "step": 7815 }, { "epoch": 3.6781176470588237, "grad_norm": 1.5912051253132067, "learning_rate": 0.00013072171726130957, "loss": 1.1016, "step": 7816 }, { "epoch": 3.6785882352941175, "grad_norm": 0.8453659433619515, "learning_rate": 0.00013067824545634718, "loss": 0.4199, "step": 7817 }, { "epoch": 3.6790588235294117, "grad_norm": 1.1194759327594064, "learning_rate": 0.00013063477530155738, "loss": 0.7695, "step": 7818 }, { "epoch": 3.679529411764706, "grad_norm": 0.69301560003706, "learning_rate": 0.00013059130680065278, "loss": 0.5781, "step": 7819 }, { "epoch": 3.68, "grad_norm": 0.8948872441064414, "learning_rate": 0.0001305478399573458, "loss": 0.6797, "step": 7820 }, { "epoch": 3.680470588235294, "grad_norm": 0.45374941756819515, "learning_rate": 0.00013050437477534862, "loss": 0.0884, "step": 7821 }, { "epoch": 3.680941176470588, "grad_norm": 1.4823163842482303, "learning_rate": 0.00013046091125837352, "loss": 0.6055, "step": 7822 }, { "epoch": 3.6814117647058824, "grad_norm": 1.1134852622097307, "learning_rate": 0.00013041744941013245, "loss": 0.8711, "step": 7823 }, { "epoch": 3.6818823529411766, "grad_norm": 1.6262738183116032, "learning_rate": 0.00013037398923433725, "loss": 1.7734, "step": 7824 }, { "epoch": 3.682352941176471, "grad_norm": 0.9656360576997457, "learning_rate": 0.00013033053073469958, "loss": 0.5938, "step": 7825 }, { "epoch": 3.6828235294117646, "grad_norm": 1.0202609106660028, "learning_rate": 0.00013028707391493112, "loss": 1.0625, "step": 7826 }, { "epoch": 3.683294117647059, "grad_norm": 1.0004175077105302, "learning_rate": 0.00013024361877874324, "loss": 1.1406, "step": 7827 }, { "epoch": 3.683764705882353, "grad_norm": 1.1015895087731526, "learning_rate": 0.00013020016532984714, "loss": 0.8047, "step": 7828 }, { "epoch": 3.684235294117647, "grad_norm": 1.3262192269681112, "learning_rate": 0.00013015671357195413, "loss": 0.9766, "step": 7829 }, { "epoch": 3.684705882352941, "grad_norm": 0.8243547232121813, "learning_rate": 0.00013011326350877508, "loss": 0.3457, "step": 7830 }, { "epoch": 3.6851764705882353, "grad_norm": 0.9125766083009247, "learning_rate": 0.00013006981514402085, "loss": 0.8672, "step": 7831 }, { "epoch": 3.6856470588235295, "grad_norm": 0.8477930058519936, "learning_rate": 0.00013002636848140217, "loss": 0.4609, "step": 7832 }, { "epoch": 3.6861176470588237, "grad_norm": 0.7284030622772305, "learning_rate": 0.00012998292352462963, "loss": 0.3691, "step": 7833 }, { "epoch": 3.6865882352941175, "grad_norm": 1.0547619067153025, "learning_rate": 0.00012993948027741357, "loss": 1.2656, "step": 7834 }, { "epoch": 3.6870588235294117, "grad_norm": 1.2443132967780892, "learning_rate": 0.0001298960387434643, "loss": 1.4375, "step": 7835 }, { "epoch": 3.687529411764706, "grad_norm": 0.5215300266994218, "learning_rate": 0.00012985259892649191, "loss": 0.1001, "step": 7836 }, { "epoch": 3.6879999999999997, "grad_norm": 0.9871291277313786, "learning_rate": 0.00012980916083020646, "loss": 0.6328, "step": 7837 }, { "epoch": 3.688470588235294, "grad_norm": 0.6806343883880269, "learning_rate": 0.0001297657244583176, "loss": 0.3496, "step": 7838 }, { "epoch": 3.688941176470588, "grad_norm": 1.2612560704286717, "learning_rate": 0.00012972228981453522, "loss": 1.0312, "step": 7839 }, { "epoch": 3.6894117647058824, "grad_norm": 1.0510405657976756, "learning_rate": 0.00012967885690256872, "loss": 0.8438, "step": 7840 }, { "epoch": 3.6898823529411766, "grad_norm": 0.8579101160164299, "learning_rate": 0.0001296354257261275, "loss": 0.5586, "step": 7841 }, { "epoch": 3.690352941176471, "grad_norm": 1.06590959507259, "learning_rate": 0.0001295919962889208, "loss": 1.0312, "step": 7842 }, { "epoch": 3.6908235294117646, "grad_norm": 0.9336959921750897, "learning_rate": 0.0001295485685946577, "loss": 0.6289, "step": 7843 }, { "epoch": 3.691294117647059, "grad_norm": 0.8086797674071842, "learning_rate": 0.0001295051426470472, "loss": 0.4336, "step": 7844 }, { "epoch": 3.691764705882353, "grad_norm": 0.7562350623069415, "learning_rate": 0.00012946171844979804, "loss": 0.4746, "step": 7845 }, { "epoch": 3.692235294117647, "grad_norm": 0.42538304344415545, "learning_rate": 0.00012941829600661886, "loss": 0.1768, "step": 7846 }, { "epoch": 3.692705882352941, "grad_norm": 0.6763072125089388, "learning_rate": 0.00012937487532121818, "loss": 0.3691, "step": 7847 }, { "epoch": 3.6931764705882353, "grad_norm": 1.791973478581268, "learning_rate": 0.00012933145639730428, "loss": 0.7695, "step": 7848 }, { "epoch": 3.6936470588235295, "grad_norm": 1.294916170367463, "learning_rate": 0.00012928803923858536, "loss": 0.793, "step": 7849 }, { "epoch": 3.6941176470588237, "grad_norm": 0.8619234759877227, "learning_rate": 0.00012924462384876953, "loss": 0.3379, "step": 7850 }, { "epoch": 3.6945882352941175, "grad_norm": 0.8200556860279312, "learning_rate": 0.00012920121023156462, "loss": 0.4492, "step": 7851 }, { "epoch": 3.6950588235294117, "grad_norm": 0.8920702188675733, "learning_rate": 0.00012915779839067838, "loss": 0.4277, "step": 7852 }, { "epoch": 3.695529411764706, "grad_norm": 0.7043318581230614, "learning_rate": 0.00012911438832981836, "loss": 0.3906, "step": 7853 }, { "epoch": 3.6959999999999997, "grad_norm": 0.9125977315351875, "learning_rate": 0.00012907098005269203, "loss": 0.6562, "step": 7854 }, { "epoch": 3.696470588235294, "grad_norm": 1.5483777844620776, "learning_rate": 0.00012902757356300667, "loss": 1.875, "step": 7855 }, { "epoch": 3.696941176470588, "grad_norm": 0.6486362098754819, "learning_rate": 0.00012898416886446937, "loss": 0.2334, "step": 7856 }, { "epoch": 3.6974117647058824, "grad_norm": 2.1616850905623672, "learning_rate": 0.00012894076596078717, "loss": 1.625, "step": 7857 }, { "epoch": 3.6978823529411766, "grad_norm": 1.0587720761613426, "learning_rate": 0.00012889736485566683, "loss": 0.5625, "step": 7858 }, { "epoch": 3.698352941176471, "grad_norm": 1.1199607346656324, "learning_rate": 0.000128853965552815, "loss": 0.8633, "step": 7859 }, { "epoch": 3.6988235294117646, "grad_norm": 1.013170933616229, "learning_rate": 0.0001288105680559383, "loss": 0.6641, "step": 7860 }, { "epoch": 3.699294117647059, "grad_norm": 0.6882776620690388, "learning_rate": 0.000128767172368743, "loss": 0.4629, "step": 7861 }, { "epoch": 3.699764705882353, "grad_norm": 0.5657302788667822, "learning_rate": 0.0001287237784949353, "loss": 0.4941, "step": 7862 }, { "epoch": 3.700235294117647, "grad_norm": 0.8289297227899904, "learning_rate": 0.0001286803864382213, "loss": 0.6523, "step": 7863 }, { "epoch": 3.700705882352941, "grad_norm": 0.9107584620881061, "learning_rate": 0.00012863699620230688, "loss": 0.5781, "step": 7864 }, { "epoch": 3.7011764705882353, "grad_norm": 0.5040184241401215, "learning_rate": 0.00012859360779089775, "loss": 0.1128, "step": 7865 }, { "epoch": 3.7016470588235295, "grad_norm": 0.7596501515258333, "learning_rate": 0.00012855022120769952, "loss": 0.8438, "step": 7866 }, { "epoch": 3.7021176470588237, "grad_norm": 0.5401516914335138, "learning_rate": 0.0001285068364564176, "loss": 0.2314, "step": 7867 }, { "epoch": 3.7025882352941175, "grad_norm": 0.5775505618561961, "learning_rate": 0.0001284634535407573, "loss": 0.2949, "step": 7868 }, { "epoch": 3.7030588235294117, "grad_norm": 0.8959891391046556, "learning_rate": 0.00012842007246442367, "loss": 0.5703, "step": 7869 }, { "epoch": 3.703529411764706, "grad_norm": 0.8581969054007682, "learning_rate": 0.0001283766932311217, "loss": 0.7227, "step": 7870 }, { "epoch": 3.7039999999999997, "grad_norm": 1.093208266018072, "learning_rate": 0.00012833331584455622, "loss": 1.3594, "step": 7871 }, { "epoch": 3.704470588235294, "grad_norm": 1.3877217657264893, "learning_rate": 0.0001282899403084318, "loss": 1.8672, "step": 7872 }, { "epoch": 3.704941176470588, "grad_norm": 0.8772532938309805, "learning_rate": 0.000128246566626453, "loss": 0.6445, "step": 7873 }, { "epoch": 3.7054117647058824, "grad_norm": 1.4460460589880053, "learning_rate": 0.0001282031948023241, "loss": 0.8555, "step": 7874 }, { "epoch": 3.7058823529411766, "grad_norm": 0.8209307913681028, "learning_rate": 0.0001281598248397493, "loss": 0.7148, "step": 7875 }, { "epoch": 3.706352941176471, "grad_norm": 1.0276143535976188, "learning_rate": 0.00012811645674243254, "loss": 1.2734, "step": 7876 }, { "epoch": 3.7068235294117646, "grad_norm": 1.7184223776473826, "learning_rate": 0.00012807309051407774, "loss": 0.6953, "step": 7877 }, { "epoch": 3.707294117647059, "grad_norm": 0.9940886186900916, "learning_rate": 0.00012802972615838856, "loss": 1.0156, "step": 7878 }, { "epoch": 3.707764705882353, "grad_norm": 1.022102776478019, "learning_rate": 0.00012798636367906849, "loss": 1.0391, "step": 7879 }, { "epoch": 3.708235294117647, "grad_norm": 0.9634586002787668, "learning_rate": 0.00012794300307982098, "loss": 0.4707, "step": 7880 }, { "epoch": 3.708705882352941, "grad_norm": 0.7790553201164268, "learning_rate": 0.0001278996443643492, "loss": 0.2617, "step": 7881 }, { "epoch": 3.7091764705882353, "grad_norm": 1.0890382057143195, "learning_rate": 0.00012785628753635617, "loss": 0.3359, "step": 7882 }, { "epoch": 3.7096470588235295, "grad_norm": 1.2116884405749173, "learning_rate": 0.0001278129325995448, "loss": 0.7812, "step": 7883 }, { "epoch": 3.7101176470588237, "grad_norm": 1.1040890198746396, "learning_rate": 0.00012776957955761784, "loss": 0.8398, "step": 7884 }, { "epoch": 3.7105882352941175, "grad_norm": 0.9172425346451788, "learning_rate": 0.0001277262284142778, "loss": 0.6914, "step": 7885 }, { "epoch": 3.7110588235294117, "grad_norm": 1.0129668000365795, "learning_rate": 0.00012768287917322706, "loss": 0.8477, "step": 7886 }, { "epoch": 3.711529411764706, "grad_norm": 0.947624120149632, "learning_rate": 0.00012763953183816794, "loss": 1.0469, "step": 7887 }, { "epoch": 3.7119999999999997, "grad_norm": 0.8766530550808463, "learning_rate": 0.00012759618641280251, "loss": 1.1172, "step": 7888 }, { "epoch": 3.712470588235294, "grad_norm": 1.1047137284896114, "learning_rate": 0.00012755284290083264, "loss": 0.6328, "step": 7889 }, { "epoch": 3.712941176470588, "grad_norm": 1.0722914406518818, "learning_rate": 0.00012750950130596006, "loss": 1.3516, "step": 7890 }, { "epoch": 3.7134117647058824, "grad_norm": 1.308819219633085, "learning_rate": 0.0001274661616318864, "loss": 1.0078, "step": 7891 }, { "epoch": 3.7138823529411766, "grad_norm": 0.6366501889787983, "learning_rate": 0.0001274228238823131, "loss": 0.4043, "step": 7892 }, { "epoch": 3.714352941176471, "grad_norm": 0.9272206893598401, "learning_rate": 0.00012737948806094135, "loss": 0.5352, "step": 7893 }, { "epoch": 3.7148235294117646, "grad_norm": 0.7774623156305195, "learning_rate": 0.00012733615417147228, "loss": 0.6094, "step": 7894 }, { "epoch": 3.715294117647059, "grad_norm": 0.8004791940852078, "learning_rate": 0.0001272928222176068, "loss": 0.3672, "step": 7895 }, { "epoch": 3.715764705882353, "grad_norm": 0.8392035881652404, "learning_rate": 0.0001272494922030457, "loss": 0.8945, "step": 7896 }, { "epoch": 3.716235294117647, "grad_norm": 0.9766713802823479, "learning_rate": 0.0001272061641314895, "loss": 0.8555, "step": 7897 }, { "epoch": 3.716705882352941, "grad_norm": 1.0722309399931462, "learning_rate": 0.00012716283800663874, "loss": 0.9141, "step": 7898 }, { "epoch": 3.7171764705882353, "grad_norm": 1.5166835816904385, "learning_rate": 0.00012711951383219362, "loss": 1.9141, "step": 7899 }, { "epoch": 3.7176470588235295, "grad_norm": 1.188858070609212, "learning_rate": 0.00012707619161185423, "loss": 0.9023, "step": 7900 }, { "epoch": 3.7181176470588237, "grad_norm": 0.5535017975351522, "learning_rate": 0.00012703287134932054, "loss": 0.4863, "step": 7901 }, { "epoch": 3.7185882352941175, "grad_norm": 1.0832721398558365, "learning_rate": 0.0001269895530482923, "loss": 0.8867, "step": 7902 }, { "epoch": 3.7190588235294118, "grad_norm": 1.0436090190854188, "learning_rate": 0.00012694623671246908, "loss": 1.0859, "step": 7903 }, { "epoch": 3.719529411764706, "grad_norm": 1.2040082567033017, "learning_rate": 0.0001269029223455503, "loss": 1.2812, "step": 7904 }, { "epoch": 3.7199999999999998, "grad_norm": 1.1105589476186213, "learning_rate": 0.00012685960995123527, "loss": 1.375, "step": 7905 }, { "epoch": 3.720470588235294, "grad_norm": 1.0158925349853718, "learning_rate": 0.00012681629953322307, "loss": 1.2266, "step": 7906 }, { "epoch": 3.720941176470588, "grad_norm": 1.0584437516124034, "learning_rate": 0.00012677299109521256, "loss": 1.3281, "step": 7907 }, { "epoch": 3.7214117647058824, "grad_norm": 0.7111047265820315, "learning_rate": 0.00012672968464090255, "loss": 0.75, "step": 7908 }, { "epoch": 3.7218823529411766, "grad_norm": 0.8464815326618295, "learning_rate": 0.00012668638017399164, "loss": 0.7812, "step": 7909 }, { "epoch": 3.722352941176471, "grad_norm": 0.6404127674757121, "learning_rate": 0.00012664307769817815, "loss": 0.1982, "step": 7910 }, { "epoch": 3.7228235294117646, "grad_norm": 1.258224083472204, "learning_rate": 0.00012659977721716045, "loss": 1.2188, "step": 7911 }, { "epoch": 3.723294117647059, "grad_norm": 0.9600410899752834, "learning_rate": 0.00012655647873463657, "loss": 0.4395, "step": 7912 }, { "epoch": 3.723764705882353, "grad_norm": 1.4849618014982005, "learning_rate": 0.00012651318225430437, "loss": 1.7891, "step": 7913 }, { "epoch": 3.724235294117647, "grad_norm": 1.063617925478089, "learning_rate": 0.00012646988777986157, "loss": 0.9414, "step": 7914 }, { "epoch": 3.724705882352941, "grad_norm": 0.9365086766937104, "learning_rate": 0.00012642659531500583, "loss": 0.543, "step": 7915 }, { "epoch": 3.7251764705882353, "grad_norm": 1.18288245045442, "learning_rate": 0.00012638330486343445, "loss": 0.793, "step": 7916 }, { "epoch": 3.7256470588235295, "grad_norm": 1.0474531539267922, "learning_rate": 0.00012634001642884465, "loss": 1.0625, "step": 7917 }, { "epoch": 3.7261176470588238, "grad_norm": 1.3389116284653222, "learning_rate": 0.0001262967300149335, "loss": 1.2031, "step": 7918 }, { "epoch": 3.7265882352941175, "grad_norm": 0.8230738049837836, "learning_rate": 0.0001262534456253979, "loss": 0.7305, "step": 7919 }, { "epoch": 3.7270588235294118, "grad_norm": 0.7824353547632439, "learning_rate": 0.0001262101632639345, "loss": 0.3008, "step": 7920 }, { "epoch": 3.727529411764706, "grad_norm": 1.2145055132064648, "learning_rate": 0.00012616688293423987, "loss": 1.5234, "step": 7921 }, { "epoch": 3.7279999999999998, "grad_norm": 0.8983666588771707, "learning_rate": 0.00012612360464001033, "loss": 0.8008, "step": 7922 }, { "epoch": 3.728470588235294, "grad_norm": 1.0199424208099188, "learning_rate": 0.00012608032838494205, "loss": 0.7031, "step": 7923 }, { "epoch": 3.728941176470588, "grad_norm": 0.9475088819605739, "learning_rate": 0.00012603705417273107, "loss": 0.332, "step": 7924 }, { "epoch": 3.7294117647058824, "grad_norm": 1.249654991359854, "learning_rate": 0.0001259937820070732, "loss": 0.957, "step": 7925 }, { "epoch": 3.7298823529411766, "grad_norm": 0.8965892270167993, "learning_rate": 0.0001259505118916641, "loss": 0.6328, "step": 7926 }, { "epoch": 3.7303529411764704, "grad_norm": 0.9445592495779854, "learning_rate": 0.0001259072438301992, "loss": 0.5781, "step": 7927 }, { "epoch": 3.7308235294117646, "grad_norm": 0.9553292344634882, "learning_rate": 0.00012586397782637392, "loss": 0.9961, "step": 7928 }, { "epoch": 3.731294117647059, "grad_norm": 1.3731960484187045, "learning_rate": 0.0001258207138838833, "loss": 1.0, "step": 7929 }, { "epoch": 3.731764705882353, "grad_norm": 1.1519252449430204, "learning_rate": 0.00012577745200642227, "loss": 0.6406, "step": 7930 }, { "epoch": 3.732235294117647, "grad_norm": 0.5519188070899881, "learning_rate": 0.00012573419219768562, "loss": 0.4863, "step": 7931 }, { "epoch": 3.732705882352941, "grad_norm": 0.8757330317321115, "learning_rate": 0.00012569093446136806, "loss": 0.5938, "step": 7932 }, { "epoch": 3.7331764705882353, "grad_norm": 0.8299618373569083, "learning_rate": 0.0001256476788011639, "loss": 0.4082, "step": 7933 }, { "epoch": 3.7336470588235295, "grad_norm": 1.0806816467198912, "learning_rate": 0.00012560442522076745, "loss": 0.5781, "step": 7934 }, { "epoch": 3.7341176470588238, "grad_norm": 0.5113456985503322, "learning_rate": 0.00012556117372387272, "loss": 0.2656, "step": 7935 }, { "epoch": 3.7345882352941175, "grad_norm": 0.8575114024054301, "learning_rate": 0.00012551792431417363, "loss": 0.373, "step": 7936 }, { "epoch": 3.7350588235294118, "grad_norm": 1.3427003750604756, "learning_rate": 0.0001254746769953639, "loss": 1.1641, "step": 7937 }, { "epoch": 3.735529411764706, "grad_norm": 1.0564845521488624, "learning_rate": 0.000125431431771137, "loss": 0.9844, "step": 7938 }, { "epoch": 3.7359999999999998, "grad_norm": 0.8854897485800081, "learning_rate": 0.0001253881886451864, "loss": 0.75, "step": 7939 }, { "epoch": 3.736470588235294, "grad_norm": 0.913819719574793, "learning_rate": 0.0001253449476212052, "loss": 0.543, "step": 7940 }, { "epoch": 3.736941176470588, "grad_norm": 0.8162903103878139, "learning_rate": 0.00012530170870288636, "loss": 0.4883, "step": 7941 }, { "epoch": 3.7374117647058824, "grad_norm": 1.1472526970512364, "learning_rate": 0.00012525847189392278, "loss": 0.8633, "step": 7942 }, { "epoch": 3.7378823529411767, "grad_norm": 0.6818471634568377, "learning_rate": 0.00012521523719800706, "loss": 0.4434, "step": 7943 }, { "epoch": 3.7383529411764704, "grad_norm": 0.8941552168480779, "learning_rate": 0.00012517200461883163, "loss": 0.9297, "step": 7944 }, { "epoch": 3.7388235294117647, "grad_norm": 1.0595188630700236, "learning_rate": 0.00012512877416008878, "loss": 0.5234, "step": 7945 }, { "epoch": 3.739294117647059, "grad_norm": 0.8957533754678696, "learning_rate": 0.00012508554582547064, "loss": 0.4023, "step": 7946 }, { "epoch": 3.739764705882353, "grad_norm": 0.7589564627447929, "learning_rate": 0.00012504231961866908, "loss": 0.5938, "step": 7947 }, { "epoch": 3.740235294117647, "grad_norm": 0.5533613796964629, "learning_rate": 0.00012499909554337582, "loss": 0.166, "step": 7948 }, { "epoch": 3.740705882352941, "grad_norm": 0.7790922467746533, "learning_rate": 0.00012495587360328246, "loss": 0.3945, "step": 7949 }, { "epoch": 3.7411764705882353, "grad_norm": 0.9935934929428443, "learning_rate": 0.00012491265380208032, "loss": 1.0078, "step": 7950 }, { "epoch": 3.7416470588235295, "grad_norm": 0.7864518071697322, "learning_rate": 0.0001248694361434606, "loss": 0.2754, "step": 7951 }, { "epoch": 3.7421176470588238, "grad_norm": 0.7794983145245381, "learning_rate": 0.0001248262206311143, "loss": 0.5078, "step": 7952 }, { "epoch": 3.7425882352941175, "grad_norm": 0.8990413063215424, "learning_rate": 0.00012478300726873225, "loss": 0.6875, "step": 7953 }, { "epoch": 3.7430588235294118, "grad_norm": 0.8016350183470755, "learning_rate": 0.00012473979606000504, "loss": 0.6523, "step": 7954 }, { "epoch": 3.743529411764706, "grad_norm": 1.2022904406911947, "learning_rate": 0.00012469658700862315, "loss": 0.8555, "step": 7955 }, { "epoch": 3.7439999999999998, "grad_norm": 0.7417981742623608, "learning_rate": 0.00012465338011827685, "loss": 0.4375, "step": 7956 }, { "epoch": 3.744470588235294, "grad_norm": 0.9457420283828623, "learning_rate": 0.00012461017539265622, "loss": 0.8945, "step": 7957 }, { "epoch": 3.744941176470588, "grad_norm": 1.0084382726011716, "learning_rate": 0.00012456697283545114, "loss": 0.7539, "step": 7958 }, { "epoch": 3.7454117647058824, "grad_norm": 0.832842782402512, "learning_rate": 0.00012452377245035132, "loss": 0.6602, "step": 7959 }, { "epoch": 3.7458823529411767, "grad_norm": 1.247499618355614, "learning_rate": 0.00012448057424104632, "loss": 1.1172, "step": 7960 }, { "epoch": 3.7463529411764704, "grad_norm": 0.8346621153052475, "learning_rate": 0.00012443737821122541, "loss": 0.5508, "step": 7961 }, { "epoch": 3.7468235294117647, "grad_norm": 0.7738898863986188, "learning_rate": 0.0001243941843645778, "loss": 0.3652, "step": 7962 }, { "epoch": 3.747294117647059, "grad_norm": 0.9241177327416487, "learning_rate": 0.00012435099270479246, "loss": 0.5273, "step": 7963 }, { "epoch": 3.7477647058823527, "grad_norm": 1.1449775942120108, "learning_rate": 0.00012430780323555816, "loss": 1.5703, "step": 7964 }, { "epoch": 3.748235294117647, "grad_norm": 1.1646552149047409, "learning_rate": 0.00012426461596056344, "loss": 1.0234, "step": 7965 }, { "epoch": 3.748705882352941, "grad_norm": 0.9021160727483729, "learning_rate": 0.00012422143088349679, "loss": 0.5391, "step": 7966 }, { "epoch": 3.7491764705882353, "grad_norm": 1.1420019385422429, "learning_rate": 0.00012417824800804638, "loss": 0.7891, "step": 7967 }, { "epoch": 3.7496470588235296, "grad_norm": 0.7507636337058798, "learning_rate": 0.0001241350673379002, "loss": 0.2969, "step": 7968 }, { "epoch": 3.7501176470588238, "grad_norm": 0.9981236572193618, "learning_rate": 0.0001240918888767462, "loss": 0.5195, "step": 7969 }, { "epoch": 3.7505882352941176, "grad_norm": 1.0926402518296017, "learning_rate": 0.00012404871262827193, "loss": 0.7422, "step": 7970 }, { "epoch": 3.751058823529412, "grad_norm": 1.050023928974607, "learning_rate": 0.00012400553859616488, "loss": 0.7695, "step": 7971 }, { "epoch": 3.751529411764706, "grad_norm": 0.8495656172945011, "learning_rate": 0.00012396236678411236, "loss": 0.7969, "step": 7972 }, { "epoch": 3.752, "grad_norm": 1.13824514184521, "learning_rate": 0.00012391919719580144, "loss": 1.2891, "step": 7973 }, { "epoch": 3.752470588235294, "grad_norm": 1.1511483602524366, "learning_rate": 0.00012387602983491894, "loss": 1.5078, "step": 7974 }, { "epoch": 3.7529411764705882, "grad_norm": 0.635821944953209, "learning_rate": 0.00012383286470515166, "loss": 0.2041, "step": 7975 }, { "epoch": 3.7534117647058824, "grad_norm": 1.1615708619092129, "learning_rate": 0.0001237897018101861, "loss": 0.5703, "step": 7976 }, { "epoch": 3.7538823529411767, "grad_norm": 0.924276653559989, "learning_rate": 0.00012374654115370856, "loss": 0.7109, "step": 7977 }, { "epoch": 3.7543529411764704, "grad_norm": 0.9904628976753851, "learning_rate": 0.0001237033827394052, "loss": 0.7227, "step": 7978 }, { "epoch": 3.7548235294117647, "grad_norm": 1.079672918245803, "learning_rate": 0.0001236602265709619, "loss": 1.1328, "step": 7979 }, { "epoch": 3.755294117647059, "grad_norm": 0.6413548046547388, "learning_rate": 0.00012361707265206448, "loss": 0.1973, "step": 7980 }, { "epoch": 3.7557647058823527, "grad_norm": 0.9356447332328259, "learning_rate": 0.00012357392098639848, "loss": 0.3926, "step": 7981 }, { "epoch": 3.756235294117647, "grad_norm": 0.5425825818979956, "learning_rate": 0.00012353077157764922, "loss": 0.3633, "step": 7982 }, { "epoch": 3.756705882352941, "grad_norm": 1.5294764726986625, "learning_rate": 0.00012348762442950187, "loss": 0.9766, "step": 7983 }, { "epoch": 3.7571764705882353, "grad_norm": 0.9094038244825884, "learning_rate": 0.00012344447954564148, "loss": 0.8398, "step": 7984 }, { "epoch": 3.7576470588235296, "grad_norm": 0.5805936310335862, "learning_rate": 0.0001234013369297528, "loss": 0.1445, "step": 7985 }, { "epoch": 3.758117647058824, "grad_norm": 0.8515431934830581, "learning_rate": 0.00012335819658552036, "loss": 0.4863, "step": 7986 }, { "epoch": 3.7585882352941176, "grad_norm": 1.3638876905217154, "learning_rate": 0.00012331505851662864, "loss": 1.5547, "step": 7987 }, { "epoch": 3.759058823529412, "grad_norm": 1.1171840432870526, "learning_rate": 0.00012327192272676183, "loss": 1.1094, "step": 7988 }, { "epoch": 3.759529411764706, "grad_norm": 1.0584830471943445, "learning_rate": 0.0001232287892196039, "loss": 1.4141, "step": 7989 }, { "epoch": 3.76, "grad_norm": 0.7710912230567475, "learning_rate": 0.00012318565799883865, "loss": 0.6875, "step": 7990 }, { "epoch": 3.760470588235294, "grad_norm": 0.5620760068052488, "learning_rate": 0.0001231425290681498, "loss": 0.1846, "step": 7991 }, { "epoch": 3.7609411764705882, "grad_norm": 1.1762612057526243, "learning_rate": 0.00012309940243122064, "loss": 0.8516, "step": 7992 }, { "epoch": 3.7614117647058825, "grad_norm": 1.2683698948812276, "learning_rate": 0.00012305627809173448, "loss": 1.5234, "step": 7993 }, { "epoch": 3.7618823529411767, "grad_norm": 0.8348468395708383, "learning_rate": 0.00012301315605337435, "loss": 0.459, "step": 7994 }, { "epoch": 3.7623529411764705, "grad_norm": 0.8611473710825639, "learning_rate": 0.00012297003631982304, "loss": 0.7188, "step": 7995 }, { "epoch": 3.7628235294117647, "grad_norm": 0.6807021669629807, "learning_rate": 0.0001229269188947632, "loss": 0.3418, "step": 7996 }, { "epoch": 3.763294117647059, "grad_norm": 0.9617771683538198, "learning_rate": 0.0001228838037818773, "loss": 0.6172, "step": 7997 }, { "epoch": 3.7637647058823527, "grad_norm": 0.8008524497938218, "learning_rate": 0.00012284069098484758, "loss": 0.8711, "step": 7998 }, { "epoch": 3.764235294117647, "grad_norm": 1.3892789433760877, "learning_rate": 0.000122797580507356, "loss": 1.9219, "step": 7999 }, { "epoch": 3.764705882352941, "grad_norm": 0.9717476557518083, "learning_rate": 0.00012275447235308453, "loss": 0.8555, "step": 8000 }, { "epoch": 3.7651764705882353, "grad_norm": 1.0845081960437677, "learning_rate": 0.00012271136652571474, "loss": 0.8789, "step": 8001 }, { "epoch": 3.7656470588235296, "grad_norm": 0.9618691448250445, "learning_rate": 0.0001226682630289281, "loss": 0.6992, "step": 8002 }, { "epoch": 3.766117647058824, "grad_norm": 1.041858010901928, "learning_rate": 0.00012262516186640587, "loss": 0.6289, "step": 8003 }, { "epoch": 3.7665882352941176, "grad_norm": 0.8371755141852468, "learning_rate": 0.0001225820630418291, "loss": 0.625, "step": 8004 }, { "epoch": 3.767058823529412, "grad_norm": 0.7894447618732586, "learning_rate": 0.00012253896655887861, "loss": 0.5664, "step": 8005 }, { "epoch": 3.767529411764706, "grad_norm": 0.8528873268230224, "learning_rate": 0.00012249587242123505, "loss": 0.4258, "step": 8006 }, { "epoch": 3.768, "grad_norm": 1.2723533851746154, "learning_rate": 0.00012245278063257892, "loss": 0.8086, "step": 8007 }, { "epoch": 3.768470588235294, "grad_norm": 0.9179763695503846, "learning_rate": 0.00012240969119659044, "loss": 0.3555, "step": 8008 }, { "epoch": 3.7689411764705882, "grad_norm": 0.9128179409560588, "learning_rate": 0.00012236660411694964, "loss": 0.4512, "step": 8009 }, { "epoch": 3.7694117647058825, "grad_norm": 0.8737838665259282, "learning_rate": 0.00012232351939733642, "loss": 0.8672, "step": 8010 }, { "epoch": 3.7698823529411767, "grad_norm": 1.003167602471004, "learning_rate": 0.0001222804370414304, "loss": 0.9062, "step": 8011 }, { "epoch": 3.7703529411764705, "grad_norm": 1.0143479447331027, "learning_rate": 0.00012223735705291098, "loss": 0.918, "step": 8012 }, { "epoch": 3.7708235294117647, "grad_norm": 0.8962173525707511, "learning_rate": 0.00012219427943545744, "loss": 1.0156, "step": 8013 }, { "epoch": 3.771294117647059, "grad_norm": 1.0773054976272116, "learning_rate": 0.00012215120419274887, "loss": 1.3984, "step": 8014 }, { "epoch": 3.7717647058823527, "grad_norm": 0.46163001030573864, "learning_rate": 0.00012210813132846406, "loss": 0.1484, "step": 8015 }, { "epoch": 3.772235294117647, "grad_norm": 0.6356984118868205, "learning_rate": 0.0001220650608462816, "loss": 0.3105, "step": 8016 }, { "epoch": 3.772705882352941, "grad_norm": 0.9585169060086095, "learning_rate": 0.00012202199274987999, "loss": 0.5898, "step": 8017 }, { "epoch": 3.7731764705882354, "grad_norm": 0.9611625387489202, "learning_rate": 0.00012197892704293743, "loss": 1.0469, "step": 8018 }, { "epoch": 3.7736470588235296, "grad_norm": 0.7480767352006404, "learning_rate": 0.00012193586372913192, "loss": 0.2852, "step": 8019 }, { "epoch": 3.774117647058824, "grad_norm": 1.0620286503253036, "learning_rate": 0.00012189280281214126, "loss": 1.0547, "step": 8020 }, { "epoch": 3.7745882352941176, "grad_norm": 0.8792807916334004, "learning_rate": 0.00012184974429564318, "loss": 1.0312, "step": 8021 }, { "epoch": 3.775058823529412, "grad_norm": 0.6276003858278031, "learning_rate": 0.000121806688183315, "loss": 0.6406, "step": 8022 }, { "epoch": 3.775529411764706, "grad_norm": 1.216489739003358, "learning_rate": 0.00012176363447883397, "loss": 0.7461, "step": 8023 }, { "epoch": 3.776, "grad_norm": 1.3998513535989916, "learning_rate": 0.00012172058318587698, "loss": 1.2656, "step": 8024 }, { "epoch": 3.776470588235294, "grad_norm": 0.9745559175461564, "learning_rate": 0.00012167753430812093, "loss": 0.9141, "step": 8025 }, { "epoch": 3.7769411764705882, "grad_norm": 0.8508714350171644, "learning_rate": 0.0001216344878492424, "loss": 0.8203, "step": 8026 }, { "epoch": 3.7774117647058825, "grad_norm": 0.6897228087590453, "learning_rate": 0.0001215914438129177, "loss": 0.416, "step": 8027 }, { "epoch": 3.7778823529411767, "grad_norm": 1.0274893938058083, "learning_rate": 0.00012154840220282306, "loss": 0.9336, "step": 8028 }, { "epoch": 3.7783529411764705, "grad_norm": 0.5020602120554829, "learning_rate": 0.00012150536302263444, "loss": 0.1807, "step": 8029 }, { "epoch": 3.7788235294117647, "grad_norm": 1.0410635935129227, "learning_rate": 0.00012146232627602753, "loss": 0.918, "step": 8030 }, { "epoch": 3.779294117647059, "grad_norm": 1.1234851953063405, "learning_rate": 0.00012141929196667798, "loss": 1.1875, "step": 8031 }, { "epoch": 3.7797647058823527, "grad_norm": 1.250329616190334, "learning_rate": 0.00012137626009826108, "loss": 1.4531, "step": 8032 }, { "epoch": 3.780235294117647, "grad_norm": 1.0715698130953408, "learning_rate": 0.00012133323067445198, "loss": 0.8477, "step": 8033 }, { "epoch": 3.780705882352941, "grad_norm": 0.7060606934505104, "learning_rate": 0.00012129020369892555, "loss": 0.3379, "step": 8034 }, { "epoch": 3.7811764705882354, "grad_norm": 0.6711361688192808, "learning_rate": 0.00012124717917535656, "loss": 0.2051, "step": 8035 }, { "epoch": 3.7816470588235296, "grad_norm": 0.563855185189457, "learning_rate": 0.00012120415710741951, "loss": 0.1387, "step": 8036 }, { "epoch": 3.782117647058824, "grad_norm": 0.8236848346736655, "learning_rate": 0.00012116113749878867, "loss": 0.3555, "step": 8037 }, { "epoch": 3.7825882352941176, "grad_norm": 1.7521537582495865, "learning_rate": 0.00012111812035313814, "loss": 1.7891, "step": 8038 }, { "epoch": 3.783058823529412, "grad_norm": 1.0787685725257936, "learning_rate": 0.00012107510567414183, "loss": 1.1328, "step": 8039 }, { "epoch": 3.783529411764706, "grad_norm": 0.8667427286209199, "learning_rate": 0.00012103209346547332, "loss": 0.875, "step": 8040 }, { "epoch": 3.784, "grad_norm": 0.5525366143779086, "learning_rate": 0.00012098908373080617, "loss": 0.2324, "step": 8041 }, { "epoch": 3.784470588235294, "grad_norm": 0.5523694079167148, "learning_rate": 0.00012094607647381355, "loss": 0.2656, "step": 8042 }, { "epoch": 3.7849411764705883, "grad_norm": 0.6744784966867096, "learning_rate": 0.00012090307169816847, "loss": 0.25, "step": 8043 }, { "epoch": 3.7854117647058825, "grad_norm": 0.9675007148087353, "learning_rate": 0.00012086006940754382, "loss": 0.7227, "step": 8044 }, { "epoch": 3.7858823529411767, "grad_norm": 0.8818885662004712, "learning_rate": 0.00012081706960561217, "loss": 0.7461, "step": 8045 }, { "epoch": 3.7863529411764705, "grad_norm": 0.8922596624996572, "learning_rate": 0.00012077407229604593, "loss": 0.3828, "step": 8046 }, { "epoch": 3.7868235294117647, "grad_norm": 1.406646484308999, "learning_rate": 0.00012073107748251724, "loss": 2.0781, "step": 8047 }, { "epoch": 3.787294117647059, "grad_norm": 1.3125513763062424, "learning_rate": 0.00012068808516869812, "loss": 0.498, "step": 8048 }, { "epoch": 3.7877647058823527, "grad_norm": 1.2814453118508318, "learning_rate": 0.00012064509535826029, "loss": 1.0781, "step": 8049 }, { "epoch": 3.788235294117647, "grad_norm": 1.1588470938167885, "learning_rate": 0.00012060210805487529, "loss": 1.0234, "step": 8050 }, { "epoch": 3.788705882352941, "grad_norm": 0.8536305994160984, "learning_rate": 0.00012055912326221448, "loss": 0.3145, "step": 8051 }, { "epoch": 3.7891764705882354, "grad_norm": 0.9905489430228911, "learning_rate": 0.00012051614098394894, "loss": 0.6641, "step": 8052 }, { "epoch": 3.7896470588235296, "grad_norm": 0.4856888563087049, "learning_rate": 0.00012047316122374957, "loss": 0.123, "step": 8053 }, { "epoch": 3.7901176470588234, "grad_norm": 1.2609597349072432, "learning_rate": 0.00012043018398528707, "loss": 1.0703, "step": 8054 }, { "epoch": 3.7905882352941176, "grad_norm": 0.9769023582985726, "learning_rate": 0.0001203872092722319, "loss": 1.0078, "step": 8055 }, { "epoch": 3.791058823529412, "grad_norm": 1.138015605638745, "learning_rate": 0.00012034423708825431, "loss": 1.1953, "step": 8056 }, { "epoch": 3.791529411764706, "grad_norm": 0.9101199391042217, "learning_rate": 0.00012030126743702431, "loss": 0.7305, "step": 8057 }, { "epoch": 3.792, "grad_norm": 1.0525908183413455, "learning_rate": 0.00012025830032221177, "loss": 0.9453, "step": 8058 }, { "epoch": 3.792470588235294, "grad_norm": 1.0518044458351916, "learning_rate": 0.00012021533574748625, "loss": 0.7305, "step": 8059 }, { "epoch": 3.7929411764705883, "grad_norm": 0.6167775403295642, "learning_rate": 0.00012017237371651714, "loss": 0.1953, "step": 8060 }, { "epoch": 3.7934117647058825, "grad_norm": 1.1410229426785454, "learning_rate": 0.00012012941423297363, "loss": 1.0859, "step": 8061 }, { "epoch": 3.7938823529411767, "grad_norm": 1.1090282341438644, "learning_rate": 0.00012008645730052468, "loss": 1.0469, "step": 8062 }, { "epoch": 3.7943529411764705, "grad_norm": 1.0422139243140411, "learning_rate": 0.00012004350292283896, "loss": 1.1719, "step": 8063 }, { "epoch": 3.7948235294117647, "grad_norm": 1.0444102633865617, "learning_rate": 0.00012000055110358502, "loss": 0.9766, "step": 8064 }, { "epoch": 3.795294117647059, "grad_norm": 0.9760888549619537, "learning_rate": 0.00011995760184643118, "loss": 0.8945, "step": 8065 }, { "epoch": 3.7957647058823527, "grad_norm": 1.2863639720542719, "learning_rate": 0.00011991465515504551, "loss": 0.5742, "step": 8066 }, { "epoch": 3.796235294117647, "grad_norm": 0.8249944993507641, "learning_rate": 0.00011987171103309586, "loss": 0.625, "step": 8067 }, { "epoch": 3.796705882352941, "grad_norm": 1.1226158667581605, "learning_rate": 0.00011982876948424986, "loss": 1.1172, "step": 8068 }, { "epoch": 3.7971764705882354, "grad_norm": 0.41583691985211896, "learning_rate": 0.00011978583051217493, "loss": 0.0894, "step": 8069 }, { "epoch": 3.7976470588235296, "grad_norm": 1.0912924770465138, "learning_rate": 0.0001197428941205383, "loss": 0.8008, "step": 8070 }, { "epoch": 3.7981176470588234, "grad_norm": 1.0519778530922739, "learning_rate": 0.00011969996031300687, "loss": 0.8594, "step": 8071 }, { "epoch": 3.7985882352941176, "grad_norm": 1.2308756463794897, "learning_rate": 0.00011965702909324751, "loss": 1.0391, "step": 8072 }, { "epoch": 3.799058823529412, "grad_norm": 0.7872010467756008, "learning_rate": 0.00011961410046492667, "loss": 0.543, "step": 8073 }, { "epoch": 3.799529411764706, "grad_norm": 0.32186415734354457, "learning_rate": 0.0001195711744317107, "loss": 0.0491, "step": 8074 }, { "epoch": 3.8, "grad_norm": 0.5200060016863434, "learning_rate": 0.00011952825099726567, "loss": 0.4551, "step": 8075 }, { "epoch": 3.800470588235294, "grad_norm": 0.6077713897502036, "learning_rate": 0.00011948533016525749, "loss": 0.6406, "step": 8076 }, { "epoch": 3.8009411764705883, "grad_norm": 1.0666455599059985, "learning_rate": 0.00011944241193935178, "loss": 0.4492, "step": 8077 }, { "epoch": 3.8014117647058825, "grad_norm": 0.786258614786132, "learning_rate": 0.00011939949632321395, "loss": 0.8789, "step": 8078 }, { "epoch": 3.8018823529411767, "grad_norm": 1.1470379335352785, "learning_rate": 0.00011935658332050925, "loss": 0.6914, "step": 8079 }, { "epoch": 3.8023529411764705, "grad_norm": 0.8740483896191654, "learning_rate": 0.00011931367293490265, "loss": 0.5703, "step": 8080 }, { "epoch": 3.8028235294117647, "grad_norm": 1.2135742432772902, "learning_rate": 0.00011927076517005885, "loss": 0.8828, "step": 8081 }, { "epoch": 3.803294117647059, "grad_norm": 0.5907977783738159, "learning_rate": 0.0001192278600296425, "loss": 0.2441, "step": 8082 }, { "epoch": 3.8037647058823527, "grad_norm": 1.1825666581904215, "learning_rate": 0.00011918495751731781, "loss": 0.7109, "step": 8083 }, { "epoch": 3.804235294117647, "grad_norm": 0.7916169365138375, "learning_rate": 0.0001191420576367489, "loss": 0.6328, "step": 8084 }, { "epoch": 3.804705882352941, "grad_norm": 0.7175947078791693, "learning_rate": 0.00011909916039159958, "loss": 0.4434, "step": 8085 }, { "epoch": 3.8051764705882354, "grad_norm": 0.8938374770894284, "learning_rate": 0.00011905626578553359, "loss": 0.4512, "step": 8086 }, { "epoch": 3.8056470588235296, "grad_norm": 0.8322647439818356, "learning_rate": 0.00011901337382221425, "loss": 0.5977, "step": 8087 }, { "epoch": 3.8061176470588234, "grad_norm": 1.1134672024161065, "learning_rate": 0.00011897048450530477, "loss": 1.0078, "step": 8088 }, { "epoch": 3.8065882352941176, "grad_norm": 1.3895440051245302, "learning_rate": 0.00011892759783846812, "loss": 1.125, "step": 8089 }, { "epoch": 3.807058823529412, "grad_norm": 0.7821122794284728, "learning_rate": 0.00011888471382536704, "loss": 0.4316, "step": 8090 }, { "epoch": 3.8075294117647056, "grad_norm": 0.8481151557209625, "learning_rate": 0.00011884183246966399, "loss": 0.4805, "step": 8091 }, { "epoch": 3.808, "grad_norm": 1.1057866395227167, "learning_rate": 0.00011879895377502129, "loss": 1.3359, "step": 8092 }, { "epoch": 3.808470588235294, "grad_norm": 0.11539909662919225, "learning_rate": 0.00011875607774510099, "loss": 0.0356, "step": 8093 }, { "epoch": 3.8089411764705883, "grad_norm": 0.8245308741823325, "learning_rate": 0.00011871320438356491, "loss": 0.334, "step": 8094 }, { "epoch": 3.8094117647058825, "grad_norm": 0.7332897931982858, "learning_rate": 0.0001186703336940746, "loss": 0.3691, "step": 8095 }, { "epoch": 3.8098823529411767, "grad_norm": 1.4485812273861158, "learning_rate": 0.00011862746568029149, "loss": 1.3281, "step": 8096 }, { "epoch": 3.8103529411764705, "grad_norm": 0.9892201448046909, "learning_rate": 0.00011858460034587669, "loss": 1.375, "step": 8097 }, { "epoch": 3.8108235294117647, "grad_norm": 1.070529806727403, "learning_rate": 0.0001185417376944911, "loss": 0.8789, "step": 8098 }, { "epoch": 3.811294117647059, "grad_norm": 0.6846533515356859, "learning_rate": 0.00011849887772979541, "loss": 1.1172, "step": 8099 }, { "epoch": 3.8117647058823527, "grad_norm": 0.7693401153284306, "learning_rate": 0.00011845602045545008, "loss": 0.4297, "step": 8100 }, { "epoch": 3.812235294117647, "grad_norm": 0.9798564332046481, "learning_rate": 0.00011841316587511531, "loss": 0.7617, "step": 8101 }, { "epoch": 3.812705882352941, "grad_norm": 0.9560461878609318, "learning_rate": 0.00011837031399245115, "loss": 0.7109, "step": 8102 }, { "epoch": 3.8131764705882354, "grad_norm": 0.42908499912859266, "learning_rate": 0.00011832746481111728, "loss": 0.0869, "step": 8103 }, { "epoch": 3.8136470588235296, "grad_norm": 0.7478110250175439, "learning_rate": 0.00011828461833477326, "loss": 0.7422, "step": 8104 }, { "epoch": 3.8141176470588234, "grad_norm": 0.977493684172217, "learning_rate": 0.00011824177456707844, "loss": 0.8828, "step": 8105 }, { "epoch": 3.8145882352941176, "grad_norm": 1.189456321862756, "learning_rate": 0.00011819893351169183, "loss": 1.1719, "step": 8106 }, { "epoch": 3.815058823529412, "grad_norm": 0.39656684021641797, "learning_rate": 0.00011815609517227228, "loss": 0.4121, "step": 8107 }, { "epoch": 3.8155294117647056, "grad_norm": 1.0417169330620497, "learning_rate": 0.00011811325955247833, "loss": 0.7305, "step": 8108 }, { "epoch": 3.816, "grad_norm": 0.941045950778676, "learning_rate": 0.00011807042665596846, "loss": 1.0625, "step": 8109 }, { "epoch": 3.816470588235294, "grad_norm": 0.9026919234676984, "learning_rate": 0.0001180275964864008, "loss": 1.5625, "step": 8110 }, { "epoch": 3.8169411764705883, "grad_norm": 0.9639713856269763, "learning_rate": 0.00011798476904743321, "loss": 0.9883, "step": 8111 }, { "epoch": 3.8174117647058825, "grad_norm": 0.8698211238458471, "learning_rate": 0.00011794194434272336, "loss": 0.7891, "step": 8112 }, { "epoch": 3.8178823529411767, "grad_norm": 1.5240794103423725, "learning_rate": 0.00011789912237592874, "loss": 1.1484, "step": 8113 }, { "epoch": 3.8183529411764705, "grad_norm": 1.1274438908497681, "learning_rate": 0.00011785630315070651, "loss": 0.7617, "step": 8114 }, { "epoch": 3.8188235294117647, "grad_norm": 1.3970775286435193, "learning_rate": 0.00011781348667071365, "loss": 1.1719, "step": 8115 }, { "epoch": 3.819294117647059, "grad_norm": 0.9916174277569232, "learning_rate": 0.0001177706729396069, "loss": 0.7695, "step": 8116 }, { "epoch": 3.8197647058823527, "grad_norm": 1.076483016964259, "learning_rate": 0.00011772786196104276, "loss": 0.625, "step": 8117 }, { "epoch": 3.820235294117647, "grad_norm": 1.3162597816710895, "learning_rate": 0.00011768505373867753, "loss": 1.7422, "step": 8118 }, { "epoch": 3.820705882352941, "grad_norm": 1.7076137350426306, "learning_rate": 0.00011764224827616717, "loss": 1.9688, "step": 8119 }, { "epoch": 3.8211764705882354, "grad_norm": 1.2178858573319609, "learning_rate": 0.00011759944557716755, "loss": 1.3438, "step": 8120 }, { "epoch": 3.8216470588235296, "grad_norm": 1.578663617995464, "learning_rate": 0.00011755664564533422, "loss": 1.5625, "step": 8121 }, { "epoch": 3.8221176470588234, "grad_norm": 0.8832910349303077, "learning_rate": 0.00011751384848432245, "loss": 0.3535, "step": 8122 }, { "epoch": 3.8225882352941176, "grad_norm": 1.0561800147940272, "learning_rate": 0.00011747105409778738, "loss": 0.8281, "step": 8123 }, { "epoch": 3.823058823529412, "grad_norm": 1.293320139162042, "learning_rate": 0.00011742826248938384, "loss": 0.875, "step": 8124 }, { "epoch": 3.8235294117647056, "grad_norm": 1.4435765214462994, "learning_rate": 0.00011738547366276645, "loss": 1.1562, "step": 8125 }, { "epoch": 3.824, "grad_norm": 1.0090331673208497, "learning_rate": 0.00011734268762158957, "loss": 0.6953, "step": 8126 }, { "epoch": 3.824470588235294, "grad_norm": 0.8744134137405276, "learning_rate": 0.00011729990436950737, "loss": 0.7422, "step": 8127 }, { "epoch": 3.8249411764705883, "grad_norm": 0.9407079161766413, "learning_rate": 0.00011725712391017374, "loss": 0.7852, "step": 8128 }, { "epoch": 3.8254117647058825, "grad_norm": 0.8464826017500304, "learning_rate": 0.00011721434624724227, "loss": 0.5234, "step": 8129 }, { "epoch": 3.8258823529411767, "grad_norm": 1.0410445057573847, "learning_rate": 0.0001171715713843665, "loss": 0.7109, "step": 8130 }, { "epoch": 3.8263529411764705, "grad_norm": 0.9864113442917143, "learning_rate": 0.00011712879932519954, "loss": 0.3301, "step": 8131 }, { "epoch": 3.8268235294117647, "grad_norm": 1.0083289110310398, "learning_rate": 0.00011708603007339433, "loss": 0.7031, "step": 8132 }, { "epoch": 3.827294117647059, "grad_norm": 1.3877938634433609, "learning_rate": 0.00011704326363260365, "loss": 1.6484, "step": 8133 }, { "epoch": 3.8277647058823527, "grad_norm": 1.3706037849663384, "learning_rate": 0.00011700050000647986, "loss": 0.9336, "step": 8134 }, { "epoch": 3.828235294117647, "grad_norm": 1.3523613628735673, "learning_rate": 0.00011695773919867525, "loss": 0.7031, "step": 8135 }, { "epoch": 3.828705882352941, "grad_norm": 0.8938448353687342, "learning_rate": 0.00011691498121284176, "loss": 0.3945, "step": 8136 }, { "epoch": 3.8291764705882354, "grad_norm": 0.49537840159253393, "learning_rate": 0.00011687222605263118, "loss": 0.498, "step": 8137 }, { "epoch": 3.8296470588235296, "grad_norm": 0.7410961074069397, "learning_rate": 0.000116829473721695, "loss": 0.6875, "step": 8138 }, { "epoch": 3.8301176470588234, "grad_norm": 0.6845625462952182, "learning_rate": 0.00011678672422368442, "loss": 0.2314, "step": 8139 }, { "epoch": 3.8305882352941176, "grad_norm": 1.4615327592612273, "learning_rate": 0.00011674397756225053, "loss": 1.4688, "step": 8140 }, { "epoch": 3.831058823529412, "grad_norm": 0.8527916605878875, "learning_rate": 0.00011670123374104409, "loss": 0.4473, "step": 8141 }, { "epoch": 3.8315294117647056, "grad_norm": 1.3019754712113554, "learning_rate": 0.00011665849276371557, "loss": 1.375, "step": 8142 }, { "epoch": 3.832, "grad_norm": 0.803719444113731, "learning_rate": 0.00011661575463391534, "loss": 0.7812, "step": 8143 }, { "epoch": 3.832470588235294, "grad_norm": 1.0643437488330139, "learning_rate": 0.0001165730193552934, "loss": 0.7109, "step": 8144 }, { "epoch": 3.8329411764705883, "grad_norm": 0.7870092955799213, "learning_rate": 0.00011653028693149961, "loss": 0.2754, "step": 8145 }, { "epoch": 3.8334117647058825, "grad_norm": 1.268601470386205, "learning_rate": 0.00011648755736618342, "loss": 1.0859, "step": 8146 }, { "epoch": 3.8338823529411767, "grad_norm": 1.3531856622711256, "learning_rate": 0.00011644483066299424, "loss": 1.7812, "step": 8147 }, { "epoch": 3.8343529411764705, "grad_norm": 1.7819117952470807, "learning_rate": 0.00011640210682558112, "loss": 1.6094, "step": 8148 }, { "epoch": 3.8348235294117647, "grad_norm": 0.9524112448094272, "learning_rate": 0.00011635938585759284, "loss": 1.3906, "step": 8149 }, { "epoch": 3.835294117647059, "grad_norm": 0.8645325737928324, "learning_rate": 0.00011631666776267803, "loss": 0.6016, "step": 8150 }, { "epoch": 3.8357647058823527, "grad_norm": 1.1890342467069341, "learning_rate": 0.00011627395254448504, "loss": 0.4512, "step": 8151 }, { "epoch": 3.836235294117647, "grad_norm": 1.4648653959661793, "learning_rate": 0.00011623124020666186, "loss": 1.0625, "step": 8152 }, { "epoch": 3.836705882352941, "grad_norm": 0.9090340674782909, "learning_rate": 0.00011618853075285644, "loss": 0.5273, "step": 8153 }, { "epoch": 3.8371764705882354, "grad_norm": 0.6646349526660557, "learning_rate": 0.00011614582418671629, "loss": 0.3848, "step": 8154 }, { "epoch": 3.8376470588235296, "grad_norm": 1.0516713589347861, "learning_rate": 0.00011610312051188888, "loss": 1.2734, "step": 8155 }, { "epoch": 3.8381176470588234, "grad_norm": 0.8047165363676336, "learning_rate": 0.00011606041973202123, "loss": 0.5156, "step": 8156 }, { "epoch": 3.8385882352941176, "grad_norm": 0.979813159787177, "learning_rate": 0.00011601772185076018, "loss": 0.707, "step": 8157 }, { "epoch": 3.839058823529412, "grad_norm": 0.9922047049176059, "learning_rate": 0.00011597502687175238, "loss": 0.5938, "step": 8158 }, { "epoch": 3.8395294117647056, "grad_norm": 0.7715327955967065, "learning_rate": 0.0001159323347986442, "loss": 0.2949, "step": 8159 }, { "epoch": 3.84, "grad_norm": 2.0844407848560564, "learning_rate": 0.00011588964563508167, "loss": 1.5938, "step": 8160 }, { "epoch": 3.840470588235294, "grad_norm": 1.4302541290517201, "learning_rate": 0.00011584695938471077, "loss": 1.1094, "step": 8161 }, { "epoch": 3.8409411764705883, "grad_norm": 0.6469692512886501, "learning_rate": 0.00011580427605117705, "loss": 0.1758, "step": 8162 }, { "epoch": 3.8414117647058825, "grad_norm": 0.9049218493543479, "learning_rate": 0.00011576159563812589, "loss": 0.4062, "step": 8163 }, { "epoch": 3.8418823529411767, "grad_norm": 1.05153863236791, "learning_rate": 0.00011571891814920236, "loss": 0.4668, "step": 8164 }, { "epoch": 3.8423529411764705, "grad_norm": 0.5211174407566758, "learning_rate": 0.00011567624358805139, "loss": 0.5117, "step": 8165 }, { "epoch": 3.8428235294117647, "grad_norm": 0.7322586768568641, "learning_rate": 0.0001156335719583176, "loss": 0.291, "step": 8166 }, { "epoch": 3.843294117647059, "grad_norm": 1.075659939857615, "learning_rate": 0.00011559090326364527, "loss": 0.9766, "step": 8167 }, { "epoch": 3.8437647058823528, "grad_norm": 0.912300103035643, "learning_rate": 0.00011554823750767863, "loss": 0.8945, "step": 8168 }, { "epoch": 3.844235294117647, "grad_norm": 0.9725945116283417, "learning_rate": 0.0001155055746940615, "loss": 0.6602, "step": 8169 }, { "epoch": 3.844705882352941, "grad_norm": 1.2072300538271115, "learning_rate": 0.00011546291482643744, "loss": 1.2344, "step": 8170 }, { "epoch": 3.8451764705882354, "grad_norm": 0.9325164665457522, "learning_rate": 0.00011542025790844988, "loss": 0.6133, "step": 8171 }, { "epoch": 3.8456470588235296, "grad_norm": 1.3054367302824266, "learning_rate": 0.00011537760394374192, "loss": 1.1484, "step": 8172 }, { "epoch": 3.8461176470588234, "grad_norm": 0.992489815002253, "learning_rate": 0.00011533495293595637, "loss": 1.0156, "step": 8173 }, { "epoch": 3.8465882352941176, "grad_norm": 0.8536658219268071, "learning_rate": 0.0001152923048887359, "loss": 0.4102, "step": 8174 }, { "epoch": 3.847058823529412, "grad_norm": 1.1227407829137996, "learning_rate": 0.00011524965980572284, "loss": 1.2891, "step": 8175 }, { "epoch": 3.8475294117647056, "grad_norm": 0.5378593645192751, "learning_rate": 0.00011520701769055927, "loss": 0.3359, "step": 8176 }, { "epoch": 3.848, "grad_norm": 0.8753941480204209, "learning_rate": 0.00011516437854688703, "loss": 0.8359, "step": 8177 }, { "epoch": 3.848470588235294, "grad_norm": 0.6655301691121572, "learning_rate": 0.00011512174237834776, "loss": 0.2578, "step": 8178 }, { "epoch": 3.8489411764705883, "grad_norm": 1.1127735128282379, "learning_rate": 0.00011507910918858276, "loss": 0.7109, "step": 8179 }, { "epoch": 3.8494117647058825, "grad_norm": 0.9318656523869772, "learning_rate": 0.0001150364789812331, "loss": 0.8828, "step": 8180 }, { "epoch": 3.8498823529411763, "grad_norm": 1.216589164597941, "learning_rate": 0.00011499385175993965, "loss": 1.3047, "step": 8181 }, { "epoch": 3.8503529411764705, "grad_norm": 0.8752855500208798, "learning_rate": 0.00011495122752834296, "loss": 0.6172, "step": 8182 }, { "epoch": 3.8508235294117648, "grad_norm": 0.926331713110355, "learning_rate": 0.00011490860629008333, "loss": 1.0703, "step": 8183 }, { "epoch": 3.851294117647059, "grad_norm": 0.8095107368003367, "learning_rate": 0.00011486598804880088, "loss": 1.1406, "step": 8184 }, { "epoch": 3.8517647058823528, "grad_norm": 0.9481810800833225, "learning_rate": 0.00011482337280813536, "loss": 0.9727, "step": 8185 }, { "epoch": 3.852235294117647, "grad_norm": 1.0490850760357424, "learning_rate": 0.00011478076057172638, "loss": 0.7383, "step": 8186 }, { "epoch": 3.852705882352941, "grad_norm": 1.6813020335582765, "learning_rate": 0.00011473815134321315, "loss": 0.9531, "step": 8187 }, { "epoch": 3.8531764705882354, "grad_norm": 1.1941559566556952, "learning_rate": 0.00011469554512623477, "loss": 1.8047, "step": 8188 }, { "epoch": 3.8536470588235296, "grad_norm": 1.0218659306785727, "learning_rate": 0.00011465294192443001, "loss": 0.6289, "step": 8189 }, { "epoch": 3.8541176470588234, "grad_norm": 1.337034164925079, "learning_rate": 0.00011461034174143736, "loss": 1.2266, "step": 8190 }, { "epoch": 3.8545882352941176, "grad_norm": 0.7911868918125554, "learning_rate": 0.00011456774458089515, "loss": 0.4883, "step": 8191 }, { "epoch": 3.855058823529412, "grad_norm": 0.7952540209567288, "learning_rate": 0.00011452515044644132, "loss": 0.8867, "step": 8192 }, { "epoch": 3.8555294117647056, "grad_norm": 0.8244267069437838, "learning_rate": 0.00011448255934171363, "loss": 0.3945, "step": 8193 }, { "epoch": 3.856, "grad_norm": 1.2293811549419198, "learning_rate": 0.00011443997127034962, "loss": 2.0625, "step": 8194 }, { "epoch": 3.856470588235294, "grad_norm": 1.128968166186706, "learning_rate": 0.0001143973862359865, "loss": 0.8672, "step": 8195 }, { "epoch": 3.8569411764705883, "grad_norm": 0.9594324463205213, "learning_rate": 0.0001143548042422612, "loss": 0.4941, "step": 8196 }, { "epoch": 3.8574117647058825, "grad_norm": 0.8729191407908686, "learning_rate": 0.0001143122252928104, "loss": 0.3848, "step": 8197 }, { "epoch": 3.8578823529411763, "grad_norm": 0.9770987818485782, "learning_rate": 0.00011426964939127063, "loss": 0.4238, "step": 8198 }, { "epoch": 3.8583529411764705, "grad_norm": 1.1348927663453106, "learning_rate": 0.00011422707654127809, "loss": 0.957, "step": 8199 }, { "epoch": 3.8588235294117648, "grad_norm": 0.9040041860134865, "learning_rate": 0.00011418450674646868, "loss": 1.3672, "step": 8200 }, { "epoch": 3.859294117647059, "grad_norm": 0.7965867417624504, "learning_rate": 0.00011414194001047806, "loss": 0.2871, "step": 8201 }, { "epoch": 3.8597647058823528, "grad_norm": 0.6311898067492079, "learning_rate": 0.00011409937633694165, "loss": 0.2773, "step": 8202 }, { "epoch": 3.860235294117647, "grad_norm": 1.0107045253605442, "learning_rate": 0.0001140568157294946, "loss": 1.0938, "step": 8203 }, { "epoch": 3.860705882352941, "grad_norm": 1.2085559863184505, "learning_rate": 0.00011401425819177179, "loss": 1.2109, "step": 8204 }, { "epoch": 3.8611764705882354, "grad_norm": 0.9888846178265014, "learning_rate": 0.00011397170372740782, "loss": 0.8203, "step": 8205 }, { "epoch": 3.8616470588235297, "grad_norm": 1.0037234350276736, "learning_rate": 0.0001139291523400371, "loss": 0.9766, "step": 8206 }, { "epoch": 3.8621176470588234, "grad_norm": 1.4057461092628136, "learning_rate": 0.00011388660403329368, "loss": 1.4141, "step": 8207 }, { "epoch": 3.8625882352941177, "grad_norm": 0.9851131759511368, "learning_rate": 0.0001138440588108114, "loss": 1.0859, "step": 8208 }, { "epoch": 3.863058823529412, "grad_norm": 0.8131473703460509, "learning_rate": 0.00011380151667622384, "loss": 0.3027, "step": 8209 }, { "epoch": 3.8635294117647057, "grad_norm": 1.1949296645875223, "learning_rate": 0.00011375897763316433, "loss": 1.0781, "step": 8210 }, { "epoch": 3.864, "grad_norm": 1.5154055846355765, "learning_rate": 0.00011371644168526586, "loss": 1.4297, "step": 8211 }, { "epoch": 3.864470588235294, "grad_norm": 1.0478803272756327, "learning_rate": 0.00011367390883616126, "loss": 0.9414, "step": 8212 }, { "epoch": 3.8649411764705883, "grad_norm": 1.0444814744698157, "learning_rate": 0.00011363137908948303, "loss": 1.3047, "step": 8213 }, { "epoch": 3.8654117647058825, "grad_norm": 0.7963273101777815, "learning_rate": 0.00011358885244886342, "loss": 0.3555, "step": 8214 }, { "epoch": 3.8658823529411763, "grad_norm": 1.2105823616404232, "learning_rate": 0.00011354632891793437, "loss": 1.0156, "step": 8215 }, { "epoch": 3.8663529411764705, "grad_norm": 0.9302882016363433, "learning_rate": 0.00011350380850032766, "loss": 0.9844, "step": 8216 }, { "epoch": 3.8668235294117648, "grad_norm": 1.200549551851389, "learning_rate": 0.00011346129119967473, "loss": 0.8906, "step": 8217 }, { "epoch": 3.8672941176470585, "grad_norm": 1.061033603970267, "learning_rate": 0.0001134187770196067, "loss": 1.2344, "step": 8218 }, { "epoch": 3.8677647058823528, "grad_norm": 1.1495739777997638, "learning_rate": 0.00011337626596375458, "loss": 0.9883, "step": 8219 }, { "epoch": 3.868235294117647, "grad_norm": 0.7702573611345971, "learning_rate": 0.000113333758035749, "loss": 0.3574, "step": 8220 }, { "epoch": 3.868705882352941, "grad_norm": 0.7015221134159508, "learning_rate": 0.00011329125323922031, "loss": 0.3203, "step": 8221 }, { "epoch": 3.8691764705882354, "grad_norm": 0.8220782591475888, "learning_rate": 0.00011324875157779865, "loss": 0.3789, "step": 8222 }, { "epoch": 3.8696470588235297, "grad_norm": 0.9585246448321088, "learning_rate": 0.00011320625305511388, "loss": 0.918, "step": 8223 }, { "epoch": 3.8701176470588234, "grad_norm": 0.7317308368457606, "learning_rate": 0.00011316375767479554, "loss": 0.5859, "step": 8224 }, { "epoch": 3.8705882352941177, "grad_norm": 1.0981601730146342, "learning_rate": 0.00011312126544047302, "loss": 0.8555, "step": 8225 }, { "epoch": 3.871058823529412, "grad_norm": 1.0674693850062633, "learning_rate": 0.0001130787763557753, "loss": 0.707, "step": 8226 }, { "epoch": 3.8715294117647057, "grad_norm": 1.2163681826316068, "learning_rate": 0.00011303629042433118, "loss": 1.4141, "step": 8227 }, { "epoch": 3.872, "grad_norm": 1.2195079320821427, "learning_rate": 0.00011299380764976915, "loss": 0.707, "step": 8228 }, { "epoch": 3.872470588235294, "grad_norm": 1.4967869882556084, "learning_rate": 0.00011295132803571747, "loss": 1.3203, "step": 8229 }, { "epoch": 3.8729411764705883, "grad_norm": 1.4037385248228964, "learning_rate": 0.00011290885158580411, "loss": 0.9609, "step": 8230 }, { "epoch": 3.8734117647058826, "grad_norm": 1.1132710413480067, "learning_rate": 0.00011286637830365672, "loss": 1.1875, "step": 8231 }, { "epoch": 3.8738823529411763, "grad_norm": 0.6641659996095337, "learning_rate": 0.0001128239081929028, "loss": 0.332, "step": 8232 }, { "epoch": 3.8743529411764706, "grad_norm": 1.2577362790972606, "learning_rate": 0.00011278144125716946, "loss": 1.6172, "step": 8233 }, { "epoch": 3.8748235294117648, "grad_norm": 1.3345250105027822, "learning_rate": 0.00011273897750008358, "loss": 1.7891, "step": 8234 }, { "epoch": 3.8752941176470586, "grad_norm": 1.118202021336586, "learning_rate": 0.0001126965169252718, "loss": 0.7617, "step": 8235 }, { "epoch": 3.875764705882353, "grad_norm": 1.096192368631524, "learning_rate": 0.00011265405953636044, "loss": 1.1875, "step": 8236 }, { "epoch": 3.876235294117647, "grad_norm": 0.930308997537824, "learning_rate": 0.0001126116053369756, "loss": 0.6562, "step": 8237 }, { "epoch": 3.8767058823529412, "grad_norm": 0.7286204530623482, "learning_rate": 0.000112569154330743, "loss": 0.5781, "step": 8238 }, { "epoch": 3.8771764705882354, "grad_norm": 0.6119304934667332, "learning_rate": 0.00011252670652128825, "loss": 0.168, "step": 8239 }, { "epoch": 3.8776470588235297, "grad_norm": 1.2287685193921893, "learning_rate": 0.00011248426191223658, "loss": 1.625, "step": 8240 }, { "epoch": 3.8781176470588234, "grad_norm": 1.0157803608630225, "learning_rate": 0.00011244182050721293, "loss": 0.5469, "step": 8241 }, { "epoch": 3.8785882352941177, "grad_norm": 0.8044948024926196, "learning_rate": 0.00011239938230984198, "loss": 0.4688, "step": 8242 }, { "epoch": 3.879058823529412, "grad_norm": 0.6008919213014771, "learning_rate": 0.00011235694732374829, "loss": 0.3398, "step": 8243 }, { "epoch": 3.8795294117647057, "grad_norm": 1.6987571283368563, "learning_rate": 0.00011231451555255592, "loss": 1.0625, "step": 8244 }, { "epoch": 3.88, "grad_norm": 0.7183802011363299, "learning_rate": 0.00011227208699988879, "loss": 0.2334, "step": 8245 }, { "epoch": 3.880470588235294, "grad_norm": 1.150309049696857, "learning_rate": 0.00011222966166937044, "loss": 0.8672, "step": 8246 }, { "epoch": 3.8809411764705883, "grad_norm": 0.8856778593951127, "learning_rate": 0.00011218723956462427, "loss": 0.9336, "step": 8247 }, { "epoch": 3.8814117647058826, "grad_norm": 1.0935376216598125, "learning_rate": 0.00011214482068927332, "loss": 0.8125, "step": 8248 }, { "epoch": 3.8818823529411763, "grad_norm": 0.6802084561361749, "learning_rate": 0.00011210240504694033, "loss": 0.543, "step": 8249 }, { "epoch": 3.8823529411764706, "grad_norm": 1.0078581258211596, "learning_rate": 0.00011205999264124786, "loss": 0.6211, "step": 8250 }, { "epoch": 3.882823529411765, "grad_norm": 0.3404192039756728, "learning_rate": 0.00011201758347581814, "loss": 0.0559, "step": 8251 }, { "epoch": 3.8832941176470586, "grad_norm": 0.445280153750344, "learning_rate": 0.00011197517755427303, "loss": 0.1196, "step": 8252 }, { "epoch": 3.883764705882353, "grad_norm": 0.9738859910974418, "learning_rate": 0.00011193277488023432, "loss": 1.0234, "step": 8253 }, { "epoch": 3.884235294117647, "grad_norm": 0.9350195672477319, "learning_rate": 0.00011189037545732335, "loss": 0.7695, "step": 8254 }, { "epoch": 3.8847058823529412, "grad_norm": 0.7824442027606408, "learning_rate": 0.00011184797928916124, "loss": 0.4766, "step": 8255 }, { "epoch": 3.8851764705882355, "grad_norm": 0.9965814452915581, "learning_rate": 0.0001118055863793688, "loss": 0.7266, "step": 8256 }, { "epoch": 3.8856470588235297, "grad_norm": 0.7343035172596847, "learning_rate": 0.00011176319673156666, "loss": 0.3008, "step": 8257 }, { "epoch": 3.8861176470588235, "grad_norm": 0.8107790475230716, "learning_rate": 0.00011172081034937507, "loss": 0.4062, "step": 8258 }, { "epoch": 3.8865882352941177, "grad_norm": 0.44721066801886844, "learning_rate": 0.00011167842723641398, "loss": 0.127, "step": 8259 }, { "epoch": 3.887058823529412, "grad_norm": 1.3385091285389978, "learning_rate": 0.00011163604739630324, "loss": 0.6602, "step": 8260 }, { "epoch": 3.8875294117647057, "grad_norm": 1.3728242500993506, "learning_rate": 0.00011159367083266219, "loss": 1.2734, "step": 8261 }, { "epoch": 3.888, "grad_norm": 0.7511351736411104, "learning_rate": 0.00011155129754911002, "loss": 0.3164, "step": 8262 }, { "epoch": 3.888470588235294, "grad_norm": 0.865955244412615, "learning_rate": 0.00011150892754926568, "loss": 0.3398, "step": 8263 }, { "epoch": 3.8889411764705883, "grad_norm": 0.7936128870720757, "learning_rate": 0.00011146656083674768, "loss": 0.3555, "step": 8264 }, { "epoch": 3.8894117647058826, "grad_norm": 1.1270205396140194, "learning_rate": 0.00011142419741517443, "loss": 0.7695, "step": 8265 }, { "epoch": 3.8898823529411763, "grad_norm": 1.0083584032725885, "learning_rate": 0.00011138183728816389, "loss": 0.8594, "step": 8266 }, { "epoch": 3.8903529411764706, "grad_norm": 0.6676751109171543, "learning_rate": 0.0001113394804593339, "loss": 0.2539, "step": 8267 }, { "epoch": 3.890823529411765, "grad_norm": 0.9894857276640106, "learning_rate": 0.00011129712693230191, "loss": 0.8203, "step": 8268 }, { "epoch": 3.8912941176470586, "grad_norm": 0.6217000689277996, "learning_rate": 0.00011125477671068506, "loss": 0.1211, "step": 8269 }, { "epoch": 3.891764705882353, "grad_norm": 0.99598316059136, "learning_rate": 0.0001112124297981004, "loss": 0.7344, "step": 8270 }, { "epoch": 3.892235294117647, "grad_norm": 1.0820939807843972, "learning_rate": 0.00011117008619816445, "loss": 0.8359, "step": 8271 }, { "epoch": 3.8927058823529412, "grad_norm": 0.8900109049915321, "learning_rate": 0.00011112774591449356, "loss": 0.9336, "step": 8272 }, { "epoch": 3.8931764705882355, "grad_norm": 1.090440267280845, "learning_rate": 0.00011108540895070388, "loss": 1.4531, "step": 8273 }, { "epoch": 3.8936470588235297, "grad_norm": 0.6250952914875749, "learning_rate": 0.00011104307531041117, "loss": 0.2197, "step": 8274 }, { "epoch": 3.8941176470588235, "grad_norm": 0.8183800304202429, "learning_rate": 0.0001110007449972309, "loss": 0.3926, "step": 8275 }, { "epoch": 3.8945882352941177, "grad_norm": 0.9312477604410098, "learning_rate": 0.00011095841801477826, "loss": 0.3145, "step": 8276 }, { "epoch": 3.895058823529412, "grad_norm": 1.3645146615794097, "learning_rate": 0.00011091609436666825, "loss": 1.0469, "step": 8277 }, { "epoch": 3.8955294117647057, "grad_norm": 0.7694780266599267, "learning_rate": 0.00011087377405651549, "loss": 0.5625, "step": 8278 }, { "epoch": 3.896, "grad_norm": 0.7856609800876821, "learning_rate": 0.00011083145708793431, "loss": 0.3496, "step": 8279 }, { "epoch": 3.896470588235294, "grad_norm": 0.4691604186469935, "learning_rate": 0.00011078914346453883, "loss": 0.1289, "step": 8280 }, { "epoch": 3.8969411764705884, "grad_norm": 0.7900749605133169, "learning_rate": 0.00011074683318994286, "loss": 1.4219, "step": 8281 }, { "epoch": 3.8974117647058826, "grad_norm": 1.3234073773132307, "learning_rate": 0.00011070452626775981, "loss": 1.7812, "step": 8282 }, { "epoch": 3.8978823529411764, "grad_norm": 0.7672087972945975, "learning_rate": 0.000110662222701603, "loss": 0.5469, "step": 8283 }, { "epoch": 3.8983529411764706, "grad_norm": 1.1439023766876286, "learning_rate": 0.00011061992249508533, "loss": 1.0156, "step": 8284 }, { "epoch": 3.898823529411765, "grad_norm": 1.0580429095790391, "learning_rate": 0.0001105776256518194, "loss": 0.8906, "step": 8285 }, { "epoch": 3.8992941176470586, "grad_norm": 0.8304729181034881, "learning_rate": 0.0001105353321754176, "loss": 0.3301, "step": 8286 }, { "epoch": 3.899764705882353, "grad_norm": 1.3178528468739636, "learning_rate": 0.000110493042069492, "loss": 1.2266, "step": 8287 }, { "epoch": 3.900235294117647, "grad_norm": 0.5682115813863925, "learning_rate": 0.00011045075533765444, "loss": 0.2324, "step": 8288 }, { "epoch": 3.9007058823529412, "grad_norm": 1.0939083322675713, "learning_rate": 0.00011040847198351635, "loss": 0.9102, "step": 8289 }, { "epoch": 3.9011764705882355, "grad_norm": 0.7830368031681146, "learning_rate": 0.00011036619201068892, "loss": 0.4023, "step": 8290 }, { "epoch": 3.9016470588235292, "grad_norm": 1.8269574126607036, "learning_rate": 0.00011032391542278313, "loss": 1.7266, "step": 8291 }, { "epoch": 3.9021176470588235, "grad_norm": 1.0719727300431063, "learning_rate": 0.00011028164222340957, "loss": 0.5078, "step": 8292 }, { "epoch": 3.9025882352941177, "grad_norm": 1.006449973580832, "learning_rate": 0.00011023937241617853, "loss": 1.0156, "step": 8293 }, { "epoch": 3.903058823529412, "grad_norm": 1.2265053821662124, "learning_rate": 0.00011019710600470016, "loss": 0.8047, "step": 8294 }, { "epoch": 3.9035294117647057, "grad_norm": 1.1699996425851227, "learning_rate": 0.00011015484299258417, "loss": 0.9297, "step": 8295 }, { "epoch": 3.904, "grad_norm": 0.6022911776533348, "learning_rate": 0.00011011258338344002, "loss": 0.248, "step": 8296 }, { "epoch": 3.904470588235294, "grad_norm": 1.2039694081340977, "learning_rate": 0.00011007032718087688, "loss": 1.5469, "step": 8297 }, { "epoch": 3.9049411764705884, "grad_norm": 1.3086576109131207, "learning_rate": 0.00011002807438850367, "loss": 0.6914, "step": 8298 }, { "epoch": 3.9054117647058826, "grad_norm": 1.0091024644061983, "learning_rate": 0.00010998582500992897, "loss": 0.9531, "step": 8299 }, { "epoch": 3.9058823529411764, "grad_norm": 0.8061250420685969, "learning_rate": 0.00010994357904876106, "loss": 0.3633, "step": 8300 }, { "epoch": 3.9063529411764706, "grad_norm": 1.668845319556603, "learning_rate": 0.000109901336508608, "loss": 1.4844, "step": 8301 }, { "epoch": 3.906823529411765, "grad_norm": 0.8484887848713093, "learning_rate": 0.00010985909739307748, "loss": 0.6602, "step": 8302 }, { "epoch": 3.9072941176470586, "grad_norm": 0.9772703960004389, "learning_rate": 0.0001098168617057769, "loss": 0.8203, "step": 8303 }, { "epoch": 3.907764705882353, "grad_norm": 0.7474992043141098, "learning_rate": 0.00010977462945031347, "loss": 0.5703, "step": 8304 }, { "epoch": 3.908235294117647, "grad_norm": 0.7883659686366368, "learning_rate": 0.00010973240063029398, "loss": 0.2676, "step": 8305 }, { "epoch": 3.9087058823529413, "grad_norm": 0.8808566356547175, "learning_rate": 0.000109690175249325, "loss": 0.6914, "step": 8306 }, { "epoch": 3.9091764705882355, "grad_norm": 0.9027873553488551, "learning_rate": 0.00010964795331101273, "loss": 0.4492, "step": 8307 }, { "epoch": 3.9096470588235293, "grad_norm": 1.1180386448805792, "learning_rate": 0.00010960573481896322, "loss": 0.7539, "step": 8308 }, { "epoch": 3.9101176470588235, "grad_norm": 0.9165908704763142, "learning_rate": 0.00010956351977678206, "loss": 0.8281, "step": 8309 }, { "epoch": 3.9105882352941177, "grad_norm": 0.8731921106938327, "learning_rate": 0.00010952130818807465, "loss": 1.0391, "step": 8310 }, { "epoch": 3.911058823529412, "grad_norm": 1.023814943132181, "learning_rate": 0.00010947910005644608, "loss": 1.1094, "step": 8311 }, { "epoch": 3.9115294117647057, "grad_norm": 0.9155649019346892, "learning_rate": 0.00010943689538550111, "loss": 0.8945, "step": 8312 }, { "epoch": 3.912, "grad_norm": 0.9600539964123918, "learning_rate": 0.00010939469417884424, "loss": 0.6992, "step": 8313 }, { "epoch": 3.912470588235294, "grad_norm": 0.4411429751739136, "learning_rate": 0.00010935249644007966, "loss": 0.1079, "step": 8314 }, { "epoch": 3.9129411764705884, "grad_norm": 1.11414032645435, "learning_rate": 0.00010931030217281125, "loss": 0.9609, "step": 8315 }, { "epoch": 3.9134117647058826, "grad_norm": 0.6169652022064476, "learning_rate": 0.00010926811138064264, "loss": 0.1855, "step": 8316 }, { "epoch": 3.9138823529411764, "grad_norm": 0.7193578580647567, "learning_rate": 0.00010922592406717706, "loss": 0.2949, "step": 8317 }, { "epoch": 3.9143529411764706, "grad_norm": 1.0334758090975023, "learning_rate": 0.0001091837402360176, "loss": 1.1875, "step": 8318 }, { "epoch": 3.914823529411765, "grad_norm": 1.0112222054911522, "learning_rate": 0.00010914155989076692, "loss": 1.0938, "step": 8319 }, { "epoch": 3.9152941176470586, "grad_norm": 1.0550176054563936, "learning_rate": 0.00010909938303502741, "loss": 0.9609, "step": 8320 }, { "epoch": 3.915764705882353, "grad_norm": 1.0326328340577533, "learning_rate": 0.00010905720967240122, "loss": 0.8398, "step": 8321 }, { "epoch": 3.916235294117647, "grad_norm": 0.6922504668558375, "learning_rate": 0.00010901503980649017, "loss": 0.3496, "step": 8322 }, { "epoch": 3.9167058823529413, "grad_norm": 0.8950442377578165, "learning_rate": 0.0001089728734408957, "loss": 0.6367, "step": 8323 }, { "epoch": 3.9171764705882355, "grad_norm": 0.8805810721111348, "learning_rate": 0.00010893071057921911, "loss": 0.334, "step": 8324 }, { "epoch": 3.9176470588235293, "grad_norm": 1.0044058905458615, "learning_rate": 0.0001088885512250613, "loss": 0.793, "step": 8325 }, { "epoch": 3.9181176470588235, "grad_norm": 0.8420148882297136, "learning_rate": 0.00010884639538202285, "loss": 0.4609, "step": 8326 }, { "epoch": 3.9185882352941177, "grad_norm": 0.8441771576225531, "learning_rate": 0.00010880424305370406, "loss": 0.3184, "step": 8327 }, { "epoch": 3.919058823529412, "grad_norm": 1.1474266065514036, "learning_rate": 0.00010876209424370502, "loss": 0.7656, "step": 8328 }, { "epoch": 3.9195294117647057, "grad_norm": 1.112917618521068, "learning_rate": 0.00010871994895562538, "loss": 1.25, "step": 8329 }, { "epoch": 3.92, "grad_norm": 1.3461488061183655, "learning_rate": 0.00010867780719306458, "loss": 0.8008, "step": 8330 }, { "epoch": 3.920470588235294, "grad_norm": 0.6712625098796244, "learning_rate": 0.00010863566895962166, "loss": 0.2676, "step": 8331 }, { "epoch": 3.9209411764705884, "grad_norm": 0.7938069910851849, "learning_rate": 0.0001085935342588956, "loss": 0.6562, "step": 8332 }, { "epoch": 3.9214117647058826, "grad_norm": 0.6310454118298636, "learning_rate": 0.00010855140309448479, "loss": 0.5273, "step": 8333 }, { "epoch": 3.9218823529411764, "grad_norm": 0.5909517171883825, "learning_rate": 0.00010850927546998744, "loss": 0.167, "step": 8334 }, { "epoch": 3.9223529411764706, "grad_norm": 1.0216780059341963, "learning_rate": 0.00010846715138900148, "loss": 0.7148, "step": 8335 }, { "epoch": 3.922823529411765, "grad_norm": 0.7898881086221627, "learning_rate": 0.00010842503085512453, "loss": 0.3887, "step": 8336 }, { "epoch": 3.9232941176470586, "grad_norm": 0.8668771889945381, "learning_rate": 0.00010838291387195387, "loss": 0.3672, "step": 8337 }, { "epoch": 3.923764705882353, "grad_norm": 0.980631288272517, "learning_rate": 0.00010834080044308646, "loss": 0.4277, "step": 8338 }, { "epoch": 3.924235294117647, "grad_norm": 0.8914789185816164, "learning_rate": 0.00010829869057211906, "loss": 0.7344, "step": 8339 }, { "epoch": 3.9247058823529413, "grad_norm": 0.9853059134718671, "learning_rate": 0.00010825658426264803, "loss": 0.7617, "step": 8340 }, { "epoch": 3.9251764705882355, "grad_norm": 1.1428229271726202, "learning_rate": 0.00010821448151826941, "loss": 1.0156, "step": 8341 }, { "epoch": 3.9256470588235293, "grad_norm": 1.156196278355313, "learning_rate": 0.00010817238234257907, "loss": 1.3047, "step": 8342 }, { "epoch": 3.9261176470588235, "grad_norm": 0.859647224222282, "learning_rate": 0.00010813028673917245, "loss": 0.3574, "step": 8343 }, { "epoch": 3.9265882352941177, "grad_norm": 0.9236441529051724, "learning_rate": 0.0001080881947116447, "loss": 0.8789, "step": 8344 }, { "epoch": 3.9270588235294115, "grad_norm": 0.9114530126020649, "learning_rate": 0.00010804610626359068, "loss": 0.457, "step": 8345 }, { "epoch": 3.9275294117647057, "grad_norm": 1.2172560115614834, "learning_rate": 0.000108004021398605, "loss": 1.4219, "step": 8346 }, { "epoch": 3.928, "grad_norm": 1.4401222699197826, "learning_rate": 0.00010796194012028185, "loss": 1.7109, "step": 8347 }, { "epoch": 3.928470588235294, "grad_norm": 0.8897955068457375, "learning_rate": 0.00010791986243221521, "loss": 1.2812, "step": 8348 }, { "epoch": 3.9289411764705884, "grad_norm": 1.080527790176006, "learning_rate": 0.00010787778833799875, "loss": 1.2109, "step": 8349 }, { "epoch": 3.9294117647058826, "grad_norm": 1.298716673061074, "learning_rate": 0.00010783571784122577, "loss": 1.625, "step": 8350 }, { "epoch": 3.9298823529411764, "grad_norm": 1.4631978854640406, "learning_rate": 0.00010779365094548926, "loss": 1.2812, "step": 8351 }, { "epoch": 3.9303529411764706, "grad_norm": 1.1322074306454932, "learning_rate": 0.00010775158765438202, "loss": 0.9336, "step": 8352 }, { "epoch": 3.930823529411765, "grad_norm": 0.9732363363489003, "learning_rate": 0.0001077095279714964, "loss": 0.5898, "step": 8353 }, { "epoch": 3.9312941176470586, "grad_norm": 0.6669792582227922, "learning_rate": 0.00010766747190042452, "loss": 0.6133, "step": 8354 }, { "epoch": 3.931764705882353, "grad_norm": 1.4097316760899326, "learning_rate": 0.00010762541944475821, "loss": 1.0859, "step": 8355 }, { "epoch": 3.932235294117647, "grad_norm": 0.9793166073997741, "learning_rate": 0.0001075833706080889, "loss": 1.1016, "step": 8356 }, { "epoch": 3.9327058823529413, "grad_norm": 0.5820622729414354, "learning_rate": 0.00010754132539400782, "loss": 0.252, "step": 8357 }, { "epoch": 3.9331764705882355, "grad_norm": 1.3843442669334844, "learning_rate": 0.00010749928380610577, "loss": 1.4609, "step": 8358 }, { "epoch": 3.9336470588235293, "grad_norm": 0.6637200867918849, "learning_rate": 0.00010745724584797338, "loss": 0.4688, "step": 8359 }, { "epoch": 3.9341176470588235, "grad_norm": 1.1003554703211247, "learning_rate": 0.00010741521152320088, "loss": 1.2656, "step": 8360 }, { "epoch": 3.9345882352941177, "grad_norm": 0.9538285593846847, "learning_rate": 0.00010737318083537813, "loss": 1.0391, "step": 8361 }, { "epoch": 3.9350588235294115, "grad_norm": 0.9365246288760849, "learning_rate": 0.00010733115378809485, "loss": 0.7812, "step": 8362 }, { "epoch": 3.9355294117647057, "grad_norm": 1.1185019603263675, "learning_rate": 0.00010728913038494038, "loss": 1.2578, "step": 8363 }, { "epoch": 3.936, "grad_norm": 0.8438459493647245, "learning_rate": 0.00010724711062950358, "loss": 0.3789, "step": 8364 }, { "epoch": 3.936470588235294, "grad_norm": 0.7841005100517411, "learning_rate": 0.00010720509452537329, "loss": 0.3262, "step": 8365 }, { "epoch": 3.9369411764705884, "grad_norm": 0.9548786790787258, "learning_rate": 0.00010716308207613784, "loss": 0.4609, "step": 8366 }, { "epoch": 3.9374117647058826, "grad_norm": 1.1182734243802868, "learning_rate": 0.00010712107328538528, "loss": 0.9805, "step": 8367 }, { "epoch": 3.9378823529411764, "grad_norm": 0.925450779128971, "learning_rate": 0.00010707906815670337, "loss": 1.1562, "step": 8368 }, { "epoch": 3.9383529411764706, "grad_norm": 1.5588096867156542, "learning_rate": 0.00010703706669367956, "loss": 0.8906, "step": 8369 }, { "epoch": 3.938823529411765, "grad_norm": 0.9256665901229036, "learning_rate": 0.00010699506889990102, "loss": 0.5938, "step": 8370 }, { "epoch": 3.9392941176470586, "grad_norm": 0.7003260644340048, "learning_rate": 0.00010695307477895449, "loss": 0.2139, "step": 8371 }, { "epoch": 3.939764705882353, "grad_norm": 1.5040206089791757, "learning_rate": 0.00010691108433442656, "loss": 1.3047, "step": 8372 }, { "epoch": 3.940235294117647, "grad_norm": 1.050599211368537, "learning_rate": 0.00010686909756990337, "loss": 0.5234, "step": 8373 }, { "epoch": 3.9407058823529413, "grad_norm": 0.793589804779773, "learning_rate": 0.00010682711448897077, "loss": 0.4727, "step": 8374 }, { "epoch": 3.9411764705882355, "grad_norm": 0.7120362685520978, "learning_rate": 0.00010678513509521435, "loss": 0.2334, "step": 8375 }, { "epoch": 3.9416470588235293, "grad_norm": 0.6582903458332148, "learning_rate": 0.00010674315939221937, "loss": 0.9297, "step": 8376 }, { "epoch": 3.9421176470588235, "grad_norm": 1.1390815539676526, "learning_rate": 0.00010670118738357078, "loss": 0.9023, "step": 8377 }, { "epoch": 3.9425882352941177, "grad_norm": 0.6185839644346361, "learning_rate": 0.00010665921907285317, "loss": 0.2637, "step": 8378 }, { "epoch": 3.9430588235294115, "grad_norm": 1.4104913664295395, "learning_rate": 0.0001066172544636508, "loss": 0.9023, "step": 8379 }, { "epoch": 3.9435294117647057, "grad_norm": 0.8381477503839413, "learning_rate": 0.00010657529355954771, "loss": 0.8828, "step": 8380 }, { "epoch": 3.944, "grad_norm": 0.6249523956887287, "learning_rate": 0.00010653333636412758, "loss": 0.2012, "step": 8381 }, { "epoch": 3.944470588235294, "grad_norm": 1.2115164141258077, "learning_rate": 0.00010649138288097366, "loss": 1.1328, "step": 8382 }, { "epoch": 3.9449411764705884, "grad_norm": 1.0980807386732891, "learning_rate": 0.00010644943311366911, "loss": 1.4766, "step": 8383 }, { "epoch": 3.9454117647058826, "grad_norm": 0.9926455609484651, "learning_rate": 0.00010640748706579658, "loss": 0.4238, "step": 8384 }, { "epoch": 3.9458823529411764, "grad_norm": 0.8670788981461929, "learning_rate": 0.00010636554474093846, "loss": 0.6602, "step": 8385 }, { "epoch": 3.9463529411764706, "grad_norm": 1.0657487848654295, "learning_rate": 0.00010632360614267683, "loss": 0.8789, "step": 8386 }, { "epoch": 3.946823529411765, "grad_norm": 0.9823148306468904, "learning_rate": 0.0001062816712745935, "loss": 0.8438, "step": 8387 }, { "epoch": 3.9472941176470586, "grad_norm": 1.6967523588327795, "learning_rate": 0.00010623974014026986, "loss": 1.1484, "step": 8388 }, { "epoch": 3.947764705882353, "grad_norm": 0.7774245436352095, "learning_rate": 0.00010619781274328703, "loss": 0.3359, "step": 8389 }, { "epoch": 3.948235294117647, "grad_norm": 1.2940855484776226, "learning_rate": 0.00010615588908722588, "loss": 1.6875, "step": 8390 }, { "epoch": 3.9487058823529413, "grad_norm": 0.8851937076207546, "learning_rate": 0.00010611396917566685, "loss": 0.5703, "step": 8391 }, { "epoch": 3.9491764705882355, "grad_norm": 1.194026135421268, "learning_rate": 0.00010607205301219008, "loss": 1.6172, "step": 8392 }, { "epoch": 3.9496470588235293, "grad_norm": 0.9898943631298917, "learning_rate": 0.00010603014060037549, "loss": 0.7305, "step": 8393 }, { "epoch": 3.9501176470588235, "grad_norm": 1.0310757557073027, "learning_rate": 0.00010598823194380256, "loss": 0.5898, "step": 8394 }, { "epoch": 3.9505882352941177, "grad_norm": 1.1497130398954378, "learning_rate": 0.0001059463270460505, "loss": 0.6797, "step": 8395 }, { "epoch": 3.9510588235294115, "grad_norm": 0.9504877267376315, "learning_rate": 0.00010590442591069816, "loss": 1.0234, "step": 8396 }, { "epoch": 3.9515294117647057, "grad_norm": 0.998646735949321, "learning_rate": 0.00010586252854132415, "loss": 0.5742, "step": 8397 }, { "epoch": 3.952, "grad_norm": 1.5442925079929055, "learning_rate": 0.0001058206349415067, "loss": 1.2109, "step": 8398 }, { "epoch": 3.952470588235294, "grad_norm": 1.0645566566950804, "learning_rate": 0.00010577874511482371, "loss": 0.5742, "step": 8399 }, { "epoch": 3.9529411764705884, "grad_norm": 1.5818089132031667, "learning_rate": 0.00010573685906485282, "loss": 0.6406, "step": 8400 }, { "epoch": 3.9534117647058826, "grad_norm": 1.1644326814691937, "learning_rate": 0.00010569497679517126, "loss": 0.9844, "step": 8401 }, { "epoch": 3.9538823529411764, "grad_norm": 0.7558358015941385, "learning_rate": 0.00010565309830935599, "loss": 0.6328, "step": 8402 }, { "epoch": 3.9543529411764706, "grad_norm": 0.3961510132081802, "learning_rate": 0.00010561122361098365, "loss": 0.0369, "step": 8403 }, { "epoch": 3.954823529411765, "grad_norm": 0.6078369354068703, "learning_rate": 0.00010556935270363055, "loss": 0.334, "step": 8404 }, { "epoch": 3.9552941176470586, "grad_norm": 0.8009225523571297, "learning_rate": 0.00010552748559087261, "loss": 0.3496, "step": 8405 }, { "epoch": 3.955764705882353, "grad_norm": 1.1397664430870948, "learning_rate": 0.00010548562227628559, "loss": 1.0938, "step": 8406 }, { "epoch": 3.956235294117647, "grad_norm": 0.9374473560367532, "learning_rate": 0.00010544376276344477, "loss": 0.8125, "step": 8407 }, { "epoch": 3.9567058823529413, "grad_norm": 1.305827099921265, "learning_rate": 0.00010540190705592516, "loss": 1.125, "step": 8408 }, { "epoch": 3.9571764705882355, "grad_norm": 1.1201524371105465, "learning_rate": 0.00010536005515730141, "loss": 0.957, "step": 8409 }, { "epoch": 3.9576470588235293, "grad_norm": 1.1792952864319437, "learning_rate": 0.00010531820707114793, "loss": 0.7148, "step": 8410 }, { "epoch": 3.9581176470588235, "grad_norm": 0.6112708411491239, "learning_rate": 0.00010527636280103872, "loss": 0.1865, "step": 8411 }, { "epoch": 3.9585882352941177, "grad_norm": 0.5201923544105376, "learning_rate": 0.00010523452235054746, "loss": 0.1377, "step": 8412 }, { "epoch": 3.9590588235294115, "grad_norm": 1.2035546079998827, "learning_rate": 0.00010519268572324763, "loss": 0.6562, "step": 8413 }, { "epoch": 3.9595294117647057, "grad_norm": 0.9232691673517547, "learning_rate": 0.00010515085292271219, "loss": 0.5859, "step": 8414 }, { "epoch": 3.96, "grad_norm": 1.264417996027459, "learning_rate": 0.00010510902395251387, "loss": 1.5234, "step": 8415 }, { "epoch": 3.960470588235294, "grad_norm": 0.7185805035994941, "learning_rate": 0.00010506719881622514, "loss": 0.3691, "step": 8416 }, { "epoch": 3.9609411764705884, "grad_norm": 1.1369786888345725, "learning_rate": 0.00010502537751741801, "loss": 1.1016, "step": 8417 }, { "epoch": 3.961411764705882, "grad_norm": 0.9485282045374536, "learning_rate": 0.00010498356005966422, "loss": 0.707, "step": 8418 }, { "epoch": 3.9618823529411764, "grad_norm": 0.8883167981952239, "learning_rate": 0.00010494174644653518, "loss": 0.8086, "step": 8419 }, { "epoch": 3.9623529411764706, "grad_norm": 0.9435793669752695, "learning_rate": 0.000104899936681602, "loss": 0.4199, "step": 8420 }, { "epoch": 3.962823529411765, "grad_norm": 1.3671080177344672, "learning_rate": 0.0001048581307684355, "loss": 0.9453, "step": 8421 }, { "epoch": 3.9632941176470586, "grad_norm": 1.1131896938319188, "learning_rate": 0.00010481632871060605, "loss": 1.0547, "step": 8422 }, { "epoch": 3.963764705882353, "grad_norm": 1.1867081646047977, "learning_rate": 0.0001047745305116837, "loss": 0.7812, "step": 8423 }, { "epoch": 3.964235294117647, "grad_norm": 1.4013419428304186, "learning_rate": 0.00010473273617523831, "loss": 1.2109, "step": 8424 }, { "epoch": 3.9647058823529413, "grad_norm": 1.3702026743688187, "learning_rate": 0.00010469094570483928, "loss": 1.5625, "step": 8425 }, { "epoch": 3.9651764705882355, "grad_norm": 0.7713858806221076, "learning_rate": 0.00010464915910405576, "loss": 0.6328, "step": 8426 }, { "epoch": 3.9656470588235293, "grad_norm": 0.7512562671260665, "learning_rate": 0.00010460737637645645, "loss": 0.5352, "step": 8427 }, { "epoch": 3.9661176470588235, "grad_norm": 0.9661412067398757, "learning_rate": 0.00010456559752560987, "loss": 1.1328, "step": 8428 }, { "epoch": 3.9665882352941177, "grad_norm": 0.752423109707709, "learning_rate": 0.00010452382255508412, "loss": 0.7852, "step": 8429 }, { "epoch": 3.9670588235294115, "grad_norm": 1.0769918991876566, "learning_rate": 0.00010448205146844696, "loss": 0.6953, "step": 8430 }, { "epoch": 3.9675294117647057, "grad_norm": 0.7364904725284123, "learning_rate": 0.00010444028426926588, "loss": 0.7773, "step": 8431 }, { "epoch": 3.968, "grad_norm": 1.3396141302176225, "learning_rate": 0.000104398520961108, "loss": 1.0938, "step": 8432 }, { "epoch": 3.968470588235294, "grad_norm": 0.882071522928664, "learning_rate": 0.00010435676154754009, "loss": 0.8008, "step": 8433 }, { "epoch": 3.9689411764705884, "grad_norm": 1.1359443134905947, "learning_rate": 0.00010431500603212863, "loss": 1.0859, "step": 8434 }, { "epoch": 3.969411764705882, "grad_norm": 0.9271996954693497, "learning_rate": 0.00010427325441843976, "loss": 0.6914, "step": 8435 }, { "epoch": 3.9698823529411764, "grad_norm": 1.099712303315322, "learning_rate": 0.00010423150671003924, "loss": 1.5156, "step": 8436 }, { "epoch": 3.9703529411764706, "grad_norm": 0.8159219518686498, "learning_rate": 0.00010418976291049251, "loss": 0.6523, "step": 8437 }, { "epoch": 3.970823529411765, "grad_norm": 1.2460081541566177, "learning_rate": 0.00010414802302336474, "loss": 1.1875, "step": 8438 }, { "epoch": 3.9712941176470586, "grad_norm": 0.9219710749909134, "learning_rate": 0.00010410628705222073, "loss": 1.3125, "step": 8439 }, { "epoch": 3.971764705882353, "grad_norm": 1.0962326801609201, "learning_rate": 0.00010406455500062487, "loss": 1.3047, "step": 8440 }, { "epoch": 3.972235294117647, "grad_norm": 1.2368572568281238, "learning_rate": 0.00010402282687214137, "loss": 0.9922, "step": 8441 }, { "epoch": 3.9727058823529413, "grad_norm": 1.381102326199941, "learning_rate": 0.00010398110267033396, "loss": 1.7422, "step": 8442 }, { "epoch": 3.9731764705882355, "grad_norm": 1.254649334486791, "learning_rate": 0.00010393938239876608, "loss": 1.1797, "step": 8443 }, { "epoch": 3.9736470588235293, "grad_norm": 1.034688391771586, "learning_rate": 0.0001038976660610009, "loss": 0.9375, "step": 8444 }, { "epoch": 3.9741176470588235, "grad_norm": 1.8711417184408323, "learning_rate": 0.00010385595366060116, "loss": 1.2734, "step": 8445 }, { "epoch": 3.9745882352941178, "grad_norm": 0.5270940186685765, "learning_rate": 0.0001038142452011293, "loss": 0.0991, "step": 8446 }, { "epoch": 3.9750588235294115, "grad_norm": 1.4798979698393362, "learning_rate": 0.00010377254068614743, "loss": 0.9766, "step": 8447 }, { "epoch": 3.9755294117647058, "grad_norm": 1.1447652017108554, "learning_rate": 0.00010373084011921736, "loss": 1.0078, "step": 8448 }, { "epoch": 3.976, "grad_norm": 0.9100023407186324, "learning_rate": 0.00010368914350390048, "loss": 0.5234, "step": 8449 }, { "epoch": 3.976470588235294, "grad_norm": 1.0121805060005238, "learning_rate": 0.0001036474508437579, "loss": 1.0469, "step": 8450 }, { "epoch": 3.9769411764705884, "grad_norm": 1.0791265545745472, "learning_rate": 0.00010360576214235036, "loss": 1.0469, "step": 8451 }, { "epoch": 3.977411764705882, "grad_norm": 0.8383040811558629, "learning_rate": 0.00010356407740323832, "loss": 0.3223, "step": 8452 }, { "epoch": 3.9778823529411764, "grad_norm": 1.038335414242489, "learning_rate": 0.00010352239662998182, "loss": 0.668, "step": 8453 }, { "epoch": 3.9783529411764706, "grad_norm": 0.921440893761817, "learning_rate": 0.00010348071982614063, "loss": 0.4414, "step": 8454 }, { "epoch": 3.978823529411765, "grad_norm": 0.9380079630193052, "learning_rate": 0.00010343904699527415, "loss": 0.8438, "step": 8455 }, { "epoch": 3.9792941176470586, "grad_norm": 1.106082121566471, "learning_rate": 0.00010339737814094144, "loss": 0.7539, "step": 8456 }, { "epoch": 3.979764705882353, "grad_norm": 1.2835343497734255, "learning_rate": 0.00010335571326670117, "loss": 1.3047, "step": 8457 }, { "epoch": 3.980235294117647, "grad_norm": 0.884771701984423, "learning_rate": 0.00010331405237611182, "loss": 0.2871, "step": 8458 }, { "epoch": 3.9807058823529413, "grad_norm": 1.2463998176348219, "learning_rate": 0.00010327239547273138, "loss": 0.6406, "step": 8459 }, { "epoch": 3.9811764705882355, "grad_norm": 1.2213355761678235, "learning_rate": 0.00010323074256011756, "loss": 0.9688, "step": 8460 }, { "epoch": 3.9816470588235293, "grad_norm": 1.1520478810207484, "learning_rate": 0.00010318909364182774, "loss": 0.8281, "step": 8461 }, { "epoch": 3.9821176470588235, "grad_norm": 0.9211853183072897, "learning_rate": 0.00010314744872141893, "loss": 0.4434, "step": 8462 }, { "epoch": 3.9825882352941178, "grad_norm": 2.112500758467078, "learning_rate": 0.00010310580780244779, "loss": 0.6562, "step": 8463 }, { "epoch": 3.9830588235294115, "grad_norm": 0.7663144556842905, "learning_rate": 0.00010306417088847067, "loss": 0.375, "step": 8464 }, { "epoch": 3.9835294117647058, "grad_norm": 0.6090800104305127, "learning_rate": 0.00010302253798304356, "loss": 0.4297, "step": 8465 }, { "epoch": 3.984, "grad_norm": 1.0004290151382254, "learning_rate": 0.00010298090908972217, "loss": 0.9258, "step": 8466 }, { "epoch": 3.984470588235294, "grad_norm": 1.1385324971409152, "learning_rate": 0.00010293928421206179, "loss": 0.6445, "step": 8467 }, { "epoch": 3.9849411764705884, "grad_norm": 0.6526357001541259, "learning_rate": 0.00010289766335361733, "loss": 0.2178, "step": 8468 }, { "epoch": 3.985411764705882, "grad_norm": 1.3215837818792975, "learning_rate": 0.00010285604651794347, "loss": 1.5703, "step": 8469 }, { "epoch": 3.9858823529411764, "grad_norm": 0.992959957560681, "learning_rate": 0.00010281443370859449, "loss": 0.7188, "step": 8470 }, { "epoch": 3.9863529411764707, "grad_norm": 1.1019081570087892, "learning_rate": 0.00010277282492912429, "loss": 0.6914, "step": 8471 }, { "epoch": 3.9868235294117644, "grad_norm": 1.267115879549761, "learning_rate": 0.0001027312201830865, "loss": 0.7852, "step": 8472 }, { "epoch": 3.9872941176470587, "grad_norm": 1.1822692152761265, "learning_rate": 0.00010268961947403437, "loss": 0.5781, "step": 8473 }, { "epoch": 3.987764705882353, "grad_norm": 1.2489098496035898, "learning_rate": 0.00010264802280552077, "loss": 1.8125, "step": 8474 }, { "epoch": 3.988235294117647, "grad_norm": 0.7987971205970243, "learning_rate": 0.0001026064301810983, "loss": 0.7188, "step": 8475 }, { "epoch": 3.9887058823529413, "grad_norm": 0.7682185927076213, "learning_rate": 0.00010256484160431917, "loss": 0.252, "step": 8476 }, { "epoch": 3.9891764705882355, "grad_norm": 1.207739867131199, "learning_rate": 0.00010252325707873523, "loss": 0.8164, "step": 8477 }, { "epoch": 3.9896470588235293, "grad_norm": 1.1585112375370914, "learning_rate": 0.00010248167660789797, "loss": 1.3281, "step": 8478 }, { "epoch": 3.9901176470588235, "grad_norm": 1.4561282390814554, "learning_rate": 0.00010244010019535864, "loss": 1.8516, "step": 8479 }, { "epoch": 3.9905882352941178, "grad_norm": 0.8876746264030888, "learning_rate": 0.00010239852784466803, "loss": 0.4883, "step": 8480 }, { "epoch": 3.9910588235294115, "grad_norm": 0.5904429111993047, "learning_rate": 0.00010235695955937661, "loss": 0.2197, "step": 8481 }, { "epoch": 3.9915294117647058, "grad_norm": 0.8452178551188849, "learning_rate": 0.00010231539534303454, "loss": 0.3203, "step": 8482 }, { "epoch": 3.992, "grad_norm": 0.9155640819633006, "learning_rate": 0.00010227383519919161, "loss": 0.9297, "step": 8483 }, { "epoch": 3.992470588235294, "grad_norm": 1.1892504364888787, "learning_rate": 0.00010223227913139722, "loss": 0.7227, "step": 8484 }, { "epoch": 3.9929411764705884, "grad_norm": 0.5298720376076441, "learning_rate": 0.0001021907271432005, "loss": 0.1758, "step": 8485 }, { "epoch": 3.993411764705882, "grad_norm": 0.8752658887205435, "learning_rate": 0.00010214917923815022, "loss": 1.0469, "step": 8486 }, { "epoch": 3.9938823529411764, "grad_norm": 0.43719029807306303, "learning_rate": 0.00010210763541979473, "loss": 0.1226, "step": 8487 }, { "epoch": 3.9943529411764707, "grad_norm": 1.4132194940954521, "learning_rate": 0.00010206609569168204, "loss": 1.1719, "step": 8488 }, { "epoch": 3.9948235294117644, "grad_norm": 0.7529844474402322, "learning_rate": 0.00010202456005735993, "loss": 0.3184, "step": 8489 }, { "epoch": 3.9952941176470587, "grad_norm": 0.9954363720205359, "learning_rate": 0.00010198302852037573, "loss": 0.5195, "step": 8490 }, { "epoch": 3.995764705882353, "grad_norm": 0.8074062059168725, "learning_rate": 0.00010194150108427637, "loss": 0.3848, "step": 8491 }, { "epoch": 3.996235294117647, "grad_norm": 1.111781185539406, "learning_rate": 0.00010189997775260857, "loss": 0.625, "step": 8492 }, { "epoch": 3.9967058823529413, "grad_norm": 1.1023094345750566, "learning_rate": 0.00010185845852891858, "loss": 1.0625, "step": 8493 }, { "epoch": 3.9971764705882356, "grad_norm": 1.539668127325887, "learning_rate": 0.00010181694341675235, "loss": 1.7109, "step": 8494 }, { "epoch": 3.9976470588235293, "grad_norm": 0.722363891383574, "learning_rate": 0.00010177543241965555, "loss": 0.2617, "step": 8495 }, { "epoch": 3.9981176470588236, "grad_norm": 0.8408603242263094, "learning_rate": 0.00010173392554117333, "loss": 0.5156, "step": 8496 }, { "epoch": 3.9985882352941178, "grad_norm": 0.9002255538985372, "learning_rate": 0.00010169242278485063, "loss": 0.3613, "step": 8497 }, { "epoch": 3.9990588235294116, "grad_norm": 1.171785006006295, "learning_rate": 0.00010165092415423193, "loss": 1.2734, "step": 8498 }, { "epoch": 3.9995294117647058, "grad_norm": 0.7990995158615389, "learning_rate": 0.00010160942965286148, "loss": 0.5703, "step": 8499 }, { "epoch": 4.0, "grad_norm": 1.2743094519293623, "learning_rate": 0.0001015679392842831, "loss": 1.0938, "step": 8500 }, { "epoch": 4.0, "eval_loss": 0.8190028667449951, "eval_runtime": 36.557, "eval_samples_per_second": 20.516, "eval_steps_per_second": 10.258, "step": 8500 }, { "epoch": 4.000470588235294, "grad_norm": 0.9707643636185237, "learning_rate": 0.00010152645305204022, "loss": 0.7578, "step": 8501 }, { "epoch": 4.0009411764705884, "grad_norm": 1.027863535419913, "learning_rate": 0.00010148497095967606, "loss": 0.6836, "step": 8502 }, { "epoch": 4.001411764705883, "grad_norm": 0.8563101952770792, "learning_rate": 0.00010144349301073334, "loss": 1.125, "step": 8503 }, { "epoch": 4.001882352941177, "grad_norm": 0.9055168323272077, "learning_rate": 0.00010140201920875445, "loss": 1.125, "step": 8504 }, { "epoch": 4.00235294117647, "grad_norm": 1.1473057919097787, "learning_rate": 0.00010136054955728154, "loss": 0.5586, "step": 8505 }, { "epoch": 4.0028235294117644, "grad_norm": 0.8061649587014806, "learning_rate": 0.00010131908405985625, "loss": 0.373, "step": 8506 }, { "epoch": 4.003294117647059, "grad_norm": 0.7315521065198163, "learning_rate": 0.00010127762272001996, "loss": 0.4746, "step": 8507 }, { "epoch": 4.003764705882353, "grad_norm": 2.164495568981309, "learning_rate": 0.00010123616554131363, "loss": 0.6914, "step": 8508 }, { "epoch": 4.004235294117647, "grad_norm": 1.0825565051856065, "learning_rate": 0.00010119471252727796, "loss": 1.2422, "step": 8509 }, { "epoch": 4.004705882352941, "grad_norm": 1.1291437179738364, "learning_rate": 0.00010115326368145327, "loss": 0.7852, "step": 8510 }, { "epoch": 4.005176470588236, "grad_norm": 0.541572609118511, "learning_rate": 0.00010111181900737945, "loss": 0.1504, "step": 8511 }, { "epoch": 4.00564705882353, "grad_norm": 1.3435716689872403, "learning_rate": 0.00010107037850859605, "loss": 1.3984, "step": 8512 }, { "epoch": 4.006117647058823, "grad_norm": 1.2856981209049612, "learning_rate": 0.00010102894218864235, "loss": 0.9688, "step": 8513 }, { "epoch": 4.006588235294117, "grad_norm": 0.9760334981529575, "learning_rate": 0.00010098751005105719, "loss": 1.1719, "step": 8514 }, { "epoch": 4.007058823529412, "grad_norm": 0.5907505528865977, "learning_rate": 0.0001009460820993791, "loss": 0.4375, "step": 8515 }, { "epoch": 4.007529411764706, "grad_norm": 1.2635128792063053, "learning_rate": 0.00010090465833714616, "loss": 0.9258, "step": 8516 }, { "epoch": 4.008, "grad_norm": 0.8254940251912378, "learning_rate": 0.00010086323876789621, "loss": 0.582, "step": 8517 }, { "epoch": 4.008470588235294, "grad_norm": 0.800897513995333, "learning_rate": 0.00010082182339516674, "loss": 0.4746, "step": 8518 }, { "epoch": 4.0089411764705885, "grad_norm": 1.5743273551032082, "learning_rate": 0.00010078041222249469, "loss": 1.6719, "step": 8519 }, { "epoch": 4.009411764705883, "grad_norm": 1.0571458914571112, "learning_rate": 0.00010073900525341692, "loss": 0.4375, "step": 8520 }, { "epoch": 4.009882352941177, "grad_norm": 0.7631282955440518, "learning_rate": 0.00010069760249146972, "loss": 0.416, "step": 8521 }, { "epoch": 4.01035294117647, "grad_norm": 0.9511868683385002, "learning_rate": 0.00010065620394018904, "loss": 0.7812, "step": 8522 }, { "epoch": 4.0108235294117645, "grad_norm": 1.0194007595741779, "learning_rate": 0.00010061480960311061, "loss": 0.5625, "step": 8523 }, { "epoch": 4.011294117647059, "grad_norm": 1.7664388631740753, "learning_rate": 0.0001005734194837697, "loss": 1.4219, "step": 8524 }, { "epoch": 4.011764705882353, "grad_norm": 0.9937895043951149, "learning_rate": 0.00010053203358570115, "loss": 0.5938, "step": 8525 }, { "epoch": 4.012235294117647, "grad_norm": 1.1631192103736008, "learning_rate": 0.00010049065191243961, "loss": 0.8047, "step": 8526 }, { "epoch": 4.012705882352941, "grad_norm": 0.7160687268750116, "learning_rate": 0.00010044927446751923, "loss": 0.6797, "step": 8527 }, { "epoch": 4.013176470588236, "grad_norm": 0.9722866896665063, "learning_rate": 0.00010040790125447386, "loss": 1.0391, "step": 8528 }, { "epoch": 4.01364705882353, "grad_norm": 0.5666310769386359, "learning_rate": 0.00010036653227683694, "loss": 0.1445, "step": 8529 }, { "epoch": 4.014117647058823, "grad_norm": 0.8836369130695056, "learning_rate": 0.00010032516753814166, "loss": 0.4688, "step": 8530 }, { "epoch": 4.014588235294117, "grad_norm": 1.1486408979121174, "learning_rate": 0.0001002838070419207, "loss": 1.0469, "step": 8531 }, { "epoch": 4.015058823529412, "grad_norm": 0.9561027493477025, "learning_rate": 0.00010024245079170645, "loss": 0.2715, "step": 8532 }, { "epoch": 4.015529411764706, "grad_norm": 0.9634308811633906, "learning_rate": 0.000100201098791031, "loss": 1.1328, "step": 8533 }, { "epoch": 4.016, "grad_norm": 0.572839183267472, "learning_rate": 0.00010015975104342598, "loss": 0.2168, "step": 8534 }, { "epoch": 4.016470588235294, "grad_norm": 0.8047056028534958, "learning_rate": 0.00010011840755242266, "loss": 0.5117, "step": 8535 }, { "epoch": 4.0169411764705885, "grad_norm": 0.8481781960210415, "learning_rate": 0.00010007706832155201, "loss": 0.2461, "step": 8536 }, { "epoch": 4.017411764705883, "grad_norm": 0.9143809248658348, "learning_rate": 0.00010003573335434462, "loss": 0.3926, "step": 8537 }, { "epoch": 4.017882352941177, "grad_norm": 0.6996100688266509, "learning_rate": 9.999440265433065e-05, "loss": 0.2988, "step": 8538 }, { "epoch": 4.01835294117647, "grad_norm": 0.8905886377637823, "learning_rate": 9.995307622503993e-05, "loss": 0.3984, "step": 8539 }, { "epoch": 4.0188235294117645, "grad_norm": 1.096480794485517, "learning_rate": 9.991175407000206e-05, "loss": 1.0312, "step": 8540 }, { "epoch": 4.019294117647059, "grad_norm": 0.8390727461659546, "learning_rate": 9.987043619274604e-05, "loss": 0.4258, "step": 8541 }, { "epoch": 4.019764705882353, "grad_norm": 1.1383443751992472, "learning_rate": 9.982912259680063e-05, "loss": 0.6289, "step": 8542 }, { "epoch": 4.020235294117647, "grad_norm": 1.6072462779376187, "learning_rate": 9.978781328569427e-05, "loss": 0.9453, "step": 8543 }, { "epoch": 4.020705882352941, "grad_norm": 1.3204900604944636, "learning_rate": 9.974650826295493e-05, "loss": 1.4453, "step": 8544 }, { "epoch": 4.021176470588236, "grad_norm": 1.106856744580465, "learning_rate": 9.970520753211027e-05, "loss": 1.0, "step": 8545 }, { "epoch": 4.02164705882353, "grad_norm": 1.1946994054288689, "learning_rate": 9.966391109668758e-05, "loss": 0.6719, "step": 8546 }, { "epoch": 4.022117647058823, "grad_norm": 1.0558182596933514, "learning_rate": 9.96226189602138e-05, "loss": 0.7305, "step": 8547 }, { "epoch": 4.022588235294117, "grad_norm": 1.4116670755069305, "learning_rate": 9.958133112621547e-05, "loss": 1.0703, "step": 8548 }, { "epoch": 4.023058823529412, "grad_norm": 1.3731205277137981, "learning_rate": 9.954004759821871e-05, "loss": 1.1484, "step": 8549 }, { "epoch": 4.023529411764706, "grad_norm": 0.6230860217694483, "learning_rate": 9.949876837974944e-05, "loss": 0.3672, "step": 8550 }, { "epoch": 4.024, "grad_norm": 1.2763712285201212, "learning_rate": 9.945749347433304e-05, "loss": 1.5156, "step": 8551 }, { "epoch": 4.024470588235294, "grad_norm": 1.1099680946288075, "learning_rate": 9.941622288549458e-05, "loss": 0.5781, "step": 8552 }, { "epoch": 4.0249411764705885, "grad_norm": 1.0415479131530851, "learning_rate": 9.937495661675878e-05, "loss": 0.4375, "step": 8553 }, { "epoch": 4.025411764705883, "grad_norm": 1.200503627107371, "learning_rate": 9.933369467165004e-05, "loss": 1.0156, "step": 8554 }, { "epoch": 4.025882352941177, "grad_norm": 1.084102882405791, "learning_rate": 9.929243705369232e-05, "loss": 0.582, "step": 8555 }, { "epoch": 4.02635294117647, "grad_norm": 1.02779634466465, "learning_rate": 9.925118376640918e-05, "loss": 0.5508, "step": 8556 }, { "epoch": 4.0268235294117645, "grad_norm": 1.0256323839860524, "learning_rate": 9.920993481332383e-05, "loss": 0.8242, "step": 8557 }, { "epoch": 4.027294117647059, "grad_norm": 1.000327802001832, "learning_rate": 9.916869019795922e-05, "loss": 0.9609, "step": 8558 }, { "epoch": 4.027764705882353, "grad_norm": 1.017160199928868, "learning_rate": 9.912744992383777e-05, "loss": 0.6758, "step": 8559 }, { "epoch": 4.028235294117647, "grad_norm": 0.9783095157285695, "learning_rate": 9.90862139944816e-05, "loss": 0.5664, "step": 8560 }, { "epoch": 4.028705882352941, "grad_norm": 1.1381580623763738, "learning_rate": 9.904498241341252e-05, "loss": 1.0, "step": 8561 }, { "epoch": 4.029176470588236, "grad_norm": 0.9665647836752482, "learning_rate": 9.900375518415188e-05, "loss": 0.5977, "step": 8562 }, { "epoch": 4.02964705882353, "grad_norm": 0.5286595276616338, "learning_rate": 9.896253231022066e-05, "loss": 0.3867, "step": 8563 }, { "epoch": 4.030117647058823, "grad_norm": 1.2673038639915621, "learning_rate": 9.892131379513953e-05, "loss": 0.9688, "step": 8564 }, { "epoch": 4.030588235294117, "grad_norm": 0.7760414833868292, "learning_rate": 9.888009964242876e-05, "loss": 0.2236, "step": 8565 }, { "epoch": 4.031058823529412, "grad_norm": 0.9870358751627074, "learning_rate": 9.883888985560822e-05, "loss": 0.6172, "step": 8566 }, { "epoch": 4.031529411764706, "grad_norm": 1.3183891349626868, "learning_rate": 9.879768443819737e-05, "loss": 1.2656, "step": 8567 }, { "epoch": 4.032, "grad_norm": 1.1224230092402425, "learning_rate": 9.875648339371546e-05, "loss": 1.1797, "step": 8568 }, { "epoch": 4.032470588235294, "grad_norm": 0.855477707781643, "learning_rate": 9.871528672568124e-05, "loss": 1.0, "step": 8569 }, { "epoch": 4.0329411764705885, "grad_norm": 0.9956579638029138, "learning_rate": 9.867409443761304e-05, "loss": 0.6562, "step": 8570 }, { "epoch": 4.033411764705883, "grad_norm": 1.0935633160756648, "learning_rate": 9.863290653302894e-05, "loss": 0.9219, "step": 8571 }, { "epoch": 4.033882352941176, "grad_norm": 0.6257003374882478, "learning_rate": 9.859172301544659e-05, "loss": 0.5078, "step": 8572 }, { "epoch": 4.03435294117647, "grad_norm": 0.9972418254040576, "learning_rate": 9.855054388838322e-05, "loss": 0.8008, "step": 8573 }, { "epoch": 4.0348235294117645, "grad_norm": 1.37521045739242, "learning_rate": 9.85093691553558e-05, "loss": 1.3516, "step": 8574 }, { "epoch": 4.035294117647059, "grad_norm": 0.6219602360448161, "learning_rate": 9.84681988198808e-05, "loss": 0.1924, "step": 8575 }, { "epoch": 4.035764705882353, "grad_norm": 0.6581350279654795, "learning_rate": 9.842703288547438e-05, "loss": 0.1504, "step": 8576 }, { "epoch": 4.036235294117647, "grad_norm": 0.726569977808688, "learning_rate": 9.838587135565229e-05, "loss": 0.2197, "step": 8577 }, { "epoch": 4.036705882352941, "grad_norm": 1.0756690875099553, "learning_rate": 9.834471423393e-05, "loss": 0.7773, "step": 8578 }, { "epoch": 4.037176470588236, "grad_norm": 1.0227161269463252, "learning_rate": 9.830356152382245e-05, "loss": 0.9336, "step": 8579 }, { "epoch": 4.03764705882353, "grad_norm": 0.8117552373829624, "learning_rate": 9.826241322884432e-05, "loss": 0.2148, "step": 8580 }, { "epoch": 4.038117647058823, "grad_norm": 1.225394590776594, "learning_rate": 9.822126935250987e-05, "loss": 1.1719, "step": 8581 }, { "epoch": 4.038588235294117, "grad_norm": 0.8820385338341898, "learning_rate": 9.818012989833301e-05, "loss": 0.3945, "step": 8582 }, { "epoch": 4.039058823529412, "grad_norm": 1.0866493189037087, "learning_rate": 9.81389948698272e-05, "loss": 0.6562, "step": 8583 }, { "epoch": 4.039529411764706, "grad_norm": 0.8438424553188237, "learning_rate": 9.809786427050562e-05, "loss": 0.2969, "step": 8584 }, { "epoch": 4.04, "grad_norm": 0.9804744958254327, "learning_rate": 9.805673810388103e-05, "loss": 0.8203, "step": 8585 }, { "epoch": 4.040470588235294, "grad_norm": 1.2773699634482854, "learning_rate": 9.801561637346575e-05, "loss": 0.5977, "step": 8586 }, { "epoch": 4.0409411764705885, "grad_norm": 0.5914569230017472, "learning_rate": 9.797449908277184e-05, "loss": 0.166, "step": 8587 }, { "epoch": 4.041411764705883, "grad_norm": 0.690394322949573, "learning_rate": 9.793338623531089e-05, "loss": 0.7734, "step": 8588 }, { "epoch": 4.041882352941176, "grad_norm": 0.6314215775114004, "learning_rate": 9.789227783459415e-05, "loss": 0.2061, "step": 8589 }, { "epoch": 4.04235294117647, "grad_norm": 1.1438236826332824, "learning_rate": 9.785117388413244e-05, "loss": 0.7852, "step": 8590 }, { "epoch": 4.0428235294117645, "grad_norm": 1.223396692958488, "learning_rate": 9.781007438743629e-05, "loss": 0.957, "step": 8591 }, { "epoch": 4.043294117647059, "grad_norm": 0.7154515278287666, "learning_rate": 9.776897934801577e-05, "loss": 0.1562, "step": 8592 }, { "epoch": 4.043764705882353, "grad_norm": 0.9737966361199885, "learning_rate": 9.77278887693806e-05, "loss": 0.7305, "step": 8593 }, { "epoch": 4.044235294117647, "grad_norm": 1.337890122093729, "learning_rate": 9.768680265504013e-05, "loss": 1.5938, "step": 8594 }, { "epoch": 4.044705882352941, "grad_norm": 0.7173067917173066, "learning_rate": 9.764572100850331e-05, "loss": 0.5547, "step": 8595 }, { "epoch": 4.045176470588236, "grad_norm": 1.040898582731196, "learning_rate": 9.760464383327868e-05, "loss": 0.7773, "step": 8596 }, { "epoch": 4.04564705882353, "grad_norm": 0.9680390176903164, "learning_rate": 9.756357113287449e-05, "loss": 0.4453, "step": 8597 }, { "epoch": 4.046117647058823, "grad_norm": 0.831595888522906, "learning_rate": 9.75225029107985e-05, "loss": 0.3613, "step": 8598 }, { "epoch": 4.046588235294117, "grad_norm": 0.9095123062162804, "learning_rate": 9.74814391705582e-05, "loss": 1.125, "step": 8599 }, { "epoch": 4.047058823529412, "grad_norm": 1.3933225907919535, "learning_rate": 9.744037991566058e-05, "loss": 1.2422, "step": 8600 }, { "epoch": 4.047529411764706, "grad_norm": 1.0817569607640447, "learning_rate": 9.739932514961232e-05, "loss": 0.5781, "step": 8601 }, { "epoch": 4.048, "grad_norm": 1.026211480527458, "learning_rate": 9.73582748759197e-05, "loss": 1.1172, "step": 8602 }, { "epoch": 4.048470588235294, "grad_norm": 0.9477886397030529, "learning_rate": 9.731722909808863e-05, "loss": 0.418, "step": 8603 }, { "epoch": 4.0489411764705885, "grad_norm": 1.1901404174761938, "learning_rate": 9.727618781962457e-05, "loss": 0.9297, "step": 8604 }, { "epoch": 4.049411764705883, "grad_norm": 0.9314286572784906, "learning_rate": 9.723515104403268e-05, "loss": 0.5781, "step": 8605 }, { "epoch": 4.049882352941176, "grad_norm": 0.45374567777791175, "learning_rate": 9.719411877481776e-05, "loss": 0.3613, "step": 8606 }, { "epoch": 4.05035294117647, "grad_norm": 0.7876548055509628, "learning_rate": 9.715309101548406e-05, "loss": 0.3184, "step": 8607 }, { "epoch": 4.0508235294117645, "grad_norm": 0.9530853621928729, "learning_rate": 9.711206776953559e-05, "loss": 0.6172, "step": 8608 }, { "epoch": 4.051294117647059, "grad_norm": 0.9956946542968999, "learning_rate": 9.707104904047596e-05, "loss": 0.6172, "step": 8609 }, { "epoch": 4.051764705882353, "grad_norm": 0.7979544874689375, "learning_rate": 9.703003483180836e-05, "loss": 0.3691, "step": 8610 }, { "epoch": 4.052235294117647, "grad_norm": 0.8133327936925223, "learning_rate": 9.698902514703558e-05, "loss": 0.4844, "step": 8611 }, { "epoch": 4.052705882352941, "grad_norm": 0.8255907327096531, "learning_rate": 9.694801998966011e-05, "loss": 0.7773, "step": 8612 }, { "epoch": 4.053176470588236, "grad_norm": 1.0784425105197883, "learning_rate": 9.690701936318395e-05, "loss": 0.7891, "step": 8613 }, { "epoch": 4.05364705882353, "grad_norm": 0.7391735056573413, "learning_rate": 9.686602327110869e-05, "loss": 0.4004, "step": 8614 }, { "epoch": 4.054117647058823, "grad_norm": 0.7240481764105062, "learning_rate": 9.682503171693573e-05, "loss": 0.5039, "step": 8615 }, { "epoch": 4.054588235294117, "grad_norm": 1.0086757748930344, "learning_rate": 9.678404470416587e-05, "loss": 0.9727, "step": 8616 }, { "epoch": 4.055058823529412, "grad_norm": 1.1101798603810624, "learning_rate": 9.674306223629963e-05, "loss": 0.75, "step": 8617 }, { "epoch": 4.055529411764706, "grad_norm": 1.0494396043983827, "learning_rate": 9.670208431683704e-05, "loss": 0.9375, "step": 8618 }, { "epoch": 4.056, "grad_norm": 1.281347281462621, "learning_rate": 9.666111094927792e-05, "loss": 0.875, "step": 8619 }, { "epoch": 4.056470588235294, "grad_norm": 1.1730163983302837, "learning_rate": 9.662014213712155e-05, "loss": 1.0938, "step": 8620 }, { "epoch": 4.0569411764705885, "grad_norm": 1.0388905391108145, "learning_rate": 9.657917788386684e-05, "loss": 0.8047, "step": 8621 }, { "epoch": 4.057411764705883, "grad_norm": 1.0612894186527693, "learning_rate": 9.65382181930124e-05, "loss": 1.1875, "step": 8622 }, { "epoch": 4.057882352941176, "grad_norm": 0.9409230002617293, "learning_rate": 9.649726306805635e-05, "loss": 0.418, "step": 8623 }, { "epoch": 4.05835294117647, "grad_norm": 1.0826125153794464, "learning_rate": 9.645631251249643e-05, "loss": 1.1484, "step": 8624 }, { "epoch": 4.0588235294117645, "grad_norm": 1.1249841884849026, "learning_rate": 9.641536652983008e-05, "loss": 1.4297, "step": 8625 }, { "epoch": 4.059294117647059, "grad_norm": 0.9895890262081282, "learning_rate": 9.637442512355428e-05, "loss": 0.8945, "step": 8626 }, { "epoch": 4.059764705882353, "grad_norm": 0.9593031160200486, "learning_rate": 9.63334882971656e-05, "loss": 0.707, "step": 8627 }, { "epoch": 4.060235294117647, "grad_norm": 0.626197254987629, "learning_rate": 9.629255605416021e-05, "loss": 0.2559, "step": 8628 }, { "epoch": 4.060705882352941, "grad_norm": 0.7453355227144316, "learning_rate": 9.6251628398034e-05, "loss": 0.2119, "step": 8629 }, { "epoch": 4.061176470588236, "grad_norm": 0.8793353479029112, "learning_rate": 9.621070533228236e-05, "loss": 0.3809, "step": 8630 }, { "epoch": 4.06164705882353, "grad_norm": 1.2959233193704336, "learning_rate": 9.616978686040028e-05, "loss": 1.1172, "step": 8631 }, { "epoch": 4.062117647058823, "grad_norm": 1.0912531286727791, "learning_rate": 9.612887298588246e-05, "loss": 1.1094, "step": 8632 }, { "epoch": 4.062588235294117, "grad_norm": 0.5262381934592075, "learning_rate": 9.608796371222316e-05, "loss": 0.1113, "step": 8633 }, { "epoch": 4.063058823529412, "grad_norm": 1.3946697376054458, "learning_rate": 9.604705904291614e-05, "loss": 0.8789, "step": 8634 }, { "epoch": 4.063529411764706, "grad_norm": 1.1092425619545532, "learning_rate": 9.600615898145496e-05, "loss": 0.9336, "step": 8635 }, { "epoch": 4.064, "grad_norm": 1.5637025910335791, "learning_rate": 9.596526353133261e-05, "loss": 1.0547, "step": 8636 }, { "epoch": 4.064470588235294, "grad_norm": 1.0165257331567314, "learning_rate": 9.59243726960418e-05, "loss": 0.9102, "step": 8637 }, { "epoch": 4.0649411764705885, "grad_norm": 0.98645632828003, "learning_rate": 9.588348647907481e-05, "loss": 1.2109, "step": 8638 }, { "epoch": 4.065411764705883, "grad_norm": 1.023341478741656, "learning_rate": 9.584260488392353e-05, "loss": 0.7383, "step": 8639 }, { "epoch": 4.065882352941176, "grad_norm": 0.48655381808836057, "learning_rate": 9.580172791407942e-05, "loss": 0.1099, "step": 8640 }, { "epoch": 4.06635294117647, "grad_norm": 1.295177649585133, "learning_rate": 9.576085557303355e-05, "loss": 1.3672, "step": 8641 }, { "epoch": 4.0668235294117645, "grad_norm": 1.396294078110444, "learning_rate": 9.571998786427665e-05, "loss": 1.2734, "step": 8642 }, { "epoch": 4.067294117647059, "grad_norm": 0.37193790291928963, "learning_rate": 9.567912479129907e-05, "loss": 0.0469, "step": 8643 }, { "epoch": 4.067764705882353, "grad_norm": 1.1230649277789615, "learning_rate": 9.563826635759068e-05, "loss": 1.1719, "step": 8644 }, { "epoch": 4.068235294117647, "grad_norm": 0.7931737388274915, "learning_rate": 9.559741256664098e-05, "loss": 0.2891, "step": 8645 }, { "epoch": 4.068705882352941, "grad_norm": 0.7661307135547251, "learning_rate": 9.555656342193907e-05, "loss": 0.2598, "step": 8646 }, { "epoch": 4.069176470588236, "grad_norm": 1.0950038655413652, "learning_rate": 9.551571892697372e-05, "loss": 1.2188, "step": 8647 }, { "epoch": 4.06964705882353, "grad_norm": 1.0053836915903709, "learning_rate": 9.547487908523322e-05, "loss": 0.8828, "step": 8648 }, { "epoch": 4.070117647058823, "grad_norm": 1.0321837579637423, "learning_rate": 9.543404390020543e-05, "loss": 1.1484, "step": 8649 }, { "epoch": 4.070588235294117, "grad_norm": 0.8090891465856426, "learning_rate": 9.5393213375378e-05, "loss": 0.2949, "step": 8650 }, { "epoch": 4.071058823529412, "grad_norm": 0.5859282817923601, "learning_rate": 9.535238751423799e-05, "loss": 0.1465, "step": 8651 }, { "epoch": 4.071529411764706, "grad_norm": 1.037443383943411, "learning_rate": 9.531156632027209e-05, "loss": 1.0469, "step": 8652 }, { "epoch": 4.072, "grad_norm": 1.0465209452440467, "learning_rate": 9.52707497969667e-05, "loss": 0.8984, "step": 8653 }, { "epoch": 4.072470588235294, "grad_norm": 0.929857419621017, "learning_rate": 9.522993794780773e-05, "loss": 0.8008, "step": 8654 }, { "epoch": 4.0729411764705885, "grad_norm": 0.8791286842864334, "learning_rate": 9.518913077628066e-05, "loss": 0.291, "step": 8655 }, { "epoch": 4.073411764705883, "grad_norm": 1.3699975913365474, "learning_rate": 9.514832828587071e-05, "loss": 1.4219, "step": 8656 }, { "epoch": 4.073882352941176, "grad_norm": 1.149687636516442, "learning_rate": 9.510753048006257e-05, "loss": 0.7891, "step": 8657 }, { "epoch": 4.07435294117647, "grad_norm": 1.1018713350767528, "learning_rate": 9.506673736234057e-05, "loss": 0.5977, "step": 8658 }, { "epoch": 4.0748235294117645, "grad_norm": 1.314612353398298, "learning_rate": 9.502594893618861e-05, "loss": 0.7227, "step": 8659 }, { "epoch": 4.075294117647059, "grad_norm": 1.5312886112840838, "learning_rate": 9.49851652050903e-05, "loss": 2.0938, "step": 8660 }, { "epoch": 4.075764705882353, "grad_norm": 1.112297327213811, "learning_rate": 9.494438617252875e-05, "loss": 0.875, "step": 8661 }, { "epoch": 4.076235294117647, "grad_norm": 0.7708730072171439, "learning_rate": 9.490361184198659e-05, "loss": 0.5078, "step": 8662 }, { "epoch": 4.076705882352941, "grad_norm": 1.2551565931856918, "learning_rate": 9.486284221694626e-05, "loss": 1.6719, "step": 8663 }, { "epoch": 4.077176470588236, "grad_norm": 1.076341463183826, "learning_rate": 9.482207730088968e-05, "loss": 0.5117, "step": 8664 }, { "epoch": 4.07764705882353, "grad_norm": 1.2117815728934904, "learning_rate": 9.47813170972983e-05, "loss": 0.9336, "step": 8665 }, { "epoch": 4.078117647058823, "grad_norm": 1.8301700510256227, "learning_rate": 9.474056160965332e-05, "loss": 0.9062, "step": 8666 }, { "epoch": 4.078588235294117, "grad_norm": 1.1069730169920013, "learning_rate": 9.469981084143541e-05, "loss": 1.0156, "step": 8667 }, { "epoch": 4.079058823529412, "grad_norm": 0.6328452793746026, "learning_rate": 9.465906479612492e-05, "loss": 0.1875, "step": 8668 }, { "epoch": 4.079529411764706, "grad_norm": 1.352494646266644, "learning_rate": 9.461832347720168e-05, "loss": 1.1641, "step": 8669 }, { "epoch": 4.08, "grad_norm": 1.4966214059299126, "learning_rate": 9.457758688814533e-05, "loss": 0.9922, "step": 8670 }, { "epoch": 4.080470588235294, "grad_norm": 0.8034481352345392, "learning_rate": 9.453685503243489e-05, "loss": 0.7148, "step": 8671 }, { "epoch": 4.0809411764705885, "grad_norm": 1.0436627359027588, "learning_rate": 9.449612791354904e-05, "loss": 0.9258, "step": 8672 }, { "epoch": 4.081411764705883, "grad_norm": 1.2243498263904715, "learning_rate": 9.445540553496614e-05, "loss": 0.8477, "step": 8673 }, { "epoch": 4.081882352941176, "grad_norm": 0.4110547579700255, "learning_rate": 9.441468790016406e-05, "loss": 0.0791, "step": 8674 }, { "epoch": 4.08235294117647, "grad_norm": 0.9532750194068493, "learning_rate": 9.437397501262026e-05, "loss": 1.0078, "step": 8675 }, { "epoch": 4.0828235294117645, "grad_norm": 1.3930616614932803, "learning_rate": 9.433326687581183e-05, "loss": 1.5781, "step": 8676 }, { "epoch": 4.083294117647059, "grad_norm": 0.6993253330216654, "learning_rate": 9.429256349321548e-05, "loss": 0.2324, "step": 8677 }, { "epoch": 4.083764705882353, "grad_norm": 0.6675308339528038, "learning_rate": 9.425186486830745e-05, "loss": 0.1787, "step": 8678 }, { "epoch": 4.084235294117647, "grad_norm": 1.2096388152373023, "learning_rate": 9.421117100456356e-05, "loss": 1.2578, "step": 8679 }, { "epoch": 4.084705882352941, "grad_norm": 0.9676875957259847, "learning_rate": 9.417048190545935e-05, "loss": 0.9609, "step": 8680 }, { "epoch": 4.085176470588236, "grad_norm": 1.0292229018229613, "learning_rate": 9.412979757446983e-05, "loss": 0.4883, "step": 8681 }, { "epoch": 4.085647058823529, "grad_norm": 1.1052963926329205, "learning_rate": 9.40891180150696e-05, "loss": 0.6406, "step": 8682 }, { "epoch": 4.086117647058823, "grad_norm": 1.057356694169745, "learning_rate": 9.404844323073296e-05, "loss": 0.5781, "step": 8683 }, { "epoch": 4.086588235294117, "grad_norm": 1.0681175108727963, "learning_rate": 9.400777322493373e-05, "loss": 0.9609, "step": 8684 }, { "epoch": 4.087058823529412, "grad_norm": 1.2198536170666816, "learning_rate": 9.396710800114526e-05, "loss": 0.7969, "step": 8685 }, { "epoch": 4.087529411764706, "grad_norm": 0.7873935019293155, "learning_rate": 9.39264475628406e-05, "loss": 0.6445, "step": 8686 }, { "epoch": 4.088, "grad_norm": 0.9060184019604564, "learning_rate": 9.388579191349237e-05, "loss": 0.7305, "step": 8687 }, { "epoch": 4.088470588235294, "grad_norm": 1.4090508510496569, "learning_rate": 9.384514105657277e-05, "loss": 1.75, "step": 8688 }, { "epoch": 4.0889411764705885, "grad_norm": 1.3588255428382585, "learning_rate": 9.380449499555358e-05, "loss": 0.75, "step": 8689 }, { "epoch": 4.089411764705883, "grad_norm": 0.9992387502259398, "learning_rate": 9.376385373390611e-05, "loss": 0.7422, "step": 8690 }, { "epoch": 4.089882352941176, "grad_norm": 0.9947074388872652, "learning_rate": 9.372321727510139e-05, "loss": 0.6523, "step": 8691 }, { "epoch": 4.09035294117647, "grad_norm": 0.5520873618359085, "learning_rate": 9.368258562260999e-05, "loss": 0.3301, "step": 8692 }, { "epoch": 4.0908235294117645, "grad_norm": 1.002498536699651, "learning_rate": 9.364195877990195e-05, "loss": 1.0938, "step": 8693 }, { "epoch": 4.091294117647059, "grad_norm": 0.44936036745750224, "learning_rate": 9.360133675044708e-05, "loss": 0.1445, "step": 8694 }, { "epoch": 4.091764705882353, "grad_norm": 1.0106053875858065, "learning_rate": 9.356071953771472e-05, "loss": 1.0234, "step": 8695 }, { "epoch": 4.092235294117647, "grad_norm": 0.6375514862056083, "learning_rate": 9.352010714517371e-05, "loss": 0.1611, "step": 8696 }, { "epoch": 4.092705882352941, "grad_norm": 1.012958191366354, "learning_rate": 9.347949957629258e-05, "loss": 0.5234, "step": 8697 }, { "epoch": 4.093176470588236, "grad_norm": 0.8380785616364474, "learning_rate": 9.343889683453945e-05, "loss": 0.334, "step": 8698 }, { "epoch": 4.09364705882353, "grad_norm": 1.2617400521990314, "learning_rate": 9.339829892338193e-05, "loss": 0.8945, "step": 8699 }, { "epoch": 4.094117647058823, "grad_norm": 0.7363224914510733, "learning_rate": 9.33577058462873e-05, "loss": 0.3125, "step": 8700 }, { "epoch": 4.094588235294117, "grad_norm": 0.7803445695233454, "learning_rate": 9.331711760672246e-05, "loss": 0.2812, "step": 8701 }, { "epoch": 4.095058823529412, "grad_norm": 1.2290301531553536, "learning_rate": 9.32765342081538e-05, "loss": 0.8477, "step": 8702 }, { "epoch": 4.095529411764706, "grad_norm": 1.0664257694617583, "learning_rate": 9.323595565404731e-05, "loss": 1.1953, "step": 8703 }, { "epoch": 4.096, "grad_norm": 0.735417330086494, "learning_rate": 9.319538194786867e-05, "loss": 0.2295, "step": 8704 }, { "epoch": 4.096470588235294, "grad_norm": 0.6310426437178848, "learning_rate": 9.315481309308303e-05, "loss": 0.1885, "step": 8705 }, { "epoch": 4.0969411764705885, "grad_norm": 1.077802942107203, "learning_rate": 9.311424909315514e-05, "loss": 0.5508, "step": 8706 }, { "epoch": 4.097411764705883, "grad_norm": 1.170412409551822, "learning_rate": 9.307368995154945e-05, "loss": 1.125, "step": 8707 }, { "epoch": 4.097882352941176, "grad_norm": 0.7324838293042377, "learning_rate": 9.303313567172985e-05, "loss": 0.2021, "step": 8708 }, { "epoch": 4.09835294117647, "grad_norm": 1.2037812706046978, "learning_rate": 9.299258625715988e-05, "loss": 1.2031, "step": 8709 }, { "epoch": 4.0988235294117645, "grad_norm": 0.8249343298982518, "learning_rate": 9.295204171130264e-05, "loss": 0.3691, "step": 8710 }, { "epoch": 4.099294117647059, "grad_norm": 0.8611426894761693, "learning_rate": 9.291150203762087e-05, "loss": 0.3809, "step": 8711 }, { "epoch": 4.099764705882353, "grad_norm": 1.0966974573855441, "learning_rate": 9.287096723957685e-05, "loss": 1.2422, "step": 8712 }, { "epoch": 4.100235294117647, "grad_norm": 1.0660350929834297, "learning_rate": 9.283043732063243e-05, "loss": 1.0234, "step": 8713 }, { "epoch": 4.100705882352941, "grad_norm": 0.8874367056813646, "learning_rate": 9.27899122842491e-05, "loss": 0.4492, "step": 8714 }, { "epoch": 4.101176470588236, "grad_norm": 0.9743728849788692, "learning_rate": 9.274939213388785e-05, "loss": 0.7734, "step": 8715 }, { "epoch": 4.101647058823529, "grad_norm": 1.1031416986721023, "learning_rate": 9.270887687300932e-05, "loss": 0.6953, "step": 8716 }, { "epoch": 4.102117647058823, "grad_norm": 1.3497015276067694, "learning_rate": 9.266836650507375e-05, "loss": 0.9648, "step": 8717 }, { "epoch": 4.102588235294117, "grad_norm": 1.0929733022274113, "learning_rate": 9.26278610335409e-05, "loss": 0.8789, "step": 8718 }, { "epoch": 4.103058823529412, "grad_norm": 1.2847765787191587, "learning_rate": 9.25873604618701e-05, "loss": 1.6719, "step": 8719 }, { "epoch": 4.103529411764706, "grad_norm": 1.0727572978008109, "learning_rate": 9.254686479352029e-05, "loss": 1.0859, "step": 8720 }, { "epoch": 4.104, "grad_norm": 0.6685819204969912, "learning_rate": 9.250637403195008e-05, "loss": 0.1641, "step": 8721 }, { "epoch": 4.104470588235294, "grad_norm": 1.7478157944331327, "learning_rate": 9.246588818061753e-05, "loss": 1.5859, "step": 8722 }, { "epoch": 4.1049411764705885, "grad_norm": 0.8291823614325339, "learning_rate": 9.242540724298032e-05, "loss": 0.2432, "step": 8723 }, { "epoch": 4.105411764705883, "grad_norm": 1.288594147029722, "learning_rate": 9.238493122249574e-05, "loss": 1.0547, "step": 8724 }, { "epoch": 4.105882352941176, "grad_norm": 0.9671722509212667, "learning_rate": 9.234446012262065e-05, "loss": 0.7188, "step": 8725 }, { "epoch": 4.10635294117647, "grad_norm": 0.6719824782859356, "learning_rate": 9.230399394681141e-05, "loss": 0.2109, "step": 8726 }, { "epoch": 4.1068235294117645, "grad_norm": 1.1683661999051849, "learning_rate": 9.226353269852413e-05, "loss": 1.0938, "step": 8727 }, { "epoch": 4.107294117647059, "grad_norm": 1.1848553661101022, "learning_rate": 9.222307638121436e-05, "loss": 0.9805, "step": 8728 }, { "epoch": 4.107764705882353, "grad_norm": 0.974111328297875, "learning_rate": 9.218262499833723e-05, "loss": 0.6406, "step": 8729 }, { "epoch": 4.108235294117647, "grad_norm": 1.1904113934990497, "learning_rate": 9.21421785533475e-05, "loss": 0.5625, "step": 8730 }, { "epoch": 4.108705882352941, "grad_norm": 0.7930308808826179, "learning_rate": 9.210173704969952e-05, "loss": 0.3066, "step": 8731 }, { "epoch": 4.109176470588236, "grad_norm": 0.9979146222785117, "learning_rate": 9.206130049084721e-05, "loss": 0.8945, "step": 8732 }, { "epoch": 4.109647058823529, "grad_norm": 0.7296125289316142, "learning_rate": 9.202086888024404e-05, "loss": 0.2168, "step": 8733 }, { "epoch": 4.110117647058823, "grad_norm": 1.1391415524324344, "learning_rate": 9.1980442221343e-05, "loss": 0.875, "step": 8734 }, { "epoch": 4.110588235294117, "grad_norm": 0.8348520893985436, "learning_rate": 9.194002051759685e-05, "loss": 0.8594, "step": 8735 }, { "epoch": 4.111058823529412, "grad_norm": 0.7441390588175983, "learning_rate": 9.189960377245768e-05, "loss": 0.2305, "step": 8736 }, { "epoch": 4.111529411764706, "grad_norm": 0.6154447366542225, "learning_rate": 9.185919198937736e-05, "loss": 0.106, "step": 8737 }, { "epoch": 4.112, "grad_norm": 0.5759865501426935, "learning_rate": 9.181878517180719e-05, "loss": 0.1187, "step": 8738 }, { "epoch": 4.112470588235294, "grad_norm": 1.2681597144537844, "learning_rate": 9.177838332319818e-05, "loss": 0.6445, "step": 8739 }, { "epoch": 4.1129411764705885, "grad_norm": 1.0282264656859792, "learning_rate": 9.173798644700079e-05, "loss": 1.1797, "step": 8740 }, { "epoch": 4.113411764705883, "grad_norm": 0.4009136285351932, "learning_rate": 9.169759454666511e-05, "loss": 0.0903, "step": 8741 }, { "epoch": 4.113882352941176, "grad_norm": 0.9622457778839454, "learning_rate": 9.165720762564086e-05, "loss": 0.4316, "step": 8742 }, { "epoch": 4.11435294117647, "grad_norm": 1.3696838873958146, "learning_rate": 9.161682568737726e-05, "loss": 0.8164, "step": 8743 }, { "epoch": 4.1148235294117645, "grad_norm": 0.8693223135501189, "learning_rate": 9.157644873532308e-05, "loss": 0.3984, "step": 8744 }, { "epoch": 4.115294117647059, "grad_norm": 0.9132931288436109, "learning_rate": 9.153607677292676e-05, "loss": 0.7305, "step": 8745 }, { "epoch": 4.115764705882353, "grad_norm": 0.7048432387638806, "learning_rate": 9.149570980363624e-05, "loss": 0.1924, "step": 8746 }, { "epoch": 4.116235294117647, "grad_norm": 1.2823423835665726, "learning_rate": 9.145534783089909e-05, "loss": 0.6602, "step": 8747 }, { "epoch": 4.116705882352941, "grad_norm": 0.6158868172671017, "learning_rate": 9.141499085816237e-05, "loss": 0.1855, "step": 8748 }, { "epoch": 4.117176470588236, "grad_norm": 0.8273066498656323, "learning_rate": 9.137463888887279e-05, "loss": 0.293, "step": 8749 }, { "epoch": 4.117647058823529, "grad_norm": 0.797461244458482, "learning_rate": 9.133429192647661e-05, "loss": 0.2832, "step": 8750 }, { "epoch": 4.118117647058823, "grad_norm": 1.2200396521907495, "learning_rate": 9.129394997441963e-05, "loss": 0.6875, "step": 8751 }, { "epoch": 4.118588235294117, "grad_norm": 0.5734934171562914, "learning_rate": 9.12536130361473e-05, "loss": 0.3613, "step": 8752 }, { "epoch": 4.119058823529412, "grad_norm": 1.0015923371972093, "learning_rate": 9.121328111510457e-05, "loss": 0.4023, "step": 8753 }, { "epoch": 4.119529411764706, "grad_norm": 0.8416166700672909, "learning_rate": 9.117295421473595e-05, "loss": 0.2969, "step": 8754 }, { "epoch": 4.12, "grad_norm": 1.1855720486566466, "learning_rate": 9.113263233848562e-05, "loss": 0.8516, "step": 8755 }, { "epoch": 4.120470588235294, "grad_norm": 1.2998986397215215, "learning_rate": 9.109231548979722e-05, "loss": 1.4219, "step": 8756 }, { "epoch": 4.1209411764705886, "grad_norm": 0.5817038573454282, "learning_rate": 9.105200367211401e-05, "loss": 0.1738, "step": 8757 }, { "epoch": 4.121411764705883, "grad_norm": 1.058560066735348, "learning_rate": 9.101169688887882e-05, "loss": 0.4922, "step": 8758 }, { "epoch": 4.121882352941176, "grad_norm": 0.8588934121158026, "learning_rate": 9.097139514353407e-05, "loss": 0.5039, "step": 8759 }, { "epoch": 4.12235294117647, "grad_norm": 1.601795111429809, "learning_rate": 9.093109843952171e-05, "loss": 0.7344, "step": 8760 }, { "epoch": 4.1228235294117646, "grad_norm": 1.302390474946749, "learning_rate": 9.089080678028325e-05, "loss": 0.8281, "step": 8761 }, { "epoch": 4.123294117647059, "grad_norm": 0.8326844351696535, "learning_rate": 9.085052016925984e-05, "loss": 0.3398, "step": 8762 }, { "epoch": 4.123764705882353, "grad_norm": 1.1881721322206675, "learning_rate": 9.081023860989213e-05, "loss": 1.0703, "step": 8763 }, { "epoch": 4.124235294117647, "grad_norm": 0.6725263556864299, "learning_rate": 9.076996210562035e-05, "loss": 0.4668, "step": 8764 }, { "epoch": 4.124705882352941, "grad_norm": 0.6381739344160996, "learning_rate": 9.072969065988435e-05, "loss": 0.21, "step": 8765 }, { "epoch": 4.125176470588236, "grad_norm": 0.9114247206618288, "learning_rate": 9.06894242761235e-05, "loss": 0.5469, "step": 8766 }, { "epoch": 4.125647058823529, "grad_norm": 0.769811210799763, "learning_rate": 9.064916295777669e-05, "loss": 0.4004, "step": 8767 }, { "epoch": 4.126117647058823, "grad_norm": 0.6644045995850073, "learning_rate": 9.06089067082825e-05, "loss": 0.3223, "step": 8768 }, { "epoch": 4.1265882352941174, "grad_norm": 0.7225866855180861, "learning_rate": 9.0568655531079e-05, "loss": 0.1787, "step": 8769 }, { "epoch": 4.127058823529412, "grad_norm": 0.9507711449900482, "learning_rate": 9.052840942960383e-05, "loss": 0.3125, "step": 8770 }, { "epoch": 4.127529411764706, "grad_norm": 1.1571890023191445, "learning_rate": 9.048816840729415e-05, "loss": 0.9141, "step": 8771 }, { "epoch": 4.128, "grad_norm": 1.1509185283615713, "learning_rate": 9.044793246758685e-05, "loss": 0.832, "step": 8772 }, { "epoch": 4.128470588235294, "grad_norm": 1.1387432954811818, "learning_rate": 9.04077016139182e-05, "loss": 0.7383, "step": 8773 }, { "epoch": 4.128941176470589, "grad_norm": 0.8801737909081936, "learning_rate": 9.03674758497241e-05, "loss": 0.5352, "step": 8774 }, { "epoch": 4.129411764705883, "grad_norm": 1.2125971217996037, "learning_rate": 9.032725517844004e-05, "loss": 0.6641, "step": 8775 }, { "epoch": 4.129882352941176, "grad_norm": 1.3554893537116077, "learning_rate": 9.028703960350113e-05, "loss": 0.4082, "step": 8776 }, { "epoch": 4.13035294117647, "grad_norm": 0.9216426421971898, "learning_rate": 9.024682912834196e-05, "loss": 0.5352, "step": 8777 }, { "epoch": 4.130823529411765, "grad_norm": 0.9709143486819519, "learning_rate": 9.020662375639663e-05, "loss": 0.3457, "step": 8778 }, { "epoch": 4.131294117647059, "grad_norm": 0.7324948654118923, "learning_rate": 9.016642349109892e-05, "loss": 0.293, "step": 8779 }, { "epoch": 4.131764705882353, "grad_norm": 0.8691650540851062, "learning_rate": 9.012622833588213e-05, "loss": 0.7305, "step": 8780 }, { "epoch": 4.132235294117647, "grad_norm": 1.3809442202389173, "learning_rate": 9.008603829417913e-05, "loss": 1.0547, "step": 8781 }, { "epoch": 4.1327058823529415, "grad_norm": 1.1261047558318016, "learning_rate": 9.00458533694223e-05, "loss": 0.6797, "step": 8782 }, { "epoch": 4.133176470588236, "grad_norm": 0.9911529774042898, "learning_rate": 9.000567356504368e-05, "loss": 0.9336, "step": 8783 }, { "epoch": 4.133647058823529, "grad_norm": 0.8213662301025136, "learning_rate": 8.996549888447484e-05, "loss": 0.8438, "step": 8784 }, { "epoch": 4.134117647058823, "grad_norm": 1.125205533881649, "learning_rate": 8.99253293311468e-05, "loss": 0.9219, "step": 8785 }, { "epoch": 4.1345882352941175, "grad_norm": 1.0306612813198464, "learning_rate": 8.988516490849037e-05, "loss": 0.459, "step": 8786 }, { "epoch": 4.135058823529412, "grad_norm": 0.5689957663276847, "learning_rate": 8.984500561993569e-05, "loss": 0.0977, "step": 8787 }, { "epoch": 4.135529411764706, "grad_norm": 1.0335658890014656, "learning_rate": 8.98048514689126e-05, "loss": 0.3379, "step": 8788 }, { "epoch": 4.136, "grad_norm": 1.063363432243913, "learning_rate": 8.976470245885041e-05, "loss": 0.8047, "step": 8789 }, { "epoch": 4.136470588235294, "grad_norm": 1.290303465861149, "learning_rate": 8.972455859317811e-05, "loss": 0.6953, "step": 8790 }, { "epoch": 4.136941176470589, "grad_norm": 1.1093612196101823, "learning_rate": 8.968441987532417e-05, "loss": 0.5781, "step": 8791 }, { "epoch": 4.137411764705883, "grad_norm": 0.988418326768851, "learning_rate": 8.964428630871657e-05, "loss": 0.6875, "step": 8792 }, { "epoch": 4.137882352941176, "grad_norm": 1.1992090782505322, "learning_rate": 8.960415789678303e-05, "loss": 1.5625, "step": 8793 }, { "epoch": 4.13835294117647, "grad_norm": 0.6089737007055163, "learning_rate": 8.95640346429506e-05, "loss": 0.6016, "step": 8794 }, { "epoch": 4.138823529411765, "grad_norm": 0.6711567096389005, "learning_rate": 8.952391655064605e-05, "loss": 0.2168, "step": 8795 }, { "epoch": 4.139294117647059, "grad_norm": 1.5366194381140508, "learning_rate": 8.948380362329567e-05, "loss": 1.8984, "step": 8796 }, { "epoch": 4.139764705882353, "grad_norm": 0.8996111826568081, "learning_rate": 8.944369586432531e-05, "loss": 0.3984, "step": 8797 }, { "epoch": 4.140235294117647, "grad_norm": 0.8418186331304487, "learning_rate": 8.940359327716032e-05, "loss": 0.4824, "step": 8798 }, { "epoch": 4.1407058823529415, "grad_norm": 1.0139927973320921, "learning_rate": 8.936349586522569e-05, "loss": 0.3809, "step": 8799 }, { "epoch": 4.141176470588236, "grad_norm": 1.054074564482328, "learning_rate": 8.932340363194595e-05, "loss": 1.4688, "step": 8800 }, { "epoch": 4.141647058823529, "grad_norm": 1.2071007033482666, "learning_rate": 8.928331658074516e-05, "loss": 1.0625, "step": 8801 }, { "epoch": 4.142117647058823, "grad_norm": 0.8321816184040484, "learning_rate": 8.92432347150469e-05, "loss": 0.2197, "step": 8802 }, { "epoch": 4.1425882352941175, "grad_norm": 0.7736995237810856, "learning_rate": 8.920315803827446e-05, "loss": 0.4941, "step": 8803 }, { "epoch": 4.143058823529412, "grad_norm": 1.2149000143681723, "learning_rate": 8.916308655385054e-05, "loss": 1.2031, "step": 8804 }, { "epoch": 4.143529411764706, "grad_norm": 0.968527914794541, "learning_rate": 8.912302026519738e-05, "loss": 0.3477, "step": 8805 }, { "epoch": 4.144, "grad_norm": 1.02426107859736, "learning_rate": 8.908295917573692e-05, "loss": 0.6523, "step": 8806 }, { "epoch": 4.144470588235294, "grad_norm": 0.9570646989345585, "learning_rate": 8.904290328889055e-05, "loss": 0.4395, "step": 8807 }, { "epoch": 4.144941176470589, "grad_norm": 0.8956775249422344, "learning_rate": 8.900285260807922e-05, "loss": 0.4512, "step": 8808 }, { "epoch": 4.145411764705882, "grad_norm": 0.9928893674361303, "learning_rate": 8.896280713672344e-05, "loss": 0.4473, "step": 8809 }, { "epoch": 4.145882352941176, "grad_norm": 2.1154242712027167, "learning_rate": 8.892276687824335e-05, "loss": 1.4688, "step": 8810 }, { "epoch": 4.14635294117647, "grad_norm": 1.2415121792290422, "learning_rate": 8.888273183605856e-05, "loss": 0.9805, "step": 8811 }, { "epoch": 4.146823529411765, "grad_norm": 1.1510795987603195, "learning_rate": 8.884270201358821e-05, "loss": 0.7305, "step": 8812 }, { "epoch": 4.147294117647059, "grad_norm": 0.9252161709212545, "learning_rate": 8.880267741425111e-05, "loss": 0.6641, "step": 8813 }, { "epoch": 4.147764705882353, "grad_norm": 1.0647139914151889, "learning_rate": 8.876265804146554e-05, "loss": 0.4277, "step": 8814 }, { "epoch": 4.148235294117647, "grad_norm": 0.8846984867487195, "learning_rate": 8.872264389864929e-05, "loss": 0.6367, "step": 8815 }, { "epoch": 4.1487058823529415, "grad_norm": 1.2594589434240668, "learning_rate": 8.868263498921987e-05, "loss": 0.8359, "step": 8816 }, { "epoch": 4.149176470588236, "grad_norm": 0.9818171813743851, "learning_rate": 8.864263131659417e-05, "loss": 0.8438, "step": 8817 }, { "epoch": 4.149647058823529, "grad_norm": 1.446627873499096, "learning_rate": 8.860263288418866e-05, "loss": 0.8047, "step": 8818 }, { "epoch": 4.150117647058823, "grad_norm": 0.8401696960827927, "learning_rate": 8.856263969541948e-05, "loss": 0.2871, "step": 8819 }, { "epoch": 4.1505882352941175, "grad_norm": 0.711490932520202, "learning_rate": 8.852265175370223e-05, "loss": 0.707, "step": 8820 }, { "epoch": 4.151058823529412, "grad_norm": 0.7600738676672406, "learning_rate": 8.848266906245207e-05, "loss": 0.5234, "step": 8821 }, { "epoch": 4.151529411764706, "grad_norm": 1.1798632836338203, "learning_rate": 8.844269162508373e-05, "loss": 0.7266, "step": 8822 }, { "epoch": 4.152, "grad_norm": 0.9128076845679867, "learning_rate": 8.840271944501143e-05, "loss": 1.0312, "step": 8823 }, { "epoch": 4.152470588235294, "grad_norm": 0.9756046819770053, "learning_rate": 8.836275252564903e-05, "loss": 1.1562, "step": 8824 }, { "epoch": 4.152941176470589, "grad_norm": 1.0112303332499197, "learning_rate": 8.832279087040992e-05, "loss": 1.0938, "step": 8825 }, { "epoch": 4.153411764705883, "grad_norm": 0.8491010634552926, "learning_rate": 8.828283448270698e-05, "loss": 0.4688, "step": 8826 }, { "epoch": 4.153882352941176, "grad_norm": 0.8576241722695027, "learning_rate": 8.824288336595267e-05, "loss": 0.5938, "step": 8827 }, { "epoch": 4.15435294117647, "grad_norm": 1.3214553840166219, "learning_rate": 8.820293752355906e-05, "loss": 0.8633, "step": 8828 }, { "epoch": 4.154823529411765, "grad_norm": 1.0595473061198029, "learning_rate": 8.816299695893773e-05, "loss": 0.7031, "step": 8829 }, { "epoch": 4.155294117647059, "grad_norm": 0.5201529850214459, "learning_rate": 8.81230616754997e-05, "loss": 0.0825, "step": 8830 }, { "epoch": 4.155764705882353, "grad_norm": 1.1774173559818368, "learning_rate": 8.808313167665575e-05, "loss": 1.1094, "step": 8831 }, { "epoch": 4.156235294117647, "grad_norm": 1.2535032907518928, "learning_rate": 8.804320696581606e-05, "loss": 0.7578, "step": 8832 }, { "epoch": 4.1567058823529415, "grad_norm": 0.8832359071347761, "learning_rate": 8.800328754639038e-05, "loss": 0.3535, "step": 8833 }, { "epoch": 4.157176470588236, "grad_norm": 1.0276212431512466, "learning_rate": 8.796337342178804e-05, "loss": 1.2109, "step": 8834 }, { "epoch": 4.157647058823529, "grad_norm": 0.8215742816447161, "learning_rate": 8.792346459541789e-05, "loss": 0.3184, "step": 8835 }, { "epoch": 4.158117647058823, "grad_norm": 0.7065616977876342, "learning_rate": 8.788356107068833e-05, "loss": 0.3945, "step": 8836 }, { "epoch": 4.1585882352941175, "grad_norm": 1.2004431243670195, "learning_rate": 8.784366285100738e-05, "loss": 0.6914, "step": 8837 }, { "epoch": 4.159058823529412, "grad_norm": 1.1280571552373386, "learning_rate": 8.780376993978246e-05, "loss": 0.7539, "step": 8838 }, { "epoch": 4.159529411764706, "grad_norm": 0.8591944900635062, "learning_rate": 8.77638823404207e-05, "loss": 0.6406, "step": 8839 }, { "epoch": 4.16, "grad_norm": 0.8646932073654627, "learning_rate": 8.772400005632859e-05, "loss": 0.4551, "step": 8840 }, { "epoch": 4.160470588235294, "grad_norm": 0.7247124446757078, "learning_rate": 8.768412309091238e-05, "loss": 0.3379, "step": 8841 }, { "epoch": 4.160941176470589, "grad_norm": 1.0003654386017364, "learning_rate": 8.764425144757771e-05, "loss": 0.7305, "step": 8842 }, { "epoch": 4.161411764705882, "grad_norm": 0.955438572085777, "learning_rate": 8.760438512972982e-05, "loss": 0.4492, "step": 8843 }, { "epoch": 4.161882352941176, "grad_norm": 1.0061975256081828, "learning_rate": 8.756452414077351e-05, "loss": 0.3906, "step": 8844 }, { "epoch": 4.16235294117647, "grad_norm": 1.1428862854689352, "learning_rate": 8.752466848411305e-05, "loss": 1.125, "step": 8845 }, { "epoch": 4.162823529411765, "grad_norm": 1.022316060624887, "learning_rate": 8.748481816315233e-05, "loss": 0.9062, "step": 8846 }, { "epoch": 4.163294117647059, "grad_norm": 0.8955662240832333, "learning_rate": 8.744497318129483e-05, "loss": 0.3027, "step": 8847 }, { "epoch": 4.163764705882353, "grad_norm": 0.7822628872122774, "learning_rate": 8.74051335419434e-05, "loss": 0.9023, "step": 8848 }, { "epoch": 4.164235294117647, "grad_norm": 0.5813851627082297, "learning_rate": 8.736529924850057e-05, "loss": 0.1406, "step": 8849 }, { "epoch": 4.1647058823529415, "grad_norm": 0.28543164689416217, "learning_rate": 8.73254703043685e-05, "loss": 0.0087, "step": 8850 }, { "epoch": 4.165176470588236, "grad_norm": 1.0291483828750225, "learning_rate": 8.72856467129486e-05, "loss": 0.7773, "step": 8851 }, { "epoch": 4.165647058823529, "grad_norm": 1.4172264337444893, "learning_rate": 8.72458284776421e-05, "loss": 0.8047, "step": 8852 }, { "epoch": 4.166117647058823, "grad_norm": 1.4532952173973002, "learning_rate": 8.72060156018497e-05, "loss": 1.6953, "step": 8853 }, { "epoch": 4.1665882352941175, "grad_norm": 1.1820760178294867, "learning_rate": 8.716620808897151e-05, "loss": 1.5703, "step": 8854 }, { "epoch": 4.167058823529412, "grad_norm": 1.1546441480447467, "learning_rate": 8.712640594240735e-05, "loss": 1.3516, "step": 8855 }, { "epoch": 4.167529411764706, "grad_norm": 1.3640810330207656, "learning_rate": 8.708660916555657e-05, "loss": 1.4453, "step": 8856 }, { "epoch": 4.168, "grad_norm": 1.0178019466817603, "learning_rate": 8.70468177618179e-05, "loss": 0.6172, "step": 8857 }, { "epoch": 4.168470588235294, "grad_norm": 0.8133897594470791, "learning_rate": 8.700703173458978e-05, "loss": 0.3281, "step": 8858 }, { "epoch": 4.168941176470589, "grad_norm": 1.1716964240738763, "learning_rate": 8.69672510872702e-05, "loss": 0.8008, "step": 8859 }, { "epoch": 4.169411764705882, "grad_norm": 1.1357154278852666, "learning_rate": 8.692747582325646e-05, "loss": 1.1641, "step": 8860 }, { "epoch": 4.169882352941176, "grad_norm": 0.7017119566364428, "learning_rate": 8.68877059459457e-05, "loss": 0.5859, "step": 8861 }, { "epoch": 4.17035294117647, "grad_norm": 0.907597550334809, "learning_rate": 8.684794145873443e-05, "loss": 0.3945, "step": 8862 }, { "epoch": 4.170823529411765, "grad_norm": 1.293240339479126, "learning_rate": 8.680818236501867e-05, "loss": 1.3125, "step": 8863 }, { "epoch": 4.171294117647059, "grad_norm": 1.1870444533180688, "learning_rate": 8.67684286681941e-05, "loss": 1.2578, "step": 8864 }, { "epoch": 4.171764705882353, "grad_norm": 0.5536751625539115, "learning_rate": 8.672868037165587e-05, "loss": 0.1523, "step": 8865 }, { "epoch": 4.172235294117647, "grad_norm": 0.9633910655147824, "learning_rate": 8.668893747879869e-05, "loss": 0.9609, "step": 8866 }, { "epoch": 4.1727058823529415, "grad_norm": 1.0236240966025236, "learning_rate": 8.664919999301683e-05, "loss": 1.1094, "step": 8867 }, { "epoch": 4.173176470588236, "grad_norm": 0.7103383681888729, "learning_rate": 8.660946791770397e-05, "loss": 0.1533, "step": 8868 }, { "epoch": 4.173647058823529, "grad_norm": 1.1365106861587868, "learning_rate": 8.656974125625349e-05, "loss": 0.9375, "step": 8869 }, { "epoch": 4.174117647058823, "grad_norm": 0.8595378470304877, "learning_rate": 8.653002001205827e-05, "loss": 0.3711, "step": 8870 }, { "epoch": 4.1745882352941175, "grad_norm": 0.9626540775284391, "learning_rate": 8.649030418851061e-05, "loss": 0.582, "step": 8871 }, { "epoch": 4.175058823529412, "grad_norm": 0.7471296134790656, "learning_rate": 8.645059378900255e-05, "loss": 0.4727, "step": 8872 }, { "epoch": 4.175529411764706, "grad_norm": 1.2639679046246821, "learning_rate": 8.641088881692542e-05, "loss": 0.7227, "step": 8873 }, { "epoch": 4.176, "grad_norm": 0.6967229088332738, "learning_rate": 8.637118927567028e-05, "loss": 0.4082, "step": 8874 }, { "epoch": 4.176470588235294, "grad_norm": 1.241777802491848, "learning_rate": 8.633149516862775e-05, "loss": 0.5391, "step": 8875 }, { "epoch": 4.176941176470589, "grad_norm": 0.9008614531336612, "learning_rate": 8.629180649918775e-05, "loss": 0.3105, "step": 8876 }, { "epoch": 4.177411764705882, "grad_norm": 1.0154207614495143, "learning_rate": 8.625212327073998e-05, "loss": 1.4453, "step": 8877 }, { "epoch": 4.177882352941176, "grad_norm": 2.1876037850455816, "learning_rate": 8.621244548667363e-05, "loss": 0.6719, "step": 8878 }, { "epoch": 4.17835294117647, "grad_norm": 0.9506576286208136, "learning_rate": 8.617277315037723e-05, "loss": 0.5781, "step": 8879 }, { "epoch": 4.178823529411765, "grad_norm": 1.2279146515563246, "learning_rate": 8.613310626523909e-05, "loss": 0.8359, "step": 8880 }, { "epoch": 4.179294117647059, "grad_norm": 0.9218011219268881, "learning_rate": 8.609344483464697e-05, "loss": 0.8008, "step": 8881 }, { "epoch": 4.179764705882353, "grad_norm": 1.0034582270653045, "learning_rate": 8.60537888619881e-05, "loss": 0.957, "step": 8882 }, { "epoch": 4.180235294117647, "grad_norm": 1.3658939730670672, "learning_rate": 8.601413835064929e-05, "loss": 0.6172, "step": 8883 }, { "epoch": 4.1807058823529415, "grad_norm": 1.0281213972235803, "learning_rate": 8.597449330401698e-05, "loss": 0.4453, "step": 8884 }, { "epoch": 4.181176470588236, "grad_norm": 0.9374047811707211, "learning_rate": 8.593485372547693e-05, "loss": 0.4414, "step": 8885 }, { "epoch": 4.181647058823529, "grad_norm": 1.159432026466391, "learning_rate": 8.589521961841462e-05, "loss": 0.9062, "step": 8886 }, { "epoch": 4.182117647058823, "grad_norm": 0.8679205291648232, "learning_rate": 8.585559098621503e-05, "loss": 0.334, "step": 8887 }, { "epoch": 4.1825882352941175, "grad_norm": 1.2824061087121117, "learning_rate": 8.581596783226254e-05, "loss": 1.0938, "step": 8888 }, { "epoch": 4.183058823529412, "grad_norm": 1.153998236123474, "learning_rate": 8.577635015994122e-05, "loss": 1.25, "step": 8889 }, { "epoch": 4.183529411764706, "grad_norm": 1.0242181681018079, "learning_rate": 8.573673797263466e-05, "loss": 0.9609, "step": 8890 }, { "epoch": 4.184, "grad_norm": 0.731610868269281, "learning_rate": 8.569713127372583e-05, "loss": 0.4707, "step": 8891 }, { "epoch": 4.184470588235294, "grad_norm": 0.8988919919700883, "learning_rate": 8.565753006659741e-05, "loss": 0.3613, "step": 8892 }, { "epoch": 4.184941176470589, "grad_norm": 1.1356123977257302, "learning_rate": 8.561793435463157e-05, "loss": 0.8203, "step": 8893 }, { "epoch": 4.185411764705882, "grad_norm": 0.7004735215726391, "learning_rate": 8.557834414120986e-05, "loss": 0.498, "step": 8894 }, { "epoch": 4.185882352941176, "grad_norm": 0.8056949908775024, "learning_rate": 8.553875942971362e-05, "loss": 0.3535, "step": 8895 }, { "epoch": 4.18635294117647, "grad_norm": 1.1185705938747605, "learning_rate": 8.549918022352343e-05, "loss": 0.5, "step": 8896 }, { "epoch": 4.186823529411765, "grad_norm": 0.7385462180103118, "learning_rate": 8.545960652601964e-05, "loss": 0.2695, "step": 8897 }, { "epoch": 4.187294117647059, "grad_norm": 0.9617033328091291, "learning_rate": 8.542003834058207e-05, "loss": 0.6992, "step": 8898 }, { "epoch": 4.187764705882353, "grad_norm": 0.8629192304677256, "learning_rate": 8.538047567058994e-05, "loss": 0.6484, "step": 8899 }, { "epoch": 4.188235294117647, "grad_norm": 1.1777380273278306, "learning_rate": 8.534091851942214e-05, "loss": 0.8828, "step": 8900 }, { "epoch": 4.1887058823529415, "grad_norm": 1.1296872000244633, "learning_rate": 8.530136689045712e-05, "loss": 0.4863, "step": 8901 }, { "epoch": 4.189176470588236, "grad_norm": 0.7486670546452677, "learning_rate": 8.526182078707261e-05, "loss": 0.498, "step": 8902 }, { "epoch": 4.189647058823529, "grad_norm": 0.7711591381791623, "learning_rate": 8.522228021264618e-05, "loss": 0.2812, "step": 8903 }, { "epoch": 4.190117647058823, "grad_norm": 0.8387393187454961, "learning_rate": 8.518274517055482e-05, "loss": 0.4395, "step": 8904 }, { "epoch": 4.1905882352941175, "grad_norm": 0.7541845761915495, "learning_rate": 8.514321566417488e-05, "loss": 0.2422, "step": 8905 }, { "epoch": 4.191058823529412, "grad_norm": 0.8757638624873016, "learning_rate": 8.510369169688243e-05, "loss": 0.3828, "step": 8906 }, { "epoch": 4.191529411764706, "grad_norm": 1.2226985411765523, "learning_rate": 8.506417327205309e-05, "loss": 0.9102, "step": 8907 }, { "epoch": 4.192, "grad_norm": 0.7906346140205168, "learning_rate": 8.502466039306175e-05, "loss": 0.7305, "step": 8908 }, { "epoch": 4.192470588235294, "grad_norm": 0.5406984580621892, "learning_rate": 8.498515306328323e-05, "loss": 0.1689, "step": 8909 }, { "epoch": 4.192941176470589, "grad_norm": 1.1342826889689364, "learning_rate": 8.494565128609146e-05, "loss": 0.7188, "step": 8910 }, { "epoch": 4.193411764705882, "grad_norm": 1.055035391372517, "learning_rate": 8.490615506486018e-05, "loss": 0.4141, "step": 8911 }, { "epoch": 4.193882352941176, "grad_norm": 0.8106596235527616, "learning_rate": 8.486666440296259e-05, "loss": 0.3047, "step": 8912 }, { "epoch": 4.19435294117647, "grad_norm": 0.9857654980817624, "learning_rate": 8.482717930377129e-05, "loss": 0.5664, "step": 8913 }, { "epoch": 4.194823529411765, "grad_norm": 0.6861168099049783, "learning_rate": 8.478769977065853e-05, "loss": 0.2002, "step": 8914 }, { "epoch": 4.195294117647059, "grad_norm": 0.6805267409065677, "learning_rate": 8.474822580699614e-05, "loss": 0.1748, "step": 8915 }, { "epoch": 4.195764705882353, "grad_norm": 1.0665783848418369, "learning_rate": 8.470875741615527e-05, "loss": 0.8438, "step": 8916 }, { "epoch": 4.196235294117647, "grad_norm": 1.1753536484329252, "learning_rate": 8.466929460150678e-05, "loss": 0.8711, "step": 8917 }, { "epoch": 4.1967058823529415, "grad_norm": 0.7917770931115097, "learning_rate": 8.462983736642101e-05, "loss": 0.4082, "step": 8918 }, { "epoch": 4.197176470588236, "grad_norm": 1.044422479692957, "learning_rate": 8.459038571426772e-05, "loss": 1.0703, "step": 8919 }, { "epoch": 4.197647058823529, "grad_norm": 1.7184865659667752, "learning_rate": 8.455093964841632e-05, "loss": 0.8555, "step": 8920 }, { "epoch": 4.198117647058823, "grad_norm": 1.2577100078848715, "learning_rate": 8.451149917223576e-05, "loss": 0.5508, "step": 8921 }, { "epoch": 4.1985882352941175, "grad_norm": 0.8242825969022576, "learning_rate": 8.447206428909434e-05, "loss": 0.3008, "step": 8922 }, { "epoch": 4.199058823529412, "grad_norm": 0.8409617031032709, "learning_rate": 8.44326350023601e-05, "loss": 0.3496, "step": 8923 }, { "epoch": 4.199529411764706, "grad_norm": 0.914865080888266, "learning_rate": 8.439321131540036e-05, "loss": 0.3164, "step": 8924 }, { "epoch": 4.2, "grad_norm": 0.9262120843439174, "learning_rate": 8.435379323158218e-05, "loss": 1.0078, "step": 8925 }, { "epoch": 4.200470588235294, "grad_norm": 1.0209392655984555, "learning_rate": 8.431438075427213e-05, "loss": 0.5781, "step": 8926 }, { "epoch": 4.200941176470589, "grad_norm": 1.9176630749724837, "learning_rate": 8.427497388683607e-05, "loss": 1.2812, "step": 8927 }, { "epoch": 4.201411764705882, "grad_norm": 0.7710773467735061, "learning_rate": 8.423557263263961e-05, "loss": 0.5391, "step": 8928 }, { "epoch": 4.201882352941176, "grad_norm": 0.9458419264616353, "learning_rate": 8.419617699504791e-05, "loss": 1.2969, "step": 8929 }, { "epoch": 4.20235294117647, "grad_norm": 0.9158945447321895, "learning_rate": 8.415678697742538e-05, "loss": 0.4688, "step": 8930 }, { "epoch": 4.202823529411765, "grad_norm": 0.9009176280148621, "learning_rate": 8.41174025831362e-05, "loss": 0.7617, "step": 8931 }, { "epoch": 4.203294117647059, "grad_norm": 0.8428756874540889, "learning_rate": 8.407802381554406e-05, "loss": 0.5039, "step": 8932 }, { "epoch": 4.203764705882353, "grad_norm": 1.5033280648290124, "learning_rate": 8.403865067801197e-05, "loss": 1.1641, "step": 8933 }, { "epoch": 4.204235294117647, "grad_norm": 0.3126740283327955, "learning_rate": 8.399928317390265e-05, "loss": 0.053, "step": 8934 }, { "epoch": 4.2047058823529415, "grad_norm": 1.1954718221504623, "learning_rate": 8.395992130657834e-05, "loss": 1.1719, "step": 8935 }, { "epoch": 4.205176470588235, "grad_norm": 0.9023154418557413, "learning_rate": 8.392056507940061e-05, "loss": 0.5664, "step": 8936 }, { "epoch": 4.205647058823529, "grad_norm": 1.2237377555733913, "learning_rate": 8.388121449573075e-05, "loss": 1.1875, "step": 8937 }, { "epoch": 4.206117647058823, "grad_norm": 1.042668739677428, "learning_rate": 8.384186955892954e-05, "loss": 0.5625, "step": 8938 }, { "epoch": 4.2065882352941175, "grad_norm": 1.0979663447200092, "learning_rate": 8.380253027235709e-05, "loss": 1.4844, "step": 8939 }, { "epoch": 4.207058823529412, "grad_norm": 0.975903031057808, "learning_rate": 8.37631966393733e-05, "loss": 0.6875, "step": 8940 }, { "epoch": 4.207529411764706, "grad_norm": 0.8875623394340986, "learning_rate": 8.372386866333743e-05, "loss": 0.7852, "step": 8941 }, { "epoch": 4.208, "grad_norm": 1.0446263392989905, "learning_rate": 8.36845463476082e-05, "loss": 0.5156, "step": 8942 }, { "epoch": 4.208470588235294, "grad_norm": 0.8273677912465677, "learning_rate": 8.364522969554399e-05, "loss": 0.3145, "step": 8943 }, { "epoch": 4.208941176470589, "grad_norm": 0.8623266974346361, "learning_rate": 8.36059187105027e-05, "loss": 0.6406, "step": 8944 }, { "epoch": 4.209411764705882, "grad_norm": 1.1221670779268247, "learning_rate": 8.356661339584156e-05, "loss": 1.2031, "step": 8945 }, { "epoch": 4.209882352941176, "grad_norm": 0.9801477014473808, "learning_rate": 8.352731375491754e-05, "loss": 0.4551, "step": 8946 }, { "epoch": 4.21035294117647, "grad_norm": 0.635171395481121, "learning_rate": 8.348801979108691e-05, "loss": 0.1963, "step": 8947 }, { "epoch": 4.210823529411765, "grad_norm": 0.8154249198094099, "learning_rate": 8.344873150770562e-05, "loss": 0.2559, "step": 8948 }, { "epoch": 4.211294117647059, "grad_norm": 0.7065395657335678, "learning_rate": 8.340944890812916e-05, "loss": 0.5, "step": 8949 }, { "epoch": 4.211764705882353, "grad_norm": 0.7582770340559609, "learning_rate": 8.337017199571235e-05, "loss": 0.707, "step": 8950 }, { "epoch": 4.212235294117647, "grad_norm": 1.1798722058299524, "learning_rate": 8.333090077380964e-05, "loss": 1.3359, "step": 8951 }, { "epoch": 4.2127058823529415, "grad_norm": 1.238738386216382, "learning_rate": 8.329163524577504e-05, "loss": 0.9219, "step": 8952 }, { "epoch": 4.213176470588236, "grad_norm": 0.7665813846112904, "learning_rate": 8.325237541496197e-05, "loss": 0.2178, "step": 8953 }, { "epoch": 4.213647058823529, "grad_norm": 0.8041235373903544, "learning_rate": 8.32131212847235e-05, "loss": 0.3359, "step": 8954 }, { "epoch": 4.214117647058823, "grad_norm": 0.6087329580702187, "learning_rate": 8.3173872858412e-05, "loss": 0.1226, "step": 8955 }, { "epoch": 4.2145882352941175, "grad_norm": 1.0878988420813143, "learning_rate": 8.313463013937953e-05, "loss": 1.0938, "step": 8956 }, { "epoch": 4.215058823529412, "grad_norm": 1.083300502217616, "learning_rate": 8.309539313097769e-05, "loss": 1.4219, "step": 8957 }, { "epoch": 4.215529411764706, "grad_norm": 0.947582752681776, "learning_rate": 8.305616183655738e-05, "loss": 0.5156, "step": 8958 }, { "epoch": 4.216, "grad_norm": 0.6493099720254517, "learning_rate": 8.30169362594692e-05, "loss": 0.5547, "step": 8959 }, { "epoch": 4.216470588235294, "grad_norm": 1.897929260506078, "learning_rate": 8.29777164030633e-05, "loss": 0.957, "step": 8960 }, { "epoch": 4.216941176470589, "grad_norm": 1.1228219036972962, "learning_rate": 8.293850227068908e-05, "loss": 1.25, "step": 8961 }, { "epoch": 4.217411764705882, "grad_norm": 1.110946351387144, "learning_rate": 8.289929386569569e-05, "loss": 0.8828, "step": 8962 }, { "epoch": 4.217882352941176, "grad_norm": 1.8011976956458189, "learning_rate": 8.28600911914318e-05, "loss": 1.3594, "step": 8963 }, { "epoch": 4.21835294117647, "grad_norm": 1.209219365288593, "learning_rate": 8.282089425124538e-05, "loss": 0.7305, "step": 8964 }, { "epoch": 4.218823529411765, "grad_norm": 0.718379690893734, "learning_rate": 8.278170304848413e-05, "loss": 0.3262, "step": 8965 }, { "epoch": 4.219294117647059, "grad_norm": 1.2384898538975284, "learning_rate": 8.274251758649518e-05, "loss": 1.1016, "step": 8966 }, { "epoch": 4.219764705882353, "grad_norm": 1.2364267361984767, "learning_rate": 8.270333786862506e-05, "loss": 0.9023, "step": 8967 }, { "epoch": 4.220235294117647, "grad_norm": 1.5250921827120958, "learning_rate": 8.266416389821997e-05, "loss": 0.9961, "step": 8968 }, { "epoch": 4.2207058823529415, "grad_norm": 1.2824042364849761, "learning_rate": 8.262499567862565e-05, "loss": 1.1328, "step": 8969 }, { "epoch": 4.221176470588235, "grad_norm": 1.2841422665256867, "learning_rate": 8.258583321318708e-05, "loss": 0.7695, "step": 8970 }, { "epoch": 4.221647058823529, "grad_norm": 1.5543231956182326, "learning_rate": 8.254667650524904e-05, "loss": 1.4688, "step": 8971 }, { "epoch": 4.222117647058823, "grad_norm": 0.8044573919503544, "learning_rate": 8.250752555815574e-05, "loss": 0.252, "step": 8972 }, { "epoch": 4.2225882352941175, "grad_norm": 1.2274195712852933, "learning_rate": 8.246838037525073e-05, "loss": 1.4297, "step": 8973 }, { "epoch": 4.223058823529412, "grad_norm": 1.222531929314953, "learning_rate": 8.242924095987736e-05, "loss": 0.7383, "step": 8974 }, { "epoch": 4.223529411764706, "grad_norm": 1.0330088611850554, "learning_rate": 8.239010731537816e-05, "loss": 0.5, "step": 8975 }, { "epoch": 4.224, "grad_norm": 1.1105167961023281, "learning_rate": 8.235097944509545e-05, "loss": 1.0234, "step": 8976 }, { "epoch": 4.224470588235294, "grad_norm": 1.7161771531770316, "learning_rate": 8.231185735237096e-05, "loss": 1.2266, "step": 8977 }, { "epoch": 4.224941176470589, "grad_norm": 1.1774371091862195, "learning_rate": 8.22727410405458e-05, "loss": 0.8438, "step": 8978 }, { "epoch": 4.225411764705882, "grad_norm": 1.4009489626146376, "learning_rate": 8.223363051296074e-05, "loss": 1.2266, "step": 8979 }, { "epoch": 4.225882352941176, "grad_norm": 1.1610663010444728, "learning_rate": 8.21945257729561e-05, "loss": 0.7305, "step": 8980 }, { "epoch": 4.22635294117647, "grad_norm": 1.097982089857593, "learning_rate": 8.215542682387148e-05, "loss": 1.1641, "step": 8981 }, { "epoch": 4.226823529411765, "grad_norm": 1.1322310045753818, "learning_rate": 8.211633366904618e-05, "loss": 0.7969, "step": 8982 }, { "epoch": 4.227294117647059, "grad_norm": 1.1727586661203804, "learning_rate": 8.207724631181904e-05, "loss": 0.4805, "step": 8983 }, { "epoch": 4.227764705882353, "grad_norm": 1.3014657980828208, "learning_rate": 8.203816475552813e-05, "loss": 0.7773, "step": 8984 }, { "epoch": 4.228235294117647, "grad_norm": 0.9711073628364507, "learning_rate": 8.19990890035113e-05, "loss": 0.9492, "step": 8985 }, { "epoch": 4.2287058823529415, "grad_norm": 1.086107287514434, "learning_rate": 8.196001905910589e-05, "loss": 0.6914, "step": 8986 }, { "epoch": 4.229176470588235, "grad_norm": 1.1215780301704832, "learning_rate": 8.192095492564851e-05, "loss": 0.8281, "step": 8987 }, { "epoch": 4.229647058823529, "grad_norm": 0.929991619638706, "learning_rate": 8.188189660647554e-05, "loss": 1.0938, "step": 8988 }, { "epoch": 4.230117647058823, "grad_norm": 0.7414366318936697, "learning_rate": 8.184284410492274e-05, "loss": 0.2754, "step": 8989 }, { "epoch": 4.2305882352941175, "grad_norm": 0.9919356661622886, "learning_rate": 8.180379742432533e-05, "loss": 0.8555, "step": 8990 }, { "epoch": 4.231058823529412, "grad_norm": 0.9272266417109514, "learning_rate": 8.176475656801811e-05, "loss": 0.4004, "step": 8991 }, { "epoch": 4.231529411764706, "grad_norm": 1.3178466129839959, "learning_rate": 8.172572153933541e-05, "loss": 0.8359, "step": 8992 }, { "epoch": 4.232, "grad_norm": 0.9846834518820665, "learning_rate": 8.168669234161094e-05, "loss": 1.1094, "step": 8993 }, { "epoch": 4.232470588235294, "grad_norm": 1.25264319125337, "learning_rate": 8.164766897817806e-05, "loss": 0.5625, "step": 8994 }, { "epoch": 4.232941176470589, "grad_norm": 0.979084558386499, "learning_rate": 8.160865145236946e-05, "loss": 1.2344, "step": 8995 }, { "epoch": 4.233411764705882, "grad_norm": 1.0793652457050025, "learning_rate": 8.15696397675175e-05, "loss": 1.1094, "step": 8996 }, { "epoch": 4.233882352941176, "grad_norm": 1.1633893169844705, "learning_rate": 8.153063392695394e-05, "loss": 1.3281, "step": 8997 }, { "epoch": 4.23435294117647, "grad_norm": 0.9113426930727221, "learning_rate": 8.14916339340101e-05, "loss": 0.5938, "step": 8998 }, { "epoch": 4.234823529411765, "grad_norm": 1.2778456026770812, "learning_rate": 8.145263979201672e-05, "loss": 0.9844, "step": 8999 }, { "epoch": 4.235294117647059, "grad_norm": 0.46761339096540133, "learning_rate": 8.141365150430421e-05, "loss": 0.1504, "step": 9000 }, { "epoch": 4.235764705882353, "grad_norm": 0.9033781393443971, "learning_rate": 8.137466907420222e-05, "loss": 0.8516, "step": 9001 }, { "epoch": 4.236235294117647, "grad_norm": 1.019264900188765, "learning_rate": 8.133569250504013e-05, "loss": 0.5234, "step": 9002 }, { "epoch": 4.2367058823529415, "grad_norm": 1.0289959856925686, "learning_rate": 8.129672180014666e-05, "loss": 0.8164, "step": 9003 }, { "epoch": 4.237176470588235, "grad_norm": 0.9215295542729465, "learning_rate": 8.125775696285013e-05, "loss": 0.5898, "step": 9004 }, { "epoch": 4.237647058823529, "grad_norm": 0.7530686339740036, "learning_rate": 8.121879799647839e-05, "loss": 0.543, "step": 9005 }, { "epoch": 4.238117647058823, "grad_norm": 1.1240477143679137, "learning_rate": 8.117984490435861e-05, "loss": 0.8555, "step": 9006 }, { "epoch": 4.2385882352941175, "grad_norm": 1.0152899983162322, "learning_rate": 8.114089768981765e-05, "loss": 1.3438, "step": 9007 }, { "epoch": 4.239058823529412, "grad_norm": 0.9826438979457071, "learning_rate": 8.110195635618184e-05, "loss": 0.5664, "step": 9008 }, { "epoch": 4.239529411764706, "grad_norm": 1.8812563466055559, "learning_rate": 8.106302090677682e-05, "loss": 0.8477, "step": 9009 }, { "epoch": 4.24, "grad_norm": 1.0613647813346345, "learning_rate": 8.102409134492795e-05, "loss": 0.6211, "step": 9010 }, { "epoch": 4.240470588235294, "grad_norm": 0.96297817766675, "learning_rate": 8.098516767396008e-05, "loss": 0.5547, "step": 9011 }, { "epoch": 4.240941176470589, "grad_norm": 0.8142282508898974, "learning_rate": 8.094624989719734e-05, "loss": 0.8672, "step": 9012 }, { "epoch": 4.241411764705882, "grad_norm": 0.5748914822035158, "learning_rate": 8.090733801796356e-05, "loss": 0.0913, "step": 9013 }, { "epoch": 4.241882352941176, "grad_norm": 1.4288109980222974, "learning_rate": 8.086843203958207e-05, "loss": 1.1562, "step": 9014 }, { "epoch": 4.24235294117647, "grad_norm": 1.0179263600078847, "learning_rate": 8.082953196537553e-05, "loss": 0.9453, "step": 9015 }, { "epoch": 4.242823529411765, "grad_norm": 1.1417914926668504, "learning_rate": 8.079063779866622e-05, "loss": 1.2812, "step": 9016 }, { "epoch": 4.243294117647059, "grad_norm": 1.192251099281915, "learning_rate": 8.075174954277598e-05, "loss": 0.5312, "step": 9017 }, { "epoch": 4.243764705882353, "grad_norm": 1.0516458348053686, "learning_rate": 8.071286720102594e-05, "loss": 0.4473, "step": 9018 }, { "epoch": 4.244235294117647, "grad_norm": 1.0318916356400578, "learning_rate": 8.067399077673688e-05, "loss": 0.8594, "step": 9019 }, { "epoch": 4.2447058823529416, "grad_norm": 0.8041902370036613, "learning_rate": 8.06351202732291e-05, "loss": 0.3145, "step": 9020 }, { "epoch": 4.245176470588235, "grad_norm": 1.0989367843754858, "learning_rate": 8.059625569382223e-05, "loss": 1.2969, "step": 9021 }, { "epoch": 4.245647058823529, "grad_norm": 0.8827765769287732, "learning_rate": 8.055739704183553e-05, "loss": 0.7344, "step": 9022 }, { "epoch": 4.246117647058823, "grad_norm": 1.3378387748751885, "learning_rate": 8.051854432058785e-05, "loss": 0.5938, "step": 9023 }, { "epoch": 4.2465882352941176, "grad_norm": 0.5726725519560133, "learning_rate": 8.047969753339717e-05, "loss": 0.4688, "step": 9024 }, { "epoch": 4.247058823529412, "grad_norm": 0.8359170982435427, "learning_rate": 8.04408566835814e-05, "loss": 0.2373, "step": 9025 }, { "epoch": 4.247529411764706, "grad_norm": 0.9648867731596366, "learning_rate": 8.040202177445761e-05, "loss": 0.7773, "step": 9026 }, { "epoch": 4.248, "grad_norm": 0.9445453412039206, "learning_rate": 8.036319280934251e-05, "loss": 0.7422, "step": 9027 }, { "epoch": 4.248470588235294, "grad_norm": 0.7955196019315163, "learning_rate": 8.032436979155237e-05, "loss": 0.2695, "step": 9028 }, { "epoch": 4.248941176470589, "grad_norm": 1.4962755968616313, "learning_rate": 8.02855527244028e-05, "loss": 1.7109, "step": 9029 }, { "epoch": 4.249411764705882, "grad_norm": 1.0977452398601966, "learning_rate": 8.024674161120892e-05, "loss": 0.582, "step": 9030 }, { "epoch": 4.249882352941176, "grad_norm": 0.7100315470136588, "learning_rate": 8.020793645528552e-05, "loss": 0.5547, "step": 9031 }, { "epoch": 4.2503529411764704, "grad_norm": 1.0409842568671281, "learning_rate": 8.016913725994661e-05, "loss": 0.6836, "step": 9032 }, { "epoch": 4.250823529411765, "grad_norm": 0.9924805279565166, "learning_rate": 8.01303440285059e-05, "loss": 0.3887, "step": 9033 }, { "epoch": 4.251294117647059, "grad_norm": 1.2933296494156095, "learning_rate": 8.009155676427659e-05, "loss": 1.0859, "step": 9034 }, { "epoch": 4.251764705882353, "grad_norm": 0.8232250950081516, "learning_rate": 8.005277547057114e-05, "loss": 0.5859, "step": 9035 }, { "epoch": 4.252235294117647, "grad_norm": 1.1438863242489683, "learning_rate": 8.001400015070178e-05, "loss": 0.6758, "step": 9036 }, { "epoch": 4.252705882352942, "grad_norm": 0.8228636444330204, "learning_rate": 7.997523080798014e-05, "loss": 0.3633, "step": 9037 }, { "epoch": 4.253176470588235, "grad_norm": 1.0798152355248962, "learning_rate": 7.993646744571719e-05, "loss": 0.9375, "step": 9038 }, { "epoch": 4.253647058823529, "grad_norm": 1.1366312526056508, "learning_rate": 7.98977100672236e-05, "loss": 0.7656, "step": 9039 }, { "epoch": 4.254117647058823, "grad_norm": 1.35021894686243, "learning_rate": 7.985895867580948e-05, "loss": 1.2266, "step": 9040 }, { "epoch": 4.254588235294118, "grad_norm": 0.9646453754687456, "learning_rate": 7.982021327478419e-05, "loss": 0.4883, "step": 9041 }, { "epoch": 4.255058823529412, "grad_norm": 1.3110660838829198, "learning_rate": 7.978147386745705e-05, "loss": 0.8398, "step": 9042 }, { "epoch": 4.255529411764706, "grad_norm": 1.1087336285130056, "learning_rate": 7.974274045713638e-05, "loss": 1.0156, "step": 9043 }, { "epoch": 4.256, "grad_norm": 1.192788124169519, "learning_rate": 7.97040130471303e-05, "loss": 0.9414, "step": 9044 }, { "epoch": 4.2564705882352944, "grad_norm": 1.045215465770891, "learning_rate": 7.966529164074636e-05, "loss": 0.6211, "step": 9045 }, { "epoch": 4.256941176470589, "grad_norm": 0.9711735271731777, "learning_rate": 7.962657624129147e-05, "loss": 0.6602, "step": 9046 }, { "epoch": 4.257411764705882, "grad_norm": 1.0325903546885238, "learning_rate": 7.958786685207212e-05, "loss": 0.8008, "step": 9047 }, { "epoch": 4.257882352941176, "grad_norm": 1.2513896542972298, "learning_rate": 7.954916347639438e-05, "loss": 1.3828, "step": 9048 }, { "epoch": 4.2583529411764705, "grad_norm": 1.2133533223193635, "learning_rate": 7.951046611756358e-05, "loss": 1.0703, "step": 9049 }, { "epoch": 4.258823529411765, "grad_norm": 0.6966354688896282, "learning_rate": 7.947177477888472e-05, "loss": 0.2129, "step": 9050 }, { "epoch": 4.259294117647059, "grad_norm": 1.3238283354151998, "learning_rate": 7.94330894636623e-05, "loss": 1.0703, "step": 9051 }, { "epoch": 4.259764705882353, "grad_norm": 0.8298892903219272, "learning_rate": 7.939441017520011e-05, "loss": 1.0469, "step": 9052 }, { "epoch": 4.260235294117647, "grad_norm": 1.1160047695400617, "learning_rate": 7.935573691680166e-05, "loss": 1.2656, "step": 9053 }, { "epoch": 4.260705882352942, "grad_norm": 1.0541625804758685, "learning_rate": 7.931706969176972e-05, "loss": 1.2266, "step": 9054 }, { "epoch": 4.261176470588235, "grad_norm": 1.2622338542111826, "learning_rate": 7.927840850340673e-05, "loss": 1.0156, "step": 9055 }, { "epoch": 4.261647058823529, "grad_norm": 0.41053541031628243, "learning_rate": 7.92397533550146e-05, "loss": 0.0532, "step": 9056 }, { "epoch": 4.262117647058823, "grad_norm": 0.7423672959602757, "learning_rate": 7.920110424989455e-05, "loss": 0.2002, "step": 9057 }, { "epoch": 4.262588235294118, "grad_norm": 1.1612265619286124, "learning_rate": 7.916246119134748e-05, "loss": 1.1406, "step": 9058 }, { "epoch": 4.263058823529412, "grad_norm": 1.0618352208863542, "learning_rate": 7.912382418267374e-05, "loss": 0.5586, "step": 9059 }, { "epoch": 4.263529411764706, "grad_norm": 0.5838958648816459, "learning_rate": 7.908519322717297e-05, "loss": 0.4746, "step": 9060 }, { "epoch": 4.264, "grad_norm": 1.0011894620090869, "learning_rate": 7.904656832814457e-05, "loss": 0.8164, "step": 9061 }, { "epoch": 4.2644705882352945, "grad_norm": 0.7395435884467386, "learning_rate": 7.900794948888728e-05, "loss": 0.293, "step": 9062 }, { "epoch": 4.264941176470588, "grad_norm": 1.0488580381296613, "learning_rate": 7.896933671269929e-05, "loss": 0.4316, "step": 9063 }, { "epoch": 4.265411764705882, "grad_norm": 0.9657356223609577, "learning_rate": 7.893073000287834e-05, "loss": 0.6133, "step": 9064 }, { "epoch": 4.265882352941176, "grad_norm": 1.3343536343993254, "learning_rate": 7.88921293627217e-05, "loss": 0.9922, "step": 9065 }, { "epoch": 4.2663529411764705, "grad_norm": 0.7023810077183703, "learning_rate": 7.885353479552592e-05, "loss": 0.1406, "step": 9066 }, { "epoch": 4.266823529411765, "grad_norm": 1.175299814925207, "learning_rate": 7.881494630458726e-05, "loss": 0.6055, "step": 9067 }, { "epoch": 4.267294117647059, "grad_norm": 0.8817594948023831, "learning_rate": 7.87763638932014e-05, "loss": 0.4277, "step": 9068 }, { "epoch": 4.267764705882353, "grad_norm": 0.5582453166897039, "learning_rate": 7.873778756466335e-05, "loss": 0.1426, "step": 9069 }, { "epoch": 4.268235294117647, "grad_norm": 0.8384198493032188, "learning_rate": 7.869921732226779e-05, "loss": 0.3008, "step": 9070 }, { "epoch": 4.268705882352942, "grad_norm": 0.5274279411375838, "learning_rate": 7.866065316930886e-05, "loss": 0.165, "step": 9071 }, { "epoch": 4.269176470588235, "grad_norm": 1.1303385025386419, "learning_rate": 7.862209510907999e-05, "loss": 1.2891, "step": 9072 }, { "epoch": 4.269647058823529, "grad_norm": 1.1535793515346662, "learning_rate": 7.858354314487432e-05, "loss": 0.8359, "step": 9073 }, { "epoch": 4.270117647058823, "grad_norm": 0.8780852026208286, "learning_rate": 7.854499727998442e-05, "loss": 0.3535, "step": 9074 }, { "epoch": 4.270588235294118, "grad_norm": 0.9710503790156612, "learning_rate": 7.850645751770217e-05, "loss": 0.9023, "step": 9075 }, { "epoch": 4.271058823529412, "grad_norm": 0.6690362513261674, "learning_rate": 7.846792386131917e-05, "loss": 0.6289, "step": 9076 }, { "epoch": 4.271529411764706, "grad_norm": 0.760572144376827, "learning_rate": 7.84293963141263e-05, "loss": 0.3281, "step": 9077 }, { "epoch": 4.272, "grad_norm": 0.9790485364636036, "learning_rate": 7.839087487941403e-05, "loss": 0.4492, "step": 9078 }, { "epoch": 4.2724705882352945, "grad_norm": 1.5936807139936073, "learning_rate": 7.835235956047234e-05, "loss": 0.9844, "step": 9079 }, { "epoch": 4.272941176470589, "grad_norm": 0.7864016910630679, "learning_rate": 7.83138503605905e-05, "loss": 1.3594, "step": 9080 }, { "epoch": 4.273411764705882, "grad_norm": 1.2491529791978158, "learning_rate": 7.827534728305747e-05, "loss": 1.2109, "step": 9081 }, { "epoch": 4.273882352941176, "grad_norm": 0.7208770110803369, "learning_rate": 7.823685033116166e-05, "loss": 0.1309, "step": 9082 }, { "epoch": 4.2743529411764705, "grad_norm": 0.9788204512692162, "learning_rate": 7.819835950819078e-05, "loss": 0.957, "step": 9083 }, { "epoch": 4.274823529411765, "grad_norm": 0.9547847918889455, "learning_rate": 7.815987481743215e-05, "loss": 0.8594, "step": 9084 }, { "epoch": 4.275294117647059, "grad_norm": 1.0062821518363227, "learning_rate": 7.812139626217266e-05, "loss": 0.375, "step": 9085 }, { "epoch": 4.275764705882353, "grad_norm": 0.5717845033868385, "learning_rate": 7.808292384569838e-05, "loss": 0.1177, "step": 9086 }, { "epoch": 4.276235294117647, "grad_norm": 4.736925164103272, "learning_rate": 7.804445757129525e-05, "loss": 0.6562, "step": 9087 }, { "epoch": 4.276705882352942, "grad_norm": 1.763601525417534, "learning_rate": 7.800599744224834e-05, "loss": 1.0859, "step": 9088 }, { "epoch": 4.277176470588235, "grad_norm": 1.0056585851748265, "learning_rate": 7.796754346184238e-05, "loss": 0.1963, "step": 9089 }, { "epoch": 4.277647058823529, "grad_norm": 0.9776345076832346, "learning_rate": 7.792909563336157e-05, "loss": 0.4688, "step": 9090 }, { "epoch": 4.278117647058823, "grad_norm": 1.0840877101521387, "learning_rate": 7.789065396008945e-05, "loss": 0.9492, "step": 9091 }, { "epoch": 4.278588235294118, "grad_norm": 0.9184237820973856, "learning_rate": 7.785221844530918e-05, "loss": 0.2559, "step": 9092 }, { "epoch": 4.279058823529412, "grad_norm": 1.2572439326893536, "learning_rate": 7.78137890923034e-05, "loss": 1.2344, "step": 9093 }, { "epoch": 4.279529411764706, "grad_norm": 1.144231229215965, "learning_rate": 7.777536590435404e-05, "loss": 0.5703, "step": 9094 }, { "epoch": 4.28, "grad_norm": 0.7625732963596707, "learning_rate": 7.773694888474267e-05, "loss": 0.2266, "step": 9095 }, { "epoch": 4.2804705882352945, "grad_norm": 0.9272287311643179, "learning_rate": 7.769853803675041e-05, "loss": 0.498, "step": 9096 }, { "epoch": 4.280941176470588, "grad_norm": 1.0092076150242715, "learning_rate": 7.766013336365755e-05, "loss": 0.8281, "step": 9097 }, { "epoch": 4.281411764705882, "grad_norm": 0.8473669514882364, "learning_rate": 7.762173486874414e-05, "loss": 0.7812, "step": 9098 }, { "epoch": 4.281882352941176, "grad_norm": 1.1164700013109687, "learning_rate": 7.758334255528963e-05, "loss": 0.5586, "step": 9099 }, { "epoch": 4.2823529411764705, "grad_norm": 0.9790498772334051, "learning_rate": 7.754495642657282e-05, "loss": 0.9023, "step": 9100 }, { "epoch": 4.282823529411765, "grad_norm": 1.0741953110398275, "learning_rate": 7.750657648587213e-05, "loss": 0.6836, "step": 9101 }, { "epoch": 4.283294117647059, "grad_norm": 1.0371873447629845, "learning_rate": 7.746820273646542e-05, "loss": 0.8242, "step": 9102 }, { "epoch": 4.283764705882353, "grad_norm": 0.7981324249984111, "learning_rate": 7.742983518162993e-05, "loss": 1.1172, "step": 9103 }, { "epoch": 4.284235294117647, "grad_norm": 1.2478807519376116, "learning_rate": 7.73914738246425e-05, "loss": 0.5586, "step": 9104 }, { "epoch": 4.284705882352942, "grad_norm": 0.7328523994891354, "learning_rate": 7.735311866877933e-05, "loss": 0.2793, "step": 9105 }, { "epoch": 4.285176470588235, "grad_norm": 0.8719719557056334, "learning_rate": 7.731476971731612e-05, "loss": 0.166, "step": 9106 }, { "epoch": 4.285647058823529, "grad_norm": 0.8724061948341227, "learning_rate": 7.727642697352817e-05, "loss": 0.6758, "step": 9107 }, { "epoch": 4.286117647058823, "grad_norm": 0.934839543041914, "learning_rate": 7.723809044069e-05, "loss": 0.8633, "step": 9108 }, { "epoch": 4.286588235294118, "grad_norm": 1.4103000538393482, "learning_rate": 7.719976012207579e-05, "loss": 0.9297, "step": 9109 }, { "epoch": 4.287058823529412, "grad_norm": 1.0790655947564312, "learning_rate": 7.716143602095924e-05, "loss": 0.6133, "step": 9110 }, { "epoch": 4.287529411764706, "grad_norm": 1.2083242692766019, "learning_rate": 7.712311814061323e-05, "loss": 0.4434, "step": 9111 }, { "epoch": 4.288, "grad_norm": 0.8722985061698197, "learning_rate": 7.708480648431043e-05, "loss": 0.4355, "step": 9112 }, { "epoch": 4.2884705882352945, "grad_norm": 0.578820319057857, "learning_rate": 7.704650105532284e-05, "loss": 0.1006, "step": 9113 }, { "epoch": 4.288941176470589, "grad_norm": 0.9446267030128022, "learning_rate": 7.700820185692184e-05, "loss": 0.9492, "step": 9114 }, { "epoch": 4.289411764705882, "grad_norm": 0.85060185680743, "learning_rate": 7.69699088923784e-05, "loss": 0.7383, "step": 9115 }, { "epoch": 4.289882352941176, "grad_norm": 1.1663501788967183, "learning_rate": 7.693162216496303e-05, "loss": 1.0469, "step": 9116 }, { "epoch": 4.2903529411764705, "grad_norm": 0.8552766150456679, "learning_rate": 7.689334167794547e-05, "loss": 0.2236, "step": 9117 }, { "epoch": 4.290823529411765, "grad_norm": 0.8563796934208193, "learning_rate": 7.685506743459511e-05, "loss": 0.2832, "step": 9118 }, { "epoch": 4.291294117647059, "grad_norm": 1.289152786669655, "learning_rate": 7.681679943818082e-05, "loss": 1.3359, "step": 9119 }, { "epoch": 4.291764705882353, "grad_norm": 1.0740821190346492, "learning_rate": 7.677853769197078e-05, "loss": 0.6992, "step": 9120 }, { "epoch": 4.292235294117647, "grad_norm": 0.7464523878688856, "learning_rate": 7.674028219923275e-05, "loss": 0.2812, "step": 9121 }, { "epoch": 4.292705882352941, "grad_norm": 0.8776961265320108, "learning_rate": 7.670203296323403e-05, "loss": 1.4219, "step": 9122 }, { "epoch": 4.293176470588235, "grad_norm": 0.7370459049124883, "learning_rate": 7.666378998724116e-05, "loss": 0.2148, "step": 9123 }, { "epoch": 4.293647058823529, "grad_norm": 1.0055322981674553, "learning_rate": 7.662555327452031e-05, "loss": 0.75, "step": 9124 }, { "epoch": 4.294117647058823, "grad_norm": 1.2294680699689287, "learning_rate": 7.65873228283372e-05, "loss": 1.5938, "step": 9125 }, { "epoch": 4.294588235294118, "grad_norm": 1.6566362623909061, "learning_rate": 7.654909865195675e-05, "loss": 1.6797, "step": 9126 }, { "epoch": 4.295058823529412, "grad_norm": 0.8262149757829746, "learning_rate": 7.65108807486436e-05, "loss": 0.3262, "step": 9127 }, { "epoch": 4.295529411764706, "grad_norm": 1.3953429251370248, "learning_rate": 7.647266912166162e-05, "loss": 1.5859, "step": 9128 }, { "epoch": 4.296, "grad_norm": 1.2792241249567073, "learning_rate": 7.643446377427438e-05, "loss": 0.8867, "step": 9129 }, { "epoch": 4.2964705882352945, "grad_norm": 0.7012601813722865, "learning_rate": 7.639626470974474e-05, "loss": 0.2334, "step": 9130 }, { "epoch": 4.296941176470588, "grad_norm": 1.1711119410632458, "learning_rate": 7.635807193133512e-05, "loss": 1.0391, "step": 9131 }, { "epoch": 4.297411764705882, "grad_norm": 0.8705826583562084, "learning_rate": 7.631988544230745e-05, "loss": 0.4297, "step": 9132 }, { "epoch": 4.297882352941176, "grad_norm": 0.8276656830653256, "learning_rate": 7.628170524592288e-05, "loss": 0.6719, "step": 9133 }, { "epoch": 4.2983529411764705, "grad_norm": 0.9439060796013908, "learning_rate": 7.624353134544229e-05, "loss": 0.875, "step": 9134 }, { "epoch": 4.298823529411765, "grad_norm": 1.0552994516428424, "learning_rate": 7.620536374412594e-05, "loss": 0.4766, "step": 9135 }, { "epoch": 4.299294117647059, "grad_norm": 1.5928458791548485, "learning_rate": 7.616720244523342e-05, "loss": 1.1953, "step": 9136 }, { "epoch": 4.299764705882353, "grad_norm": 1.3516915656253505, "learning_rate": 7.612904745202399e-05, "loss": 1.3828, "step": 9137 }, { "epoch": 4.300235294117647, "grad_norm": 1.2299942889281483, "learning_rate": 7.609089876775628e-05, "loss": 1.5625, "step": 9138 }, { "epoch": 4.300705882352942, "grad_norm": 1.060823931668094, "learning_rate": 7.605275639568828e-05, "loss": 0.373, "step": 9139 }, { "epoch": 4.301176470588235, "grad_norm": 0.76042381611171, "learning_rate": 7.60146203390776e-05, "loss": 0.2402, "step": 9140 }, { "epoch": 4.301647058823529, "grad_norm": 0.9980089732346121, "learning_rate": 7.59764906011813e-05, "loss": 0.9609, "step": 9141 }, { "epoch": 4.302117647058823, "grad_norm": 0.943738689045311, "learning_rate": 7.593836718525572e-05, "loss": 0.6055, "step": 9142 }, { "epoch": 4.302588235294118, "grad_norm": 1.0919133645152042, "learning_rate": 7.590025009455684e-05, "loss": 0.6992, "step": 9143 }, { "epoch": 4.303058823529412, "grad_norm": 0.9904889484922641, "learning_rate": 7.586213933234013e-05, "loss": 1.125, "step": 9144 }, { "epoch": 4.303529411764706, "grad_norm": 0.9272954356056174, "learning_rate": 7.58240349018603e-05, "loss": 0.5312, "step": 9145 }, { "epoch": 4.304, "grad_norm": 1.2455124721326443, "learning_rate": 7.578593680637173e-05, "loss": 0.6484, "step": 9146 }, { "epoch": 4.3044705882352945, "grad_norm": 0.8059272223014243, "learning_rate": 7.574784504912823e-05, "loss": 0.2275, "step": 9147 }, { "epoch": 4.304941176470588, "grad_norm": 1.0779973005976176, "learning_rate": 7.57097596333829e-05, "loss": 0.3574, "step": 9148 }, { "epoch": 4.305411764705882, "grad_norm": 0.9062223563267163, "learning_rate": 7.567168056238851e-05, "loss": 0.4316, "step": 9149 }, { "epoch": 4.305882352941176, "grad_norm": 1.1836267710179882, "learning_rate": 7.563360783939722e-05, "loss": 1.0391, "step": 9150 }, { "epoch": 4.3063529411764705, "grad_norm": 1.0286995249944428, "learning_rate": 7.559554146766054e-05, "loss": 0.5117, "step": 9151 }, { "epoch": 4.306823529411765, "grad_norm": 1.2069236068105464, "learning_rate": 7.555748145042957e-05, "loss": 0.8359, "step": 9152 }, { "epoch": 4.307294117647059, "grad_norm": 1.0229433738263367, "learning_rate": 7.551942779095486e-05, "loss": 0.5391, "step": 9153 }, { "epoch": 4.307764705882353, "grad_norm": 0.4029639160777665, "learning_rate": 7.54813804924863e-05, "loss": 0.3438, "step": 9154 }, { "epoch": 4.308235294117647, "grad_norm": 1.3179282943832709, "learning_rate": 7.544333955827342e-05, "loss": 1.2031, "step": 9155 }, { "epoch": 4.308705882352941, "grad_norm": 1.3495091453600823, "learning_rate": 7.540530499156496e-05, "loss": 0.9102, "step": 9156 }, { "epoch": 4.309176470588235, "grad_norm": 1.134196523818692, "learning_rate": 7.536727679560937e-05, "loss": 0.668, "step": 9157 }, { "epoch": 4.309647058823529, "grad_norm": 1.0543610060168451, "learning_rate": 7.532925497365444e-05, "loss": 0.8125, "step": 9158 }, { "epoch": 4.310117647058823, "grad_norm": 1.2003668754953352, "learning_rate": 7.529123952894734e-05, "loss": 0.7383, "step": 9159 }, { "epoch": 4.310588235294118, "grad_norm": 1.0542655192463615, "learning_rate": 7.525323046473485e-05, "loss": 0.6211, "step": 9160 }, { "epoch": 4.311058823529412, "grad_norm": 0.9226209553670701, "learning_rate": 7.521522778426315e-05, "loss": 0.4434, "step": 9161 }, { "epoch": 4.311529411764706, "grad_norm": 0.5915731532617495, "learning_rate": 7.517723149077776e-05, "loss": 0.1631, "step": 9162 }, { "epoch": 4.312, "grad_norm": 1.317783690298047, "learning_rate": 7.513924158752383e-05, "loss": 1.7422, "step": 9163 }, { "epoch": 4.3124705882352945, "grad_norm": 0.7439873550165353, "learning_rate": 7.510125807774589e-05, "loss": 0.3633, "step": 9164 }, { "epoch": 4.312941176470588, "grad_norm": 0.9886174270766707, "learning_rate": 7.506328096468782e-05, "loss": 1.2031, "step": 9165 }, { "epoch": 4.313411764705882, "grad_norm": 1.0641116606377943, "learning_rate": 7.502531025159313e-05, "loss": 0.832, "step": 9166 }, { "epoch": 4.313882352941176, "grad_norm": 0.9763495380459373, "learning_rate": 7.498734594170474e-05, "loss": 0.2949, "step": 9167 }, { "epoch": 4.3143529411764705, "grad_norm": 1.0210058725901894, "learning_rate": 7.49493880382649e-05, "loss": 1.0156, "step": 9168 }, { "epoch": 4.314823529411765, "grad_norm": 0.5550300238274307, "learning_rate": 7.491143654451544e-05, "loss": 0.1807, "step": 9169 }, { "epoch": 4.315294117647059, "grad_norm": 1.2397101715519045, "learning_rate": 7.487349146369766e-05, "loss": 1.0078, "step": 9170 }, { "epoch": 4.315764705882353, "grad_norm": 0.6023270307319327, "learning_rate": 7.483555279905215e-05, "loss": 0.3789, "step": 9171 }, { "epoch": 4.316235294117647, "grad_norm": 0.980916573545189, "learning_rate": 7.479762055381909e-05, "loss": 0.8789, "step": 9172 }, { "epoch": 4.316705882352942, "grad_norm": 0.7523081160997848, "learning_rate": 7.475969473123819e-05, "loss": 0.3496, "step": 9173 }, { "epoch": 4.317176470588235, "grad_norm": 0.44649990919268867, "learning_rate": 7.472177533454833e-05, "loss": 0.1035, "step": 9174 }, { "epoch": 4.317647058823529, "grad_norm": 1.4215790426774002, "learning_rate": 7.46838623669881e-05, "loss": 1.0469, "step": 9175 }, { "epoch": 4.318117647058823, "grad_norm": 0.7909219050236501, "learning_rate": 7.464595583179544e-05, "loss": 0.2334, "step": 9176 }, { "epoch": 4.318588235294118, "grad_norm": 1.4251922257117218, "learning_rate": 7.460805573220779e-05, "loss": 0.6758, "step": 9177 }, { "epoch": 4.319058823529412, "grad_norm": 1.0092358023949917, "learning_rate": 7.4570162071462e-05, "loss": 0.5117, "step": 9178 }, { "epoch": 4.319529411764706, "grad_norm": 1.1842906772101138, "learning_rate": 7.453227485279432e-05, "loss": 0.7344, "step": 9179 }, { "epoch": 4.32, "grad_norm": 1.142441536659276, "learning_rate": 7.449439407944052e-05, "loss": 1.1016, "step": 9180 }, { "epoch": 4.3204705882352945, "grad_norm": 0.7999140693432959, "learning_rate": 7.445651975463588e-05, "loss": 0.2773, "step": 9181 }, { "epoch": 4.320941176470588, "grad_norm": 2.2220034559625303, "learning_rate": 7.441865188161494e-05, "loss": 0.8984, "step": 9182 }, { "epoch": 4.321411764705882, "grad_norm": 0.899371233049772, "learning_rate": 7.43807904636119e-05, "loss": 0.4629, "step": 9183 }, { "epoch": 4.321882352941176, "grad_norm": 1.4626387617873178, "learning_rate": 7.434293550386021e-05, "loss": 1.3672, "step": 9184 }, { "epoch": 4.3223529411764705, "grad_norm": 0.9416351315069896, "learning_rate": 7.430508700559292e-05, "loss": 0.6172, "step": 9185 }, { "epoch": 4.322823529411765, "grad_norm": 0.7718049231197983, "learning_rate": 7.426724497204254e-05, "loss": 0.2021, "step": 9186 }, { "epoch": 4.323294117647059, "grad_norm": 0.8624551297241985, "learning_rate": 7.422940940644085e-05, "loss": 0.2695, "step": 9187 }, { "epoch": 4.323764705882353, "grad_norm": 1.1298640840306413, "learning_rate": 7.419158031201924e-05, "loss": 0.9883, "step": 9188 }, { "epoch": 4.324235294117647, "grad_norm": 1.188645654873683, "learning_rate": 7.415375769200859e-05, "loss": 0.6406, "step": 9189 }, { "epoch": 4.324705882352941, "grad_norm": 0.9222218373109542, "learning_rate": 7.411594154963897e-05, "loss": 0.4258, "step": 9190 }, { "epoch": 4.325176470588235, "grad_norm": 0.9747823522042165, "learning_rate": 7.407813188814017e-05, "loss": 0.7383, "step": 9191 }, { "epoch": 4.325647058823529, "grad_norm": 0.9745837214276941, "learning_rate": 7.404032871074135e-05, "loss": 0.2891, "step": 9192 }, { "epoch": 4.326117647058823, "grad_norm": 1.1354233643025629, "learning_rate": 7.400253202067101e-05, "loss": 1.1953, "step": 9193 }, { "epoch": 4.326588235294118, "grad_norm": 1.1920201577523246, "learning_rate": 7.396474182115717e-05, "loss": 0.9531, "step": 9194 }, { "epoch": 4.327058823529412, "grad_norm": 1.1023031085100914, "learning_rate": 7.392695811542739e-05, "loss": 0.6719, "step": 9195 }, { "epoch": 4.327529411764706, "grad_norm": 1.0648544195963603, "learning_rate": 7.388918090670847e-05, "loss": 0.7812, "step": 9196 }, { "epoch": 4.328, "grad_norm": 1.242624886151853, "learning_rate": 7.385141019822683e-05, "loss": 1.4531, "step": 9197 }, { "epoch": 4.3284705882352945, "grad_norm": 0.9458869100900751, "learning_rate": 7.381364599320831e-05, "loss": 0.5273, "step": 9198 }, { "epoch": 4.328941176470588, "grad_norm": 0.928698763836304, "learning_rate": 7.377588829487806e-05, "loss": 0.7969, "step": 9199 }, { "epoch": 4.329411764705882, "grad_norm": 1.1115383654618411, "learning_rate": 7.373813710646083e-05, "loss": 0.6445, "step": 9200 }, { "epoch": 4.329882352941176, "grad_norm": 1.1236391343049754, "learning_rate": 7.370039243118082e-05, "loss": 0.4766, "step": 9201 }, { "epoch": 4.3303529411764705, "grad_norm": 0.369939127648161, "learning_rate": 7.366265427226148e-05, "loss": 0.0771, "step": 9202 }, { "epoch": 4.330823529411765, "grad_norm": 0.67774784198328, "learning_rate": 7.36249226329259e-05, "loss": 0.1865, "step": 9203 }, { "epoch": 4.331294117647059, "grad_norm": 0.968346607352328, "learning_rate": 7.358719751639659e-05, "loss": 0.3848, "step": 9204 }, { "epoch": 4.331764705882353, "grad_norm": 1.1843110239775962, "learning_rate": 7.354947892589536e-05, "loss": 0.7109, "step": 9205 }, { "epoch": 4.332235294117647, "grad_norm": 1.143723737076886, "learning_rate": 7.351176686464368e-05, "loss": 0.8164, "step": 9206 }, { "epoch": 4.332705882352942, "grad_norm": 1.1619898413191718, "learning_rate": 7.347406133586223e-05, "loss": 0.7227, "step": 9207 }, { "epoch": 4.333176470588235, "grad_norm": 0.5254765335305827, "learning_rate": 7.34363623427713e-05, "loss": 0.1348, "step": 9208 }, { "epoch": 4.333647058823529, "grad_norm": 1.2418076731339165, "learning_rate": 7.339866988859061e-05, "loss": 1.0625, "step": 9209 }, { "epoch": 4.334117647058823, "grad_norm": 1.0625399542197846, "learning_rate": 7.33609839765392e-05, "loss": 0.4824, "step": 9210 }, { "epoch": 4.334588235294118, "grad_norm": 1.3250327509199105, "learning_rate": 7.332330460983565e-05, "loss": 0.6094, "step": 9211 }, { "epoch": 4.335058823529412, "grad_norm": 0.899125202122191, "learning_rate": 7.328563179169806e-05, "loss": 1.4688, "step": 9212 }, { "epoch": 4.335529411764706, "grad_norm": 1.689422846886229, "learning_rate": 7.324796552534373e-05, "loss": 0.6758, "step": 9213 }, { "epoch": 4.336, "grad_norm": 0.9311679629117171, "learning_rate": 7.321030581398961e-05, "loss": 0.5039, "step": 9214 }, { "epoch": 4.3364705882352945, "grad_norm": 2.196618364377388, "learning_rate": 7.317265266085209e-05, "loss": 1.5547, "step": 9215 }, { "epoch": 4.336941176470588, "grad_norm": 0.7355537369541663, "learning_rate": 7.31350060691468e-05, "loss": 0.2539, "step": 9216 }, { "epoch": 4.337411764705882, "grad_norm": 0.9616270708946955, "learning_rate": 7.309736604208902e-05, "loss": 0.7031, "step": 9217 }, { "epoch": 4.337882352941176, "grad_norm": 0.9195116392921238, "learning_rate": 7.305973258289343e-05, "loss": 0.7578, "step": 9218 }, { "epoch": 4.3383529411764705, "grad_norm": 1.7280517161069453, "learning_rate": 7.302210569477397e-05, "loss": 1.3594, "step": 9219 }, { "epoch": 4.338823529411765, "grad_norm": 1.4246999640572842, "learning_rate": 7.298448538094436e-05, "loss": 0.7422, "step": 9220 }, { "epoch": 4.339294117647059, "grad_norm": 1.115411811152956, "learning_rate": 7.294687164461741e-05, "loss": 0.6836, "step": 9221 }, { "epoch": 4.339764705882353, "grad_norm": 0.8037121107607259, "learning_rate": 7.290926448900557e-05, "loss": 0.6953, "step": 9222 }, { "epoch": 4.340235294117647, "grad_norm": 1.0386966574031804, "learning_rate": 7.287166391732071e-05, "loss": 1.1094, "step": 9223 }, { "epoch": 4.340705882352941, "grad_norm": 0.9526950845494807, "learning_rate": 7.283406993277401e-05, "loss": 0.4199, "step": 9224 }, { "epoch": 4.341176470588235, "grad_norm": 1.095337553420702, "learning_rate": 7.279648253857626e-05, "loss": 1.2734, "step": 9225 }, { "epoch": 4.341647058823529, "grad_norm": 1.062868746230188, "learning_rate": 7.275890173793761e-05, "loss": 1.0703, "step": 9226 }, { "epoch": 4.342117647058823, "grad_norm": 0.7958833803900448, "learning_rate": 7.272132753406757e-05, "loss": 1.0703, "step": 9227 }, { "epoch": 4.342588235294118, "grad_norm": 1.1546066736281744, "learning_rate": 7.26837599301752e-05, "loss": 0.9414, "step": 9228 }, { "epoch": 4.343058823529412, "grad_norm": 1.0351164978576208, "learning_rate": 7.264619892946904e-05, "loss": 1.0625, "step": 9229 }, { "epoch": 4.343529411764706, "grad_norm": 1.1645688100778888, "learning_rate": 7.260864453515683e-05, "loss": 0.7188, "step": 9230 }, { "epoch": 4.344, "grad_norm": 2.3708500202270213, "learning_rate": 7.2571096750446e-05, "loss": 0.8242, "step": 9231 }, { "epoch": 4.3444705882352945, "grad_norm": 1.0732955565205824, "learning_rate": 7.253355557854334e-05, "loss": 1.125, "step": 9232 }, { "epoch": 4.344941176470588, "grad_norm": 1.8041742011371438, "learning_rate": 7.249602102265496e-05, "loss": 1.0312, "step": 9233 }, { "epoch": 4.345411764705882, "grad_norm": 0.9015488340625516, "learning_rate": 7.245849308598658e-05, "loss": 0.2891, "step": 9234 }, { "epoch": 4.345882352941176, "grad_norm": 0.8568082494466803, "learning_rate": 7.242097177174318e-05, "loss": 0.4863, "step": 9235 }, { "epoch": 4.3463529411764705, "grad_norm": 1.1409621796377731, "learning_rate": 7.238345708312931e-05, "loss": 0.9766, "step": 9236 }, { "epoch": 4.346823529411765, "grad_norm": 1.0177728974427007, "learning_rate": 7.234594902334897e-05, "loss": 0.8867, "step": 9237 }, { "epoch": 4.347294117647059, "grad_norm": 0.6467166951883092, "learning_rate": 7.230844759560544e-05, "loss": 0.4102, "step": 9238 }, { "epoch": 4.347764705882353, "grad_norm": 1.28544569769154, "learning_rate": 7.227095280310152e-05, "loss": 1.4141, "step": 9239 }, { "epoch": 4.348235294117647, "grad_norm": 1.1967027854857053, "learning_rate": 7.223346464903958e-05, "loss": 1.1953, "step": 9240 }, { "epoch": 4.348705882352942, "grad_norm": 1.2054757692627962, "learning_rate": 7.219598313662113e-05, "loss": 0.5859, "step": 9241 }, { "epoch": 4.349176470588235, "grad_norm": 0.9130031266184971, "learning_rate": 7.215850826904734e-05, "loss": 1.25, "step": 9242 }, { "epoch": 4.349647058823529, "grad_norm": 0.8350426719386511, "learning_rate": 7.212104004951883e-05, "loss": 1.0156, "step": 9243 }, { "epoch": 4.350117647058823, "grad_norm": 0.8803855779891757, "learning_rate": 7.208357848123543e-05, "loss": 0.5664, "step": 9244 }, { "epoch": 4.350588235294118, "grad_norm": 0.8308069171978213, "learning_rate": 7.204612356739658e-05, "loss": 0.707, "step": 9245 }, { "epoch": 4.351058823529412, "grad_norm": 1.0488662551981474, "learning_rate": 7.200867531120124e-05, "loss": 0.6523, "step": 9246 }, { "epoch": 4.351529411764706, "grad_norm": 0.48238509652387596, "learning_rate": 7.197123371584749e-05, "loss": 0.0879, "step": 9247 }, { "epoch": 4.352, "grad_norm": 0.9435267187522325, "learning_rate": 7.193379878453312e-05, "loss": 0.5234, "step": 9248 }, { "epoch": 4.352470588235294, "grad_norm": 1.3014834644472264, "learning_rate": 7.189637052045529e-05, "loss": 0.6797, "step": 9249 }, { "epoch": 4.352941176470588, "grad_norm": 1.029327638135908, "learning_rate": 7.185894892681048e-05, "loss": 0.8828, "step": 9250 }, { "epoch": 4.353411764705882, "grad_norm": 1.4008170289068076, "learning_rate": 7.182153400679468e-05, "loss": 1.9609, "step": 9251 }, { "epoch": 4.353882352941176, "grad_norm": 1.1406200482144375, "learning_rate": 7.17841257636034e-05, "loss": 1.3359, "step": 9252 }, { "epoch": 4.3543529411764705, "grad_norm": 0.7028876438480002, "learning_rate": 7.17467242004314e-05, "loss": 0.2285, "step": 9253 }, { "epoch": 4.354823529411765, "grad_norm": 0.7648996747754663, "learning_rate": 7.170932932047296e-05, "loss": 0.4297, "step": 9254 }, { "epoch": 4.355294117647059, "grad_norm": 1.338863471362973, "learning_rate": 7.167194112692186e-05, "loss": 1.2266, "step": 9255 }, { "epoch": 4.355764705882353, "grad_norm": 1.4205375434900829, "learning_rate": 7.163455962297116e-05, "loss": 0.6797, "step": 9256 }, { "epoch": 4.356235294117647, "grad_norm": 1.4261780393314143, "learning_rate": 7.159718481181346e-05, "loss": 0.9336, "step": 9257 }, { "epoch": 4.356705882352941, "grad_norm": 1.2974021613947688, "learning_rate": 7.155981669664072e-05, "loss": 0.5469, "step": 9258 }, { "epoch": 4.357176470588235, "grad_norm": 1.0879316690369272, "learning_rate": 7.152245528064437e-05, "loss": 0.8047, "step": 9259 }, { "epoch": 4.357647058823529, "grad_norm": 1.813506732529234, "learning_rate": 7.148510056701532e-05, "loss": 0.9141, "step": 9260 }, { "epoch": 4.358117647058823, "grad_norm": 1.1699529520352512, "learning_rate": 7.144775255894375e-05, "loss": 0.793, "step": 9261 }, { "epoch": 4.358588235294118, "grad_norm": 1.2551530968735913, "learning_rate": 7.14104112596194e-05, "loss": 1.3203, "step": 9262 }, { "epoch": 4.359058823529412, "grad_norm": 1.288597679915873, "learning_rate": 7.137307667223141e-05, "loss": 0.5508, "step": 9263 }, { "epoch": 4.359529411764706, "grad_norm": 1.1962611887368635, "learning_rate": 7.133574879996833e-05, "loss": 0.9609, "step": 9264 }, { "epoch": 4.36, "grad_norm": 1.354603248028014, "learning_rate": 7.129842764601821e-05, "loss": 1.2266, "step": 9265 }, { "epoch": 4.3604705882352945, "grad_norm": 0.9495908352916521, "learning_rate": 7.126111321356834e-05, "loss": 1.1484, "step": 9266 }, { "epoch": 4.360941176470588, "grad_norm": 0.7899318314612198, "learning_rate": 7.122380550580562e-05, "loss": 0.3496, "step": 9267 }, { "epoch": 4.361411764705882, "grad_norm": 1.0722371428225044, "learning_rate": 7.118650452591634e-05, "loss": 0.5977, "step": 9268 }, { "epoch": 4.361882352941176, "grad_norm": 0.9213211821375883, "learning_rate": 7.114921027708613e-05, "loss": 0.5469, "step": 9269 }, { "epoch": 4.3623529411764705, "grad_norm": 1.178886760777981, "learning_rate": 7.111192276250011e-05, "loss": 0.8672, "step": 9270 }, { "epoch": 4.362823529411765, "grad_norm": 1.4823870908042625, "learning_rate": 7.107464198534287e-05, "loss": 0.9531, "step": 9271 }, { "epoch": 4.363294117647059, "grad_norm": 1.2529825512696102, "learning_rate": 7.103736794879829e-05, "loss": 0.4355, "step": 9272 }, { "epoch": 4.363764705882353, "grad_norm": 0.788285397396213, "learning_rate": 7.100010065604978e-05, "loss": 0.3496, "step": 9273 }, { "epoch": 4.364235294117647, "grad_norm": 0.9159388011210197, "learning_rate": 7.096284011028024e-05, "loss": 0.4844, "step": 9274 }, { "epoch": 4.364705882352941, "grad_norm": 0.9922992038618168, "learning_rate": 7.092558631467177e-05, "loss": 0.3906, "step": 9275 }, { "epoch": 4.365176470588235, "grad_norm": 0.7928807931538565, "learning_rate": 7.088833927240608e-05, "loss": 0.4629, "step": 9276 }, { "epoch": 4.365647058823529, "grad_norm": 1.3985361644909335, "learning_rate": 7.085109898666433e-05, "loss": 0.6016, "step": 9277 }, { "epoch": 4.366117647058823, "grad_norm": 0.9564269110097783, "learning_rate": 7.081386546062687e-05, "loss": 0.2471, "step": 9278 }, { "epoch": 4.366588235294118, "grad_norm": 0.9857517530143132, "learning_rate": 7.077663869747372e-05, "loss": 0.8984, "step": 9279 }, { "epoch": 4.367058823529412, "grad_norm": 0.7650903037777592, "learning_rate": 7.073941870038427e-05, "loss": 0.6055, "step": 9280 }, { "epoch": 4.367529411764706, "grad_norm": 0.9548479333970931, "learning_rate": 7.070220547253717e-05, "loss": 0.3848, "step": 9281 }, { "epoch": 4.368, "grad_norm": 1.1424637641839719, "learning_rate": 7.066499901711068e-05, "loss": 0.4141, "step": 9282 }, { "epoch": 4.368470588235294, "grad_norm": 0.9888263959589559, "learning_rate": 7.062779933728247e-05, "loss": 0.5664, "step": 9283 }, { "epoch": 4.368941176470588, "grad_norm": 1.2258222634753417, "learning_rate": 7.059060643622944e-05, "loss": 0.875, "step": 9284 }, { "epoch": 4.369411764705882, "grad_norm": 1.3875230304834527, "learning_rate": 7.05534203171282e-05, "loss": 1.2734, "step": 9285 }, { "epoch": 4.369882352941176, "grad_norm": 1.2062270338070074, "learning_rate": 7.051624098315449e-05, "loss": 0.7578, "step": 9286 }, { "epoch": 4.3703529411764706, "grad_norm": 0.6060608027540958, "learning_rate": 7.047906843748365e-05, "loss": 0.1543, "step": 9287 }, { "epoch": 4.370823529411765, "grad_norm": 1.123340307258744, "learning_rate": 7.044190268329049e-05, "loss": 0.8555, "step": 9288 }, { "epoch": 4.371294117647059, "grad_norm": 0.4192631727184428, "learning_rate": 7.0404743723749e-05, "loss": 0.0708, "step": 9289 }, { "epoch": 4.371764705882353, "grad_norm": 1.1002636691560361, "learning_rate": 7.036759156203281e-05, "loss": 0.7109, "step": 9290 }, { "epoch": 4.372235294117647, "grad_norm": 0.8520891557453173, "learning_rate": 7.033044620131495e-05, "loss": 0.3047, "step": 9291 }, { "epoch": 4.372705882352941, "grad_norm": 0.8346728593801053, "learning_rate": 7.02933076447677e-05, "loss": 0.2793, "step": 9292 }, { "epoch": 4.373176470588235, "grad_norm": 1.06648523241271, "learning_rate": 7.025617589556297e-05, "loss": 0.793, "step": 9293 }, { "epoch": 4.373647058823529, "grad_norm": 1.1203656998744869, "learning_rate": 7.021905095687198e-05, "loss": 0.9062, "step": 9294 }, { "epoch": 4.374117647058823, "grad_norm": 1.1816380600820315, "learning_rate": 7.018193283186531e-05, "loss": 1.0859, "step": 9295 }, { "epoch": 4.374588235294118, "grad_norm": 0.5812856140061093, "learning_rate": 7.01448215237131e-05, "loss": 0.1108, "step": 9296 }, { "epoch": 4.375058823529412, "grad_norm": 1.3123439624270385, "learning_rate": 7.010771703558488e-05, "loss": 0.7773, "step": 9297 }, { "epoch": 4.375529411764706, "grad_norm": 1.1024531694440378, "learning_rate": 7.007061937064941e-05, "loss": 0.6719, "step": 9298 }, { "epoch": 4.376, "grad_norm": 1.0968198796267612, "learning_rate": 7.00335285320751e-05, "loss": 0.6094, "step": 9299 }, { "epoch": 4.376470588235295, "grad_norm": 0.9725923209191267, "learning_rate": 6.999644452302975e-05, "loss": 0.7188, "step": 9300 }, { "epoch": 4.376941176470588, "grad_norm": 0.989482806400012, "learning_rate": 6.995936734668042e-05, "loss": 0.7305, "step": 9301 }, { "epoch": 4.377411764705882, "grad_norm": 1.3861504301153478, "learning_rate": 6.992229700619369e-05, "loss": 1.5312, "step": 9302 }, { "epoch": 4.377882352941176, "grad_norm": 1.1088957262236745, "learning_rate": 6.988523350473563e-05, "loss": 0.9219, "step": 9303 }, { "epoch": 4.378352941176471, "grad_norm": 1.2387417205158213, "learning_rate": 6.984817684547155e-05, "loss": 2.0938, "step": 9304 }, { "epoch": 4.378823529411765, "grad_norm": 1.3160722479331886, "learning_rate": 6.981112703156628e-05, "loss": 1.0938, "step": 9305 }, { "epoch": 4.379294117647059, "grad_norm": 0.7027923086903218, "learning_rate": 6.977408406618418e-05, "loss": 0.2578, "step": 9306 }, { "epoch": 4.379764705882353, "grad_norm": 1.3095629675607625, "learning_rate": 6.973704795248871e-05, "loss": 1.1406, "step": 9307 }, { "epoch": 4.3802352941176474, "grad_norm": 1.1123848003692087, "learning_rate": 6.970001869364306e-05, "loss": 1.3359, "step": 9308 }, { "epoch": 4.380705882352941, "grad_norm": 0.6500595994246023, "learning_rate": 6.966299629280968e-05, "loss": 0.1904, "step": 9309 }, { "epoch": 4.381176470588235, "grad_norm": 1.105575852847113, "learning_rate": 6.962598075315046e-05, "loss": 0.582, "step": 9310 }, { "epoch": 4.381647058823529, "grad_norm": 0.5706016345598481, "learning_rate": 6.958897207782677e-05, "loss": 0.2539, "step": 9311 }, { "epoch": 4.3821176470588235, "grad_norm": 0.6883219348477265, "learning_rate": 6.955197026999925e-05, "loss": 0.3027, "step": 9312 }, { "epoch": 4.382588235294118, "grad_norm": 1.0517337132458466, "learning_rate": 6.951497533282809e-05, "loss": 0.6914, "step": 9313 }, { "epoch": 4.383058823529412, "grad_norm": 1.1312580960525689, "learning_rate": 6.947798726947277e-05, "loss": 0.5586, "step": 9314 }, { "epoch": 4.383529411764706, "grad_norm": 0.9864301655369618, "learning_rate": 6.94410060830923e-05, "loss": 0.5352, "step": 9315 }, { "epoch": 4.384, "grad_norm": 0.9405531883789932, "learning_rate": 6.940403177684509e-05, "loss": 1.0469, "step": 9316 }, { "epoch": 4.384470588235294, "grad_norm": 0.7925684768300557, "learning_rate": 6.936706435388882e-05, "loss": 0.2969, "step": 9317 }, { "epoch": 4.384941176470588, "grad_norm": 1.171520148676119, "learning_rate": 6.93301038173808e-05, "loss": 1.1719, "step": 9318 }, { "epoch": 4.385411764705882, "grad_norm": 2.251733507346521, "learning_rate": 6.929315017047762e-05, "loss": 2.0, "step": 9319 }, { "epoch": 4.385882352941176, "grad_norm": 1.0440258861428637, "learning_rate": 6.925620341633522e-05, "loss": 1.0781, "step": 9320 }, { "epoch": 4.386352941176471, "grad_norm": 1.1070269125767203, "learning_rate": 6.921926355810911e-05, "loss": 0.6641, "step": 9321 }, { "epoch": 4.386823529411765, "grad_norm": 1.1273724810786097, "learning_rate": 6.918233059895418e-05, "loss": 0.9023, "step": 9322 }, { "epoch": 4.387294117647059, "grad_norm": 1.4459186049710155, "learning_rate": 6.914540454202457e-05, "loss": 1.5703, "step": 9323 }, { "epoch": 4.387764705882353, "grad_norm": 1.1554470110101085, "learning_rate": 6.910848539047399e-05, "loss": 0.8008, "step": 9324 }, { "epoch": 4.3882352941176475, "grad_norm": 0.8437043861361754, "learning_rate": 6.907157314745558e-05, "loss": 0.2236, "step": 9325 }, { "epoch": 4.388705882352941, "grad_norm": 1.0420706906506911, "learning_rate": 6.903466781612173e-05, "loss": 0.9844, "step": 9326 }, { "epoch": 4.389176470588235, "grad_norm": 1.1558933397355553, "learning_rate": 6.899776939962438e-05, "loss": 0.8672, "step": 9327 }, { "epoch": 4.389647058823529, "grad_norm": 0.8811237857687023, "learning_rate": 6.89608779011149e-05, "loss": 1.0703, "step": 9328 }, { "epoch": 4.3901176470588235, "grad_norm": 0.9976373529632512, "learning_rate": 6.892399332374387e-05, "loss": 0.3555, "step": 9329 }, { "epoch": 4.390588235294118, "grad_norm": 1.5113728410595273, "learning_rate": 6.888711567066149e-05, "loss": 1.2812, "step": 9330 }, { "epoch": 4.391058823529412, "grad_norm": 0.9885555987637252, "learning_rate": 6.885024494501735e-05, "loss": 0.6055, "step": 9331 }, { "epoch": 4.391529411764706, "grad_norm": 0.6447653588743564, "learning_rate": 6.881338114996027e-05, "loss": 0.1934, "step": 9332 }, { "epoch": 4.392, "grad_norm": 1.5959555660201785, "learning_rate": 6.877652428863866e-05, "loss": 1.4375, "step": 9333 }, { "epoch": 4.392470588235295, "grad_norm": 0.9172996114671034, "learning_rate": 6.873967436420033e-05, "loss": 0.6836, "step": 9334 }, { "epoch": 4.392941176470588, "grad_norm": 0.6426700572934739, "learning_rate": 6.870283137979233e-05, "loss": 0.5781, "step": 9335 }, { "epoch": 4.393411764705882, "grad_norm": 0.8844032085315802, "learning_rate": 6.866599533856135e-05, "loss": 0.6797, "step": 9336 }, { "epoch": 4.393882352941176, "grad_norm": 0.9000576205429253, "learning_rate": 6.862916624365325e-05, "loss": 0.4238, "step": 9337 }, { "epoch": 4.394352941176471, "grad_norm": 1.2511473924044718, "learning_rate": 6.859234409821347e-05, "loss": 1.0156, "step": 9338 }, { "epoch": 4.394823529411765, "grad_norm": 1.1568646151102417, "learning_rate": 6.855552890538688e-05, "loss": 0.5547, "step": 9339 }, { "epoch": 4.395294117647059, "grad_norm": 0.9662203454885095, "learning_rate": 6.851872066831751e-05, "loss": 0.8828, "step": 9340 }, { "epoch": 4.395764705882353, "grad_norm": 0.7524254012695514, "learning_rate": 6.848191939014909e-05, "loss": 0.6992, "step": 9341 }, { "epoch": 4.3962352941176475, "grad_norm": 1.8226451441685305, "learning_rate": 6.844512507402465e-05, "loss": 1.4766, "step": 9342 }, { "epoch": 4.396705882352941, "grad_norm": 1.3326234188875972, "learning_rate": 6.84083377230865e-05, "loss": 1.0312, "step": 9343 }, { "epoch": 4.397176470588235, "grad_norm": 0.6463234572762424, "learning_rate": 6.837155734047652e-05, "loss": 0.4727, "step": 9344 }, { "epoch": 4.397647058823529, "grad_norm": 1.542068115450802, "learning_rate": 6.833478392933597e-05, "loss": 0.9141, "step": 9345 }, { "epoch": 4.3981176470588235, "grad_norm": 1.3305113060943554, "learning_rate": 6.829801749280537e-05, "loss": 1.5312, "step": 9346 }, { "epoch": 4.398588235294118, "grad_norm": 1.2180724426905065, "learning_rate": 6.826125803402484e-05, "loss": 0.8594, "step": 9347 }, { "epoch": 4.399058823529412, "grad_norm": 0.8422067859660968, "learning_rate": 6.822450555613385e-05, "loss": 0.2471, "step": 9348 }, { "epoch": 4.399529411764706, "grad_norm": 1.314245569356504, "learning_rate": 6.818776006227114e-05, "loss": 1.1328, "step": 9349 }, { "epoch": 4.4, "grad_norm": 0.9254631275785138, "learning_rate": 6.815102155557501e-05, "loss": 0.6758, "step": 9350 }, { "epoch": 4.400470588235294, "grad_norm": 0.9255609727111355, "learning_rate": 6.811429003918317e-05, "loss": 0.459, "step": 9351 }, { "epoch": 4.400941176470588, "grad_norm": 0.9203530902021979, "learning_rate": 6.807756551623248e-05, "loss": 0.377, "step": 9352 }, { "epoch": 4.401411764705882, "grad_norm": 1.2755394878379405, "learning_rate": 6.804084798985964e-05, "loss": 0.9688, "step": 9353 }, { "epoch": 4.401882352941176, "grad_norm": 1.1537895578408708, "learning_rate": 6.800413746320034e-05, "loss": 1.1328, "step": 9354 }, { "epoch": 4.402352941176471, "grad_norm": 0.9701930067365926, "learning_rate": 6.796743393938987e-05, "loss": 0.3398, "step": 9355 }, { "epoch": 4.402823529411765, "grad_norm": 0.8779385832968299, "learning_rate": 6.793073742156298e-05, "loss": 0.4043, "step": 9356 }, { "epoch": 4.403294117647059, "grad_norm": 0.6748429145507098, "learning_rate": 6.78940479128536e-05, "loss": 0.166, "step": 9357 }, { "epoch": 4.403764705882353, "grad_norm": 1.0153872086474325, "learning_rate": 6.785736541639526e-05, "loss": 0.7656, "step": 9358 }, { "epoch": 4.4042352941176475, "grad_norm": 0.761469679481906, "learning_rate": 6.782068993532085e-05, "loss": 0.582, "step": 9359 }, { "epoch": 4.404705882352941, "grad_norm": 1.0504121372429147, "learning_rate": 6.778402147276259e-05, "loss": 0.875, "step": 9360 }, { "epoch": 4.405176470588235, "grad_norm": 0.8021443164777684, "learning_rate": 6.774736003185215e-05, "loss": 0.4805, "step": 9361 }, { "epoch": 4.405647058823529, "grad_norm": 0.319050632176454, "learning_rate": 6.771070561572067e-05, "loss": 0.05, "step": 9362 }, { "epoch": 4.4061176470588235, "grad_norm": 1.3490678811229961, "learning_rate": 6.767405822749848e-05, "loss": 1.1641, "step": 9363 }, { "epoch": 4.406588235294118, "grad_norm": 0.8801648950686102, "learning_rate": 6.76374178703156e-05, "loss": 0.2656, "step": 9364 }, { "epoch": 4.407058823529412, "grad_norm": 0.9779458562904664, "learning_rate": 6.760078454730115e-05, "loss": 0.6523, "step": 9365 }, { "epoch": 4.407529411764706, "grad_norm": 1.0859906889780124, "learning_rate": 6.756415826158388e-05, "loss": 0.8945, "step": 9366 }, { "epoch": 4.408, "grad_norm": 1.004334612311986, "learning_rate": 6.752753901629189e-05, "loss": 0.8633, "step": 9367 }, { "epoch": 4.408470588235295, "grad_norm": 1.1169522668122305, "learning_rate": 6.749092681455255e-05, "loss": 0.8516, "step": 9368 }, { "epoch": 4.408941176470588, "grad_norm": 0.5232033333527518, "learning_rate": 6.745432165949274e-05, "loss": 0.1089, "step": 9369 }, { "epoch": 4.409411764705882, "grad_norm": 0.849903205764083, "learning_rate": 6.741772355423882e-05, "loss": 0.7617, "step": 9370 }, { "epoch": 4.409882352941176, "grad_norm": 1.1821312722799584, "learning_rate": 6.73811325019163e-05, "loss": 0.6602, "step": 9371 }, { "epoch": 4.410352941176471, "grad_norm": 1.023110747367104, "learning_rate": 6.734454850565031e-05, "loss": 0.8984, "step": 9372 }, { "epoch": 4.410823529411765, "grad_norm": 1.3559479518830695, "learning_rate": 6.730797156856537e-05, "loss": 0.3516, "step": 9373 }, { "epoch": 4.411294117647059, "grad_norm": 1.7936263758047573, "learning_rate": 6.727140169378521e-05, "loss": 1.125, "step": 9374 }, { "epoch": 4.411764705882353, "grad_norm": 1.5742919861566522, "learning_rate": 6.72348388844331e-05, "loss": 1.3281, "step": 9375 }, { "epoch": 4.412235294117647, "grad_norm": 0.8104645398333389, "learning_rate": 6.719828314363178e-05, "loss": 0.3652, "step": 9376 }, { "epoch": 4.412705882352941, "grad_norm": 0.9297028753266856, "learning_rate": 6.716173447450317e-05, "loss": 0.4863, "step": 9377 }, { "epoch": 4.413176470588235, "grad_norm": 1.7872332836072231, "learning_rate": 6.712519288016875e-05, "loss": 1.0234, "step": 9378 }, { "epoch": 4.413647058823529, "grad_norm": 0.9561120836989576, "learning_rate": 6.70886583637494e-05, "loss": 0.4297, "step": 9379 }, { "epoch": 4.4141176470588235, "grad_norm": 1.0430115091790824, "learning_rate": 6.705213092836527e-05, "loss": 0.3809, "step": 9380 }, { "epoch": 4.414588235294118, "grad_norm": 0.763605248654153, "learning_rate": 6.701561057713599e-05, "loss": 0.3184, "step": 9381 }, { "epoch": 4.415058823529412, "grad_norm": 0.9215457186550299, "learning_rate": 6.697909731318069e-05, "loss": 0.1641, "step": 9382 }, { "epoch": 4.415529411764706, "grad_norm": 0.958471263558013, "learning_rate": 6.694259113961761e-05, "loss": 0.8594, "step": 9383 }, { "epoch": 4.416, "grad_norm": 1.065177236922496, "learning_rate": 6.690609205956464e-05, "loss": 1.0625, "step": 9384 }, { "epoch": 4.416470588235294, "grad_norm": 1.3909082292882498, "learning_rate": 6.686960007613906e-05, "loss": 1.6328, "step": 9385 }, { "epoch": 4.416941176470588, "grad_norm": 1.2024170176378104, "learning_rate": 6.683311519245733e-05, "loss": 1.1328, "step": 9386 }, { "epoch": 4.417411764705882, "grad_norm": 1.3896280000274273, "learning_rate": 6.679663741163553e-05, "loss": 1.0547, "step": 9387 }, { "epoch": 4.417882352941176, "grad_norm": 1.039202464224088, "learning_rate": 6.676016673678896e-05, "loss": 1.2344, "step": 9388 }, { "epoch": 4.418352941176471, "grad_norm": 0.9066889368038369, "learning_rate": 6.672370317103245e-05, "loss": 0.4824, "step": 9389 }, { "epoch": 4.418823529411765, "grad_norm": 0.6438560394416581, "learning_rate": 6.668724671748018e-05, "loss": 0.1807, "step": 9390 }, { "epoch": 4.419294117647059, "grad_norm": 0.9754558980836896, "learning_rate": 6.665079737924566e-05, "loss": 0.8945, "step": 9391 }, { "epoch": 4.419764705882353, "grad_norm": 0.9370711995338724, "learning_rate": 6.661435515944186e-05, "loss": 0.3262, "step": 9392 }, { "epoch": 4.4202352941176475, "grad_norm": 0.9840947595567952, "learning_rate": 6.65779200611812e-05, "loss": 0.5977, "step": 9393 }, { "epoch": 4.420705882352941, "grad_norm": 1.0517600008382944, "learning_rate": 6.654149208757529e-05, "loss": 0.9297, "step": 9394 }, { "epoch": 4.421176470588235, "grad_norm": 1.2489599426531675, "learning_rate": 6.65050712417353e-05, "loss": 0.9375, "step": 9395 }, { "epoch": 4.421647058823529, "grad_norm": 0.9947297911351848, "learning_rate": 6.646865752677186e-05, "loss": 1.1797, "step": 9396 }, { "epoch": 4.4221176470588235, "grad_norm": 0.9373502345597384, "learning_rate": 6.643225094579462e-05, "loss": 0.3613, "step": 9397 }, { "epoch": 4.422588235294118, "grad_norm": 1.0149222433144582, "learning_rate": 6.63958515019132e-05, "loss": 0.4824, "step": 9398 }, { "epoch": 4.423058823529412, "grad_norm": 1.2773178224252857, "learning_rate": 6.635945919823607e-05, "loss": 1.2188, "step": 9399 }, { "epoch": 4.423529411764706, "grad_norm": 0.706205769600173, "learning_rate": 6.632307403787138e-05, "loss": 0.2969, "step": 9400 }, { "epoch": 4.424, "grad_norm": 0.6738117667473571, "learning_rate": 6.628669602392665e-05, "loss": 0.0991, "step": 9401 }, { "epoch": 4.424470588235294, "grad_norm": 1.5095410205017878, "learning_rate": 6.625032515950865e-05, "loss": 1.1328, "step": 9402 }, { "epoch": 4.424941176470588, "grad_norm": 0.7410604411800765, "learning_rate": 6.621396144772365e-05, "loss": 0.2256, "step": 9403 }, { "epoch": 4.425411764705882, "grad_norm": 1.106431638122732, "learning_rate": 6.617760489167738e-05, "loss": 1.1719, "step": 9404 }, { "epoch": 4.425882352941176, "grad_norm": 1.0295793802181226, "learning_rate": 6.614125549447475e-05, "loss": 0.498, "step": 9405 }, { "epoch": 4.426352941176471, "grad_norm": 0.7774500788817968, "learning_rate": 6.610491325922022e-05, "loss": 0.4746, "step": 9406 }, { "epoch": 4.426823529411765, "grad_norm": 1.2321819571911432, "learning_rate": 6.606857818901765e-05, "loss": 0.7734, "step": 9407 }, { "epoch": 4.427294117647059, "grad_norm": 0.42433835187493674, "learning_rate": 6.603225028697013e-05, "loss": 0.0703, "step": 9408 }, { "epoch": 4.427764705882353, "grad_norm": 1.417575253177317, "learning_rate": 6.59959295561803e-05, "loss": 0.9141, "step": 9409 }, { "epoch": 4.428235294117647, "grad_norm": 1.1087703834037967, "learning_rate": 6.595961599975016e-05, "loss": 0.4863, "step": 9410 }, { "epoch": 4.428705882352941, "grad_norm": 0.6663687869919951, "learning_rate": 6.5923309620781e-05, "loss": 0.5742, "step": 9411 }, { "epoch": 4.429176470588235, "grad_norm": 0.7925176481279054, "learning_rate": 6.588701042237358e-05, "loss": 0.4922, "step": 9412 }, { "epoch": 4.429647058823529, "grad_norm": 0.9191453057822834, "learning_rate": 6.585071840762809e-05, "loss": 0.7148, "step": 9413 }, { "epoch": 4.4301176470588235, "grad_norm": 0.7740716008609618, "learning_rate": 6.581443357964395e-05, "loss": 0.25, "step": 9414 }, { "epoch": 4.430588235294118, "grad_norm": 0.844199869471466, "learning_rate": 6.577815594152017e-05, "loss": 0.3359, "step": 9415 }, { "epoch": 4.431058823529412, "grad_norm": 0.6801535091697873, "learning_rate": 6.574188549635493e-05, "loss": 0.8008, "step": 9416 }, { "epoch": 4.431529411764706, "grad_norm": 1.0788867078518203, "learning_rate": 6.570562224724593e-05, "loss": 0.8047, "step": 9417 }, { "epoch": 4.432, "grad_norm": 1.0358960741299126, "learning_rate": 6.566936619729033e-05, "loss": 0.7812, "step": 9418 }, { "epoch": 4.432470588235294, "grad_norm": 0.9296366077628643, "learning_rate": 6.563311734958444e-05, "loss": 0.2383, "step": 9419 }, { "epoch": 4.432941176470588, "grad_norm": 0.9297045835710196, "learning_rate": 6.559687570722413e-05, "loss": 1.1328, "step": 9420 }, { "epoch": 4.433411764705882, "grad_norm": 0.7382833903854717, "learning_rate": 6.556064127330467e-05, "loss": 0.7539, "step": 9421 }, { "epoch": 4.433882352941176, "grad_norm": 0.9000420588082586, "learning_rate": 6.552441405092058e-05, "loss": 0.793, "step": 9422 }, { "epoch": 4.434352941176471, "grad_norm": 0.991787576357928, "learning_rate": 6.54881940431659e-05, "loss": 0.3809, "step": 9423 }, { "epoch": 4.434823529411765, "grad_norm": 0.951624690757917, "learning_rate": 6.5451981253134e-05, "loss": 0.6758, "step": 9424 }, { "epoch": 4.435294117647059, "grad_norm": 1.0893706906395708, "learning_rate": 6.541577568391758e-05, "loss": 0.7227, "step": 9425 }, { "epoch": 4.435764705882353, "grad_norm": 1.2804784744073503, "learning_rate": 6.537957733860879e-05, "loss": 0.6953, "step": 9426 }, { "epoch": 4.4362352941176475, "grad_norm": 1.0115581832236613, "learning_rate": 6.53433862202992e-05, "loss": 0.8242, "step": 9427 }, { "epoch": 4.436705882352941, "grad_norm": 1.175006930874567, "learning_rate": 6.530720233207962e-05, "loss": 0.8359, "step": 9428 }, { "epoch": 4.437176470588235, "grad_norm": 0.8090925129720801, "learning_rate": 6.527102567704036e-05, "loss": 0.2344, "step": 9429 }, { "epoch": 4.437647058823529, "grad_norm": 0.3455908298019852, "learning_rate": 6.523485625827119e-05, "loss": 0.0518, "step": 9430 }, { "epoch": 4.4381176470588235, "grad_norm": 1.1940240883641833, "learning_rate": 6.519869407886098e-05, "loss": 1.0859, "step": 9431 }, { "epoch": 4.438588235294118, "grad_norm": 0.9119725708987757, "learning_rate": 6.516253914189825e-05, "loss": 0.5469, "step": 9432 }, { "epoch": 4.439058823529412, "grad_norm": 1.0006191772023822, "learning_rate": 6.512639145047086e-05, "loss": 0.8281, "step": 9433 }, { "epoch": 4.439529411764706, "grad_norm": 1.1405813104350344, "learning_rate": 6.509025100766591e-05, "loss": 1.5156, "step": 9434 }, { "epoch": 4.44, "grad_norm": 0.5589222709738851, "learning_rate": 6.505411781656997e-05, "loss": 0.1367, "step": 9435 }, { "epoch": 4.440470588235294, "grad_norm": 0.8324504350458126, "learning_rate": 6.50179918802691e-05, "loss": 0.8125, "step": 9436 }, { "epoch": 4.440941176470588, "grad_norm": 2.797425601185487, "learning_rate": 6.49818732018485e-05, "loss": 1.6484, "step": 9437 }, { "epoch": 4.441411764705882, "grad_norm": 0.9593981907485565, "learning_rate": 6.494576178439298e-05, "loss": 0.8164, "step": 9438 }, { "epoch": 4.441882352941176, "grad_norm": 0.7455908156520245, "learning_rate": 6.490965763098654e-05, "loss": 0.2021, "step": 9439 }, { "epoch": 4.442352941176471, "grad_norm": 1.4426479213550087, "learning_rate": 6.487356074471272e-05, "loss": 1.1484, "step": 9440 }, { "epoch": 4.442823529411765, "grad_norm": 0.7069120072190759, "learning_rate": 6.483747112865436e-05, "loss": 0.4141, "step": 9441 }, { "epoch": 4.443294117647059, "grad_norm": 1.0882516066150187, "learning_rate": 6.480138878589368e-05, "loss": 0.9727, "step": 9442 }, { "epoch": 4.443764705882353, "grad_norm": 1.0499563774151515, "learning_rate": 6.476531371951229e-05, "loss": 0.498, "step": 9443 }, { "epoch": 4.444235294117647, "grad_norm": 0.8202240754685157, "learning_rate": 6.472924593259122e-05, "loss": 0.5, "step": 9444 }, { "epoch": 4.444705882352941, "grad_norm": 0.8149066921766631, "learning_rate": 6.469318542821078e-05, "loss": 0.2773, "step": 9445 }, { "epoch": 4.445176470588235, "grad_norm": 0.940543611460591, "learning_rate": 6.465713220945075e-05, "loss": 0.5195, "step": 9446 }, { "epoch": 4.445647058823529, "grad_norm": 1.2218334796208292, "learning_rate": 6.46210862793902e-05, "loss": 0.9609, "step": 9447 }, { "epoch": 4.4461176470588235, "grad_norm": 0.896113806159783, "learning_rate": 6.458504764110764e-05, "loss": 0.4941, "step": 9448 }, { "epoch": 4.446588235294118, "grad_norm": 1.227581898996693, "learning_rate": 6.454901629768105e-05, "loss": 1.3359, "step": 9449 }, { "epoch": 4.447058823529412, "grad_norm": 0.794785189514836, "learning_rate": 6.451299225218754e-05, "loss": 0.4199, "step": 9450 }, { "epoch": 4.447529411764706, "grad_norm": 1.3465156544214048, "learning_rate": 6.447697550770381e-05, "loss": 1.0469, "step": 9451 }, { "epoch": 4.448, "grad_norm": 1.2237683299889812, "learning_rate": 6.44409660673059e-05, "loss": 0.75, "step": 9452 }, { "epoch": 4.448470588235294, "grad_norm": 0.9777174476291016, "learning_rate": 6.440496393406912e-05, "loss": 0.3945, "step": 9453 }, { "epoch": 4.448941176470588, "grad_norm": 1.1197511327402296, "learning_rate": 6.436896911106824e-05, "loss": 0.5586, "step": 9454 }, { "epoch": 4.449411764705882, "grad_norm": 1.121391461160654, "learning_rate": 6.433298160137749e-05, "loss": 1.1406, "step": 9455 }, { "epoch": 4.449882352941176, "grad_norm": 0.9955991375120173, "learning_rate": 6.429700140807023e-05, "loss": 0.3398, "step": 9456 }, { "epoch": 4.450352941176471, "grad_norm": 0.7866583051049687, "learning_rate": 6.426102853421942e-05, "loss": 0.3086, "step": 9457 }, { "epoch": 4.450823529411765, "grad_norm": 0.8779964329109063, "learning_rate": 6.42250629828974e-05, "loss": 0.5039, "step": 9458 }, { "epoch": 4.451294117647059, "grad_norm": 1.298182466175527, "learning_rate": 6.418910475717565e-05, "loss": 1.6562, "step": 9459 }, { "epoch": 4.451764705882353, "grad_norm": 0.6771507785883536, "learning_rate": 6.415315386012525e-05, "loss": 0.1592, "step": 9460 }, { "epoch": 4.4522352941176475, "grad_norm": 1.2276405531992989, "learning_rate": 6.411721029481665e-05, "loss": 0.8945, "step": 9461 }, { "epoch": 4.452705882352941, "grad_norm": 0.9792303970063195, "learning_rate": 6.40812740643195e-05, "loss": 0.4277, "step": 9462 }, { "epoch": 4.453176470588235, "grad_norm": 0.9352596541608981, "learning_rate": 6.404534517170296e-05, "loss": 0.5, "step": 9463 }, { "epoch": 4.453647058823529, "grad_norm": 0.7713963405966938, "learning_rate": 6.40094236200356e-05, "loss": 0.3184, "step": 9464 }, { "epoch": 4.4541176470588235, "grad_norm": 0.4583451472909143, "learning_rate": 6.397350941238519e-05, "loss": 0.0806, "step": 9465 }, { "epoch": 4.454588235294118, "grad_norm": 0.9194382558676466, "learning_rate": 6.39376025518191e-05, "loss": 0.2188, "step": 9466 }, { "epoch": 4.455058823529412, "grad_norm": 1.097251885608037, "learning_rate": 6.390170304140382e-05, "loss": 1.0469, "step": 9467 }, { "epoch": 4.455529411764706, "grad_norm": 1.0490531179762466, "learning_rate": 6.386581088420542e-05, "loss": 0.5547, "step": 9468 }, { "epoch": 4.456, "grad_norm": 1.0308123471347461, "learning_rate": 6.382992608328929e-05, "loss": 0.5156, "step": 9469 }, { "epoch": 4.456470588235294, "grad_norm": 0.5827986840914069, "learning_rate": 6.37940486417201e-05, "loss": 0.103, "step": 9470 }, { "epoch": 4.456941176470588, "grad_norm": 1.1437767353849457, "learning_rate": 6.375817856256199e-05, "loss": 0.5352, "step": 9471 }, { "epoch": 4.457411764705882, "grad_norm": 1.75359342974985, "learning_rate": 6.37223158488785e-05, "loss": 1.75, "step": 9472 }, { "epoch": 4.457882352941176, "grad_norm": 1.090999779881211, "learning_rate": 6.368646050373236e-05, "loss": 0.707, "step": 9473 }, { "epoch": 4.458352941176471, "grad_norm": 0.8208133017330425, "learning_rate": 6.365061253018586e-05, "loss": 0.2324, "step": 9474 }, { "epoch": 4.458823529411765, "grad_norm": 0.9592828410252163, "learning_rate": 6.361477193130065e-05, "loss": 0.7148, "step": 9475 }, { "epoch": 4.459294117647059, "grad_norm": 0.44608964758329706, "learning_rate": 6.35789387101376e-05, "loss": 0.0747, "step": 9476 }, { "epoch": 4.459764705882353, "grad_norm": 0.8709511384837983, "learning_rate": 6.354311286975705e-05, "loss": 0.2559, "step": 9477 }, { "epoch": 4.460235294117647, "grad_norm": 1.1481647743585903, "learning_rate": 6.35072944132188e-05, "loss": 0.8945, "step": 9478 }, { "epoch": 4.460705882352941, "grad_norm": 0.9880256205290541, "learning_rate": 6.34714833435818e-05, "loss": 0.7852, "step": 9479 }, { "epoch": 4.461176470588235, "grad_norm": 1.3493420477372846, "learning_rate": 6.343567966390452e-05, "loss": 0.9453, "step": 9480 }, { "epoch": 4.461647058823529, "grad_norm": 1.309664289535105, "learning_rate": 6.339988337724487e-05, "loss": 0.6328, "step": 9481 }, { "epoch": 4.4621176470588235, "grad_norm": 1.161636583937931, "learning_rate": 6.336409448665988e-05, "loss": 0.9219, "step": 9482 }, { "epoch": 4.462588235294118, "grad_norm": 1.1469331030617809, "learning_rate": 6.332831299520617e-05, "loss": 0.5547, "step": 9483 }, { "epoch": 4.463058823529412, "grad_norm": 0.837785122355324, "learning_rate": 6.32925389059397e-05, "loss": 0.8672, "step": 9484 }, { "epoch": 4.463529411764706, "grad_norm": 1.695876925825217, "learning_rate": 6.325677222191568e-05, "loss": 0.9258, "step": 9485 }, { "epoch": 4.464, "grad_norm": 1.6280896829848006, "learning_rate": 6.322101294618875e-05, "loss": 0.6445, "step": 9486 }, { "epoch": 4.464470588235294, "grad_norm": 1.3758854644094325, "learning_rate": 6.318526108181299e-05, "loss": 1.4219, "step": 9487 }, { "epoch": 4.464941176470588, "grad_norm": 1.3700693387718215, "learning_rate": 6.314951663184175e-05, "loss": 1.3359, "step": 9488 }, { "epoch": 4.465411764705882, "grad_norm": 1.0257936361676554, "learning_rate": 6.311377959932783e-05, "loss": 0.7734, "step": 9489 }, { "epoch": 4.465882352941176, "grad_norm": 1.1865138709749337, "learning_rate": 6.307804998732328e-05, "loss": 1.1797, "step": 9490 }, { "epoch": 4.466352941176471, "grad_norm": 0.9509810486327085, "learning_rate": 6.304232779887958e-05, "loss": 0.4199, "step": 9491 }, { "epoch": 4.466823529411765, "grad_norm": 1.1411365563996474, "learning_rate": 6.300661303704768e-05, "loss": 0.8789, "step": 9492 }, { "epoch": 4.467294117647059, "grad_norm": 1.3137543650502612, "learning_rate": 6.297090570487766e-05, "loss": 1.0, "step": 9493 }, { "epoch": 4.467764705882353, "grad_norm": 0.9549962195282942, "learning_rate": 6.293520580541925e-05, "loss": 0.4648, "step": 9494 }, { "epoch": 4.4682352941176475, "grad_norm": 0.7136487999679952, "learning_rate": 6.289951334172123e-05, "loss": 0.3535, "step": 9495 }, { "epoch": 4.468705882352941, "grad_norm": 0.8715734523168434, "learning_rate": 6.2863828316832e-05, "loss": 0.3398, "step": 9496 }, { "epoch": 4.469176470588235, "grad_norm": 1.3381296622793495, "learning_rate": 6.282815073379932e-05, "loss": 0.9414, "step": 9497 }, { "epoch": 4.469647058823529, "grad_norm": 0.8113699421020758, "learning_rate": 6.279248059567006e-05, "loss": 0.3691, "step": 9498 }, { "epoch": 4.4701176470588235, "grad_norm": 0.797319421227134, "learning_rate": 6.275681790549072e-05, "loss": 0.498, "step": 9499 }, { "epoch": 4.470588235294118, "grad_norm": 1.3966685014953917, "learning_rate": 6.27211626663071e-05, "loss": 1.1484, "step": 9500 }, { "epoch": 4.471058823529412, "grad_norm": 1.40034584000138, "learning_rate": 6.268551488116423e-05, "loss": 1.3906, "step": 9501 }, { "epoch": 4.471529411764706, "grad_norm": 1.0117289673628587, "learning_rate": 6.264987455310667e-05, "loss": 0.7578, "step": 9502 }, { "epoch": 4.4719999999999995, "grad_norm": 1.6394925866499572, "learning_rate": 6.261424168517832e-05, "loss": 1.7344, "step": 9503 }, { "epoch": 4.472470588235294, "grad_norm": 0.49390093663119466, "learning_rate": 6.257861628042233e-05, "loss": 0.291, "step": 9504 }, { "epoch": 4.472941176470588, "grad_norm": 1.32478335167386, "learning_rate": 6.254299834188128e-05, "loss": 0.8984, "step": 9505 }, { "epoch": 4.473411764705882, "grad_norm": 0.9100155682392506, "learning_rate": 6.25073878725972e-05, "loss": 0.3379, "step": 9506 }, { "epoch": 4.473882352941176, "grad_norm": 1.1538634887436185, "learning_rate": 6.24717848756113e-05, "loss": 1.1562, "step": 9507 }, { "epoch": 4.474352941176471, "grad_norm": 0.4144178064147282, "learning_rate": 6.24361893539643e-05, "loss": 0.105, "step": 9508 }, { "epoch": 4.474823529411765, "grad_norm": 1.1921946506972676, "learning_rate": 6.240060131069628e-05, "loss": 0.8477, "step": 9509 }, { "epoch": 4.475294117647059, "grad_norm": 0.8374170556128511, "learning_rate": 6.23650207488465e-05, "loss": 0.6602, "step": 9510 }, { "epoch": 4.475764705882353, "grad_norm": 0.6071635349268784, "learning_rate": 6.232944767145384e-05, "loss": 0.123, "step": 9511 }, { "epoch": 4.476235294117647, "grad_norm": 0.7813052361661279, "learning_rate": 6.229388208155639e-05, "loss": 0.3027, "step": 9512 }, { "epoch": 4.476705882352941, "grad_norm": 0.9685893482087954, "learning_rate": 6.225832398219158e-05, "loss": 0.7852, "step": 9513 }, { "epoch": 4.477176470588235, "grad_norm": 0.8726099227929239, "learning_rate": 6.222277337639624e-05, "loss": 0.5898, "step": 9514 }, { "epoch": 4.477647058823529, "grad_norm": 1.1647787285148627, "learning_rate": 6.218723026720668e-05, "loss": 1.4062, "step": 9515 }, { "epoch": 4.4781176470588235, "grad_norm": 1.1604855906315652, "learning_rate": 6.21516946576583e-05, "loss": 0.8828, "step": 9516 }, { "epoch": 4.478588235294118, "grad_norm": 0.7897255340500486, "learning_rate": 6.211616655078616e-05, "loss": 0.2793, "step": 9517 }, { "epoch": 4.479058823529412, "grad_norm": 0.8941978011536796, "learning_rate": 6.208064594962441e-05, "loss": 0.8867, "step": 9518 }, { "epoch": 4.479529411764706, "grad_norm": 1.1316770712026893, "learning_rate": 6.204513285720673e-05, "loss": 0.7188, "step": 9519 }, { "epoch": 4.48, "grad_norm": 0.8889035127012069, "learning_rate": 6.200962727656616e-05, "loss": 0.2598, "step": 9520 }, { "epoch": 4.480470588235294, "grad_norm": 0.9658296204598684, "learning_rate": 6.197412921073498e-05, "loss": 0.4082, "step": 9521 }, { "epoch": 4.480941176470588, "grad_norm": 2.2589687796972457, "learning_rate": 6.193863866274492e-05, "loss": 1.7031, "step": 9522 }, { "epoch": 4.481411764705882, "grad_norm": 0.9199281296266785, "learning_rate": 6.190315563562711e-05, "loss": 0.3984, "step": 9523 }, { "epoch": 4.481882352941176, "grad_norm": 1.1110215991480024, "learning_rate": 6.186768013241188e-05, "loss": 0.875, "step": 9524 }, { "epoch": 4.482352941176471, "grad_norm": 0.6962505375062932, "learning_rate": 6.183221215612904e-05, "loss": 0.2129, "step": 9525 }, { "epoch": 4.482823529411765, "grad_norm": 0.8227785108011249, "learning_rate": 6.179675170980779e-05, "loss": 0.7227, "step": 9526 }, { "epoch": 4.483294117647059, "grad_norm": 0.8449224833220679, "learning_rate": 6.176129879647654e-05, "loss": 0.6992, "step": 9527 }, { "epoch": 4.483764705882353, "grad_norm": 1.0137235510555773, "learning_rate": 6.172585341916317e-05, "loss": 0.6055, "step": 9528 }, { "epoch": 4.484235294117647, "grad_norm": 1.1568851564353206, "learning_rate": 6.169041558089496e-05, "loss": 0.9961, "step": 9529 }, { "epoch": 4.484705882352941, "grad_norm": 1.4083830981718877, "learning_rate": 6.16549852846983e-05, "loss": 0.7578, "step": 9530 }, { "epoch": 4.485176470588235, "grad_norm": 0.6164065148456311, "learning_rate": 6.161956253359935e-05, "loss": 0.5586, "step": 9531 }, { "epoch": 4.485647058823529, "grad_norm": 1.1115648788929722, "learning_rate": 6.15841473306232e-05, "loss": 0.9141, "step": 9532 }, { "epoch": 4.4861176470588235, "grad_norm": 0.8958002351597102, "learning_rate": 6.154873967879455e-05, "loss": 0.3535, "step": 9533 }, { "epoch": 4.486588235294118, "grad_norm": 1.1364440953801052, "learning_rate": 6.151333958113743e-05, "loss": 0.9883, "step": 9534 }, { "epoch": 4.487058823529412, "grad_norm": 1.1179462077408866, "learning_rate": 6.147794704067508e-05, "loss": 0.5312, "step": 9535 }, { "epoch": 4.487529411764706, "grad_norm": 1.1065403371768194, "learning_rate": 6.144256206043025e-05, "loss": 1.1797, "step": 9536 }, { "epoch": 4.4879999999999995, "grad_norm": 1.1649013329741615, "learning_rate": 6.140718464342505e-05, "loss": 0.5547, "step": 9537 }, { "epoch": 4.488470588235294, "grad_norm": 0.9481477156824852, "learning_rate": 6.137181479268077e-05, "loss": 0.7852, "step": 9538 }, { "epoch": 4.488941176470588, "grad_norm": 0.9875860368122524, "learning_rate": 6.133645251121823e-05, "loss": 0.3945, "step": 9539 }, { "epoch": 4.489411764705882, "grad_norm": 1.0687058971666676, "learning_rate": 6.130109780205759e-05, "loss": 0.377, "step": 9540 }, { "epoch": 4.489882352941176, "grad_norm": 1.1096897680642464, "learning_rate": 6.126575066821821e-05, "loss": 0.8711, "step": 9541 }, { "epoch": 4.490352941176471, "grad_norm": 0.7647621525652526, "learning_rate": 6.123041111271896e-05, "loss": 0.3438, "step": 9542 }, { "epoch": 4.490823529411765, "grad_norm": 0.9813630134515848, "learning_rate": 6.119507913857807e-05, "loss": 0.8477, "step": 9543 }, { "epoch": 4.491294117647059, "grad_norm": 1.0117066625720925, "learning_rate": 6.115975474881296e-05, "loss": 0.3203, "step": 9544 }, { "epoch": 4.491764705882353, "grad_norm": 1.1163076050234204, "learning_rate": 6.11244379464406e-05, "loss": 0.7617, "step": 9545 }, { "epoch": 4.492235294117647, "grad_norm": 1.294068583462835, "learning_rate": 6.108912873447712e-05, "loss": 1.5859, "step": 9546 }, { "epoch": 4.492705882352941, "grad_norm": 0.9419940726900972, "learning_rate": 6.105382711593814e-05, "loss": 0.9023, "step": 9547 }, { "epoch": 4.493176470588235, "grad_norm": 1.1535483224975192, "learning_rate": 6.101853309383866e-05, "loss": 1.2266, "step": 9548 }, { "epoch": 4.493647058823529, "grad_norm": 1.0100670389056765, "learning_rate": 6.0983246671192866e-05, "loss": 0.3789, "step": 9549 }, { "epoch": 4.4941176470588236, "grad_norm": 0.8707882914607694, "learning_rate": 6.0947967851014405e-05, "loss": 0.6914, "step": 9550 }, { "epoch": 4.494588235294118, "grad_norm": 1.2563794248788915, "learning_rate": 6.091269663631636e-05, "loss": 0.9688, "step": 9551 }, { "epoch": 4.495058823529412, "grad_norm": 0.9039304687405441, "learning_rate": 6.0877433030110935e-05, "loss": 0.457, "step": 9552 }, { "epoch": 4.495529411764706, "grad_norm": 0.8673917610512971, "learning_rate": 6.084217703540988e-05, "loss": 0.3047, "step": 9553 }, { "epoch": 4.496, "grad_norm": 1.2919872789041256, "learning_rate": 6.0806928655224266e-05, "loss": 1.4531, "step": 9554 }, { "epoch": 4.496470588235294, "grad_norm": 0.7440353083266442, "learning_rate": 6.0771687892564394e-05, "loss": 0.2637, "step": 9555 }, { "epoch": 4.496941176470588, "grad_norm": 1.0303028946331205, "learning_rate": 6.0736454750440025e-05, "loss": 1.1328, "step": 9556 }, { "epoch": 4.497411764705882, "grad_norm": 0.9075756363621961, "learning_rate": 6.070122923186032e-05, "loss": 0.3379, "step": 9557 }, { "epoch": 4.497882352941176, "grad_norm": 0.8412605667577542, "learning_rate": 6.0666011339833584e-05, "loss": 0.2734, "step": 9558 }, { "epoch": 4.498352941176471, "grad_norm": 0.8858745047906592, "learning_rate": 6.063080107736767e-05, "loss": 0.5391, "step": 9559 }, { "epoch": 4.498823529411765, "grad_norm": 1.2705631940951243, "learning_rate": 6.0595598447469726e-05, "loss": 0.4883, "step": 9560 }, { "epoch": 4.499294117647059, "grad_norm": 1.0511785633140613, "learning_rate": 6.056040345314618e-05, "loss": 0.7734, "step": 9561 }, { "epoch": 4.499764705882353, "grad_norm": 1.360421875313191, "learning_rate": 6.0525216097402865e-05, "loss": 0.875, "step": 9562 }, { "epoch": 4.5002352941176476, "grad_norm": 1.0379458275922941, "learning_rate": 6.049003638324501e-05, "loss": 1.0234, "step": 9563 }, { "epoch": 4.500705882352941, "grad_norm": 1.0507985725097313, "learning_rate": 6.0454864313677054e-05, "loss": 0.4531, "step": 9564 }, { "epoch": 4.501176470588235, "grad_norm": 0.9403407836836646, "learning_rate": 6.041969989170289e-05, "loss": 1.0938, "step": 9565 }, { "epoch": 4.501647058823529, "grad_norm": 1.146818518627676, "learning_rate": 6.0384543120325796e-05, "loss": 0.6914, "step": 9566 }, { "epoch": 4.502117647058824, "grad_norm": 1.0684560463507538, "learning_rate": 6.034939400254824e-05, "loss": 0.4941, "step": 9567 }, { "epoch": 4.502588235294118, "grad_norm": 1.091619943031689, "learning_rate": 6.031425254137222e-05, "loss": 1.4844, "step": 9568 }, { "epoch": 4.503058823529412, "grad_norm": 1.2130946614777167, "learning_rate": 6.0279118739798874e-05, "loss": 0.625, "step": 9569 }, { "epoch": 4.503529411764706, "grad_norm": 0.7014375745449447, "learning_rate": 6.0243992600828875e-05, "loss": 0.1777, "step": 9570 }, { "epoch": 4.504, "grad_norm": 1.0257668462019, "learning_rate": 6.020887412746221e-05, "loss": 0.8359, "step": 9571 }, { "epoch": 4.504470588235294, "grad_norm": 0.961360756968971, "learning_rate": 6.017376332269807e-05, "loss": 0.7109, "step": 9572 }, { "epoch": 4.504941176470588, "grad_norm": 0.9880852594840831, "learning_rate": 6.0138660189535126e-05, "loss": 0.5312, "step": 9573 }, { "epoch": 4.505411764705882, "grad_norm": 1.0253266711082916, "learning_rate": 6.0103564730971375e-05, "loss": 1.6172, "step": 9574 }, { "epoch": 4.5058823529411764, "grad_norm": 0.8772155097485688, "learning_rate": 6.006847695000415e-05, "loss": 0.5039, "step": 9575 }, { "epoch": 4.506352941176471, "grad_norm": 0.9636737351879179, "learning_rate": 6.0033396849630136e-05, "loss": 0.4004, "step": 9576 }, { "epoch": 4.506823529411765, "grad_norm": 1.0379685037214263, "learning_rate": 5.999832443284527e-05, "loss": 0.9102, "step": 9577 }, { "epoch": 4.507294117647059, "grad_norm": 1.0218483610065017, "learning_rate": 5.9963259702644963e-05, "loss": 0.793, "step": 9578 }, { "epoch": 4.507764705882353, "grad_norm": 0.7007224215163156, "learning_rate": 5.992820266202394e-05, "loss": 0.1514, "step": 9579 }, { "epoch": 4.508235294117647, "grad_norm": 0.9696834237108434, "learning_rate": 5.989315331397616e-05, "loss": 1.2031, "step": 9580 }, { "epoch": 4.508705882352941, "grad_norm": 0.9986987186149706, "learning_rate": 5.985811166149507e-05, "loss": 0.8438, "step": 9581 }, { "epoch": 4.509176470588235, "grad_norm": 1.168303578474267, "learning_rate": 5.982307770757343e-05, "loss": 0.7227, "step": 9582 }, { "epoch": 4.509647058823529, "grad_norm": 1.1802354930371877, "learning_rate": 5.978805145520322e-05, "loss": 0.8672, "step": 9583 }, { "epoch": 4.510117647058824, "grad_norm": 1.2224263209900827, "learning_rate": 5.97530329073759e-05, "loss": 1.2734, "step": 9584 }, { "epoch": 4.510588235294118, "grad_norm": 0.5693414488377071, "learning_rate": 5.971802206708229e-05, "loss": 0.3594, "step": 9585 }, { "epoch": 4.511058823529412, "grad_norm": 0.9491808886762874, "learning_rate": 5.968301893731237e-05, "loss": 0.2773, "step": 9586 }, { "epoch": 4.511529411764706, "grad_norm": 1.4029284424739088, "learning_rate": 5.964802352105564e-05, "loss": 0.8398, "step": 9587 }, { "epoch": 4.5120000000000005, "grad_norm": 1.275681643389193, "learning_rate": 5.961303582130092e-05, "loss": 0.6602, "step": 9588 }, { "epoch": 4.512470588235294, "grad_norm": 1.0116592112678635, "learning_rate": 5.957805584103623e-05, "loss": 0.6133, "step": 9589 }, { "epoch": 4.512941176470588, "grad_norm": 1.1220641538829952, "learning_rate": 5.954308358324909e-05, "loss": 1.0547, "step": 9590 }, { "epoch": 4.513411764705882, "grad_norm": 1.1308795190631522, "learning_rate": 5.950811905092636e-05, "loss": 0.7266, "step": 9591 }, { "epoch": 4.5138823529411765, "grad_norm": 0.5361039199528412, "learning_rate": 5.947316224705407e-05, "loss": 0.0952, "step": 9592 }, { "epoch": 4.514352941176471, "grad_norm": 1.0651809372927594, "learning_rate": 5.943821317461775e-05, "loss": 0.7461, "step": 9593 }, { "epoch": 4.514823529411765, "grad_norm": 1.2207306475788187, "learning_rate": 5.940327183660228e-05, "loss": 0.4316, "step": 9594 }, { "epoch": 4.515294117647059, "grad_norm": 1.0427662317501678, "learning_rate": 5.936833823599172e-05, "loss": 0.3652, "step": 9595 }, { "epoch": 4.5157647058823525, "grad_norm": 1.0059517825905206, "learning_rate": 5.933341237576968e-05, "loss": 1.3203, "step": 9596 }, { "epoch": 4.516235294117647, "grad_norm": 1.3243423339197995, "learning_rate": 5.9298494258918875e-05, "loss": 1.0859, "step": 9597 }, { "epoch": 4.516705882352941, "grad_norm": 0.8029917795728131, "learning_rate": 5.926358388842156e-05, "loss": 0.416, "step": 9598 }, { "epoch": 4.517176470588235, "grad_norm": 1.1231712207091993, "learning_rate": 5.922868126725929e-05, "loss": 0.7812, "step": 9599 }, { "epoch": 4.517647058823529, "grad_norm": 1.0204973291046084, "learning_rate": 5.919378639841281e-05, "loss": 0.832, "step": 9600 }, { "epoch": 4.518117647058824, "grad_norm": 1.5074250523042882, "learning_rate": 5.915889928486239e-05, "loss": 1.25, "step": 9601 }, { "epoch": 4.518588235294118, "grad_norm": 1.0452481933382838, "learning_rate": 5.912401992958757e-05, "loss": 0.9453, "step": 9602 }, { "epoch": 4.519058823529412, "grad_norm": 0.8615327244694504, "learning_rate": 5.908914833556715e-05, "loss": 0.4961, "step": 9603 }, { "epoch": 4.519529411764706, "grad_norm": 1.17350701976901, "learning_rate": 5.9054284505779396e-05, "loss": 0.8281, "step": 9604 }, { "epoch": 4.52, "grad_norm": 0.8853282591595635, "learning_rate": 5.901942844320187e-05, "loss": 0.75, "step": 9605 }, { "epoch": 4.520470588235294, "grad_norm": 0.9764311779072259, "learning_rate": 5.898458015081136e-05, "loss": 0.4492, "step": 9606 }, { "epoch": 4.520941176470588, "grad_norm": 1.2254866996055174, "learning_rate": 5.8949739631584136e-05, "loss": 0.6484, "step": 9607 }, { "epoch": 4.521411764705882, "grad_norm": 0.8233675456929621, "learning_rate": 5.8914906888495797e-05, "loss": 0.8086, "step": 9608 }, { "epoch": 4.5218823529411765, "grad_norm": 1.1622508608014772, "learning_rate": 5.888008192452112e-05, "loss": 1.1172, "step": 9609 }, { "epoch": 4.522352941176471, "grad_norm": 0.5918405144422918, "learning_rate": 5.8845264742634394e-05, "loss": 0.4863, "step": 9610 }, { "epoch": 4.522823529411765, "grad_norm": 0.9977865693455751, "learning_rate": 5.881045534580922e-05, "loss": 1.2031, "step": 9611 }, { "epoch": 4.523294117647059, "grad_norm": 1.2150443771892148, "learning_rate": 5.87756537370184e-05, "loss": 0.416, "step": 9612 }, { "epoch": 4.523764705882353, "grad_norm": 1.4744973825214138, "learning_rate": 5.874085991923419e-05, "loss": 1.0312, "step": 9613 }, { "epoch": 4.524235294117647, "grad_norm": 0.941975677234985, "learning_rate": 5.8706073895428225e-05, "loss": 0.8906, "step": 9614 }, { "epoch": 4.524705882352941, "grad_norm": 0.6453907047386057, "learning_rate": 5.867129566857129e-05, "loss": 0.2305, "step": 9615 }, { "epoch": 4.525176470588235, "grad_norm": 1.362251481379989, "learning_rate": 5.8636525241633673e-05, "loss": 0.7188, "step": 9616 }, { "epoch": 4.525647058823529, "grad_norm": 1.1320180608406947, "learning_rate": 5.860176261758499e-05, "loss": 0.8672, "step": 9617 }, { "epoch": 4.526117647058824, "grad_norm": 1.0487942850396614, "learning_rate": 5.8567007799394015e-05, "loss": 0.8633, "step": 9618 }, { "epoch": 4.526588235294118, "grad_norm": 0.8666770080863905, "learning_rate": 5.853226079002906e-05, "loss": 0.2559, "step": 9619 }, { "epoch": 4.527058823529412, "grad_norm": 1.0881109184823565, "learning_rate": 5.849752159245769e-05, "loss": 0.4062, "step": 9620 }, { "epoch": 4.527529411764706, "grad_norm": 1.204617166632808, "learning_rate": 5.84627902096468e-05, "loss": 0.7734, "step": 9621 }, { "epoch": 4.5280000000000005, "grad_norm": 1.0850832382239282, "learning_rate": 5.842806664456264e-05, "loss": 1.2266, "step": 9622 }, { "epoch": 4.528470588235294, "grad_norm": 1.349709326990976, "learning_rate": 5.839335090017072e-05, "loss": 1.125, "step": 9623 }, { "epoch": 4.528941176470588, "grad_norm": 1.165888930503381, "learning_rate": 5.835864297943594e-05, "loss": 0.4453, "step": 9624 }, { "epoch": 4.529411764705882, "grad_norm": 1.2964854049547905, "learning_rate": 5.8323942885322605e-05, "loss": 0.9883, "step": 9625 }, { "epoch": 4.5298823529411765, "grad_norm": 0.9921963179527674, "learning_rate": 5.828925062079416e-05, "loss": 0.9805, "step": 9626 }, { "epoch": 4.530352941176471, "grad_norm": 1.3497829946117414, "learning_rate": 5.8254566188813615e-05, "loss": 1.4141, "step": 9627 }, { "epoch": 4.530823529411765, "grad_norm": 0.9119226237507735, "learning_rate": 5.821988959234308e-05, "loss": 0.5234, "step": 9628 }, { "epoch": 4.531294117647059, "grad_norm": 1.288471977324008, "learning_rate": 5.8185220834344124e-05, "loss": 0.8398, "step": 9629 }, { "epoch": 4.5317647058823525, "grad_norm": 0.8661478331433572, "learning_rate": 5.815055991777773e-05, "loss": 0.4258, "step": 9630 }, { "epoch": 4.532235294117647, "grad_norm": 0.9991550938303772, "learning_rate": 5.811590684560398e-05, "loss": 0.5742, "step": 9631 }, { "epoch": 4.532705882352941, "grad_norm": 0.40089199526053226, "learning_rate": 5.808126162078249e-05, "loss": 0.0386, "step": 9632 }, { "epoch": 4.533176470588235, "grad_norm": 0.8875663483038642, "learning_rate": 5.8046624246272146e-05, "loss": 0.6562, "step": 9633 }, { "epoch": 4.533647058823529, "grad_norm": 1.1236881569214063, "learning_rate": 5.801199472503109e-05, "loss": 0.7539, "step": 9634 }, { "epoch": 4.534117647058824, "grad_norm": 1.1566915563217781, "learning_rate": 5.797737306001686e-05, "loss": 0.832, "step": 9635 }, { "epoch": 4.534588235294118, "grad_norm": 1.1231989838166856, "learning_rate": 5.794275925418641e-05, "loss": 1.0625, "step": 9636 }, { "epoch": 4.535058823529412, "grad_norm": 1.0785017732382096, "learning_rate": 5.79081533104958e-05, "loss": 0.3613, "step": 9637 }, { "epoch": 4.535529411764706, "grad_norm": 0.9508224971854953, "learning_rate": 5.787355523190059e-05, "loss": 0.3281, "step": 9638 }, { "epoch": 4.536, "grad_norm": 1.4998772307745933, "learning_rate": 5.783896502135572e-05, "loss": 1.6406, "step": 9639 }, { "epoch": 4.536470588235294, "grad_norm": 0.559633393753288, "learning_rate": 5.780438268181522e-05, "loss": 0.1187, "step": 9640 }, { "epoch": 4.536941176470588, "grad_norm": 1.1912273837554594, "learning_rate": 5.776980821623264e-05, "loss": 1.0703, "step": 9641 }, { "epoch": 4.537411764705882, "grad_norm": 1.1554637005253106, "learning_rate": 5.7735241627560905e-05, "loss": 0.6602, "step": 9642 }, { "epoch": 4.5378823529411765, "grad_norm": 0.8588947527445201, "learning_rate": 5.770068291875202e-05, "loss": 0.7734, "step": 9643 }, { "epoch": 4.538352941176471, "grad_norm": 0.7342031190704787, "learning_rate": 5.766613209275754e-05, "loss": 0.3145, "step": 9644 }, { "epoch": 4.538823529411765, "grad_norm": 1.4850514698117536, "learning_rate": 5.763158915252833e-05, "loss": 0.9453, "step": 9645 }, { "epoch": 4.539294117647059, "grad_norm": 0.7356842986196352, "learning_rate": 5.759705410101443e-05, "loss": 0.4121, "step": 9646 }, { "epoch": 4.539764705882353, "grad_norm": 0.8759491545400974, "learning_rate": 5.756252694116538e-05, "loss": 0.3828, "step": 9647 }, { "epoch": 4.540235294117647, "grad_norm": 1.5914664227944022, "learning_rate": 5.752800767592988e-05, "loss": 1.4922, "step": 9648 }, { "epoch": 4.540705882352941, "grad_norm": 0.9144048840282343, "learning_rate": 5.74934963082561e-05, "loss": 0.3496, "step": 9649 }, { "epoch": 4.541176470588235, "grad_norm": 0.8844403006510932, "learning_rate": 5.745899284109154e-05, "loss": 0.8359, "step": 9650 }, { "epoch": 4.541647058823529, "grad_norm": 1.1268934373033261, "learning_rate": 5.742449727738285e-05, "loss": 1.0625, "step": 9651 }, { "epoch": 4.542117647058824, "grad_norm": 0.9689125216587928, "learning_rate": 5.739000962007616e-05, "loss": 0.7344, "step": 9652 }, { "epoch": 4.542588235294118, "grad_norm": 1.3343097812371982, "learning_rate": 5.7355529872116957e-05, "loss": 1.4531, "step": 9653 }, { "epoch": 4.543058823529412, "grad_norm": 1.362562279241983, "learning_rate": 5.732105803644986e-05, "loss": 0.7344, "step": 9654 }, { "epoch": 4.543529411764706, "grad_norm": 0.59312673916374, "learning_rate": 5.7286594116019e-05, "loss": 0.127, "step": 9655 }, { "epoch": 4.5440000000000005, "grad_norm": 1.2951789014600001, "learning_rate": 5.725213811376782e-05, "loss": 0.8984, "step": 9656 }, { "epoch": 4.544470588235294, "grad_norm": 1.0368814223138665, "learning_rate": 5.7217690032638916e-05, "loss": 0.7227, "step": 9657 }, { "epoch": 4.544941176470588, "grad_norm": 0.539300081071827, "learning_rate": 5.718324987557438e-05, "loss": 0.4492, "step": 9658 }, { "epoch": 4.545411764705882, "grad_norm": 1.1771335054625565, "learning_rate": 5.714881764551561e-05, "loss": 1.1719, "step": 9659 }, { "epoch": 4.5458823529411765, "grad_norm": 0.8061046428597781, "learning_rate": 5.7114393345403194e-05, "loss": 0.1699, "step": 9660 }, { "epoch": 4.546352941176471, "grad_norm": 1.621908971303414, "learning_rate": 5.707997697817719e-05, "loss": 1.9219, "step": 9661 }, { "epoch": 4.546823529411765, "grad_norm": 0.6480494013739896, "learning_rate": 5.704556854677698e-05, "loss": 0.5195, "step": 9662 }, { "epoch": 4.547294117647059, "grad_norm": 1.06332075151289, "learning_rate": 5.701116805414104e-05, "loss": 0.7266, "step": 9663 }, { "epoch": 4.5477647058823525, "grad_norm": 1.3066853092920356, "learning_rate": 5.6976775503207574e-05, "loss": 0.9922, "step": 9664 }, { "epoch": 4.548235294117647, "grad_norm": 0.6057694794758729, "learning_rate": 5.6942390896913686e-05, "loss": 0.5, "step": 9665 }, { "epoch": 4.548705882352941, "grad_norm": 0.5190749684720861, "learning_rate": 5.690801423819608e-05, "loss": 0.0708, "step": 9666 }, { "epoch": 4.549176470588235, "grad_norm": 0.4909940307353963, "learning_rate": 5.687364552999071e-05, "loss": 0.084, "step": 9667 }, { "epoch": 4.549647058823529, "grad_norm": 0.9475972965981669, "learning_rate": 5.683928477523275e-05, "loss": 0.6367, "step": 9668 }, { "epoch": 4.550117647058824, "grad_norm": 0.9462587682547878, "learning_rate": 5.680493197685681e-05, "loss": 0.7188, "step": 9669 }, { "epoch": 4.550588235294118, "grad_norm": 0.9237532160072282, "learning_rate": 5.677058713779687e-05, "loss": 1.0547, "step": 9670 }, { "epoch": 4.551058823529412, "grad_norm": 2.4107024690664876, "learning_rate": 5.6736250260986e-05, "loss": 1.3281, "step": 9671 }, { "epoch": 4.551529411764706, "grad_norm": 1.5760101313877044, "learning_rate": 5.6701921349356826e-05, "loss": 1.4453, "step": 9672 }, { "epoch": 4.552, "grad_norm": 0.9826427031642827, "learning_rate": 5.666760040584126e-05, "loss": 0.9336, "step": 9673 }, { "epoch": 4.552470588235294, "grad_norm": 0.7108746038947831, "learning_rate": 5.663328743337034e-05, "loss": 0.4922, "step": 9674 }, { "epoch": 4.552941176470588, "grad_norm": 0.8378061478032451, "learning_rate": 5.659898243487463e-05, "loss": 0.6289, "step": 9675 }, { "epoch": 4.553411764705882, "grad_norm": 0.9495097582348463, "learning_rate": 5.6564685413284014e-05, "loss": 0.3086, "step": 9676 }, { "epoch": 4.5538823529411765, "grad_norm": 0.8045746457416871, "learning_rate": 5.6530396371527515e-05, "loss": 0.2363, "step": 9677 }, { "epoch": 4.554352941176471, "grad_norm": 1.0716696999313466, "learning_rate": 5.649611531253366e-05, "loss": 0.793, "step": 9678 }, { "epoch": 4.554823529411765, "grad_norm": 1.0800464564268375, "learning_rate": 5.646184223923017e-05, "loss": 1.1016, "step": 9679 }, { "epoch": 4.555294117647059, "grad_norm": 0.6689605314306586, "learning_rate": 5.642757715454413e-05, "loss": 0.1533, "step": 9680 }, { "epoch": 4.555764705882353, "grad_norm": 1.1148138167426254, "learning_rate": 5.639332006140203e-05, "loss": 1.0781, "step": 9681 }, { "epoch": 4.556235294117647, "grad_norm": 1.4700298751286194, "learning_rate": 5.635907096272949e-05, "loss": 1.5, "step": 9682 }, { "epoch": 4.556705882352941, "grad_norm": 0.9969084517351801, "learning_rate": 5.632482986145159e-05, "loss": 0.9961, "step": 9683 }, { "epoch": 4.557176470588235, "grad_norm": 0.9584974275286918, "learning_rate": 5.6290596760492745e-05, "loss": 0.2969, "step": 9684 }, { "epoch": 4.557647058823529, "grad_norm": 0.8614840042622787, "learning_rate": 5.625637166277654e-05, "loss": 0.2812, "step": 9685 }, { "epoch": 4.558117647058824, "grad_norm": 1.1356568167642496, "learning_rate": 5.622215457122598e-05, "loss": 0.9922, "step": 9686 }, { "epoch": 4.558588235294118, "grad_norm": 0.9856347629586468, "learning_rate": 5.618794548876346e-05, "loss": 0.5352, "step": 9687 }, { "epoch": 4.559058823529412, "grad_norm": 0.9742218782541346, "learning_rate": 5.6153744418310505e-05, "loss": 0.6797, "step": 9688 }, { "epoch": 4.559529411764705, "grad_norm": 1.1333467035487015, "learning_rate": 5.611955136278808e-05, "loss": 0.8789, "step": 9689 }, { "epoch": 4.5600000000000005, "grad_norm": 1.116180588492265, "learning_rate": 5.608536632511649e-05, "loss": 1.4062, "step": 9690 }, { "epoch": 4.560470588235294, "grad_norm": 1.2913196409948775, "learning_rate": 5.6051189308215225e-05, "loss": 0.8242, "step": 9691 }, { "epoch": 4.560941176470588, "grad_norm": 0.35651053644297004, "learning_rate": 5.6017020315003195e-05, "loss": 0.0369, "step": 9692 }, { "epoch": 4.561411764705882, "grad_norm": 1.029430945545152, "learning_rate": 5.598285934839868e-05, "loss": 0.9375, "step": 9693 }, { "epoch": 4.5618823529411765, "grad_norm": 0.7577510390099056, "learning_rate": 5.594870641131907e-05, "loss": 0.21, "step": 9694 }, { "epoch": 4.562352941176471, "grad_norm": 1.0422273277197254, "learning_rate": 5.591456150668126e-05, "loss": 0.8945, "step": 9695 }, { "epoch": 4.562823529411765, "grad_norm": 1.1403468095792382, "learning_rate": 5.588042463740143e-05, "loss": 0.8984, "step": 9696 }, { "epoch": 4.563294117647059, "grad_norm": 1.167704492434485, "learning_rate": 5.584629580639494e-05, "loss": 1.4531, "step": 9697 }, { "epoch": 4.5637647058823525, "grad_norm": 1.1634987758700388, "learning_rate": 5.581217501657666e-05, "loss": 1.6406, "step": 9698 }, { "epoch": 4.564235294117647, "grad_norm": 1.2183697900930666, "learning_rate": 5.577806227086057e-05, "loss": 1.1406, "step": 9699 }, { "epoch": 4.564705882352941, "grad_norm": 1.203103218214644, "learning_rate": 5.57439575721601e-05, "loss": 0.793, "step": 9700 }, { "epoch": 4.565176470588235, "grad_norm": 1.1115331691340673, "learning_rate": 5.570986092338806e-05, "loss": 0.5078, "step": 9701 }, { "epoch": 4.565647058823529, "grad_norm": 0.6540230489905952, "learning_rate": 5.5675772327456313e-05, "loss": 0.4082, "step": 9702 }, { "epoch": 4.566117647058824, "grad_norm": 0.970821302035401, "learning_rate": 5.5641691787276286e-05, "loss": 0.5352, "step": 9703 }, { "epoch": 4.566588235294118, "grad_norm": 1.33348656853619, "learning_rate": 5.560761930575864e-05, "loss": 1.3594, "step": 9704 }, { "epoch": 4.567058823529412, "grad_norm": 1.3497451251115362, "learning_rate": 5.557355488581325e-05, "loss": 1.3047, "step": 9705 }, { "epoch": 4.567529411764706, "grad_norm": 0.9546519624647811, "learning_rate": 5.553949853034943e-05, "loss": 0.7227, "step": 9706 }, { "epoch": 4.568, "grad_norm": 1.0972055710156663, "learning_rate": 5.550545024227582e-05, "loss": 0.8125, "step": 9707 }, { "epoch": 4.568470588235294, "grad_norm": 0.86509891355876, "learning_rate": 5.5471410024500156e-05, "loss": 0.3652, "step": 9708 }, { "epoch": 4.568941176470588, "grad_norm": 1.3333239263212873, "learning_rate": 5.5437377879929826e-05, "loss": 0.6758, "step": 9709 }, { "epoch": 4.569411764705882, "grad_norm": 0.8239128640431951, "learning_rate": 5.5403353811471194e-05, "loss": 0.8164, "step": 9710 }, { "epoch": 4.5698823529411765, "grad_norm": 0.6311123350787922, "learning_rate": 5.5369337822030154e-05, "loss": 0.1807, "step": 9711 }, { "epoch": 4.570352941176471, "grad_norm": 1.0696850123022874, "learning_rate": 5.533532991451187e-05, "loss": 1.0859, "step": 9712 }, { "epoch": 4.570823529411765, "grad_norm": 0.9079944580465124, "learning_rate": 5.53013300918207e-05, "loss": 0.4238, "step": 9713 }, { "epoch": 4.571294117647059, "grad_norm": 0.676282381869602, "learning_rate": 5.526733835686041e-05, "loss": 0.5938, "step": 9714 }, { "epoch": 4.571764705882353, "grad_norm": 1.2557280268965776, "learning_rate": 5.523335471253416e-05, "loss": 1.2969, "step": 9715 }, { "epoch": 4.572235294117647, "grad_norm": 1.4541393641106677, "learning_rate": 5.519937916174419e-05, "loss": 1.0391, "step": 9716 }, { "epoch": 4.572705882352941, "grad_norm": 1.4164900111196133, "learning_rate": 5.516541170739223e-05, "loss": 1.1797, "step": 9717 }, { "epoch": 4.573176470588235, "grad_norm": 0.8738902142034091, "learning_rate": 5.5131452352379314e-05, "loss": 0.6562, "step": 9718 }, { "epoch": 4.573647058823529, "grad_norm": 1.109904682299728, "learning_rate": 5.509750109960566e-05, "loss": 0.4004, "step": 9719 }, { "epoch": 4.574117647058824, "grad_norm": 1.0464348780521495, "learning_rate": 5.5063557951970885e-05, "loss": 0.8906, "step": 9720 }, { "epoch": 4.574588235294118, "grad_norm": 0.5446936027432939, "learning_rate": 5.502962291237399e-05, "loss": 0.4414, "step": 9721 }, { "epoch": 4.575058823529412, "grad_norm": 0.6952965888550902, "learning_rate": 5.4995695983713076e-05, "loss": 0.2637, "step": 9722 }, { "epoch": 4.575529411764705, "grad_norm": 1.274408078410462, "learning_rate": 5.496177716888571e-05, "loss": 1.5078, "step": 9723 }, { "epoch": 4.576, "grad_norm": 1.4290063022622705, "learning_rate": 5.4927866470788786e-05, "loss": 1.1484, "step": 9724 }, { "epoch": 4.576470588235294, "grad_norm": 0.8044745618397822, "learning_rate": 5.489396389231834e-05, "loss": 0.2139, "step": 9725 }, { "epoch": 4.576941176470588, "grad_norm": 1.0430788451750461, "learning_rate": 5.4860069436369916e-05, "loss": 0.5547, "step": 9726 }, { "epoch": 4.577411764705882, "grad_norm": 1.0730575664811495, "learning_rate": 5.4826183105838184e-05, "loss": 0.6016, "step": 9727 }, { "epoch": 4.5778823529411765, "grad_norm": 0.9802868065110949, "learning_rate": 5.4792304903617226e-05, "loss": 0.7656, "step": 9728 }, { "epoch": 4.578352941176471, "grad_norm": 1.225962906443054, "learning_rate": 5.4758434832600475e-05, "loss": 0.7656, "step": 9729 }, { "epoch": 4.578823529411765, "grad_norm": 0.6298794895442801, "learning_rate": 5.47245728956805e-05, "loss": 0.4355, "step": 9730 }, { "epoch": 4.579294117647059, "grad_norm": 0.9447302776780656, "learning_rate": 5.469071909574933e-05, "loss": 0.4473, "step": 9731 }, { "epoch": 4.5797647058823525, "grad_norm": 1.604604699367537, "learning_rate": 5.465687343569828e-05, "loss": 2.3906, "step": 9732 }, { "epoch": 4.580235294117647, "grad_norm": 1.2384148240143529, "learning_rate": 5.462303591841783e-05, "loss": 1.2109, "step": 9733 }, { "epoch": 4.580705882352941, "grad_norm": 0.905109592606381, "learning_rate": 5.458920654679795e-05, "loss": 0.3379, "step": 9734 }, { "epoch": 4.581176470588235, "grad_norm": 1.3884104442970686, "learning_rate": 5.4555385323727876e-05, "loss": 1.0625, "step": 9735 }, { "epoch": 4.581647058823529, "grad_norm": 0.9749549899579757, "learning_rate": 5.4521572252095985e-05, "loss": 1.2656, "step": 9736 }, { "epoch": 4.582117647058824, "grad_norm": 0.9144503105826194, "learning_rate": 5.4487767334790145e-05, "loss": 0.2451, "step": 9737 }, { "epoch": 4.582588235294118, "grad_norm": 0.915144634796087, "learning_rate": 5.445397057469751e-05, "loss": 0.8203, "step": 9738 }, { "epoch": 4.583058823529412, "grad_norm": 0.8357276204247608, "learning_rate": 5.4420181974704386e-05, "loss": 0.3008, "step": 9739 }, { "epoch": 4.583529411764706, "grad_norm": 1.2605575729479241, "learning_rate": 5.4386401537696536e-05, "loss": 0.9453, "step": 9740 }, { "epoch": 4.584, "grad_norm": 0.9625651796380623, "learning_rate": 5.435262926655903e-05, "loss": 0.3965, "step": 9741 }, { "epoch": 4.584470588235294, "grad_norm": 0.9166390601105224, "learning_rate": 5.43188651641761e-05, "loss": 0.375, "step": 9742 }, { "epoch": 4.584941176470588, "grad_norm": 1.1842639568746685, "learning_rate": 5.428510923343136e-05, "loss": 1.0781, "step": 9743 }, { "epoch": 4.585411764705882, "grad_norm": 0.9990332829856636, "learning_rate": 5.425136147720785e-05, "loss": 0.8633, "step": 9744 }, { "epoch": 4.5858823529411765, "grad_norm": 1.2221433668020605, "learning_rate": 5.421762189838765e-05, "loss": 0.8828, "step": 9745 }, { "epoch": 4.586352941176471, "grad_norm": 1.0300811438243245, "learning_rate": 5.4183890499852364e-05, "loss": 0.3086, "step": 9746 }, { "epoch": 4.586823529411765, "grad_norm": 1.068231327961651, "learning_rate": 5.4150167284482844e-05, "loss": 0.7891, "step": 9747 }, { "epoch": 4.587294117647059, "grad_norm": 0.9650786021861792, "learning_rate": 5.4116452255159126e-05, "loss": 0.6992, "step": 9748 }, { "epoch": 4.587764705882353, "grad_norm": 1.5576856288772136, "learning_rate": 5.408274541476075e-05, "loss": 0.8672, "step": 9749 }, { "epoch": 4.588235294117647, "grad_norm": 1.144686083975884, "learning_rate": 5.4049046766166335e-05, "loss": 1.0859, "step": 9750 }, { "epoch": 4.588705882352941, "grad_norm": 0.840937831674147, "learning_rate": 5.401535631225396e-05, "loss": 0.3516, "step": 9751 }, { "epoch": 4.589176470588235, "grad_norm": 1.1468617463862414, "learning_rate": 5.398167405590096e-05, "loss": 1.3984, "step": 9752 }, { "epoch": 4.589647058823529, "grad_norm": 1.322430468648046, "learning_rate": 5.3947999999983967e-05, "loss": 0.6484, "step": 9753 }, { "epoch": 4.590117647058824, "grad_norm": 0.8487588319300923, "learning_rate": 5.3914334147378894e-05, "loss": 0.2324, "step": 9754 }, { "epoch": 4.590588235294118, "grad_norm": 1.2699426408437002, "learning_rate": 5.388067650096106e-05, "loss": 0.918, "step": 9755 }, { "epoch": 4.591058823529412, "grad_norm": 0.9372448841944372, "learning_rate": 5.384702706360483e-05, "loss": 0.6445, "step": 9756 }, { "epoch": 4.591529411764705, "grad_norm": 1.2568184849605628, "learning_rate": 5.3813385838184185e-05, "loss": 1.0312, "step": 9757 }, { "epoch": 4.592, "grad_norm": 0.7109039689869682, "learning_rate": 5.3779752827572126e-05, "loss": 0.3535, "step": 9758 }, { "epoch": 4.592470588235294, "grad_norm": 1.1005195568358999, "learning_rate": 5.374612803464115e-05, "loss": 0.6406, "step": 9759 }, { "epoch": 4.592941176470588, "grad_norm": 0.9823927357093958, "learning_rate": 5.3712511462262994e-05, "loss": 0.4473, "step": 9760 }, { "epoch": 4.593411764705882, "grad_norm": 0.9281632534453429, "learning_rate": 5.36789031133086e-05, "loss": 0.373, "step": 9761 }, { "epoch": 4.5938823529411765, "grad_norm": 1.3774388771564836, "learning_rate": 5.364530299064834e-05, "loss": 0.5156, "step": 9762 }, { "epoch": 4.594352941176471, "grad_norm": 1.2465530824001223, "learning_rate": 5.361171109715186e-05, "loss": 1.2734, "step": 9763 }, { "epoch": 4.594823529411765, "grad_norm": 1.086658675081502, "learning_rate": 5.357812743568799e-05, "loss": 0.6602, "step": 9764 }, { "epoch": 4.595294117647059, "grad_norm": 1.481409401404394, "learning_rate": 5.354455200912498e-05, "loss": 1.6328, "step": 9765 }, { "epoch": 4.5957647058823525, "grad_norm": 1.2976623138012462, "learning_rate": 5.35109848203304e-05, "loss": 0.7031, "step": 9766 }, { "epoch": 4.596235294117647, "grad_norm": 1.2933638329800687, "learning_rate": 5.3477425872170926e-05, "loss": 1.9219, "step": 9767 }, { "epoch": 4.596705882352941, "grad_norm": 0.8484179442155624, "learning_rate": 5.3443875167512736e-05, "loss": 0.8594, "step": 9768 }, { "epoch": 4.597176470588235, "grad_norm": 1.265812034134047, "learning_rate": 5.341033270922126e-05, "loss": 0.7773, "step": 9769 }, { "epoch": 4.597647058823529, "grad_norm": 0.6547309392813203, "learning_rate": 5.337679850016109e-05, "loss": 0.3008, "step": 9770 }, { "epoch": 4.598117647058824, "grad_norm": 0.7924238619197014, "learning_rate": 5.3343272543196264e-05, "loss": 0.6445, "step": 9771 }, { "epoch": 4.598588235294118, "grad_norm": 0.5810669971331208, "learning_rate": 5.33097548411901e-05, "loss": 0.0962, "step": 9772 }, { "epoch": 4.599058823529412, "grad_norm": 1.1682982059367213, "learning_rate": 5.3276245397005094e-05, "loss": 1.1094, "step": 9773 }, { "epoch": 4.599529411764706, "grad_norm": 1.2176983802777108, "learning_rate": 5.324274421350316e-05, "loss": 1.0312, "step": 9774 }, { "epoch": 4.6, "grad_norm": 1.1696322221141169, "learning_rate": 5.3209251293545514e-05, "loss": 1.0, "step": 9775 }, { "epoch": 4.600470588235294, "grad_norm": 1.3958044998034653, "learning_rate": 5.317576663999251e-05, "loss": 1.2109, "step": 9776 }, { "epoch": 4.600941176470588, "grad_norm": 1.445219414110646, "learning_rate": 5.314229025570401e-05, "loss": 0.9102, "step": 9777 }, { "epoch": 4.601411764705882, "grad_norm": 1.0880080428148742, "learning_rate": 5.3108822143538944e-05, "loss": 0.6445, "step": 9778 }, { "epoch": 4.6018823529411765, "grad_norm": 0.7093897201647066, "learning_rate": 5.307536230635571e-05, "loss": 0.2754, "step": 9779 }, { "epoch": 4.602352941176471, "grad_norm": 1.5179522156304504, "learning_rate": 5.304191074701201e-05, "loss": 1.1875, "step": 9780 }, { "epoch": 4.602823529411765, "grad_norm": 1.6660763408499, "learning_rate": 5.3008467468364675e-05, "loss": 1.3203, "step": 9781 }, { "epoch": 4.603294117647059, "grad_norm": 1.383079930455923, "learning_rate": 5.297503247326993e-05, "loss": 1.0859, "step": 9782 }, { "epoch": 4.603764705882353, "grad_norm": 1.0552788194291238, "learning_rate": 5.294160576458339e-05, "loss": 0.3398, "step": 9783 }, { "epoch": 4.604235294117647, "grad_norm": 0.7219243071903325, "learning_rate": 5.290818734515972e-05, "loss": 0.2021, "step": 9784 }, { "epoch": 4.604705882352941, "grad_norm": 0.9588091246371239, "learning_rate": 5.287477721785309e-05, "loss": 0.8477, "step": 9785 }, { "epoch": 4.605176470588235, "grad_norm": 1.1467643003861459, "learning_rate": 5.284137538551693e-05, "loss": 0.6172, "step": 9786 }, { "epoch": 4.605647058823529, "grad_norm": 1.1452943881430646, "learning_rate": 5.2807981851003825e-05, "loss": 0.8398, "step": 9787 }, { "epoch": 4.606117647058824, "grad_norm": 1.2147793725139764, "learning_rate": 5.2774596617165793e-05, "loss": 0.8633, "step": 9788 }, { "epoch": 4.606588235294118, "grad_norm": 1.8207074074401446, "learning_rate": 5.2741219686854146e-05, "loss": 0.6211, "step": 9789 }, { "epoch": 4.607058823529412, "grad_norm": 1.3224128536118556, "learning_rate": 5.270785106291934e-05, "loss": 0.4434, "step": 9790 }, { "epoch": 4.607529411764705, "grad_norm": 0.8936422936952755, "learning_rate": 5.2674490748211265e-05, "loss": 0.6992, "step": 9791 }, { "epoch": 4.608, "grad_norm": 1.2403696234811057, "learning_rate": 5.2641138745579126e-05, "loss": 0.8984, "step": 9792 }, { "epoch": 4.608470588235294, "grad_norm": 0.4011145536356327, "learning_rate": 5.2607795057871216e-05, "loss": 0.377, "step": 9793 }, { "epoch": 4.608941176470588, "grad_norm": 0.5450315381091303, "learning_rate": 5.257445968793531e-05, "loss": 0.1943, "step": 9794 }, { "epoch": 4.609411764705882, "grad_norm": 1.1045392077530274, "learning_rate": 5.254113263861848e-05, "loss": 0.3848, "step": 9795 }, { "epoch": 4.6098823529411765, "grad_norm": 0.8689622066839654, "learning_rate": 5.250781391276684e-05, "loss": 0.2832, "step": 9796 }, { "epoch": 4.610352941176471, "grad_norm": 1.3811188272312103, "learning_rate": 5.24745035132262e-05, "loss": 0.7891, "step": 9797 }, { "epoch": 4.610823529411765, "grad_norm": 0.85605025541013, "learning_rate": 5.2441201442841264e-05, "loss": 0.3535, "step": 9798 }, { "epoch": 4.611294117647059, "grad_norm": 0.7640677891828669, "learning_rate": 5.240790770445625e-05, "loss": 0.248, "step": 9799 }, { "epoch": 4.6117647058823525, "grad_norm": 1.6903225900099188, "learning_rate": 5.237462230091467e-05, "loss": 1.3672, "step": 9800 }, { "epoch": 4.612235294117647, "grad_norm": 1.3030065716131216, "learning_rate": 5.234134523505912e-05, "loss": 1.3984, "step": 9801 }, { "epoch": 4.612705882352941, "grad_norm": 1.2734537402070958, "learning_rate": 5.230807650973172e-05, "loss": 0.6602, "step": 9802 }, { "epoch": 4.613176470588235, "grad_norm": 0.517257096198073, "learning_rate": 5.227481612777379e-05, "loss": 0.166, "step": 9803 }, { "epoch": 4.613647058823529, "grad_norm": 1.2045633210882978, "learning_rate": 5.224156409202587e-05, "loss": 0.9141, "step": 9804 }, { "epoch": 4.614117647058824, "grad_norm": 1.0838926930540524, "learning_rate": 5.2208320405327865e-05, "loss": 0.6211, "step": 9805 }, { "epoch": 4.614588235294118, "grad_norm": 0.8088182519891768, "learning_rate": 5.2175085070519024e-05, "loss": 0.7852, "step": 9806 }, { "epoch": 4.615058823529412, "grad_norm": 0.08060363170157914, "learning_rate": 5.2141858090437696e-05, "loss": 0.0173, "step": 9807 }, { "epoch": 4.615529411764706, "grad_norm": 1.2667129974072673, "learning_rate": 5.210863946792172e-05, "loss": 0.707, "step": 9808 }, { "epoch": 4.616, "grad_norm": 0.9147418688652268, "learning_rate": 5.207542920580805e-05, "loss": 0.2441, "step": 9809 }, { "epoch": 4.616470588235294, "grad_norm": 1.0961913500177314, "learning_rate": 5.204222730693303e-05, "loss": 1.125, "step": 9810 }, { "epoch": 4.616941176470588, "grad_norm": 1.3537432579508566, "learning_rate": 5.2009033774132345e-05, "loss": 1.5625, "step": 9811 }, { "epoch": 4.617411764705882, "grad_norm": 1.0346485137753743, "learning_rate": 5.197584861024076e-05, "loss": 0.8789, "step": 9812 }, { "epoch": 4.6178823529411765, "grad_norm": 0.8372301268597925, "learning_rate": 5.194267181809251e-05, "loss": 0.5586, "step": 9813 }, { "epoch": 4.618352941176471, "grad_norm": 0.8742377680885555, "learning_rate": 5.1909503400521104e-05, "loss": 0.3301, "step": 9814 }, { "epoch": 4.618823529411765, "grad_norm": 0.690657740490874, "learning_rate": 5.187634336035918e-05, "loss": 0.2422, "step": 9815 }, { "epoch": 4.619294117647058, "grad_norm": 1.3198632262354686, "learning_rate": 5.184319170043884e-05, "loss": 0.6602, "step": 9816 }, { "epoch": 4.619764705882353, "grad_norm": 1.1435257392574154, "learning_rate": 5.1810048423591424e-05, "loss": 0.8555, "step": 9817 }, { "epoch": 4.620235294117647, "grad_norm": 0.7891654431234533, "learning_rate": 5.177691353264746e-05, "loss": 0.207, "step": 9818 }, { "epoch": 4.620705882352941, "grad_norm": 0.33198406742135445, "learning_rate": 5.174378703043684e-05, "loss": 0.0403, "step": 9819 }, { "epoch": 4.621176470588235, "grad_norm": 1.3288371354087793, "learning_rate": 5.171066891978881e-05, "loss": 1.3906, "step": 9820 }, { "epoch": 4.621647058823529, "grad_norm": 0.9626714101681872, "learning_rate": 5.167755920353171e-05, "loss": 0.9453, "step": 9821 }, { "epoch": 4.622117647058824, "grad_norm": 1.4005339457132955, "learning_rate": 5.164445788449334e-05, "loss": 1.1328, "step": 9822 }, { "epoch": 4.622588235294118, "grad_norm": 1.3742206216336568, "learning_rate": 5.161136496550072e-05, "loss": 0.875, "step": 9823 }, { "epoch": 4.623058823529412, "grad_norm": 0.5096124085158076, "learning_rate": 5.15782804493801e-05, "loss": 0.4121, "step": 9824 }, { "epoch": 4.623529411764705, "grad_norm": 0.8717434170701184, "learning_rate": 5.1545204338957066e-05, "loss": 0.291, "step": 9825 }, { "epoch": 4.624, "grad_norm": 0.9821651555281503, "learning_rate": 5.151213663705655e-05, "loss": 0.9219, "step": 9826 }, { "epoch": 4.624470588235294, "grad_norm": 0.8019669721536842, "learning_rate": 5.14790773465026e-05, "loss": 0.1797, "step": 9827 }, { "epoch": 4.624941176470588, "grad_norm": 1.189124305892655, "learning_rate": 5.144602647011874e-05, "loss": 0.7617, "step": 9828 }, { "epoch": 4.625411764705882, "grad_norm": 1.6666656507903064, "learning_rate": 5.141298401072758e-05, "loss": 1.3359, "step": 9829 }, { "epoch": 4.625882352941177, "grad_norm": 0.9717166697871018, "learning_rate": 5.137994997115115e-05, "loss": 0.6562, "step": 9830 }, { "epoch": 4.626352941176471, "grad_norm": 0.8934110992342975, "learning_rate": 5.134692435421077e-05, "loss": 0.2471, "step": 9831 }, { "epoch": 4.626823529411765, "grad_norm": 0.889738610838554, "learning_rate": 5.1313907162726883e-05, "loss": 0.332, "step": 9832 }, { "epoch": 4.627294117647059, "grad_norm": 1.0622452542131684, "learning_rate": 5.128089839951937e-05, "loss": 0.8867, "step": 9833 }, { "epoch": 4.627764705882353, "grad_norm": 1.2362710047609513, "learning_rate": 5.124789806740742e-05, "loss": 1.3438, "step": 9834 }, { "epoch": 4.628235294117647, "grad_norm": 1.239685374666195, "learning_rate": 5.12149061692093e-05, "loss": 1.3047, "step": 9835 }, { "epoch": 4.628705882352941, "grad_norm": 1.198925035442909, "learning_rate": 5.118192270774271e-05, "loss": 0.5703, "step": 9836 }, { "epoch": 4.629176470588235, "grad_norm": 1.0846745966583566, "learning_rate": 5.114894768582468e-05, "loss": 0.5977, "step": 9837 }, { "epoch": 4.6296470588235294, "grad_norm": 1.1243104409783284, "learning_rate": 5.111598110627133e-05, "loss": 0.543, "step": 9838 }, { "epoch": 4.630117647058824, "grad_norm": 1.3466686484995134, "learning_rate": 5.108302297189822e-05, "loss": 1.0938, "step": 9839 }, { "epoch": 4.630588235294118, "grad_norm": 1.0864568230849343, "learning_rate": 5.105007328552019e-05, "loss": 0.5352, "step": 9840 }, { "epoch": 4.631058823529412, "grad_norm": 1.0170142193177643, "learning_rate": 5.101713204995114e-05, "loss": 0.8438, "step": 9841 }, { "epoch": 4.631529411764706, "grad_norm": 1.3943370948048568, "learning_rate": 5.0984199268004615e-05, "loss": 0.7656, "step": 9842 }, { "epoch": 4.632, "grad_norm": 0.7837174423842105, "learning_rate": 5.095127494249309e-05, "loss": 0.2949, "step": 9843 }, { "epoch": 4.632470588235294, "grad_norm": 0.9156323272533092, "learning_rate": 5.091835907622854e-05, "loss": 0.3633, "step": 9844 }, { "epoch": 4.632941176470588, "grad_norm": 1.105520440065371, "learning_rate": 5.0885451672022156e-05, "loss": 0.5117, "step": 9845 }, { "epoch": 4.633411764705882, "grad_norm": 0.8689280213455902, "learning_rate": 5.0852552732684296e-05, "loss": 0.9336, "step": 9846 }, { "epoch": 4.633882352941177, "grad_norm": 0.7870447772871231, "learning_rate": 5.081966226102474e-05, "loss": 0.5547, "step": 9847 }, { "epoch": 4.634352941176471, "grad_norm": 0.7929865013367641, "learning_rate": 5.078678025985257e-05, "loss": 0.3438, "step": 9848 }, { "epoch": 4.634823529411765, "grad_norm": 1.2280894932513482, "learning_rate": 5.0753906731975954e-05, "loss": 1.4531, "step": 9849 }, { "epoch": 4.635294117647058, "grad_norm": 1.0745155292536048, "learning_rate": 5.07210416802025e-05, "loss": 0.7969, "step": 9850 }, { "epoch": 4.635764705882353, "grad_norm": 0.990277445789599, "learning_rate": 5.0688185107339084e-05, "loss": 0.4062, "step": 9851 }, { "epoch": 4.636235294117647, "grad_norm": 0.830794906000249, "learning_rate": 5.065533701619175e-05, "loss": 0.625, "step": 9852 }, { "epoch": 4.636705882352941, "grad_norm": 0.9818381361608313, "learning_rate": 5.06224974095659e-05, "loss": 0.3965, "step": 9853 }, { "epoch": 4.637176470588235, "grad_norm": 1.1223806929319324, "learning_rate": 5.058966629026629e-05, "loss": 0.918, "step": 9854 }, { "epoch": 4.6376470588235295, "grad_norm": 1.0297246784852772, "learning_rate": 5.0556843661096714e-05, "loss": 0.5039, "step": 9855 }, { "epoch": 4.638117647058824, "grad_norm": 0.9548964208574112, "learning_rate": 5.052402952486046e-05, "loss": 0.75, "step": 9856 }, { "epoch": 4.638588235294118, "grad_norm": 1.0207178683722355, "learning_rate": 5.0491223884360054e-05, "loss": 0.6484, "step": 9857 }, { "epoch": 4.639058823529412, "grad_norm": 1.0254228294701582, "learning_rate": 5.0458426742397174e-05, "loss": 0.4941, "step": 9858 }, { "epoch": 4.6395294117647055, "grad_norm": 0.8059366206480423, "learning_rate": 5.042563810177296e-05, "loss": 0.6328, "step": 9859 }, { "epoch": 4.64, "grad_norm": 1.3649199015037343, "learning_rate": 5.039285796528759e-05, "loss": 0.7852, "step": 9860 }, { "epoch": 4.640470588235294, "grad_norm": 1.3023038872290906, "learning_rate": 5.0360086335740734e-05, "loss": 1.0547, "step": 9861 }, { "epoch": 4.640941176470588, "grad_norm": 1.337713514328821, "learning_rate": 5.032732321593128e-05, "loss": 0.9766, "step": 9862 }, { "epoch": 4.641411764705882, "grad_norm": 0.8053989082967302, "learning_rate": 5.029456860865727e-05, "loss": 0.5703, "step": 9863 }, { "epoch": 4.641882352941177, "grad_norm": 1.0208494054190298, "learning_rate": 5.0261822516716146e-05, "loss": 0.3652, "step": 9864 }, { "epoch": 4.642352941176471, "grad_norm": 1.210182406532632, "learning_rate": 5.022908494290465e-05, "loss": 0.5625, "step": 9865 }, { "epoch": 4.642823529411765, "grad_norm": 0.9674250700164749, "learning_rate": 5.0196355890018605e-05, "loss": 0.4531, "step": 9866 }, { "epoch": 4.643294117647059, "grad_norm": 1.0170614048697157, "learning_rate": 5.016363536085333e-05, "loss": 0.8945, "step": 9867 }, { "epoch": 4.643764705882353, "grad_norm": 0.9613055322138583, "learning_rate": 5.0130923358203316e-05, "loss": 0.208, "step": 9868 }, { "epoch": 4.644235294117647, "grad_norm": 1.094472958935388, "learning_rate": 5.0098219884862265e-05, "loss": 0.7109, "step": 9869 }, { "epoch": 4.644705882352941, "grad_norm": 0.8424251427288296, "learning_rate": 5.006552494362325e-05, "loss": 0.2734, "step": 9870 }, { "epoch": 4.645176470588235, "grad_norm": 1.0297869984982642, "learning_rate": 5.0032838537278627e-05, "loss": 0.707, "step": 9871 }, { "epoch": 4.6456470588235295, "grad_norm": 1.6229477154232796, "learning_rate": 5.000016066861988e-05, "loss": 1.4297, "step": 9872 }, { "epoch": 4.646117647058824, "grad_norm": 1.319664124083472, "learning_rate": 4.9967491340437904e-05, "loss": 0.9609, "step": 9873 }, { "epoch": 4.646588235294118, "grad_norm": 1.2514222141208104, "learning_rate": 4.993483055552289e-05, "loss": 1.1094, "step": 9874 }, { "epoch": 4.647058823529412, "grad_norm": 1.2646200960595568, "learning_rate": 4.990217831666409e-05, "loss": 1.2578, "step": 9875 }, { "epoch": 4.647529411764706, "grad_norm": 0.9753214549705851, "learning_rate": 4.986953462665026e-05, "loss": 0.5039, "step": 9876 }, { "epoch": 4.648, "grad_norm": 1.1366731659994782, "learning_rate": 4.9836899488269345e-05, "loss": 1.0, "step": 9877 }, { "epoch": 4.648470588235294, "grad_norm": 1.1884132083797283, "learning_rate": 4.9804272904308465e-05, "loss": 0.793, "step": 9878 }, { "epoch": 4.648941176470588, "grad_norm": 0.9961387238388529, "learning_rate": 4.97716548775542e-05, "loss": 0.9961, "step": 9879 }, { "epoch": 4.649411764705882, "grad_norm": 0.9896215071129274, "learning_rate": 4.9739045410792164e-05, "loss": 0.4609, "step": 9880 }, { "epoch": 4.649882352941177, "grad_norm": 0.6459526453156289, "learning_rate": 4.970644450680742e-05, "loss": 0.1699, "step": 9881 }, { "epoch": 4.650352941176471, "grad_norm": 1.3941655078102357, "learning_rate": 4.9673852168384316e-05, "loss": 1.7266, "step": 9882 }, { "epoch": 4.650823529411765, "grad_norm": 0.8602906508454647, "learning_rate": 4.964126839830628e-05, "loss": 0.4355, "step": 9883 }, { "epoch": 4.651294117647058, "grad_norm": 0.8557247857983608, "learning_rate": 4.960869319935616e-05, "loss": 0.4551, "step": 9884 }, { "epoch": 4.651764705882353, "grad_norm": 1.1787322071550497, "learning_rate": 4.9576126574316086e-05, "loss": 0.6328, "step": 9885 }, { "epoch": 4.652235294117647, "grad_norm": 1.4028020729828332, "learning_rate": 4.9543568525967376e-05, "loss": 1.0703, "step": 9886 }, { "epoch": 4.652705882352941, "grad_norm": 1.1184580673040918, "learning_rate": 4.9511019057090684e-05, "loss": 1.0, "step": 9887 }, { "epoch": 4.653176470588235, "grad_norm": 1.0721457194061983, "learning_rate": 4.947847817046582e-05, "loss": 0.9492, "step": 9888 }, { "epoch": 4.6536470588235295, "grad_norm": 1.2438127574306785, "learning_rate": 4.9445945868871974e-05, "loss": 0.6523, "step": 9889 }, { "epoch": 4.654117647058824, "grad_norm": 1.1848226936796176, "learning_rate": 4.941342215508761e-05, "loss": 0.707, "step": 9890 }, { "epoch": 4.654588235294118, "grad_norm": 0.8351693376420685, "learning_rate": 4.9380907031890315e-05, "loss": 0.3496, "step": 9891 }, { "epoch": 4.655058823529412, "grad_norm": 0.8275389994823127, "learning_rate": 4.934840050205711e-05, "loss": 0.4395, "step": 9892 }, { "epoch": 4.6555294117647055, "grad_norm": 0.8000231218735305, "learning_rate": 4.931590256836424e-05, "loss": 0.2852, "step": 9893 }, { "epoch": 4.656, "grad_norm": 0.9956686618062329, "learning_rate": 4.928341323358708e-05, "loss": 0.3711, "step": 9894 }, { "epoch": 4.656470588235294, "grad_norm": 1.0746626041649394, "learning_rate": 4.925093250050045e-05, "loss": 1.2031, "step": 9895 }, { "epoch": 4.656941176470588, "grad_norm": 1.0401431242601904, "learning_rate": 4.921846037187841e-05, "loss": 0.4551, "step": 9896 }, { "epoch": 4.657411764705882, "grad_norm": 1.3648410355171652, "learning_rate": 4.918599685049414e-05, "loss": 0.6602, "step": 9897 }, { "epoch": 4.657882352941177, "grad_norm": 1.1039341708226174, "learning_rate": 4.915354193912021e-05, "loss": 1.375, "step": 9898 }, { "epoch": 4.658352941176471, "grad_norm": 0.8136543588689901, "learning_rate": 4.9121095640528516e-05, "loss": 0.2559, "step": 9899 }, { "epoch": 4.658823529411765, "grad_norm": 1.1448488583733547, "learning_rate": 4.908865795748999e-05, "loss": 0.9844, "step": 9900 }, { "epoch": 4.659294117647059, "grad_norm": 1.2733545651065166, "learning_rate": 4.905622889277506e-05, "loss": 1.2188, "step": 9901 }, { "epoch": 4.659764705882353, "grad_norm": 0.7485816759280473, "learning_rate": 4.902380844915337e-05, "loss": 0.7148, "step": 9902 }, { "epoch": 4.660235294117647, "grad_norm": 1.3498887427061566, "learning_rate": 4.899139662939366e-05, "loss": 0.4902, "step": 9903 }, { "epoch": 4.660705882352941, "grad_norm": 1.017235776990581, "learning_rate": 4.895899343626412e-05, "loss": 0.5469, "step": 9904 }, { "epoch": 4.661176470588235, "grad_norm": 0.9698359478524031, "learning_rate": 4.892659887253219e-05, "loss": 0.8008, "step": 9905 }, { "epoch": 4.6616470588235295, "grad_norm": 1.0945956342173664, "learning_rate": 4.8894212940964445e-05, "loss": 0.9844, "step": 9906 }, { "epoch": 4.662117647058824, "grad_norm": 1.1433899763934212, "learning_rate": 4.8861835644326866e-05, "loss": 0.4043, "step": 9907 }, { "epoch": 4.662588235294118, "grad_norm": 1.4577393134482421, "learning_rate": 4.882946698538457e-05, "loss": 0.8984, "step": 9908 }, { "epoch": 4.663058823529412, "grad_norm": 0.9506679880708945, "learning_rate": 4.879710696690202e-05, "loss": 0.373, "step": 9909 }, { "epoch": 4.663529411764706, "grad_norm": 0.31273371158870406, "learning_rate": 4.876475559164299e-05, "loss": 0.0166, "step": 9910 }, { "epoch": 4.664, "grad_norm": 1.2112624274486574, "learning_rate": 4.873241286237033e-05, "loss": 1.3047, "step": 9911 }, { "epoch": 4.664470588235294, "grad_norm": 0.6698672072600703, "learning_rate": 4.8700078781846326e-05, "loss": 0.4453, "step": 9912 }, { "epoch": 4.664941176470588, "grad_norm": 0.988691003763104, "learning_rate": 4.866775335283253e-05, "loss": 0.4902, "step": 9913 }, { "epoch": 4.665411764705882, "grad_norm": 0.6759295875035874, "learning_rate": 4.863543657808956e-05, "loss": 0.1924, "step": 9914 }, { "epoch": 4.665882352941177, "grad_norm": 1.1650625885262358, "learning_rate": 4.860312846037753e-05, "loss": 1.0625, "step": 9915 }, { "epoch": 4.666352941176471, "grad_norm": 1.0805451911891137, "learning_rate": 4.857082900245569e-05, "loss": 0.4141, "step": 9916 }, { "epoch": 4.666823529411765, "grad_norm": 0.9513240131727732, "learning_rate": 4.8538538207082534e-05, "loss": 0.4277, "step": 9917 }, { "epoch": 4.667294117647058, "grad_norm": 0.8987868641266206, "learning_rate": 4.850625607701588e-05, "loss": 0.6641, "step": 9918 }, { "epoch": 4.667764705882353, "grad_norm": 1.082161205639246, "learning_rate": 4.8473982615012816e-05, "loss": 0.4062, "step": 9919 }, { "epoch": 4.668235294117647, "grad_norm": 1.0352641079684166, "learning_rate": 4.844171782382959e-05, "loss": 1.1094, "step": 9920 }, { "epoch": 4.668705882352941, "grad_norm": 0.6020900818246494, "learning_rate": 4.840946170622179e-05, "loss": 0.168, "step": 9921 }, { "epoch": 4.669176470588235, "grad_norm": 0.9890989294540544, "learning_rate": 4.8377214264944296e-05, "loss": 0.6875, "step": 9922 }, { "epoch": 4.6696470588235295, "grad_norm": 0.8632237353028336, "learning_rate": 4.834497550275114e-05, "loss": 0.9766, "step": 9923 }, { "epoch": 4.670117647058824, "grad_norm": 1.4405546605344082, "learning_rate": 4.831274542239566e-05, "loss": 1.0391, "step": 9924 }, { "epoch": 4.670588235294118, "grad_norm": 0.8453444106387176, "learning_rate": 4.8280524026630565e-05, "loss": 0.3828, "step": 9925 }, { "epoch": 4.671058823529412, "grad_norm": 0.6924529893341839, "learning_rate": 4.824831131820759e-05, "loss": 0.1885, "step": 9926 }, { "epoch": 4.6715294117647055, "grad_norm": 1.2541638670333872, "learning_rate": 4.821610729987792e-05, "loss": 0.7812, "step": 9927 }, { "epoch": 4.672, "grad_norm": 1.3750293059969143, "learning_rate": 4.818391197439199e-05, "loss": 1.0078, "step": 9928 }, { "epoch": 4.672470588235294, "grad_norm": 1.0139833588345932, "learning_rate": 4.815172534449933e-05, "loss": 0.9648, "step": 9929 }, { "epoch": 4.672941176470588, "grad_norm": 0.8955967429182066, "learning_rate": 4.811954741294889e-05, "loss": 0.5625, "step": 9930 }, { "epoch": 4.673411764705882, "grad_norm": 0.6715810362062873, "learning_rate": 4.808737818248882e-05, "loss": 0.3066, "step": 9931 }, { "epoch": 4.673882352941177, "grad_norm": 1.2015428961278338, "learning_rate": 4.805521765586653e-05, "loss": 0.793, "step": 9932 }, { "epoch": 4.674352941176471, "grad_norm": 1.0062346950973426, "learning_rate": 4.802306583582873e-05, "loss": 0.8359, "step": 9933 }, { "epoch": 4.674823529411765, "grad_norm": 0.7122177129001619, "learning_rate": 4.7990922725121266e-05, "loss": 0.2988, "step": 9934 }, { "epoch": 4.675294117647059, "grad_norm": 1.0481201722524052, "learning_rate": 4.795878832648935e-05, "loss": 0.5547, "step": 9935 }, { "epoch": 4.675764705882353, "grad_norm": 0.8908559894170878, "learning_rate": 4.7926662642677466e-05, "loss": 0.4121, "step": 9936 }, { "epoch": 4.676235294117647, "grad_norm": 0.6361210722553093, "learning_rate": 4.7894545676429206e-05, "loss": 0.1147, "step": 9937 }, { "epoch": 4.676705882352941, "grad_norm": 1.1044163289502995, "learning_rate": 4.7862437430487635e-05, "loss": 1.1406, "step": 9938 }, { "epoch": 4.677176470588235, "grad_norm": 0.9132671413385204, "learning_rate": 4.783033790759481e-05, "loss": 0.3145, "step": 9939 }, { "epoch": 4.6776470588235295, "grad_norm": 0.85333495508321, "learning_rate": 4.77982471104923e-05, "loss": 0.4492, "step": 9940 }, { "epoch": 4.678117647058824, "grad_norm": 0.9382110993736913, "learning_rate": 4.776616504192082e-05, "loss": 0.5391, "step": 9941 }, { "epoch": 4.678588235294118, "grad_norm": 1.2938769978679858, "learning_rate": 4.7734091704620244e-05, "loss": 1.4766, "step": 9942 }, { "epoch": 4.679058823529411, "grad_norm": 0.756931894355088, "learning_rate": 4.7702027101329846e-05, "loss": 0.21, "step": 9943 }, { "epoch": 4.679529411764706, "grad_norm": 0.9849822045707598, "learning_rate": 4.7669971234788165e-05, "loss": 0.375, "step": 9944 }, { "epoch": 4.68, "grad_norm": 1.4749169792373356, "learning_rate": 4.763792410773281e-05, "loss": 1.3672, "step": 9945 }, { "epoch": 4.680470588235294, "grad_norm": 1.565814968128294, "learning_rate": 4.760588572290082e-05, "loss": 1.4297, "step": 9946 }, { "epoch": 4.680941176470588, "grad_norm": 1.1026552363574915, "learning_rate": 4.7573856083028475e-05, "loss": 0.9453, "step": 9947 }, { "epoch": 4.681411764705882, "grad_norm": 1.1121761914649506, "learning_rate": 4.7541835190851175e-05, "loss": 0.8203, "step": 9948 }, { "epoch": 4.681882352941177, "grad_norm": 1.2061652862790837, "learning_rate": 4.7509823049103706e-05, "loss": 0.5586, "step": 9949 }, { "epoch": 4.682352941176471, "grad_norm": 1.1791556549187077, "learning_rate": 4.74778196605201e-05, "loss": 0.668, "step": 9950 }, { "epoch": 4.682823529411765, "grad_norm": 0.7675458504315859, "learning_rate": 4.744582502783352e-05, "loss": 0.2021, "step": 9951 }, { "epoch": 4.683294117647058, "grad_norm": 1.116014854543022, "learning_rate": 4.7413839153776504e-05, "loss": 0.7617, "step": 9952 }, { "epoch": 4.683764705882353, "grad_norm": 0.961253721873898, "learning_rate": 4.7381862041080865e-05, "loss": 0.6914, "step": 9953 }, { "epoch": 4.684235294117647, "grad_norm": 1.0604349225241012, "learning_rate": 4.7349893692477495e-05, "loss": 0.8477, "step": 9954 }, { "epoch": 4.684705882352941, "grad_norm": 1.0912456090207552, "learning_rate": 4.7317934110696685e-05, "loss": 1.0312, "step": 9955 }, { "epoch": 4.685176470588235, "grad_norm": 1.107861752761007, "learning_rate": 4.728598329846802e-05, "loss": 1.1797, "step": 9956 }, { "epoch": 4.6856470588235295, "grad_norm": 1.1055749639233228, "learning_rate": 4.725404125852013e-05, "loss": 0.5742, "step": 9957 }, { "epoch": 4.686117647058824, "grad_norm": 0.9282806929313804, "learning_rate": 4.7222107993581135e-05, "loss": 0.4668, "step": 9958 }, { "epoch": 4.686588235294118, "grad_norm": 1.0983162047273343, "learning_rate": 4.7190183506378205e-05, "loss": 0.8945, "step": 9959 }, { "epoch": 4.687058823529412, "grad_norm": 1.3870462572330462, "learning_rate": 4.7158267799637864e-05, "loss": 1.2969, "step": 9960 }, { "epoch": 4.6875294117647055, "grad_norm": 1.4213614420041019, "learning_rate": 4.712636087608594e-05, "loss": 1.3359, "step": 9961 }, { "epoch": 4.688, "grad_norm": 1.267629434094682, "learning_rate": 4.709446273844736e-05, "loss": 1.0938, "step": 9962 }, { "epoch": 4.688470588235294, "grad_norm": 1.1653624545976764, "learning_rate": 4.7062573389446395e-05, "loss": 0.5508, "step": 9963 }, { "epoch": 4.688941176470588, "grad_norm": 1.2597002133653616, "learning_rate": 4.70306928318066e-05, "loss": 0.875, "step": 9964 }, { "epoch": 4.689411764705882, "grad_norm": 0.6679447087488397, "learning_rate": 4.6998821068250675e-05, "loss": 0.4277, "step": 9965 }, { "epoch": 4.689882352941177, "grad_norm": 1.1091218816139923, "learning_rate": 4.696695810150065e-05, "loss": 0.5547, "step": 9966 }, { "epoch": 4.690352941176471, "grad_norm": 1.2937450382407885, "learning_rate": 4.693510393427781e-05, "loss": 0.7305, "step": 9967 }, { "epoch": 4.690823529411765, "grad_norm": 0.6880505143508353, "learning_rate": 4.690325856930258e-05, "loss": 0.3281, "step": 9968 }, { "epoch": 4.691294117647059, "grad_norm": 0.3576132331015597, "learning_rate": 4.6871422009294764e-05, "loss": 0.0771, "step": 9969 }, { "epoch": 4.691764705882353, "grad_norm": 1.2637117905431237, "learning_rate": 4.6839594256973406e-05, "loss": 1.6875, "step": 9970 }, { "epoch": 4.692235294117647, "grad_norm": 1.2054482727492275, "learning_rate": 4.680777531505662e-05, "loss": 0.9258, "step": 9971 }, { "epoch": 4.692705882352941, "grad_norm": 0.8502585489448924, "learning_rate": 4.6775965186262e-05, "loss": 0.4023, "step": 9972 }, { "epoch": 4.693176470588235, "grad_norm": 0.8073931999446005, "learning_rate": 4.6744163873306314e-05, "loss": 0.4844, "step": 9973 }, { "epoch": 4.6936470588235295, "grad_norm": 0.8797523710257595, "learning_rate": 4.671237137890539e-05, "loss": 0.5859, "step": 9974 }, { "epoch": 4.694117647058824, "grad_norm": 1.295922812093841, "learning_rate": 4.668058770577468e-05, "loss": 0.9102, "step": 9975 }, { "epoch": 4.694588235294118, "grad_norm": 1.0579174778082023, "learning_rate": 4.6648812856628515e-05, "loss": 0.457, "step": 9976 }, { "epoch": 4.695058823529411, "grad_norm": 0.8253398857366874, "learning_rate": 4.6617046834180656e-05, "loss": 0.332, "step": 9977 }, { "epoch": 4.6955294117647055, "grad_norm": 0.8013416158145192, "learning_rate": 4.658528964114413e-05, "loss": 0.2402, "step": 9978 }, { "epoch": 4.696, "grad_norm": 1.0058291725005593, "learning_rate": 4.655354128023107e-05, "loss": 1.3359, "step": 9979 }, { "epoch": 4.696470588235294, "grad_norm": 0.9606355763091647, "learning_rate": 4.6521801754152995e-05, "loss": 0.373, "step": 9980 }, { "epoch": 4.696941176470588, "grad_norm": 0.9406028389328566, "learning_rate": 4.649007106562066e-05, "loss": 0.4102, "step": 9981 }, { "epoch": 4.697411764705882, "grad_norm": 0.8143099571017949, "learning_rate": 4.645834921734392e-05, "loss": 0.4805, "step": 9982 }, { "epoch": 4.697882352941177, "grad_norm": 0.9538865537272554, "learning_rate": 4.642663621203201e-05, "loss": 0.8359, "step": 9983 }, { "epoch": 4.698352941176471, "grad_norm": 1.2077919218901583, "learning_rate": 4.639493205239346e-05, "loss": 0.6992, "step": 9984 }, { "epoch": 4.698823529411765, "grad_norm": 1.2536399572825163, "learning_rate": 4.636323674113583e-05, "loss": 1.1797, "step": 9985 }, { "epoch": 4.699294117647058, "grad_norm": 1.271003741887641, "learning_rate": 4.6331550280966126e-05, "loss": 0.8906, "step": 9986 }, { "epoch": 4.699764705882353, "grad_norm": 0.9744072924872854, "learning_rate": 4.629987267459056e-05, "loss": 0.8008, "step": 9987 }, { "epoch": 4.700235294117647, "grad_norm": 1.125667870719958, "learning_rate": 4.626820392471446e-05, "loss": 1.1094, "step": 9988 }, { "epoch": 4.700705882352941, "grad_norm": 1.1008499056425247, "learning_rate": 4.6236544034042604e-05, "loss": 0.7852, "step": 9989 }, { "epoch": 4.701176470588235, "grad_norm": 1.8570642931334804, "learning_rate": 4.62048930052788e-05, "loss": 1.6953, "step": 9990 }, { "epoch": 4.7016470588235295, "grad_norm": 0.9677858418744336, "learning_rate": 4.6173250841126245e-05, "loss": 0.6406, "step": 9991 }, { "epoch": 4.702117647058824, "grad_norm": 0.6881449570715115, "learning_rate": 4.614161754428737e-05, "loss": 0.1807, "step": 9992 }, { "epoch": 4.702588235294118, "grad_norm": 0.6569906177065742, "learning_rate": 4.6109993117463756e-05, "loss": 0.2002, "step": 9993 }, { "epoch": 4.703058823529412, "grad_norm": 0.8497575900973702, "learning_rate": 4.607837756335629e-05, "loss": 0.3164, "step": 9994 }, { "epoch": 4.7035294117647055, "grad_norm": 1.0158279258384009, "learning_rate": 4.604677088466516e-05, "loss": 0.668, "step": 9995 }, { "epoch": 4.704, "grad_norm": 1.3025463586491526, "learning_rate": 4.6015173084089635e-05, "loss": 0.6602, "step": 9996 }, { "epoch": 4.704470588235294, "grad_norm": 0.7707353102025619, "learning_rate": 4.598358416432838e-05, "loss": 0.5039, "step": 9997 }, { "epoch": 4.704941176470588, "grad_norm": 0.9554349911753612, "learning_rate": 4.595200412807927e-05, "loss": 0.9336, "step": 9998 }, { "epoch": 4.705411764705882, "grad_norm": 1.1299569607048903, "learning_rate": 4.592043297803932e-05, "loss": 0.9648, "step": 9999 }, { "epoch": 4.705882352941177, "grad_norm": 1.1694658138149236, "learning_rate": 4.58888707169049e-05, "loss": 0.7617, "step": 10000 }, { "epoch": 4.706352941176471, "grad_norm": 1.4027214104875456, "learning_rate": 4.585731734737164e-05, "loss": 0.5781, "step": 10001 }, { "epoch": 4.706823529411765, "grad_norm": 1.4324130114229505, "learning_rate": 4.5825772872134234e-05, "loss": 0.9492, "step": 10002 }, { "epoch": 4.707294117647059, "grad_norm": 1.3539039227213843, "learning_rate": 4.579423729388681e-05, "loss": 1.0938, "step": 10003 }, { "epoch": 4.707764705882353, "grad_norm": 1.06061689359041, "learning_rate": 4.576271061532268e-05, "loss": 0.9453, "step": 10004 }, { "epoch": 4.708235294117647, "grad_norm": 0.9690041035911839, "learning_rate": 4.573119283913431e-05, "loss": 0.3496, "step": 10005 }, { "epoch": 4.708705882352941, "grad_norm": 1.0052507399277815, "learning_rate": 4.569968396801351e-05, "loss": 0.5469, "step": 10006 }, { "epoch": 4.709176470588235, "grad_norm": 0.7864231405402864, "learning_rate": 4.566818400465133e-05, "loss": 0.1777, "step": 10007 }, { "epoch": 4.7096470588235295, "grad_norm": 0.6325135693136771, "learning_rate": 4.563669295173793e-05, "loss": 0.1592, "step": 10008 }, { "epoch": 4.710117647058824, "grad_norm": 1.3468116231750156, "learning_rate": 4.560521081196289e-05, "loss": 1.375, "step": 10009 }, { "epoch": 4.710588235294118, "grad_norm": 1.1322255963367345, "learning_rate": 4.5573737588014875e-05, "loss": 1.2266, "step": 10010 }, { "epoch": 4.711058823529411, "grad_norm": 1.303360251958833, "learning_rate": 4.554227328258185e-05, "loss": 0.9062, "step": 10011 }, { "epoch": 4.7115294117647055, "grad_norm": 1.2254781142989983, "learning_rate": 4.551081789835113e-05, "loss": 1.375, "step": 10012 }, { "epoch": 4.712, "grad_norm": 1.1526033806114648, "learning_rate": 4.547937143800901e-05, "loss": 0.8984, "step": 10013 }, { "epoch": 4.712470588235294, "grad_norm": 0.5140714196670728, "learning_rate": 4.544793390424124e-05, "loss": 0.1309, "step": 10014 }, { "epoch": 4.712941176470588, "grad_norm": 0.9305631244263262, "learning_rate": 4.541650529973279e-05, "loss": 0.6133, "step": 10015 }, { "epoch": 4.713411764705882, "grad_norm": 0.7433300104631666, "learning_rate": 4.5385085627167717e-05, "loss": 0.2676, "step": 10016 }, { "epoch": 4.713882352941177, "grad_norm": 0.9493155972536302, "learning_rate": 4.535367488922947e-05, "loss": 0.6719, "step": 10017 }, { "epoch": 4.714352941176471, "grad_norm": 0.8219167855150552, "learning_rate": 4.5322273088600714e-05, "loss": 0.3457, "step": 10018 }, { "epoch": 4.714823529411765, "grad_norm": 0.5799097532223844, "learning_rate": 4.5290880227963186e-05, "loss": 0.1348, "step": 10019 }, { "epoch": 4.715294117647058, "grad_norm": 1.2828113351396067, "learning_rate": 4.525949630999818e-05, "loss": 1.4922, "step": 10020 }, { "epoch": 4.715764705882353, "grad_norm": 1.3676117602440276, "learning_rate": 4.522812133738589e-05, "loss": 1.3672, "step": 10021 }, { "epoch": 4.716235294117647, "grad_norm": 1.814091729761177, "learning_rate": 4.519675531280594e-05, "loss": 1.0391, "step": 10022 }, { "epoch": 4.716705882352941, "grad_norm": 1.2231356137134222, "learning_rate": 4.5165398238937194e-05, "loss": 0.7812, "step": 10023 }, { "epoch": 4.717176470588235, "grad_norm": 1.2108143085414087, "learning_rate": 4.51340501184576e-05, "loss": 0.8438, "step": 10024 }, { "epoch": 4.7176470588235295, "grad_norm": 1.3397092175832663, "learning_rate": 4.510271095404449e-05, "loss": 0.3359, "step": 10025 }, { "epoch": 4.718117647058824, "grad_norm": 1.3153458265153701, "learning_rate": 4.507138074837443e-05, "loss": 1.1875, "step": 10026 }, { "epoch": 4.718588235294118, "grad_norm": 0.8185442829561232, "learning_rate": 4.504005950412308e-05, "loss": 0.2988, "step": 10027 }, { "epoch": 4.719058823529412, "grad_norm": 1.0105006576781252, "learning_rate": 4.5008747223965456e-05, "loss": 0.5273, "step": 10028 }, { "epoch": 4.7195294117647055, "grad_norm": 0.8014824899433857, "learning_rate": 4.4977443910575866e-05, "loss": 0.2256, "step": 10029 }, { "epoch": 4.72, "grad_norm": 0.47570344201714787, "learning_rate": 4.494614956662765e-05, "loss": 0.0444, "step": 10030 }, { "epoch": 4.720470588235294, "grad_norm": 0.8559501799648964, "learning_rate": 4.491486419479355e-05, "loss": 0.3789, "step": 10031 }, { "epoch": 4.720941176470588, "grad_norm": 1.4371838837657247, "learning_rate": 4.488358779774551e-05, "loss": 1.2734, "step": 10032 }, { "epoch": 4.721411764705882, "grad_norm": 1.5118429524846095, "learning_rate": 4.485232037815465e-05, "loss": 1.5547, "step": 10033 }, { "epoch": 4.721882352941177, "grad_norm": 1.0319872757021582, "learning_rate": 4.4821061938691355e-05, "loss": 0.5312, "step": 10034 }, { "epoch": 4.722352941176471, "grad_norm": 0.8360094879001484, "learning_rate": 4.478981248202533e-05, "loss": 0.2695, "step": 10035 }, { "epoch": 4.722823529411765, "grad_norm": 1.107728833613424, "learning_rate": 4.4758572010825324e-05, "loss": 0.8672, "step": 10036 }, { "epoch": 4.723294117647059, "grad_norm": 1.5784788379691317, "learning_rate": 4.472734052775947e-05, "loss": 0.6523, "step": 10037 }, { "epoch": 4.723764705882353, "grad_norm": 1.0915990114412857, "learning_rate": 4.4696118035495136e-05, "loss": 1.0859, "step": 10038 }, { "epoch": 4.724235294117647, "grad_norm": 0.7636713790459413, "learning_rate": 4.4664904536698796e-05, "loss": 0.249, "step": 10039 }, { "epoch": 4.724705882352941, "grad_norm": 1.1507892872205874, "learning_rate": 4.463370003403632e-05, "loss": 0.6797, "step": 10040 }, { "epoch": 4.725176470588235, "grad_norm": 0.9047571404505447, "learning_rate": 4.4602504530172636e-05, "loss": 0.3887, "step": 10041 }, { "epoch": 4.7256470588235295, "grad_norm": 0.981201503634019, "learning_rate": 4.457131802777202e-05, "loss": 0.6133, "step": 10042 }, { "epoch": 4.726117647058824, "grad_norm": 1.1199420837430074, "learning_rate": 4.454014052949804e-05, "loss": 1.1562, "step": 10043 }, { "epoch": 4.726588235294118, "grad_norm": 0.8328291505343108, "learning_rate": 4.4508972038013274e-05, "loss": 0.498, "step": 10044 }, { "epoch": 4.727058823529411, "grad_norm": 1.2660689212093856, "learning_rate": 4.4477812555979724e-05, "loss": 1.1172, "step": 10045 }, { "epoch": 4.7275294117647055, "grad_norm": 1.0853714711277203, "learning_rate": 4.4446662086058634e-05, "loss": 1.0547, "step": 10046 }, { "epoch": 4.728, "grad_norm": 1.09519440671709, "learning_rate": 4.4415520630910265e-05, "loss": 0.4785, "step": 10047 }, { "epoch": 4.728470588235294, "grad_norm": 0.8428753586285922, "learning_rate": 4.438438819319434e-05, "loss": 0.375, "step": 10048 }, { "epoch": 4.728941176470588, "grad_norm": 0.9213604429515767, "learning_rate": 4.435326477556975e-05, "loss": 1.4453, "step": 10049 }, { "epoch": 4.729411764705882, "grad_norm": 1.0126276775406435, "learning_rate": 4.432215038069449e-05, "loss": 0.3965, "step": 10050 }, { "epoch": 4.729882352941177, "grad_norm": 0.395565103058148, "learning_rate": 4.4291045011225934e-05, "loss": 0.1055, "step": 10051 }, { "epoch": 4.730352941176471, "grad_norm": 1.2621173932569423, "learning_rate": 4.4259948669820676e-05, "loss": 0.9258, "step": 10052 }, { "epoch": 4.730823529411765, "grad_norm": 1.0506140328720184, "learning_rate": 4.422886135913441e-05, "loss": 0.8906, "step": 10053 }, { "epoch": 4.731294117647058, "grad_norm": 1.1737346887329163, "learning_rate": 4.419778308182219e-05, "loss": 0.6758, "step": 10054 }, { "epoch": 4.731764705882353, "grad_norm": 1.1112543060372484, "learning_rate": 4.41667138405383e-05, "loss": 0.8867, "step": 10055 }, { "epoch": 4.732235294117647, "grad_norm": 1.0729300486307713, "learning_rate": 4.413565363793611e-05, "loss": 0.6211, "step": 10056 }, { "epoch": 4.732705882352941, "grad_norm": 1.0717626641474367, "learning_rate": 4.4104602476668364e-05, "loss": 0.7891, "step": 10057 }, { "epoch": 4.733176470588235, "grad_norm": 1.29008300026647, "learning_rate": 4.4073560359387035e-05, "loss": 0.6484, "step": 10058 }, { "epoch": 4.7336470588235295, "grad_norm": 1.1107433132724014, "learning_rate": 4.4042527288743155e-05, "loss": 0.8398, "step": 10059 }, { "epoch": 4.734117647058824, "grad_norm": 1.1462084637962788, "learning_rate": 4.401150326738722e-05, "loss": 0.8672, "step": 10060 }, { "epoch": 4.734588235294118, "grad_norm": 1.5589649248931605, "learning_rate": 4.3980488297968726e-05, "loss": 1.5, "step": 10061 }, { "epoch": 4.735058823529412, "grad_norm": 0.5978901059447638, "learning_rate": 4.394948238313655e-05, "loss": 0.1514, "step": 10062 }, { "epoch": 4.7355294117647055, "grad_norm": 1.4232577269291995, "learning_rate": 4.391848552553875e-05, "loss": 1.0469, "step": 10063 }, { "epoch": 4.736, "grad_norm": 1.4169390403690503, "learning_rate": 4.3887497727822616e-05, "loss": 0.5664, "step": 10064 }, { "epoch": 4.736470588235294, "grad_norm": 0.9442448679845432, "learning_rate": 4.385651899263467e-05, "loss": 0.7734, "step": 10065 }, { "epoch": 4.736941176470588, "grad_norm": 1.1464386800391158, "learning_rate": 4.3825549322620657e-05, "loss": 0.4648, "step": 10066 }, { "epoch": 4.737411764705882, "grad_norm": 1.104844252966655, "learning_rate": 4.3794588720425486e-05, "loss": 0.3223, "step": 10067 }, { "epoch": 4.737882352941177, "grad_norm": 1.4647693103875425, "learning_rate": 4.3763637188693404e-05, "loss": 0.8867, "step": 10068 }, { "epoch": 4.738352941176471, "grad_norm": 1.5843620063917672, "learning_rate": 4.3732694730067766e-05, "loss": 1.5, "step": 10069 }, { "epoch": 4.738823529411764, "grad_norm": 0.9709950841074784, "learning_rate": 4.370176134719121e-05, "loss": 0.4492, "step": 10070 }, { "epoch": 4.739294117647058, "grad_norm": 1.0753931317901986, "learning_rate": 4.36708370427057e-05, "loss": 0.5469, "step": 10071 }, { "epoch": 4.739764705882353, "grad_norm": 1.1732839189937239, "learning_rate": 4.363992181925218e-05, "loss": 1.1172, "step": 10072 }, { "epoch": 4.740235294117647, "grad_norm": 1.0785885769402546, "learning_rate": 4.3609015679471036e-05, "loss": 0.9609, "step": 10073 }, { "epoch": 4.740705882352941, "grad_norm": 0.9456885262500386, "learning_rate": 4.357811862600185e-05, "loss": 0.4844, "step": 10074 }, { "epoch": 4.741176470588235, "grad_norm": 0.6852082156750852, "learning_rate": 4.3547230661483276e-05, "loss": 0.4375, "step": 10075 }, { "epoch": 4.7416470588235295, "grad_norm": 1.0167688984999579, "learning_rate": 4.351635178855336e-05, "loss": 0.7656, "step": 10076 }, { "epoch": 4.742117647058824, "grad_norm": 1.5854814330032119, "learning_rate": 4.348548200984934e-05, "loss": 0.7344, "step": 10077 }, { "epoch": 4.742588235294118, "grad_norm": 1.163564926476791, "learning_rate": 4.345462132800756e-05, "loss": 0.5273, "step": 10078 }, { "epoch": 4.743058823529411, "grad_norm": 1.74023530700241, "learning_rate": 4.342376974566372e-05, "loss": 1.6641, "step": 10079 }, { "epoch": 4.7435294117647056, "grad_norm": 2.1930677164954835, "learning_rate": 4.339292726545275e-05, "loss": 0.668, "step": 10080 }, { "epoch": 4.744, "grad_norm": 1.0564149952540625, "learning_rate": 4.3362093890008645e-05, "loss": 0.7891, "step": 10081 }, { "epoch": 4.744470588235294, "grad_norm": 1.1678586024395299, "learning_rate": 4.333126962196478e-05, "loss": 0.9727, "step": 10082 }, { "epoch": 4.744941176470588, "grad_norm": 1.1023232165859316, "learning_rate": 4.330045446395374e-05, "loss": 0.6719, "step": 10083 }, { "epoch": 4.745411764705882, "grad_norm": 1.0522534080790695, "learning_rate": 4.3269648418607194e-05, "loss": 0.5742, "step": 10084 }, { "epoch": 4.745882352941177, "grad_norm": 1.5221128343437802, "learning_rate": 4.32388514885562e-05, "loss": 1.6172, "step": 10085 }, { "epoch": 4.746352941176471, "grad_norm": 1.384652766040262, "learning_rate": 4.320806367643098e-05, "loss": 0.8672, "step": 10086 }, { "epoch": 4.746823529411765, "grad_norm": 1.047374692594572, "learning_rate": 4.3177284984860895e-05, "loss": 0.8477, "step": 10087 }, { "epoch": 4.747294117647058, "grad_norm": 0.6172538526031964, "learning_rate": 4.314651541647468e-05, "loss": 0.1963, "step": 10088 }, { "epoch": 4.747764705882353, "grad_norm": 1.0812858533881575, "learning_rate": 4.3115754973900104e-05, "loss": 0.5273, "step": 10089 }, { "epoch": 4.748235294117647, "grad_norm": 1.2640770305512368, "learning_rate": 4.308500365976434e-05, "loss": 1.0, "step": 10090 }, { "epoch": 4.748705882352941, "grad_norm": 1.046066333873181, "learning_rate": 4.305426147669371e-05, "loss": 0.9688, "step": 10091 }, { "epoch": 4.749176470588235, "grad_norm": 0.9278138479527998, "learning_rate": 4.302352842731368e-05, "loss": 0.8711, "step": 10092 }, { "epoch": 4.7496470588235296, "grad_norm": 0.947434211422205, "learning_rate": 4.299280451424903e-05, "loss": 0.7031, "step": 10093 }, { "epoch": 4.750117647058824, "grad_norm": 0.8079100220101803, "learning_rate": 4.296208974012379e-05, "loss": 0.3008, "step": 10094 }, { "epoch": 4.750588235294118, "grad_norm": 1.531805861690418, "learning_rate": 4.2931384107561065e-05, "loss": 1.4531, "step": 10095 }, { "epoch": 4.751058823529412, "grad_norm": 0.8113803080830857, "learning_rate": 4.290068761918329e-05, "loss": 0.418, "step": 10096 }, { "epoch": 4.751529411764706, "grad_norm": 1.182882418355771, "learning_rate": 4.287000027761217e-05, "loss": 1.0, "step": 10097 }, { "epoch": 4.752, "grad_norm": 1.0972622842223803, "learning_rate": 4.2839322085468445e-05, "loss": 0.5664, "step": 10098 }, { "epoch": 4.752470588235294, "grad_norm": 1.3333228855680033, "learning_rate": 4.280865304537223e-05, "loss": 0.7812, "step": 10099 }, { "epoch": 4.752941176470588, "grad_norm": 1.1125193249602694, "learning_rate": 4.277799315994286e-05, "loss": 1.0312, "step": 10100 }, { "epoch": 4.7534117647058824, "grad_norm": 0.9260542922156264, "learning_rate": 4.2747342431798767e-05, "loss": 0.3477, "step": 10101 }, { "epoch": 4.753882352941177, "grad_norm": 0.7633676669847347, "learning_rate": 4.271670086355768e-05, "loss": 0.5391, "step": 10102 }, { "epoch": 4.754352941176471, "grad_norm": 1.1713294881340923, "learning_rate": 4.268606845783663e-05, "loss": 0.543, "step": 10103 }, { "epoch": 4.754823529411764, "grad_norm": 1.3338907228727543, "learning_rate": 4.265544521725165e-05, "loss": 1.3828, "step": 10104 }, { "epoch": 4.7552941176470584, "grad_norm": 1.1059734724703558, "learning_rate": 4.2624831144418166e-05, "loss": 0.6992, "step": 10105 }, { "epoch": 4.755764705882353, "grad_norm": 0.8163965834485988, "learning_rate": 4.259422624195082e-05, "loss": 0.2061, "step": 10106 }, { "epoch": 4.756235294117647, "grad_norm": 1.0994320708925631, "learning_rate": 4.2563630512463284e-05, "loss": 1.1172, "step": 10107 }, { "epoch": 4.756705882352941, "grad_norm": 0.6237031194211369, "learning_rate": 4.253304395856876e-05, "loss": 0.418, "step": 10108 }, { "epoch": 4.757176470588235, "grad_norm": 1.0860848216055663, "learning_rate": 4.2502466582879375e-05, "loss": 0.5508, "step": 10109 }, { "epoch": 4.75764705882353, "grad_norm": 0.7388347462316625, "learning_rate": 4.2471898388006606e-05, "loss": 0.2412, "step": 10110 }, { "epoch": 4.758117647058824, "grad_norm": 0.9751552787533858, "learning_rate": 4.244133937656118e-05, "loss": 1.1328, "step": 10111 }, { "epoch": 4.758588235294118, "grad_norm": 1.3487937630188134, "learning_rate": 4.2410789551152894e-05, "loss": 0.6172, "step": 10112 }, { "epoch": 4.759058823529411, "grad_norm": 1.0169291163022276, "learning_rate": 4.238024891439092e-05, "loss": 0.9492, "step": 10113 }, { "epoch": 4.759529411764706, "grad_norm": 0.9862252128006327, "learning_rate": 4.234971746888359e-05, "loss": 0.832, "step": 10114 }, { "epoch": 4.76, "grad_norm": 0.9477974492889492, "learning_rate": 4.231919521723836e-05, "loss": 0.6445, "step": 10115 }, { "epoch": 4.760470588235294, "grad_norm": 1.1789965480381888, "learning_rate": 4.228868216206206e-05, "loss": 1.1094, "step": 10116 }, { "epoch": 4.760941176470588, "grad_norm": 0.9910501959453002, "learning_rate": 4.225817830596064e-05, "loss": 0.9297, "step": 10117 }, { "epoch": 4.7614117647058825, "grad_norm": 0.81416208967071, "learning_rate": 4.2227683651539236e-05, "loss": 0.2256, "step": 10118 }, { "epoch": 4.761882352941177, "grad_norm": 1.2115880893863105, "learning_rate": 4.21971982014023e-05, "loss": 0.5898, "step": 10119 }, { "epoch": 4.762352941176471, "grad_norm": 1.0506103268633036, "learning_rate": 4.2166721958153365e-05, "loss": 0.9336, "step": 10120 }, { "epoch": 4.762823529411765, "grad_norm": 1.443273299991942, "learning_rate": 4.2136254924395305e-05, "loss": 1.0312, "step": 10121 }, { "epoch": 4.7632941176470585, "grad_norm": 0.8244801978363048, "learning_rate": 4.210579710273018e-05, "loss": 0.2168, "step": 10122 }, { "epoch": 4.763764705882353, "grad_norm": 1.2298637549542009, "learning_rate": 4.207534849575916e-05, "loss": 0.9336, "step": 10123 }, { "epoch": 4.764235294117647, "grad_norm": 1.2057728981296272, "learning_rate": 4.204490910608274e-05, "loss": 0.8203, "step": 10124 }, { "epoch": 4.764705882352941, "grad_norm": 0.9892491688219011, "learning_rate": 4.201447893630065e-05, "loss": 0.7148, "step": 10125 }, { "epoch": 4.765176470588235, "grad_norm": 0.7262104206326941, "learning_rate": 4.1984057989011674e-05, "loss": 0.4512, "step": 10126 }, { "epoch": 4.76564705882353, "grad_norm": 1.4082724463804324, "learning_rate": 4.195364626681396e-05, "loss": 0.7539, "step": 10127 }, { "epoch": 4.766117647058824, "grad_norm": 1.8499158132862221, "learning_rate": 4.192324377230487e-05, "loss": 1.2656, "step": 10128 }, { "epoch": 4.766588235294118, "grad_norm": 0.8051256486878342, "learning_rate": 4.1892850508080824e-05, "loss": 0.2715, "step": 10129 }, { "epoch": 4.767058823529412, "grad_norm": 1.1515142786844261, "learning_rate": 4.18624664767376e-05, "loss": 0.625, "step": 10130 }, { "epoch": 4.767529411764706, "grad_norm": 1.268778155837265, "learning_rate": 4.1832091680870214e-05, "loss": 1.2109, "step": 10131 }, { "epoch": 4.768, "grad_norm": 1.1954370932816942, "learning_rate": 4.1801726123072696e-05, "loss": 0.8555, "step": 10132 }, { "epoch": 4.768470588235294, "grad_norm": 1.1236174867080209, "learning_rate": 4.1771369805938486e-05, "loss": 0.7344, "step": 10133 }, { "epoch": 4.768941176470588, "grad_norm": 0.8490235277059105, "learning_rate": 4.1741022732060204e-05, "loss": 0.2715, "step": 10134 }, { "epoch": 4.7694117647058825, "grad_norm": 1.0965342108884195, "learning_rate": 4.171068490402954e-05, "loss": 0.8477, "step": 10135 }, { "epoch": 4.769882352941177, "grad_norm": 1.1738347316451587, "learning_rate": 4.168035632443754e-05, "loss": 0.5586, "step": 10136 }, { "epoch": 4.770352941176471, "grad_norm": 1.4109954815471846, "learning_rate": 4.165003699587445e-05, "loss": 0.5859, "step": 10137 }, { "epoch": 4.770823529411764, "grad_norm": 0.7242193554675675, "learning_rate": 4.161972692092962e-05, "loss": 0.2148, "step": 10138 }, { "epoch": 4.7712941176470585, "grad_norm": 0.7959441649096624, "learning_rate": 4.158942610219175e-05, "loss": 0.2637, "step": 10139 }, { "epoch": 4.771764705882353, "grad_norm": 1.3261030479760625, "learning_rate": 4.155913454224862e-05, "loss": 0.6836, "step": 10140 }, { "epoch": 4.772235294117647, "grad_norm": 0.9616688293604879, "learning_rate": 4.1528852243687276e-05, "loss": 0.6602, "step": 10141 }, { "epoch": 4.772705882352941, "grad_norm": 0.9836327064464836, "learning_rate": 4.1498579209094046e-05, "loss": 0.4395, "step": 10142 }, { "epoch": 4.773176470588235, "grad_norm": 0.6885802932742748, "learning_rate": 4.146831544105431e-05, "loss": 0.3301, "step": 10143 }, { "epoch": 4.77364705882353, "grad_norm": 1.119733177643782, "learning_rate": 4.143806094215277e-05, "loss": 0.6133, "step": 10144 }, { "epoch": 4.774117647058824, "grad_norm": 1.0957413893233006, "learning_rate": 4.140781571497336e-05, "loss": 0.9922, "step": 10145 }, { "epoch": 4.774588235294118, "grad_norm": 1.0851583095168225, "learning_rate": 4.137757976209911e-05, "loss": 0.8633, "step": 10146 }, { "epoch": 4.775058823529411, "grad_norm": 1.2342124666640846, "learning_rate": 4.134735308611232e-05, "loss": 1.3125, "step": 10147 }, { "epoch": 4.775529411764706, "grad_norm": 0.7151191436890102, "learning_rate": 4.131713568959457e-05, "loss": 0.3887, "step": 10148 }, { "epoch": 4.776, "grad_norm": 0.8549251465349028, "learning_rate": 4.128692757512646e-05, "loss": 0.3301, "step": 10149 }, { "epoch": 4.776470588235294, "grad_norm": 1.0981786851426123, "learning_rate": 4.125672874528797e-05, "loss": 0.5039, "step": 10150 }, { "epoch": 4.776941176470588, "grad_norm": 1.2077865413950668, "learning_rate": 4.122653920265828e-05, "loss": 0.5586, "step": 10151 }, { "epoch": 4.7774117647058825, "grad_norm": 1.1935114053545635, "learning_rate": 4.119635894981558e-05, "loss": 1.3594, "step": 10152 }, { "epoch": 4.777882352941177, "grad_norm": 0.8708318310092573, "learning_rate": 4.1166187989337596e-05, "loss": 0.4395, "step": 10153 }, { "epoch": 4.778352941176471, "grad_norm": 1.387317928897105, "learning_rate": 4.113602632380094e-05, "loss": 1.0625, "step": 10154 }, { "epoch": 4.778823529411765, "grad_norm": 1.880257181339141, "learning_rate": 4.110587395578161e-05, "loss": 1.0234, "step": 10155 }, { "epoch": 4.7792941176470585, "grad_norm": 1.1146994654459312, "learning_rate": 4.1075730887854805e-05, "loss": 0.6484, "step": 10156 }, { "epoch": 4.779764705882353, "grad_norm": 0.9786973686455684, "learning_rate": 4.104559712259482e-05, "loss": 0.3613, "step": 10157 }, { "epoch": 4.780235294117647, "grad_norm": 0.5902152016685521, "learning_rate": 4.101547266257526e-05, "loss": 0.3945, "step": 10158 }, { "epoch": 4.780705882352941, "grad_norm": 1.2337164947313428, "learning_rate": 4.098535751036894e-05, "loss": 1.3516, "step": 10159 }, { "epoch": 4.781176470588235, "grad_norm": 1.0235163181091025, "learning_rate": 4.0955251668547754e-05, "loss": 0.7188, "step": 10160 }, { "epoch": 4.78164705882353, "grad_norm": 1.1655230978381956, "learning_rate": 4.092515513968294e-05, "loss": 1.1172, "step": 10161 }, { "epoch": 4.782117647058824, "grad_norm": 0.8878898185750378, "learning_rate": 4.0895067926344935e-05, "loss": 0.3086, "step": 10162 }, { "epoch": 4.782588235294118, "grad_norm": 1.1233069554758137, "learning_rate": 4.0864990031103235e-05, "loss": 0.6758, "step": 10163 }, { "epoch": 4.783058823529412, "grad_norm": 0.9260601281078229, "learning_rate": 4.0834921456526684e-05, "loss": 0.3418, "step": 10164 }, { "epoch": 4.783529411764706, "grad_norm": 0.4275039958888151, "learning_rate": 4.080486220518336e-05, "loss": 0.0649, "step": 10165 }, { "epoch": 4.784, "grad_norm": 1.3280722951157753, "learning_rate": 4.0774812279640325e-05, "loss": 1.1875, "step": 10166 }, { "epoch": 4.784470588235294, "grad_norm": 0.7517776467447715, "learning_rate": 4.074477168246409e-05, "loss": 0.293, "step": 10167 }, { "epoch": 4.784941176470588, "grad_norm": 0.5516441077744227, "learning_rate": 4.071474041622029e-05, "loss": 0.1592, "step": 10168 }, { "epoch": 4.7854117647058825, "grad_norm": 1.0951779150979908, "learning_rate": 4.0684718483473656e-05, "loss": 0.459, "step": 10169 }, { "epoch": 4.785882352941177, "grad_norm": 1.2710817966097907, "learning_rate": 4.06547058867883e-05, "loss": 1.1328, "step": 10170 }, { "epoch": 4.786352941176471, "grad_norm": 0.9033819614199834, "learning_rate": 4.0624702628727345e-05, "loss": 0.5273, "step": 10171 }, { "epoch": 4.786823529411764, "grad_norm": 1.2349257593477083, "learning_rate": 4.0594708711853267e-05, "loss": 0.8125, "step": 10172 }, { "epoch": 4.7872941176470585, "grad_norm": 0.9801702027670817, "learning_rate": 4.0564724138727734e-05, "loss": 0.3418, "step": 10173 }, { "epoch": 4.787764705882353, "grad_norm": 1.2310676574663058, "learning_rate": 4.0534748911911505e-05, "loss": 0.9062, "step": 10174 }, { "epoch": 4.788235294117647, "grad_norm": 1.0193160819682816, "learning_rate": 4.0504783033964645e-05, "loss": 0.582, "step": 10175 }, { "epoch": 4.788705882352941, "grad_norm": 1.1344407293898349, "learning_rate": 4.047482650744641e-05, "loss": 0.9258, "step": 10176 }, { "epoch": 4.789176470588235, "grad_norm": 0.966657724959951, "learning_rate": 4.044487933491518e-05, "loss": 0.4902, "step": 10177 }, { "epoch": 4.78964705882353, "grad_norm": 1.2092468938491003, "learning_rate": 4.041494151892862e-05, "loss": 0.9609, "step": 10178 }, { "epoch": 4.790117647058824, "grad_norm": 1.3602318421108837, "learning_rate": 4.038501306204361e-05, "loss": 1.0, "step": 10179 }, { "epoch": 4.790588235294118, "grad_norm": 1.1511989478032372, "learning_rate": 4.0355093966816095e-05, "loss": 0.5352, "step": 10180 }, { "epoch": 4.791058823529411, "grad_norm": 1.4369248331405229, "learning_rate": 4.032518423580135e-05, "loss": 0.8281, "step": 10181 }, { "epoch": 4.791529411764706, "grad_norm": 0.7330039398755234, "learning_rate": 4.029528387155389e-05, "loss": 0.123, "step": 10182 }, { "epoch": 4.792, "grad_norm": 1.2116982266795786, "learning_rate": 4.026539287662722e-05, "loss": 0.8672, "step": 10183 }, { "epoch": 4.792470588235294, "grad_norm": 1.2694759086355198, "learning_rate": 4.0235511253574256e-05, "loss": 0.7305, "step": 10184 }, { "epoch": 4.792941176470588, "grad_norm": 0.826885456272845, "learning_rate": 4.020563900494707e-05, "loss": 0.2373, "step": 10185 }, { "epoch": 4.7934117647058825, "grad_norm": 1.513279440860728, "learning_rate": 4.01757761332968e-05, "loss": 1.0625, "step": 10186 }, { "epoch": 4.793882352941177, "grad_norm": 1.1567314952682035, "learning_rate": 4.0145922641173934e-05, "loss": 0.3652, "step": 10187 }, { "epoch": 4.794352941176471, "grad_norm": 0.45079351184958594, "learning_rate": 4.011607853112815e-05, "loss": 0.0679, "step": 10188 }, { "epoch": 4.794823529411765, "grad_norm": 0.9874746360770282, "learning_rate": 4.008624380570819e-05, "loss": 0.8555, "step": 10189 }, { "epoch": 4.7952941176470585, "grad_norm": 1.7552486958671865, "learning_rate": 4.0056418467462184e-05, "loss": 1.1641, "step": 10190 }, { "epoch": 4.795764705882353, "grad_norm": 0.9553548637256211, "learning_rate": 4.002660251893727e-05, "loss": 0.8867, "step": 10191 }, { "epoch": 4.796235294117647, "grad_norm": 0.9263190020993063, "learning_rate": 3.9996795962679896e-05, "loss": 0.7969, "step": 10192 }, { "epoch": 4.796705882352941, "grad_norm": 1.1730329783479858, "learning_rate": 3.9966998801235766e-05, "loss": 1.0938, "step": 10193 }, { "epoch": 4.797176470588235, "grad_norm": 1.2434468927524345, "learning_rate": 3.9937211037149616e-05, "loss": 0.5469, "step": 10194 }, { "epoch": 4.79764705882353, "grad_norm": 0.8031372484692914, "learning_rate": 3.9907432672965485e-05, "loss": 0.2949, "step": 10195 }, { "epoch": 4.798117647058824, "grad_norm": 1.3101946915662859, "learning_rate": 3.987766371122661e-05, "loss": 0.6445, "step": 10196 }, { "epoch": 4.798588235294117, "grad_norm": 1.0603917938853091, "learning_rate": 3.984790415447539e-05, "loss": 0.5117, "step": 10197 }, { "epoch": 4.799058823529411, "grad_norm": 0.9471810005101504, "learning_rate": 3.9818154005253514e-05, "loss": 0.4922, "step": 10198 }, { "epoch": 4.799529411764706, "grad_norm": 0.6540951778036879, "learning_rate": 3.978841326610166e-05, "loss": 0.1963, "step": 10199 }, { "epoch": 4.8, "grad_norm": 1.1796546517726996, "learning_rate": 3.97586819395599e-05, "loss": 0.9023, "step": 10200 }, { "epoch": 4.800470588235294, "grad_norm": 0.8434572601476812, "learning_rate": 3.972896002816749e-05, "loss": 0.3105, "step": 10201 }, { "epoch": 4.800941176470588, "grad_norm": 1.2650190100682823, "learning_rate": 3.969924753446271e-05, "loss": 0.6211, "step": 10202 }, { "epoch": 4.8014117647058825, "grad_norm": 1.3072185092789004, "learning_rate": 3.966954446098322e-05, "loss": 0.7148, "step": 10203 }, { "epoch": 4.801882352941177, "grad_norm": 1.217925649267241, "learning_rate": 3.9639850810265845e-05, "loss": 1.2344, "step": 10204 }, { "epoch": 4.802352941176471, "grad_norm": 1.3144837093297188, "learning_rate": 3.961016658484648e-05, "loss": 0.8711, "step": 10205 }, { "epoch": 4.802823529411764, "grad_norm": 1.1580363134844844, "learning_rate": 3.9580491787260346e-05, "loss": 1.0547, "step": 10206 }, { "epoch": 4.8032941176470585, "grad_norm": 1.356668990534313, "learning_rate": 3.955082642004187e-05, "loss": 0.8828, "step": 10207 }, { "epoch": 4.803764705882353, "grad_norm": 1.1325546872416734, "learning_rate": 3.952117048572452e-05, "loss": 0.5312, "step": 10208 }, { "epoch": 4.804235294117647, "grad_norm": 1.0362801205950063, "learning_rate": 3.94915239868411e-05, "loss": 1.4297, "step": 10209 }, { "epoch": 4.804705882352941, "grad_norm": 0.9888357035695443, "learning_rate": 3.946188692592362e-05, "loss": 0.6094, "step": 10210 }, { "epoch": 4.805176470588235, "grad_norm": 2.0558452137583703, "learning_rate": 3.943225930550315e-05, "loss": 1.0781, "step": 10211 }, { "epoch": 4.80564705882353, "grad_norm": 1.0291275136360571, "learning_rate": 3.9402641128110054e-05, "loss": 0.875, "step": 10212 }, { "epoch": 4.806117647058824, "grad_norm": 0.5573713983544516, "learning_rate": 3.937303239627392e-05, "loss": 0.5586, "step": 10213 }, { "epoch": 4.806588235294118, "grad_norm": 1.4136269895496638, "learning_rate": 3.934343311252341e-05, "loss": 1.2734, "step": 10214 }, { "epoch": 4.807058823529411, "grad_norm": 0.9295948545716923, "learning_rate": 3.9313843279386486e-05, "loss": 1.2734, "step": 10215 }, { "epoch": 4.807529411764706, "grad_norm": 1.2824669936959323, "learning_rate": 3.9284262899390285e-05, "loss": 0.9883, "step": 10216 }, { "epoch": 4.808, "grad_norm": 0.6281439711670073, "learning_rate": 3.925469197506107e-05, "loss": 0.252, "step": 10217 }, { "epoch": 4.808470588235294, "grad_norm": 0.7362720432810738, "learning_rate": 3.922513050892434e-05, "loss": 0.2324, "step": 10218 }, { "epoch": 4.808941176470588, "grad_norm": 1.152538181355955, "learning_rate": 3.919557850350486e-05, "loss": 0.918, "step": 10219 }, { "epoch": 4.8094117647058825, "grad_norm": 1.6420360886827865, "learning_rate": 3.916603596132642e-05, "loss": 0.3809, "step": 10220 }, { "epoch": 4.809882352941177, "grad_norm": 1.213758997177962, "learning_rate": 3.91365028849122e-05, "loss": 1.1328, "step": 10221 }, { "epoch": 4.810352941176471, "grad_norm": 0.6555902386185948, "learning_rate": 3.910697927678437e-05, "loss": 0.2178, "step": 10222 }, { "epoch": 4.810823529411765, "grad_norm": 0.871804561808368, "learning_rate": 3.9077465139464446e-05, "loss": 0.2793, "step": 10223 }, { "epoch": 4.8112941176470585, "grad_norm": 1.0856090858046075, "learning_rate": 3.9047960475473095e-05, "loss": 0.9961, "step": 10224 }, { "epoch": 4.811764705882353, "grad_norm": 0.6608518624079945, "learning_rate": 3.901846528733011e-05, "loss": 0.2119, "step": 10225 }, { "epoch": 4.812235294117647, "grad_norm": 1.1118524359158983, "learning_rate": 3.898897957755454e-05, "loss": 0.4316, "step": 10226 }, { "epoch": 4.812705882352941, "grad_norm": 1.3426075387250556, "learning_rate": 3.895950334866468e-05, "loss": 1.0625, "step": 10227 }, { "epoch": 4.813176470588235, "grad_norm": 0.9830598249943203, "learning_rate": 3.893003660317783e-05, "loss": 0.7969, "step": 10228 }, { "epoch": 4.81364705882353, "grad_norm": 1.2401506350541363, "learning_rate": 3.8900579343610646e-05, "loss": 0.9141, "step": 10229 }, { "epoch": 4.814117647058824, "grad_norm": 1.5172111216011612, "learning_rate": 3.887113157247897e-05, "loss": 1.4062, "step": 10230 }, { "epoch": 4.814588235294117, "grad_norm": 1.0029770270506049, "learning_rate": 3.884169329229769e-05, "loss": 0.8789, "step": 10231 }, { "epoch": 4.815058823529411, "grad_norm": 0.7366961062140722, "learning_rate": 3.881226450558104e-05, "loss": 0.1689, "step": 10232 }, { "epoch": 4.815529411764706, "grad_norm": 1.1915859490975922, "learning_rate": 3.878284521484242e-05, "loss": 1.2656, "step": 10233 }, { "epoch": 4.816, "grad_norm": 1.2176551454528641, "learning_rate": 3.8753435422594295e-05, "loss": 0.5117, "step": 10234 }, { "epoch": 4.816470588235294, "grad_norm": 0.7062881783326015, "learning_rate": 3.8724035131348437e-05, "loss": 0.4004, "step": 10235 }, { "epoch": 4.816941176470588, "grad_norm": 1.1309651980534419, "learning_rate": 3.8694644343615836e-05, "loss": 1.0156, "step": 10236 }, { "epoch": 4.8174117647058825, "grad_norm": 1.0380680210349944, "learning_rate": 3.866526306190651e-05, "loss": 0.5195, "step": 10237 }, { "epoch": 4.817882352941177, "grad_norm": 0.6608733799520983, "learning_rate": 3.863589128872981e-05, "loss": 0.2354, "step": 10238 }, { "epoch": 4.818352941176471, "grad_norm": 1.2768229931712134, "learning_rate": 3.8606529026594276e-05, "loss": 1.1875, "step": 10239 }, { "epoch": 4.818823529411764, "grad_norm": 0.8984845689608696, "learning_rate": 3.85771762780075e-05, "loss": 0.4727, "step": 10240 }, { "epoch": 4.8192941176470585, "grad_norm": 1.0688637746384113, "learning_rate": 3.854783304547639e-05, "loss": 0.9727, "step": 10241 }, { "epoch": 4.819764705882353, "grad_norm": 1.262098895041028, "learning_rate": 3.851849933150703e-05, "loss": 0.5703, "step": 10242 }, { "epoch": 4.820235294117647, "grad_norm": 0.9668474776687138, "learning_rate": 3.848917513860463e-05, "loss": 0.3926, "step": 10243 }, { "epoch": 4.820705882352941, "grad_norm": 1.4582677369826054, "learning_rate": 3.8459860469273666e-05, "loss": 1.1562, "step": 10244 }, { "epoch": 4.821176470588235, "grad_norm": 1.1120373751909218, "learning_rate": 3.8430555326017684e-05, "loss": 0.6289, "step": 10245 }, { "epoch": 4.82164705882353, "grad_norm": 0.6865507711805633, "learning_rate": 3.840125971133952e-05, "loss": 0.2324, "step": 10246 }, { "epoch": 4.822117647058824, "grad_norm": 0.8145443128108102, "learning_rate": 3.8371973627741204e-05, "loss": 0.3008, "step": 10247 }, { "epoch": 4.822588235294118, "grad_norm": 0.8035434956016309, "learning_rate": 3.8342697077723845e-05, "loss": 0.6602, "step": 10248 }, { "epoch": 4.823058823529411, "grad_norm": 0.7024609406748834, "learning_rate": 3.8313430063787856e-05, "loss": 0.2285, "step": 10249 }, { "epoch": 4.823529411764706, "grad_norm": 0.9959501852154281, "learning_rate": 3.8284172588432716e-05, "loss": 0.416, "step": 10250 }, { "epoch": 4.824, "grad_norm": 1.2927219671998575, "learning_rate": 3.8254924654157214e-05, "loss": 1.1562, "step": 10251 }, { "epoch": 4.824470588235294, "grad_norm": 1.2284523111372083, "learning_rate": 3.8225686263459296e-05, "loss": 0.6719, "step": 10252 }, { "epoch": 4.824941176470588, "grad_norm": 1.1811295308013237, "learning_rate": 3.819645741883597e-05, "loss": 0.7109, "step": 10253 }, { "epoch": 4.8254117647058825, "grad_norm": 0.9280439652474081, "learning_rate": 3.816723812278358e-05, "loss": 0.7617, "step": 10254 }, { "epoch": 4.825882352941177, "grad_norm": 1.0668611327102369, "learning_rate": 3.813802837779762e-05, "loss": 0.6172, "step": 10255 }, { "epoch": 4.826352941176471, "grad_norm": 0.7283845193807253, "learning_rate": 3.810882818637268e-05, "loss": 0.2217, "step": 10256 }, { "epoch": 4.826823529411765, "grad_norm": 0.5960682377623835, "learning_rate": 3.807963755100263e-05, "loss": 0.4707, "step": 10257 }, { "epoch": 4.8272941176470585, "grad_norm": 1.0430637168038674, "learning_rate": 3.8050456474180554e-05, "loss": 0.7422, "step": 10258 }, { "epoch": 4.827764705882353, "grad_norm": 1.0576104711782273, "learning_rate": 3.802128495839856e-05, "loss": 0.8633, "step": 10259 }, { "epoch": 4.828235294117647, "grad_norm": 0.9419850706215719, "learning_rate": 3.799212300614809e-05, "loss": 0.252, "step": 10260 }, { "epoch": 4.828705882352941, "grad_norm": 0.9882447161102981, "learning_rate": 3.7962970619919745e-05, "loss": 0.8125, "step": 10261 }, { "epoch": 4.829176470588235, "grad_norm": 0.9312814156483098, "learning_rate": 3.7933827802203207e-05, "loss": 0.5781, "step": 10262 }, { "epoch": 4.82964705882353, "grad_norm": 1.7899272956270769, "learning_rate": 3.790469455548745e-05, "loss": 1.3984, "step": 10263 }, { "epoch": 4.830117647058824, "grad_norm": 1.3058611801989732, "learning_rate": 3.787557088226065e-05, "loss": 1.625, "step": 10264 }, { "epoch": 4.830588235294117, "grad_norm": 1.2642492543891684, "learning_rate": 3.7846456785010014e-05, "loss": 1.1562, "step": 10265 }, { "epoch": 4.831058823529411, "grad_norm": 1.040560444196723, "learning_rate": 3.7817352266222086e-05, "loss": 0.7812, "step": 10266 }, { "epoch": 4.831529411764706, "grad_norm": 1.3127540268269977, "learning_rate": 3.778825732838257e-05, "loss": 1.4766, "step": 10267 }, { "epoch": 4.832, "grad_norm": 1.058042968972149, "learning_rate": 3.775917197397622e-05, "loss": 0.3203, "step": 10268 }, { "epoch": 4.832470588235294, "grad_norm": 0.764498019481354, "learning_rate": 3.773009620548718e-05, "loss": 0.6875, "step": 10269 }, { "epoch": 4.832941176470588, "grad_norm": 1.1128776778628278, "learning_rate": 3.770103002539857e-05, "loss": 0.5742, "step": 10270 }, { "epoch": 4.8334117647058825, "grad_norm": 1.166066795244883, "learning_rate": 3.767197343619279e-05, "loss": 1.1406, "step": 10271 }, { "epoch": 4.833882352941177, "grad_norm": 1.093999801674081, "learning_rate": 3.7642926440351514e-05, "loss": 0.4922, "step": 10272 }, { "epoch": 4.834352941176471, "grad_norm": 1.2231739111216513, "learning_rate": 3.761388904035537e-05, "loss": 0.8594, "step": 10273 }, { "epoch": 4.834823529411764, "grad_norm": 1.3753204128542647, "learning_rate": 3.758486123868435e-05, "loss": 1.2656, "step": 10274 }, { "epoch": 4.8352941176470585, "grad_norm": 1.0514025232568647, "learning_rate": 3.755584303781763e-05, "loss": 0.5352, "step": 10275 }, { "epoch": 4.835764705882353, "grad_norm": 1.5923388868740105, "learning_rate": 3.7526834440233395e-05, "loss": 1.3906, "step": 10276 }, { "epoch": 4.836235294117647, "grad_norm": 0.9150595059094839, "learning_rate": 3.749783544840918e-05, "loss": 0.3184, "step": 10277 }, { "epoch": 4.836705882352941, "grad_norm": 1.0340891384259467, "learning_rate": 3.746884606482169e-05, "loss": 0.7578, "step": 10278 }, { "epoch": 4.837176470588235, "grad_norm": 0.6363517958304793, "learning_rate": 3.743986629194666e-05, "loss": 0.4609, "step": 10279 }, { "epoch": 4.83764705882353, "grad_norm": 1.4454803526454747, "learning_rate": 3.741089613225914e-05, "loss": 0.6719, "step": 10280 }, { "epoch": 4.838117647058824, "grad_norm": 0.9433113617251817, "learning_rate": 3.7381935588233385e-05, "loss": 0.4121, "step": 10281 }, { "epoch": 4.838588235294118, "grad_norm": 1.2258567228050232, "learning_rate": 3.735298466234269e-05, "loss": 1.0156, "step": 10282 }, { "epoch": 4.839058823529411, "grad_norm": 0.995960599884674, "learning_rate": 3.732404335705962e-05, "loss": 0.6172, "step": 10283 }, { "epoch": 4.839529411764706, "grad_norm": 0.8543019919881548, "learning_rate": 3.729511167485597e-05, "loss": 1.0547, "step": 10284 }, { "epoch": 4.84, "grad_norm": 0.48055847065581137, "learning_rate": 3.72661896182025e-05, "loss": 0.1079, "step": 10285 }, { "epoch": 4.840470588235294, "grad_norm": 1.050407966826206, "learning_rate": 3.7237277189569505e-05, "loss": 1.3359, "step": 10286 }, { "epoch": 4.840941176470588, "grad_norm": 1.8154036383012824, "learning_rate": 3.720837439142608e-05, "loss": 0.8398, "step": 10287 }, { "epoch": 4.8414117647058825, "grad_norm": 1.1121104503601855, "learning_rate": 3.717948122624072e-05, "loss": 0.5742, "step": 10288 }, { "epoch": 4.841882352941177, "grad_norm": 0.3768753774749177, "learning_rate": 3.7150597696481076e-05, "loss": 0.0596, "step": 10289 }, { "epoch": 4.842352941176471, "grad_norm": 1.088403019901758, "learning_rate": 3.712172380461387e-05, "loss": 0.7188, "step": 10290 }, { "epoch": 4.842823529411765, "grad_norm": 1.1070932289412265, "learning_rate": 3.709285955310513e-05, "loss": 0.4434, "step": 10291 }, { "epoch": 4.8432941176470585, "grad_norm": 0.8120941259337378, "learning_rate": 3.7064004944420026e-05, "loss": 0.2061, "step": 10292 }, { "epoch": 4.843764705882353, "grad_norm": 1.0721022511922704, "learning_rate": 3.703515998102281e-05, "loss": 0.5234, "step": 10293 }, { "epoch": 4.844235294117647, "grad_norm": 1.0735552806246969, "learning_rate": 3.7006324665377025e-05, "loss": 0.8242, "step": 10294 }, { "epoch": 4.844705882352941, "grad_norm": 0.5423225860212404, "learning_rate": 3.6977498999945384e-05, "loss": 0.1357, "step": 10295 }, { "epoch": 4.845176470588235, "grad_norm": 1.055021630788767, "learning_rate": 3.694868298718965e-05, "loss": 0.9219, "step": 10296 }, { "epoch": 4.84564705882353, "grad_norm": 1.2245312404332842, "learning_rate": 3.6919876629570905e-05, "loss": 0.6055, "step": 10297 }, { "epoch": 4.846117647058824, "grad_norm": 0.8783729346417658, "learning_rate": 3.68910799295494e-05, "loss": 0.416, "step": 10298 }, { "epoch": 4.846588235294117, "grad_norm": 1.0684800769247482, "learning_rate": 3.686229288958442e-05, "loss": 0.457, "step": 10299 }, { "epoch": 4.847058823529411, "grad_norm": 1.0347036083581178, "learning_rate": 3.6833515512134606e-05, "loss": 0.9297, "step": 10300 }, { "epoch": 4.847529411764706, "grad_norm": 1.0870417418562346, "learning_rate": 3.6804747799657594e-05, "loss": 0.8125, "step": 10301 }, { "epoch": 4.848, "grad_norm": 0.9524820341697329, "learning_rate": 3.677598975461036e-05, "loss": 0.6172, "step": 10302 }, { "epoch": 4.848470588235294, "grad_norm": 0.92743554254267, "learning_rate": 3.6747241379448986e-05, "loss": 0.6875, "step": 10303 }, { "epoch": 4.848941176470588, "grad_norm": 1.1244561250683942, "learning_rate": 3.671850267662866e-05, "loss": 0.7422, "step": 10304 }, { "epoch": 4.8494117647058825, "grad_norm": 0.842214337093607, "learning_rate": 3.668977364860386e-05, "loss": 0.2988, "step": 10305 }, { "epoch": 4.849882352941177, "grad_norm": 0.6747819811865735, "learning_rate": 3.666105429782821e-05, "loss": 0.1875, "step": 10306 }, { "epoch": 4.850352941176471, "grad_norm": 0.7175425071558189, "learning_rate": 3.6632344626754405e-05, "loss": 0.2891, "step": 10307 }, { "epoch": 4.850823529411764, "grad_norm": 0.9469000241147774, "learning_rate": 3.660364463783443e-05, "loss": 0.375, "step": 10308 }, { "epoch": 4.8512941176470585, "grad_norm": 1.1803409168028207, "learning_rate": 3.657495433351945e-05, "loss": 0.5156, "step": 10309 }, { "epoch": 4.851764705882353, "grad_norm": 0.973502021186789, "learning_rate": 3.6546273716259685e-05, "loss": 0.5781, "step": 10310 }, { "epoch": 4.852235294117647, "grad_norm": 1.1837034467563603, "learning_rate": 3.6517602788504613e-05, "loss": 0.8828, "step": 10311 }, { "epoch": 4.852705882352941, "grad_norm": 1.1742691830286902, "learning_rate": 3.648894155270295e-05, "loss": 0.5, "step": 10312 }, { "epoch": 4.853176470588235, "grad_norm": 0.8158599815049147, "learning_rate": 3.64602900113024e-05, "loss": 0.3105, "step": 10313 }, { "epoch": 4.85364705882353, "grad_norm": 0.914520014469363, "learning_rate": 3.643164816674999e-05, "loss": 0.6445, "step": 10314 }, { "epoch": 4.854117647058824, "grad_norm": 1.146856364482452, "learning_rate": 3.6403016021491924e-05, "loss": 1.2734, "step": 10315 }, { "epoch": 4.854588235294118, "grad_norm": 1.230862939417421, "learning_rate": 3.637439357797343e-05, "loss": 1.3516, "step": 10316 }, { "epoch": 4.855058823529411, "grad_norm": 1.0001742096470672, "learning_rate": 3.634578083863905e-05, "loss": 0.6406, "step": 10317 }, { "epoch": 4.855529411764706, "grad_norm": 1.4142294452379638, "learning_rate": 3.631717780593252e-05, "loss": 1.1797, "step": 10318 }, { "epoch": 4.856, "grad_norm": 1.065761819986334, "learning_rate": 3.628858448229655e-05, "loss": 0.8984, "step": 10319 }, { "epoch": 4.856470588235294, "grad_norm": 0.9192602398649877, "learning_rate": 3.6260000870173266e-05, "loss": 1.2109, "step": 10320 }, { "epoch": 4.856941176470588, "grad_norm": 1.0660840112725887, "learning_rate": 3.6231426972003766e-05, "loss": 0.4453, "step": 10321 }, { "epoch": 4.8574117647058825, "grad_norm": 0.9622340481635382, "learning_rate": 3.620286279022843e-05, "loss": 0.9023, "step": 10322 }, { "epoch": 4.857882352941177, "grad_norm": 1.0159886973379326, "learning_rate": 3.617430832728682e-05, "loss": 0.6055, "step": 10323 }, { "epoch": 4.85835294117647, "grad_norm": 0.9334368157793715, "learning_rate": 3.614576358561757e-05, "loss": 0.4473, "step": 10324 }, { "epoch": 4.858823529411764, "grad_norm": 0.906176314687298, "learning_rate": 3.611722856765855e-05, "loss": 0.6133, "step": 10325 }, { "epoch": 4.8592941176470585, "grad_norm": 1.180466358398113, "learning_rate": 3.608870327584683e-05, "loss": 1.0, "step": 10326 }, { "epoch": 4.859764705882353, "grad_norm": 1.2007800909886381, "learning_rate": 3.606018771261856e-05, "loss": 0.7734, "step": 10327 }, { "epoch": 4.860235294117647, "grad_norm": 1.266489602648346, "learning_rate": 3.6031681880409106e-05, "loss": 0.5742, "step": 10328 }, { "epoch": 4.860705882352941, "grad_norm": 1.008024931728483, "learning_rate": 3.6003185781653097e-05, "loss": 0.3672, "step": 10329 }, { "epoch": 4.861176470588235, "grad_norm": 1.1024145096641595, "learning_rate": 3.5974699418784084e-05, "loss": 0.9883, "step": 10330 }, { "epoch": 4.86164705882353, "grad_norm": 1.1672439005485191, "learning_rate": 3.5946222794235105e-05, "loss": 1.125, "step": 10331 }, { "epoch": 4.862117647058824, "grad_norm": 1.2951057652624047, "learning_rate": 3.5917755910438094e-05, "loss": 1.2031, "step": 10332 }, { "epoch": 4.862588235294117, "grad_norm": 1.467441404122617, "learning_rate": 3.58892987698243e-05, "loss": 1.2656, "step": 10333 }, { "epoch": 4.863058823529411, "grad_norm": 0.9732216585283121, "learning_rate": 3.5860851374824134e-05, "loss": 0.9336, "step": 10334 }, { "epoch": 4.863529411764706, "grad_norm": 0.7384575172478431, "learning_rate": 3.5832413727867075e-05, "loss": 0.6836, "step": 10335 }, { "epoch": 4.864, "grad_norm": 0.9616048269647142, "learning_rate": 3.580398583138188e-05, "loss": 0.6641, "step": 10336 }, { "epoch": 4.864470588235294, "grad_norm": 0.9831683040180337, "learning_rate": 3.577556768779646e-05, "loss": 0.5039, "step": 10337 }, { "epoch": 4.864941176470588, "grad_norm": 0.7083834263168473, "learning_rate": 3.574715929953778e-05, "loss": 0.1855, "step": 10338 }, { "epoch": 4.8654117647058825, "grad_norm": 1.0151856814297953, "learning_rate": 3.571876066903211e-05, "loss": 1.0469, "step": 10339 }, { "epoch": 4.865882352941177, "grad_norm": 0.9848348854916122, "learning_rate": 3.569037179870487e-05, "loss": 0.6016, "step": 10340 }, { "epoch": 4.866352941176471, "grad_norm": 0.870281157054939, "learning_rate": 3.5661992690980506e-05, "loss": 0.707, "step": 10341 }, { "epoch": 4.866823529411764, "grad_norm": 0.7409540418623204, "learning_rate": 3.56336233482828e-05, "loss": 0.1904, "step": 10342 }, { "epoch": 4.8672941176470585, "grad_norm": 0.8437483859139224, "learning_rate": 3.560526377303467e-05, "loss": 0.3105, "step": 10343 }, { "epoch": 4.867764705882353, "grad_norm": 0.9091632452986498, "learning_rate": 3.557691396765807e-05, "loss": 0.4766, "step": 10344 }, { "epoch": 4.868235294117647, "grad_norm": 0.928560885657385, "learning_rate": 3.554857393457426e-05, "loss": 0.3867, "step": 10345 }, { "epoch": 4.868705882352941, "grad_norm": 1.5341646208757036, "learning_rate": 3.5520243676203664e-05, "loss": 1.2188, "step": 10346 }, { "epoch": 4.869176470588235, "grad_norm": 1.2021766594875096, "learning_rate": 3.5491923194965735e-05, "loss": 0.5742, "step": 10347 }, { "epoch": 4.86964705882353, "grad_norm": 1.472063748106608, "learning_rate": 3.546361249327922e-05, "loss": 1.375, "step": 10348 }, { "epoch": 4.870117647058824, "grad_norm": 1.4302285040902352, "learning_rate": 3.5435311573562024e-05, "loss": 1.1172, "step": 10349 }, { "epoch": 4.870588235294118, "grad_norm": 1.1912858792563281, "learning_rate": 3.540702043823113e-05, "loss": 0.5664, "step": 10350 }, { "epoch": 4.871058823529411, "grad_norm": 0.98188611461137, "learning_rate": 3.537873908970282e-05, "loss": 0.3633, "step": 10351 }, { "epoch": 4.871529411764706, "grad_norm": 1.0057665185918714, "learning_rate": 3.535046753039235e-05, "loss": 0.2227, "step": 10352 }, { "epoch": 4.872, "grad_norm": 1.0600962706316623, "learning_rate": 3.53222057627143e-05, "loss": 0.8984, "step": 10353 }, { "epoch": 4.872470588235294, "grad_norm": 0.9906025600657475, "learning_rate": 3.5293953789082425e-05, "loss": 0.4375, "step": 10354 }, { "epoch": 4.872941176470588, "grad_norm": 0.944912698799107, "learning_rate": 3.5265711611909476e-05, "loss": 0.6367, "step": 10355 }, { "epoch": 4.8734117647058826, "grad_norm": 0.9967803697672011, "learning_rate": 3.5237479233607525e-05, "loss": 0.625, "step": 10356 }, { "epoch": 4.873882352941177, "grad_norm": 1.228967709149967, "learning_rate": 3.520925665658779e-05, "loss": 0.8555, "step": 10357 }, { "epoch": 4.87435294117647, "grad_norm": 1.1430685199931911, "learning_rate": 3.518104388326055e-05, "loss": 0.5703, "step": 10358 }, { "epoch": 4.874823529411764, "grad_norm": 0.9028088556514576, "learning_rate": 3.5152840916035344e-05, "loss": 0.4707, "step": 10359 }, { "epoch": 4.875294117647059, "grad_norm": 0.9040232437279673, "learning_rate": 3.5124647757320883e-05, "loss": 0.3281, "step": 10360 }, { "epoch": 4.875764705882353, "grad_norm": 1.0884702049818233, "learning_rate": 3.509646440952492e-05, "loss": 1.0938, "step": 10361 }, { "epoch": 4.876235294117647, "grad_norm": 0.916852578635627, "learning_rate": 3.506829087505451e-05, "loss": 0.3242, "step": 10362 }, { "epoch": 4.876705882352941, "grad_norm": 1.3044463582086359, "learning_rate": 3.504012715631583e-05, "loss": 0.7891, "step": 10363 }, { "epoch": 4.8771764705882354, "grad_norm": 0.517775085147618, "learning_rate": 3.501197325571411e-05, "loss": 0.1348, "step": 10364 }, { "epoch": 4.87764705882353, "grad_norm": 0.9631034227065409, "learning_rate": 3.49838291756539e-05, "loss": 0.2891, "step": 10365 }, { "epoch": 4.878117647058824, "grad_norm": 0.748523636212281, "learning_rate": 3.495569491853887e-05, "loss": 0.9492, "step": 10366 }, { "epoch": 4.878588235294117, "grad_norm": 1.193187714557557, "learning_rate": 3.492757048677175e-05, "loss": 1.1719, "step": 10367 }, { "epoch": 4.8790588235294114, "grad_norm": 0.9951327127290192, "learning_rate": 3.4899455882754526e-05, "loss": 0.5625, "step": 10368 }, { "epoch": 4.879529411764706, "grad_norm": 1.227541455591367, "learning_rate": 3.4871351108888375e-05, "loss": 0.625, "step": 10369 }, { "epoch": 4.88, "grad_norm": 1.014956368272633, "learning_rate": 3.4843256167573514e-05, "loss": 0.9883, "step": 10370 }, { "epoch": 4.880470588235294, "grad_norm": 0.7796480644916272, "learning_rate": 3.481517106120944e-05, "loss": 0.2559, "step": 10371 }, { "epoch": 4.880941176470588, "grad_norm": 0.9130816429530144, "learning_rate": 3.478709579219472e-05, "loss": 0.625, "step": 10372 }, { "epoch": 4.881411764705883, "grad_norm": 0.839749575778997, "learning_rate": 3.4759030362927115e-05, "loss": 0.2559, "step": 10373 }, { "epoch": 4.881882352941177, "grad_norm": 1.4675363103446415, "learning_rate": 3.4730974775803574e-05, "loss": 1.5391, "step": 10374 }, { "epoch": 4.882352941176471, "grad_norm": 0.8266176203100327, "learning_rate": 3.4702929033220174e-05, "loss": 0.5664, "step": 10375 }, { "epoch": 4.882823529411764, "grad_norm": 0.8509703707829287, "learning_rate": 3.467489313757217e-05, "loss": 0.6133, "step": 10376 }, { "epoch": 4.883294117647059, "grad_norm": 0.8647326147734193, "learning_rate": 3.4646867091254e-05, "loss": 0.7383, "step": 10377 }, { "epoch": 4.883764705882353, "grad_norm": 1.0848712591131735, "learning_rate": 3.4618850896659135e-05, "loss": 0.7852, "step": 10378 }, { "epoch": 4.884235294117647, "grad_norm": 0.9108660687799165, "learning_rate": 3.4590844556180395e-05, "loss": 0.5742, "step": 10379 }, { "epoch": 4.884705882352941, "grad_norm": 1.0494503237839874, "learning_rate": 3.456284807220958e-05, "loss": 0.4512, "step": 10380 }, { "epoch": 4.8851764705882355, "grad_norm": 0.5686420067962054, "learning_rate": 3.453486144713772e-05, "loss": 0.1602, "step": 10381 }, { "epoch": 4.88564705882353, "grad_norm": 1.313657344331523, "learning_rate": 3.450688468335512e-05, "loss": 1.0234, "step": 10382 }, { "epoch": 4.886117647058824, "grad_norm": 1.1089364451190586, "learning_rate": 3.4478917783251e-05, "loss": 0.8398, "step": 10383 }, { "epoch": 4.886588235294118, "grad_norm": 1.5547301599400862, "learning_rate": 3.445096074921393e-05, "loss": 2.2188, "step": 10384 }, { "epoch": 4.8870588235294115, "grad_norm": 1.0683870634828045, "learning_rate": 3.442301358363163e-05, "loss": 0.4453, "step": 10385 }, { "epoch": 4.887529411764706, "grad_norm": 1.2966330903825156, "learning_rate": 3.4395076288890825e-05, "loss": 0.8867, "step": 10386 }, { "epoch": 4.888, "grad_norm": 0.7862795544800283, "learning_rate": 3.436714886737753e-05, "loss": 0.2119, "step": 10387 }, { "epoch": 4.888470588235294, "grad_norm": 1.3768306619122586, "learning_rate": 3.4339231321476965e-05, "loss": 1.0938, "step": 10388 }, { "epoch": 4.888941176470588, "grad_norm": 0.5832309698211253, "learning_rate": 3.43113236535733e-05, "loss": 0.0845, "step": 10389 }, { "epoch": 4.889411764705883, "grad_norm": 0.48664975432179663, "learning_rate": 3.428342586605005e-05, "loss": 0.0669, "step": 10390 }, { "epoch": 4.889882352941177, "grad_norm": 1.176847894262301, "learning_rate": 3.425553796128985e-05, "loss": 1.3906, "step": 10391 }, { "epoch": 4.89035294117647, "grad_norm": 0.7837751704107732, "learning_rate": 3.42276599416744e-05, "loss": 0.2559, "step": 10392 }, { "epoch": 4.890823529411764, "grad_norm": 0.9871228871297918, "learning_rate": 3.419979180958464e-05, "loss": 0.3535, "step": 10393 }, { "epoch": 4.891294117647059, "grad_norm": 1.2892406055844472, "learning_rate": 3.417193356740071e-05, "loss": 1.4375, "step": 10394 }, { "epoch": 4.891764705882353, "grad_norm": 1.3839148739135592, "learning_rate": 3.414408521750173e-05, "loss": 0.5547, "step": 10395 }, { "epoch": 4.892235294117647, "grad_norm": 0.9385219015643129, "learning_rate": 3.411624676226616e-05, "loss": 0.4941, "step": 10396 }, { "epoch": 4.892705882352941, "grad_norm": 1.2884835526551504, "learning_rate": 3.408841820407157e-05, "loss": 0.9727, "step": 10397 }, { "epoch": 4.8931764705882355, "grad_norm": 1.2501984097639134, "learning_rate": 3.406059954529454e-05, "loss": 1.3594, "step": 10398 }, { "epoch": 4.89364705882353, "grad_norm": 1.4346031389750804, "learning_rate": 3.403279078831101e-05, "loss": 0.7539, "step": 10399 }, { "epoch": 4.894117647058824, "grad_norm": 1.275474841481246, "learning_rate": 3.4004991935496004e-05, "loss": 1.2578, "step": 10400 }, { "epoch": 4.894588235294117, "grad_norm": 1.5437926907609771, "learning_rate": 3.397720298922358e-05, "loss": 1.3125, "step": 10401 }, { "epoch": 4.8950588235294115, "grad_norm": 0.9521980321969117, "learning_rate": 3.394942395186716e-05, "loss": 1.1641, "step": 10402 }, { "epoch": 4.895529411764706, "grad_norm": 1.464575324375756, "learning_rate": 3.392165482579912e-05, "loss": 1.5312, "step": 10403 }, { "epoch": 4.896, "grad_norm": 0.969569561378303, "learning_rate": 3.389389561339112e-05, "loss": 0.4922, "step": 10404 }, { "epoch": 4.896470588235294, "grad_norm": 1.3578744694570661, "learning_rate": 3.3866146317013976e-05, "loss": 0.9648, "step": 10405 }, { "epoch": 4.896941176470588, "grad_norm": 0.8775982955729726, "learning_rate": 3.3838406939037525e-05, "loss": 0.3672, "step": 10406 }, { "epoch": 4.897411764705883, "grad_norm": 1.707984247807497, "learning_rate": 3.381067748183087e-05, "loss": 1.3438, "step": 10407 }, { "epoch": 4.897882352941177, "grad_norm": 1.04940737572046, "learning_rate": 3.3782957947762336e-05, "loss": 1.0703, "step": 10408 }, { "epoch": 4.898352941176471, "grad_norm": 0.8283900851997732, "learning_rate": 3.375524833919918e-05, "loss": 0.5391, "step": 10409 }, { "epoch": 4.898823529411764, "grad_norm": 1.2517541738461455, "learning_rate": 3.372754865850799e-05, "loss": 0.9883, "step": 10410 }, { "epoch": 4.899294117647059, "grad_norm": 0.5953107043425911, "learning_rate": 3.3699858908054506e-05, "loss": 0.5117, "step": 10411 }, { "epoch": 4.899764705882353, "grad_norm": 1.0375402840442922, "learning_rate": 3.367217909020348e-05, "loss": 0.5625, "step": 10412 }, { "epoch": 4.900235294117647, "grad_norm": 0.911583841659469, "learning_rate": 3.364450920731894e-05, "loss": 0.498, "step": 10413 }, { "epoch": 4.900705882352941, "grad_norm": 1.1638165501365798, "learning_rate": 3.361684926176408e-05, "loss": 1.0391, "step": 10414 }, { "epoch": 4.9011764705882355, "grad_norm": 1.1778127210789202, "learning_rate": 3.35891992559011e-05, "loss": 1.0, "step": 10415 }, { "epoch": 4.90164705882353, "grad_norm": 1.1247413847379104, "learning_rate": 3.35615591920915e-05, "loss": 0.6016, "step": 10416 }, { "epoch": 4.902117647058824, "grad_norm": 0.8732496902792847, "learning_rate": 3.353392907269591e-05, "loss": 0.4941, "step": 10417 }, { "epoch": 4.902588235294118, "grad_norm": 0.8604393024111433, "learning_rate": 3.3506308900073964e-05, "loss": 0.3223, "step": 10418 }, { "epoch": 4.9030588235294115, "grad_norm": 0.8207629793505788, "learning_rate": 3.3478698676584714e-05, "loss": 0.3418, "step": 10419 }, { "epoch": 4.903529411764706, "grad_norm": 1.442050676043313, "learning_rate": 3.34510984045861e-05, "loss": 0.875, "step": 10420 }, { "epoch": 4.904, "grad_norm": 0.9743610015376458, "learning_rate": 3.3423508086435326e-05, "loss": 0.7461, "step": 10421 }, { "epoch": 4.904470588235294, "grad_norm": 0.9687957111869676, "learning_rate": 3.339592772448883e-05, "loss": 0.3203, "step": 10422 }, { "epoch": 4.904941176470588, "grad_norm": 1.3535314581175062, "learning_rate": 3.3368357321102e-05, "loss": 0.8242, "step": 10423 }, { "epoch": 4.905411764705883, "grad_norm": 1.34802052814424, "learning_rate": 3.334079687862952e-05, "loss": 0.8047, "step": 10424 }, { "epoch": 4.905882352941177, "grad_norm": 1.4681943168246656, "learning_rate": 3.331324639942526e-05, "loss": 1.8125, "step": 10425 }, { "epoch": 4.90635294117647, "grad_norm": 1.127902010407947, "learning_rate": 3.328570588584204e-05, "loss": 0.6719, "step": 10426 }, { "epoch": 4.906823529411764, "grad_norm": 0.7961801439494669, "learning_rate": 3.325817534023202e-05, "loss": 0.3398, "step": 10427 }, { "epoch": 4.907294117647059, "grad_norm": 0.9475077367429383, "learning_rate": 3.32306547649465e-05, "loss": 0.4863, "step": 10428 }, { "epoch": 4.907764705882353, "grad_norm": 0.8998378167085295, "learning_rate": 3.320314416233574e-05, "loss": 0.3516, "step": 10429 }, { "epoch": 4.908235294117647, "grad_norm": 1.0255184994283242, "learning_rate": 3.317564353474943e-05, "loss": 0.5469, "step": 10430 }, { "epoch": 4.908705882352941, "grad_norm": 0.9966723100530221, "learning_rate": 3.314815288453613e-05, "loss": 0.4199, "step": 10431 }, { "epoch": 4.9091764705882355, "grad_norm": 0.9352295589296227, "learning_rate": 3.312067221404371e-05, "loss": 0.707, "step": 10432 }, { "epoch": 4.90964705882353, "grad_norm": 0.9844940366669159, "learning_rate": 3.309320152561923e-05, "loss": 0.6797, "step": 10433 }, { "epoch": 4.910117647058824, "grad_norm": 1.111911386737002, "learning_rate": 3.3065740821608725e-05, "loss": 0.5, "step": 10434 }, { "epoch": 4.910588235294117, "grad_norm": 0.5481698377175624, "learning_rate": 3.303829010435751e-05, "loss": 0.0801, "step": 10435 }, { "epoch": 4.9110588235294115, "grad_norm": 0.9417914176357632, "learning_rate": 3.3010849376210056e-05, "loss": 0.5352, "step": 10436 }, { "epoch": 4.911529411764706, "grad_norm": 1.1752200245985267, "learning_rate": 3.298341863950986e-05, "loss": 0.4336, "step": 10437 }, { "epoch": 4.912, "grad_norm": 1.665890716607605, "learning_rate": 3.295599789659967e-05, "loss": 1.1406, "step": 10438 }, { "epoch": 4.912470588235294, "grad_norm": 1.2778389885418673, "learning_rate": 3.2928587149821385e-05, "loss": 0.9258, "step": 10439 }, { "epoch": 4.912941176470588, "grad_norm": 1.109472748614366, "learning_rate": 3.290118640151597e-05, "loss": 0.4648, "step": 10440 }, { "epoch": 4.913411764705883, "grad_norm": 1.4400133274467228, "learning_rate": 3.287379565402358e-05, "loss": 0.9062, "step": 10441 }, { "epoch": 4.913882352941177, "grad_norm": 1.5624415541983994, "learning_rate": 3.284641490968359e-05, "loss": 0.7812, "step": 10442 }, { "epoch": 4.914352941176471, "grad_norm": 1.0370328457511238, "learning_rate": 3.281904417083436e-05, "loss": 0.2754, "step": 10443 }, { "epoch": 4.914823529411764, "grad_norm": 0.9380358626253528, "learning_rate": 3.279168343981355e-05, "loss": 0.8242, "step": 10444 }, { "epoch": 4.915294117647059, "grad_norm": 1.0602479309449797, "learning_rate": 3.2764332718957894e-05, "loss": 1.0859, "step": 10445 }, { "epoch": 4.915764705882353, "grad_norm": 1.2120662838587934, "learning_rate": 3.2736992010603226e-05, "loss": 0.5039, "step": 10446 }, { "epoch": 4.916235294117647, "grad_norm": 1.2201348984327718, "learning_rate": 3.270966131708461e-05, "loss": 0.582, "step": 10447 }, { "epoch": 4.916705882352941, "grad_norm": 1.0688185988721615, "learning_rate": 3.268234064073628e-05, "loss": 1.3984, "step": 10448 }, { "epoch": 4.9171764705882355, "grad_norm": 0.5758957135849682, "learning_rate": 3.265502998389143e-05, "loss": 0.4414, "step": 10449 }, { "epoch": 4.91764705882353, "grad_norm": 0.9954217769030322, "learning_rate": 3.262772934888265e-05, "loss": 0.6133, "step": 10450 }, { "epoch": 4.918117647058823, "grad_norm": 0.5604898409615405, "learning_rate": 3.260043873804147e-05, "loss": 0.3516, "step": 10451 }, { "epoch": 4.918588235294117, "grad_norm": 0.89116936969975, "learning_rate": 3.257315815369863e-05, "loss": 0.2451, "step": 10452 }, { "epoch": 4.9190588235294115, "grad_norm": 0.884228470747516, "learning_rate": 3.2545887598184114e-05, "loss": 0.5039, "step": 10453 }, { "epoch": 4.919529411764706, "grad_norm": 1.2615813564388751, "learning_rate": 3.251862707382686e-05, "loss": 0.75, "step": 10454 }, { "epoch": 4.92, "grad_norm": 1.052201990592991, "learning_rate": 3.249137658295511e-05, "loss": 0.416, "step": 10455 }, { "epoch": 4.920470588235294, "grad_norm": 1.1958957757062152, "learning_rate": 3.246413612789622e-05, "loss": 0.5977, "step": 10456 }, { "epoch": 4.920941176470588, "grad_norm": 0.599778729167651, "learning_rate": 3.243690571097656e-05, "loss": 0.1279, "step": 10457 }, { "epoch": 4.921411764705883, "grad_norm": 0.9139966325458837, "learning_rate": 3.24096853345218e-05, "loss": 0.5078, "step": 10458 }, { "epoch": 4.921882352941177, "grad_norm": 1.133361620680562, "learning_rate": 3.2382475000856725e-05, "loss": 0.7891, "step": 10459 }, { "epoch": 4.92235294117647, "grad_norm": 1.6002836432798855, "learning_rate": 3.2355274712305154e-05, "loss": 0.5977, "step": 10460 }, { "epoch": 4.922823529411764, "grad_norm": 0.7560822250577997, "learning_rate": 3.232808447119018e-05, "loss": 0.5508, "step": 10461 }, { "epoch": 4.923294117647059, "grad_norm": 0.8161503974566492, "learning_rate": 3.2300904279834e-05, "loss": 0.3945, "step": 10462 }, { "epoch": 4.923764705882353, "grad_norm": 1.5279036251492149, "learning_rate": 3.227373414055782e-05, "loss": 1.7266, "step": 10463 }, { "epoch": 4.924235294117647, "grad_norm": 0.9903436621106798, "learning_rate": 3.224657405568229e-05, "loss": 0.6836, "step": 10464 }, { "epoch": 4.924705882352941, "grad_norm": 1.219549589589693, "learning_rate": 3.221942402752685e-05, "loss": 0.9922, "step": 10465 }, { "epoch": 4.9251764705882355, "grad_norm": 1.1300328926650873, "learning_rate": 3.2192284058410324e-05, "loss": 0.3672, "step": 10466 }, { "epoch": 4.92564705882353, "grad_norm": 0.7022021918936181, "learning_rate": 3.216515415065061e-05, "loss": 0.1611, "step": 10467 }, { "epoch": 4.926117647058824, "grad_norm": 0.5040571843103093, "learning_rate": 3.213803430656468e-05, "loss": 0.1152, "step": 10468 }, { "epoch": 4.926588235294117, "grad_norm": 0.7707415294265785, "learning_rate": 3.2110924528468725e-05, "loss": 0.3164, "step": 10469 }, { "epoch": 4.9270588235294115, "grad_norm": 0.9063046445647999, "learning_rate": 3.208382481867809e-05, "loss": 0.6875, "step": 10470 }, { "epoch": 4.927529411764706, "grad_norm": 0.7036370229189589, "learning_rate": 3.205673517950716e-05, "loss": 0.3008, "step": 10471 }, { "epoch": 4.928, "grad_norm": 1.1737894537375733, "learning_rate": 3.202965561326955e-05, "loss": 0.5625, "step": 10472 }, { "epoch": 4.928470588235294, "grad_norm": 1.0471471071591558, "learning_rate": 3.200258612227803e-05, "loss": 0.8828, "step": 10473 }, { "epoch": 4.928941176470588, "grad_norm": 0.7321280163561747, "learning_rate": 3.19755267088444e-05, "loss": 0.1514, "step": 10474 }, { "epoch": 4.929411764705883, "grad_norm": 0.9768878548322065, "learning_rate": 3.194847737527969e-05, "loss": 0.4102, "step": 10475 }, { "epoch": 4.929882352941177, "grad_norm": 1.1447727107274066, "learning_rate": 3.1921438123894086e-05, "loss": 0.5039, "step": 10476 }, { "epoch": 4.930352941176471, "grad_norm": 0.8894937792996224, "learning_rate": 3.189440895699681e-05, "loss": 0.457, "step": 10477 }, { "epoch": 4.930823529411764, "grad_norm": 0.8322929086580884, "learning_rate": 3.1867389876896296e-05, "loss": 0.2871, "step": 10478 }, { "epoch": 4.931294117647059, "grad_norm": 0.9919658628095149, "learning_rate": 3.184038088590019e-05, "loss": 0.3711, "step": 10479 }, { "epoch": 4.931764705882353, "grad_norm": 0.8713398026923962, "learning_rate": 3.181338198631506e-05, "loss": 0.3047, "step": 10480 }, { "epoch": 4.932235294117647, "grad_norm": 1.566063365240534, "learning_rate": 3.178639318044687e-05, "loss": 1.4219, "step": 10481 }, { "epoch": 4.932705882352941, "grad_norm": 1.2452981906916778, "learning_rate": 3.175941447060048e-05, "loss": 1.2656, "step": 10482 }, { "epoch": 4.9331764705882355, "grad_norm": 0.7090872265530846, "learning_rate": 3.173244585908007e-05, "loss": 0.582, "step": 10483 }, { "epoch": 4.93364705882353, "grad_norm": 1.135665906850415, "learning_rate": 3.170548734818893e-05, "loss": 0.875, "step": 10484 }, { "epoch": 4.934117647058823, "grad_norm": 1.4518240838632561, "learning_rate": 3.1678538940229356e-05, "loss": 0.9805, "step": 10485 }, { "epoch": 4.934588235294117, "grad_norm": 0.7062559805648575, "learning_rate": 3.16516006375029e-05, "loss": 0.377, "step": 10486 }, { "epoch": 4.9350588235294115, "grad_norm": 1.0262705098888132, "learning_rate": 3.1624672442310304e-05, "loss": 1.2734, "step": 10487 }, { "epoch": 4.935529411764706, "grad_norm": 1.1374279804588237, "learning_rate": 3.159775435695126e-05, "loss": 0.5391, "step": 10488 }, { "epoch": 4.936, "grad_norm": 0.8222403298091046, "learning_rate": 3.1570846383724775e-05, "loss": 0.3926, "step": 10489 }, { "epoch": 4.936470588235294, "grad_norm": 0.9740785715704322, "learning_rate": 3.154394852492892e-05, "loss": 0.75, "step": 10490 }, { "epoch": 4.936941176470588, "grad_norm": 1.1869836649183334, "learning_rate": 3.1517060782860865e-05, "loss": 0.8438, "step": 10491 }, { "epoch": 4.937411764705883, "grad_norm": 1.0553010598747525, "learning_rate": 3.1490183159816956e-05, "loss": 0.4434, "step": 10492 }, { "epoch": 4.937882352941177, "grad_norm": 1.9804916755547988, "learning_rate": 3.146331565809273e-05, "loss": 1.3594, "step": 10493 }, { "epoch": 4.93835294117647, "grad_norm": 0.4137081116783055, "learning_rate": 3.143645827998274e-05, "loss": 0.1445, "step": 10494 }, { "epoch": 4.938823529411764, "grad_norm": 1.1213185057938548, "learning_rate": 3.140961102778075e-05, "loss": 0.7266, "step": 10495 }, { "epoch": 4.939294117647059, "grad_norm": 1.239833560608919, "learning_rate": 3.138277390377971e-05, "loss": 0.6914, "step": 10496 }, { "epoch": 4.939764705882353, "grad_norm": 0.893347332549574, "learning_rate": 3.135594691027155e-05, "loss": 0.5547, "step": 10497 }, { "epoch": 4.940235294117647, "grad_norm": 0.7900245456753714, "learning_rate": 3.132913004954747e-05, "loss": 0.6758, "step": 10498 }, { "epoch": 4.940705882352941, "grad_norm": 0.9676852585718178, "learning_rate": 3.130232332389781e-05, "loss": 0.1729, "step": 10499 }, { "epoch": 4.9411764705882355, "grad_norm": 0.8024575291532915, "learning_rate": 3.1275526735611896e-05, "loss": 0.6133, "step": 10500 }, { "epoch": 4.94164705882353, "grad_norm": 0.9309069152184574, "learning_rate": 3.1248740286978375e-05, "loss": 0.5039, "step": 10501 }, { "epoch": 4.942117647058824, "grad_norm": 0.9220408845929601, "learning_rate": 3.1221963980284876e-05, "loss": 0.3281, "step": 10502 }, { "epoch": 4.942588235294117, "grad_norm": 1.171169067402076, "learning_rate": 3.119519781781825e-05, "loss": 0.7969, "step": 10503 }, { "epoch": 4.9430588235294115, "grad_norm": 0.7376466543579226, "learning_rate": 3.1168441801864505e-05, "loss": 0.2969, "step": 10504 }, { "epoch": 4.943529411764706, "grad_norm": 1.1978747431781815, "learning_rate": 3.1141695934708664e-05, "loss": 0.6016, "step": 10505 }, { "epoch": 4.944, "grad_norm": 1.166049535731581, "learning_rate": 3.1114960218634986e-05, "loss": 1.2734, "step": 10506 }, { "epoch": 4.944470588235294, "grad_norm": 0.43366514572376574, "learning_rate": 3.1088234655926815e-05, "loss": 0.1328, "step": 10507 }, { "epoch": 4.944941176470588, "grad_norm": 1.4144982235708548, "learning_rate": 3.106151924886668e-05, "loss": 0.8867, "step": 10508 }, { "epoch": 4.945411764705883, "grad_norm": 1.2973709543484342, "learning_rate": 3.103481399973622e-05, "loss": 0.9453, "step": 10509 }, { "epoch": 4.945882352941177, "grad_norm": 1.3566247612025633, "learning_rate": 3.100811891081614e-05, "loss": 0.7383, "step": 10510 }, { "epoch": 4.946352941176471, "grad_norm": 1.1077612246219273, "learning_rate": 3.098143398438635e-05, "loss": 0.6719, "step": 10511 }, { "epoch": 4.946823529411764, "grad_norm": 0.9219607136384055, "learning_rate": 3.0954759222725916e-05, "loss": 0.4121, "step": 10512 }, { "epoch": 4.947294117647059, "grad_norm": 1.5448925081151714, "learning_rate": 3.092809462811292e-05, "loss": 1.2969, "step": 10513 }, { "epoch": 4.947764705882353, "grad_norm": 1.0279163132190166, "learning_rate": 3.090144020282469e-05, "loss": 0.9062, "step": 10514 }, { "epoch": 4.948235294117647, "grad_norm": 0.9780241520197315, "learning_rate": 3.087479594913767e-05, "loss": 0.3516, "step": 10515 }, { "epoch": 4.948705882352941, "grad_norm": 1.125995095134603, "learning_rate": 3.0848161869327356e-05, "loss": 0.4199, "step": 10516 }, { "epoch": 4.9491764705882355, "grad_norm": 1.201567509437772, "learning_rate": 3.0821537965668454e-05, "loss": 0.5859, "step": 10517 }, { "epoch": 4.94964705882353, "grad_norm": 1.5864039289298286, "learning_rate": 3.079492424043481e-05, "loss": 1.0938, "step": 10518 }, { "epoch": 4.950117647058823, "grad_norm": 0.871692677776361, "learning_rate": 3.0768320695899294e-05, "loss": 0.4883, "step": 10519 }, { "epoch": 4.950588235294117, "grad_norm": 0.8106738905513939, "learning_rate": 3.0741727334334026e-05, "loss": 0.2021, "step": 10520 }, { "epoch": 4.9510588235294115, "grad_norm": 1.003582539608932, "learning_rate": 3.071514415801024e-05, "loss": 0.5156, "step": 10521 }, { "epoch": 4.951529411764706, "grad_norm": 1.2130194488515602, "learning_rate": 3.068857116919819e-05, "loss": 0.5352, "step": 10522 }, { "epoch": 4.952, "grad_norm": 1.329015997588595, "learning_rate": 3.066200837016737e-05, "loss": 1.2266, "step": 10523 }, { "epoch": 4.952470588235294, "grad_norm": 0.9435028573095722, "learning_rate": 3.063545576318644e-05, "loss": 0.9297, "step": 10524 }, { "epoch": 4.952941176470588, "grad_norm": 0.5986337830596712, "learning_rate": 3.060891335052302e-05, "loss": 0.1768, "step": 10525 }, { "epoch": 4.953411764705883, "grad_norm": 0.49501668663272286, "learning_rate": 3.058238113444401e-05, "loss": 0.103, "step": 10526 }, { "epoch": 4.953882352941177, "grad_norm": 1.1352174669044974, "learning_rate": 3.0555859117215446e-05, "loss": 0.5977, "step": 10527 }, { "epoch": 4.95435294117647, "grad_norm": 0.8788527243775023, "learning_rate": 3.052934730110232e-05, "loss": 0.3398, "step": 10528 }, { "epoch": 4.954823529411764, "grad_norm": 1.045354156977259, "learning_rate": 3.050284568836895e-05, "loss": 1.0078, "step": 10529 }, { "epoch": 4.955294117647059, "grad_norm": 1.0323099474268744, "learning_rate": 3.0476354281278726e-05, "loss": 0.7539, "step": 10530 }, { "epoch": 4.955764705882353, "grad_norm": 0.8863597364182098, "learning_rate": 3.044987308209407e-05, "loss": 0.8477, "step": 10531 }, { "epoch": 4.956235294117647, "grad_norm": 1.097385154357012, "learning_rate": 3.0423402093076683e-05, "loss": 0.8711, "step": 10532 }, { "epoch": 4.956705882352941, "grad_norm": 1.1072676487417732, "learning_rate": 3.039694131648723e-05, "loss": 1.3438, "step": 10533 }, { "epoch": 4.9571764705882355, "grad_norm": 1.2246670068495864, "learning_rate": 3.0370490754585636e-05, "loss": 0.7539, "step": 10534 }, { "epoch": 4.95764705882353, "grad_norm": 1.2135814508829121, "learning_rate": 3.034405040963095e-05, "loss": 0.793, "step": 10535 }, { "epoch": 4.958117647058824, "grad_norm": 1.2172584953475025, "learning_rate": 3.0317620283881243e-05, "loss": 0.6289, "step": 10536 }, { "epoch": 4.958588235294117, "grad_norm": 1.0297872597369355, "learning_rate": 3.0291200379593782e-05, "loss": 0.3887, "step": 10537 }, { "epoch": 4.9590588235294115, "grad_norm": 1.5179784342626435, "learning_rate": 3.0264790699025027e-05, "loss": 1.0938, "step": 10538 }, { "epoch": 4.959529411764706, "grad_norm": 1.1296824418799718, "learning_rate": 3.0238391244430405e-05, "loss": 0.7148, "step": 10539 }, { "epoch": 4.96, "grad_norm": 0.5862207688673715, "learning_rate": 3.0212002018064614e-05, "loss": 0.1299, "step": 10540 }, { "epoch": 4.960470588235294, "grad_norm": 0.9854387402218577, "learning_rate": 3.0185623022181415e-05, "loss": 0.6523, "step": 10541 }, { "epoch": 4.960941176470588, "grad_norm": 1.4865171272938642, "learning_rate": 3.015925425903368e-05, "loss": 1.4531, "step": 10542 }, { "epoch": 4.961411764705883, "grad_norm": 1.83263757934723, "learning_rate": 3.0132895730873448e-05, "loss": 0.9062, "step": 10543 }, { "epoch": 4.961882352941177, "grad_norm": 0.9743283931126253, "learning_rate": 3.010654743995189e-05, "loss": 0.5508, "step": 10544 }, { "epoch": 4.962352941176471, "grad_norm": 1.3669965348966535, "learning_rate": 3.0080209388519222e-05, "loss": 0.7773, "step": 10545 }, { "epoch": 4.962823529411764, "grad_norm": 0.8703775977828686, "learning_rate": 3.005388157882488e-05, "loss": 0.3242, "step": 10546 }, { "epoch": 4.963294117647059, "grad_norm": 1.019861531730919, "learning_rate": 3.002756401311741e-05, "loss": 0.7812, "step": 10547 }, { "epoch": 4.963764705882353, "grad_norm": 1.1883199039995207, "learning_rate": 3.0001256693644405e-05, "loss": 0.6094, "step": 10548 }, { "epoch": 4.964235294117647, "grad_norm": 1.129670088936685, "learning_rate": 2.997495962265266e-05, "loss": 0.8594, "step": 10549 }, { "epoch": 4.964705882352941, "grad_norm": 0.7636954063487763, "learning_rate": 2.9948672802388135e-05, "loss": 0.2031, "step": 10550 }, { "epoch": 4.9651764705882355, "grad_norm": 1.7429260916733285, "learning_rate": 2.9922396235095762e-05, "loss": 1.0156, "step": 10551 }, { "epoch": 4.96564705882353, "grad_norm": 1.4372818322297396, "learning_rate": 2.9896129923019717e-05, "loss": 0.5703, "step": 10552 }, { "epoch": 4.966117647058823, "grad_norm": 1.2389961794922588, "learning_rate": 2.9869873868403287e-05, "loss": 0.7383, "step": 10553 }, { "epoch": 4.966588235294117, "grad_norm": 1.264870121412185, "learning_rate": 2.9843628073488863e-05, "loss": 0.8398, "step": 10554 }, { "epoch": 4.9670588235294115, "grad_norm": 1.2312550044362418, "learning_rate": 2.9817392540518014e-05, "loss": 0.9297, "step": 10555 }, { "epoch": 4.967529411764706, "grad_norm": 1.194620842572901, "learning_rate": 2.9791167271731287e-05, "loss": 0.6328, "step": 10556 }, { "epoch": 4.968, "grad_norm": 1.2471297677962896, "learning_rate": 2.9764952269368487e-05, "loss": 0.7461, "step": 10557 }, { "epoch": 4.968470588235294, "grad_norm": 1.1593940446357927, "learning_rate": 2.973874753566857e-05, "loss": 1.0391, "step": 10558 }, { "epoch": 4.968941176470588, "grad_norm": 1.0200628192662382, "learning_rate": 2.9712553072869437e-05, "loss": 0.5, "step": 10559 }, { "epoch": 4.969411764705883, "grad_norm": 0.9750801675612079, "learning_rate": 2.968636888320833e-05, "loss": 0.5625, "step": 10560 }, { "epoch": 4.969882352941177, "grad_norm": 1.3026646604931609, "learning_rate": 2.9660194968921403e-05, "loss": 0.5625, "step": 10561 }, { "epoch": 4.97035294117647, "grad_norm": 1.0804052558539936, "learning_rate": 2.963403133224409e-05, "loss": 0.8477, "step": 10562 }, { "epoch": 4.970823529411764, "grad_norm": 0.9827632158430611, "learning_rate": 2.9607877975410936e-05, "loss": 0.8398, "step": 10563 }, { "epoch": 4.971294117647059, "grad_norm": 1.2923682772115361, "learning_rate": 2.958173490065549e-05, "loss": 1.1641, "step": 10564 }, { "epoch": 4.971764705882353, "grad_norm": 0.8707439788328022, "learning_rate": 2.955560211021053e-05, "loss": 0.2832, "step": 10565 }, { "epoch": 4.972235294117647, "grad_norm": 0.9879410603657139, "learning_rate": 2.952947960630797e-05, "loss": 0.7305, "step": 10566 }, { "epoch": 4.972705882352941, "grad_norm": 1.1665848662184823, "learning_rate": 2.9503367391178713e-05, "loss": 0.9062, "step": 10567 }, { "epoch": 4.9731764705882355, "grad_norm": 0.7746444008319499, "learning_rate": 2.94772654670529e-05, "loss": 0.5703, "step": 10568 }, { "epoch": 4.97364705882353, "grad_norm": 1.0763969200069123, "learning_rate": 2.9451173836159826e-05, "loss": 1.0859, "step": 10569 }, { "epoch": 4.974117647058824, "grad_norm": 1.2097258321948938, "learning_rate": 2.9425092500727764e-05, "loss": 0.6602, "step": 10570 }, { "epoch": 4.974588235294117, "grad_norm": 1.4823930346662768, "learning_rate": 2.9399021462984206e-05, "loss": 1.1875, "step": 10571 }, { "epoch": 4.9750588235294115, "grad_norm": 1.135136327443449, "learning_rate": 2.9372960725155804e-05, "loss": 0.5156, "step": 10572 }, { "epoch": 4.975529411764706, "grad_norm": 0.7880954341076972, "learning_rate": 2.934691028946819e-05, "loss": 0.207, "step": 10573 }, { "epoch": 4.976, "grad_norm": 1.0327987752851149, "learning_rate": 2.9320870158146235e-05, "loss": 0.4316, "step": 10574 }, { "epoch": 4.976470588235294, "grad_norm": 0.8433433836606643, "learning_rate": 2.9294840333413932e-05, "loss": 0.2129, "step": 10575 }, { "epoch": 4.976941176470588, "grad_norm": 1.6836107534755005, "learning_rate": 2.926882081749428e-05, "loss": 1.1797, "step": 10576 }, { "epoch": 4.977411764705883, "grad_norm": 0.6024554113588108, "learning_rate": 2.9242811612609522e-05, "loss": 0.1367, "step": 10577 }, { "epoch": 4.977882352941176, "grad_norm": 1.1506745970778764, "learning_rate": 2.9216812720980998e-05, "loss": 0.6406, "step": 10578 }, { "epoch": 4.97835294117647, "grad_norm": 1.1714420841832114, "learning_rate": 2.919082414482908e-05, "loss": 0.6055, "step": 10579 }, { "epoch": 4.978823529411764, "grad_norm": 0.8868219413645245, "learning_rate": 2.9164845886373323e-05, "loss": 0.249, "step": 10580 }, { "epoch": 4.979294117647059, "grad_norm": 1.3056563695876517, "learning_rate": 2.9138877947832474e-05, "loss": 0.8516, "step": 10581 }, { "epoch": 4.979764705882353, "grad_norm": 0.9311088323707466, "learning_rate": 2.911292033142422e-05, "loss": 0.8125, "step": 10582 }, { "epoch": 4.980235294117647, "grad_norm": 2.2329804406927116, "learning_rate": 2.9086973039365562e-05, "loss": 0.5859, "step": 10583 }, { "epoch": 4.980705882352941, "grad_norm": 1.3132320518496354, "learning_rate": 2.906103607387244e-05, "loss": 0.8867, "step": 10584 }, { "epoch": 4.9811764705882355, "grad_norm": 1.050710458416062, "learning_rate": 2.9035109437160047e-05, "loss": 0.5703, "step": 10585 }, { "epoch": 4.98164705882353, "grad_norm": 1.169558206651821, "learning_rate": 2.9009193131442665e-05, "loss": 0.8359, "step": 10586 }, { "epoch": 4.982117647058823, "grad_norm": 1.5496670882432282, "learning_rate": 2.8983287158933622e-05, "loss": 1.0938, "step": 10587 }, { "epoch": 4.982588235294117, "grad_norm": 0.8463533749610602, "learning_rate": 2.895739152184543e-05, "loss": 0.2422, "step": 10588 }, { "epoch": 4.9830588235294115, "grad_norm": 1.2186934501305982, "learning_rate": 2.8931506222389756e-05, "loss": 0.4785, "step": 10589 }, { "epoch": 4.983529411764706, "grad_norm": 0.8302760813989584, "learning_rate": 2.8905631262777245e-05, "loss": 0.4062, "step": 10590 }, { "epoch": 4.984, "grad_norm": 0.9126015290060676, "learning_rate": 2.8879766645217784e-05, "loss": 0.3184, "step": 10591 }, { "epoch": 4.984470588235294, "grad_norm": 1.2902479897955226, "learning_rate": 2.8853912371920384e-05, "loss": 0.6406, "step": 10592 }, { "epoch": 4.984941176470588, "grad_norm": 1.4614342989334153, "learning_rate": 2.882806844509305e-05, "loss": 1.2031, "step": 10593 }, { "epoch": 4.985411764705883, "grad_norm": 0.4539807242075802, "learning_rate": 2.8802234866943012e-05, "loss": 0.1006, "step": 10594 }, { "epoch": 4.985882352941177, "grad_norm": 1.224066547858588, "learning_rate": 2.8776411639676644e-05, "loss": 0.8867, "step": 10595 }, { "epoch": 4.98635294117647, "grad_norm": 0.5380853372382796, "learning_rate": 2.8750598765499226e-05, "loss": 0.0864, "step": 10596 }, { "epoch": 4.986823529411764, "grad_norm": 1.0051890531018737, "learning_rate": 2.8724796246615485e-05, "loss": 0.4922, "step": 10597 }, { "epoch": 4.987294117647059, "grad_norm": 0.9410408069937803, "learning_rate": 2.8699004085228948e-05, "loss": 0.4297, "step": 10598 }, { "epoch": 4.987764705882353, "grad_norm": 0.8946807391771795, "learning_rate": 2.867322228354243e-05, "loss": 0.8594, "step": 10599 }, { "epoch": 4.988235294117647, "grad_norm": 0.5318358122257386, "learning_rate": 2.8647450843757897e-05, "loss": 0.2695, "step": 10600 }, { "epoch": 4.988705882352941, "grad_norm": 1.8179103672857084, "learning_rate": 2.862168976807624e-05, "loss": 1.0078, "step": 10601 }, { "epoch": 4.9891764705882355, "grad_norm": 1.4926834949818648, "learning_rate": 2.859593905869763e-05, "loss": 1.0625, "step": 10602 }, { "epoch": 4.98964705882353, "grad_norm": 1.114138672020175, "learning_rate": 2.8570198717821362e-05, "loss": 1.1484, "step": 10603 }, { "epoch": 4.990117647058824, "grad_norm": 1.1523363115978125, "learning_rate": 2.8544468747645676e-05, "loss": 0.875, "step": 10604 }, { "epoch": 4.990588235294117, "grad_norm": 1.183181757360133, "learning_rate": 2.8518749150368114e-05, "loss": 0.9688, "step": 10605 }, { "epoch": 4.9910588235294115, "grad_norm": 1.4043742025159387, "learning_rate": 2.849303992818527e-05, "loss": 0.9297, "step": 10606 }, { "epoch": 4.991529411764706, "grad_norm": 0.8917240904565321, "learning_rate": 2.846734108329275e-05, "loss": 0.5898, "step": 10607 }, { "epoch": 4.992, "grad_norm": 0.9369842044376279, "learning_rate": 2.8441652617885437e-05, "loss": 1.0938, "step": 10608 }, { "epoch": 4.992470588235294, "grad_norm": 1.1755203592919758, "learning_rate": 2.8415974534157255e-05, "loss": 1.0625, "step": 10609 }, { "epoch": 4.992941176470588, "grad_norm": 0.9302997919998447, "learning_rate": 2.8390306834301186e-05, "loss": 0.8672, "step": 10610 }, { "epoch": 4.993411764705883, "grad_norm": 0.7507121386681596, "learning_rate": 2.836464952050944e-05, "loss": 0.1582, "step": 10611 }, { "epoch": 4.993882352941176, "grad_norm": 0.9516228324270906, "learning_rate": 2.8339002594973204e-05, "loss": 0.9023, "step": 10612 }, { "epoch": 4.99435294117647, "grad_norm": 1.231169380063962, "learning_rate": 2.8313366059882887e-05, "loss": 1.1406, "step": 10613 }, { "epoch": 4.994823529411764, "grad_norm": 1.2642607319010604, "learning_rate": 2.828773991742802e-05, "loss": 0.8633, "step": 10614 }, { "epoch": 4.995294117647059, "grad_norm": 1.6085331624400239, "learning_rate": 2.8262124169797106e-05, "loss": 0.7383, "step": 10615 }, { "epoch": 4.995764705882353, "grad_norm": 1.148548839312798, "learning_rate": 2.823651881917793e-05, "loss": 0.75, "step": 10616 }, { "epoch": 4.996235294117647, "grad_norm": 0.9228354316891494, "learning_rate": 2.8210923867757317e-05, "loss": 0.3672, "step": 10617 }, { "epoch": 4.996705882352941, "grad_norm": 1.0521483623825052, "learning_rate": 2.8185339317721146e-05, "loss": 0.6797, "step": 10618 }, { "epoch": 4.9971764705882356, "grad_norm": 0.753373411194691, "learning_rate": 2.8159765171254495e-05, "loss": 0.4434, "step": 10619 }, { "epoch": 4.99764705882353, "grad_norm": 1.1029758764971695, "learning_rate": 2.813420143054157e-05, "loss": 0.6562, "step": 10620 }, { "epoch": 4.998117647058823, "grad_norm": 1.4491916843492776, "learning_rate": 2.810864809776554e-05, "loss": 1.0781, "step": 10621 }, { "epoch": 4.998588235294117, "grad_norm": 0.7408528709348926, "learning_rate": 2.8083105175108845e-05, "loss": 1.0078, "step": 10622 }, { "epoch": 4.9990588235294116, "grad_norm": 1.3849788874119258, "learning_rate": 2.8057572664752986e-05, "loss": 0.9062, "step": 10623 }, { "epoch": 4.999529411764706, "grad_norm": 1.0957536379674466, "learning_rate": 2.8032050568878524e-05, "loss": 1.3125, "step": 10624 }, { "epoch": 5.0, "grad_norm": 1.1306240136717958, "learning_rate": 2.800653888966519e-05, "loss": 0.5898, "step": 10625 }, { "epoch": 5.0, "eval_loss": 0.7748522162437439, "eval_runtime": 36.7329, "eval_samples_per_second": 20.418, "eval_steps_per_second": 10.209, "step": 10625 }, { "epoch": 5.000470588235294, "grad_norm": 1.5811657935660475, "learning_rate": 2.7981037629291852e-05, "loss": 0.9297, "step": 10626 }, { "epoch": 5.0009411764705884, "grad_norm": 1.206540471092074, "learning_rate": 2.7955546789936363e-05, "loss": 0.625, "step": 10627 }, { "epoch": 5.001411764705883, "grad_norm": 1.1964383304776345, "learning_rate": 2.7930066373775794e-05, "loss": 1.1875, "step": 10628 }, { "epoch": 5.001882352941177, "grad_norm": 0.8387874807662378, "learning_rate": 2.7904596382986343e-05, "loss": 0.3867, "step": 10629 }, { "epoch": 5.00235294117647, "grad_norm": 1.122275276189397, "learning_rate": 2.7879136819743203e-05, "loss": 1.2109, "step": 10630 }, { "epoch": 5.0028235294117644, "grad_norm": 1.1531535815656877, "learning_rate": 2.7853687686220766e-05, "loss": 1.0703, "step": 10631 }, { "epoch": 5.003294117647059, "grad_norm": 0.8880435544980564, "learning_rate": 2.782824898459258e-05, "loss": 0.3711, "step": 10632 }, { "epoch": 5.003764705882353, "grad_norm": 1.3518253583382254, "learning_rate": 2.7802820717031132e-05, "loss": 0.6445, "step": 10633 }, { "epoch": 5.004235294117647, "grad_norm": 0.6827559158062767, "learning_rate": 2.7777402885708212e-05, "loss": 0.5469, "step": 10634 }, { "epoch": 5.004705882352941, "grad_norm": 0.8778528085452546, "learning_rate": 2.7751995492794544e-05, "loss": 0.291, "step": 10635 }, { "epoch": 5.005176470588236, "grad_norm": 0.8674000970075872, "learning_rate": 2.7726598540460076e-05, "loss": 0.6758, "step": 10636 }, { "epoch": 5.00564705882353, "grad_norm": 1.227464147678, "learning_rate": 2.770121203087389e-05, "loss": 1.3438, "step": 10637 }, { "epoch": 5.006117647058823, "grad_norm": 0.499058929154308, "learning_rate": 2.7675835966204024e-05, "loss": 0.0664, "step": 10638 }, { "epoch": 5.006588235294117, "grad_norm": 0.8441646820147543, "learning_rate": 2.7650470348617772e-05, "loss": 0.4941, "step": 10639 }, { "epoch": 5.007058823529412, "grad_norm": 1.0131259242496369, "learning_rate": 2.7625115180281464e-05, "loss": 0.875, "step": 10640 }, { "epoch": 5.007529411764706, "grad_norm": 0.9850661978124147, "learning_rate": 2.7599770463360576e-05, "loss": 0.7617, "step": 10641 }, { "epoch": 5.008, "grad_norm": 1.2349968640428275, "learning_rate": 2.7574436200019688e-05, "loss": 0.8086, "step": 10642 }, { "epoch": 5.008470588235294, "grad_norm": 1.2184606422445603, "learning_rate": 2.7549112392422405e-05, "loss": 1.125, "step": 10643 }, { "epoch": 5.0089411764705885, "grad_norm": 0.934629951377413, "learning_rate": 2.7523799042731548e-05, "loss": 0.668, "step": 10644 }, { "epoch": 5.009411764705883, "grad_norm": 0.8998064622073593, "learning_rate": 2.749849615310902e-05, "loss": 0.8555, "step": 10645 }, { "epoch": 5.009882352941177, "grad_norm": 1.0071300361030708, "learning_rate": 2.7473203725715754e-05, "loss": 1.2422, "step": 10646 }, { "epoch": 5.01035294117647, "grad_norm": 0.9954155270529169, "learning_rate": 2.7447921762711894e-05, "loss": 0.3125, "step": 10647 }, { "epoch": 5.0108235294117645, "grad_norm": 0.7517524960781867, "learning_rate": 2.7422650266256663e-05, "loss": 0.6836, "step": 10648 }, { "epoch": 5.011294117647059, "grad_norm": 0.9895566130613938, "learning_rate": 2.7397389238508282e-05, "loss": 0.5156, "step": 10649 }, { "epoch": 5.011764705882353, "grad_norm": 1.3939541056056552, "learning_rate": 2.7372138681624244e-05, "loss": 0.5938, "step": 10650 }, { "epoch": 5.012235294117647, "grad_norm": 1.5266961743213672, "learning_rate": 2.7346898597761073e-05, "loss": 0.4238, "step": 10651 }, { "epoch": 5.012705882352941, "grad_norm": 1.1132576810560089, "learning_rate": 2.7321668989074347e-05, "loss": 0.7734, "step": 10652 }, { "epoch": 5.013176470588236, "grad_norm": 1.1097104241430684, "learning_rate": 2.7296449857718806e-05, "loss": 0.9258, "step": 10653 }, { "epoch": 5.01364705882353, "grad_norm": 1.1393917751400202, "learning_rate": 2.727124120584836e-05, "loss": 1.1172, "step": 10654 }, { "epoch": 5.014117647058823, "grad_norm": 0.8343741934664544, "learning_rate": 2.724604303561584e-05, "loss": 0.7266, "step": 10655 }, { "epoch": 5.014588235294117, "grad_norm": 0.9515908033940932, "learning_rate": 2.7220855349173353e-05, "loss": 0.4238, "step": 10656 }, { "epoch": 5.015058823529412, "grad_norm": 0.8812114823318719, "learning_rate": 2.71956781486721e-05, "loss": 0.5547, "step": 10657 }, { "epoch": 5.015529411764706, "grad_norm": 0.991404218041153, "learning_rate": 2.7170511436262243e-05, "loss": 0.7305, "step": 10658 }, { "epoch": 5.016, "grad_norm": 1.3588302007544428, "learning_rate": 2.7145355214093178e-05, "loss": 1.1562, "step": 10659 }, { "epoch": 5.016470588235294, "grad_norm": 1.0840116411830607, "learning_rate": 2.7120209484313422e-05, "loss": 0.543, "step": 10660 }, { "epoch": 5.0169411764705885, "grad_norm": 1.2455344693974018, "learning_rate": 2.7095074249070455e-05, "loss": 1.2266, "step": 10661 }, { "epoch": 5.017411764705883, "grad_norm": 1.7912804425316875, "learning_rate": 2.7069949510511023e-05, "loss": 1.4922, "step": 10662 }, { "epoch": 5.017882352941177, "grad_norm": 1.3130175911178652, "learning_rate": 2.7044835270780836e-05, "loss": 0.7617, "step": 10663 }, { "epoch": 5.01835294117647, "grad_norm": 0.758405905922057, "learning_rate": 2.701973153202482e-05, "loss": 0.3594, "step": 10664 }, { "epoch": 5.0188235294117645, "grad_norm": 1.4554603995723332, "learning_rate": 2.6994638296386977e-05, "loss": 1.0312, "step": 10665 }, { "epoch": 5.019294117647059, "grad_norm": 0.6596568204867415, "learning_rate": 2.6969555566010316e-05, "loss": 0.1826, "step": 10666 }, { "epoch": 5.019764705882353, "grad_norm": 0.8770415224023248, "learning_rate": 2.694448334303708e-05, "loss": 0.3145, "step": 10667 }, { "epoch": 5.020235294117647, "grad_norm": 1.4916383407844809, "learning_rate": 2.6919421629608584e-05, "loss": 1.1719, "step": 10668 }, { "epoch": 5.020705882352941, "grad_norm": 1.0061087677846057, "learning_rate": 2.6894370427865157e-05, "loss": 0.8398, "step": 10669 }, { "epoch": 5.021176470588236, "grad_norm": 1.4472613725751038, "learning_rate": 2.6869329739946312e-05, "loss": 1.1719, "step": 10670 }, { "epoch": 5.02164705882353, "grad_norm": 1.0743948800437528, "learning_rate": 2.6844299567990713e-05, "loss": 0.8281, "step": 10671 }, { "epoch": 5.022117647058823, "grad_norm": 1.1208051655898346, "learning_rate": 2.6819279914135956e-05, "loss": 0.4375, "step": 10672 }, { "epoch": 5.022588235294117, "grad_norm": 0.6495760366829811, "learning_rate": 2.6794270780518895e-05, "loss": 0.064, "step": 10673 }, { "epoch": 5.023058823529412, "grad_norm": 0.7848447878373624, "learning_rate": 2.6769272169275474e-05, "loss": 0.332, "step": 10674 }, { "epoch": 5.023529411764706, "grad_norm": 1.108584091970126, "learning_rate": 2.67442840825406e-05, "loss": 0.957, "step": 10675 }, { "epoch": 5.024, "grad_norm": 0.9025535968169832, "learning_rate": 2.671930652244845e-05, "loss": 0.5547, "step": 10676 }, { "epoch": 5.024470588235294, "grad_norm": 1.116143856645854, "learning_rate": 2.6694339491132237e-05, "loss": 0.9219, "step": 10677 }, { "epoch": 5.0249411764705885, "grad_norm": 0.5873953034130838, "learning_rate": 2.666938299072422e-05, "loss": 0.0986, "step": 10678 }, { "epoch": 5.025411764705883, "grad_norm": 0.8428499859428569, "learning_rate": 2.664443702335581e-05, "loss": 0.373, "step": 10679 }, { "epoch": 5.025882352941177, "grad_norm": 1.1180682230577674, "learning_rate": 2.6619501591157583e-05, "loss": 0.6133, "step": 10680 }, { "epoch": 5.02635294117647, "grad_norm": 1.164477609014356, "learning_rate": 2.6594576696259055e-05, "loss": 0.8125, "step": 10681 }, { "epoch": 5.0268235294117645, "grad_norm": 1.1041841810257693, "learning_rate": 2.6569662340789012e-05, "loss": 0.4766, "step": 10682 }, { "epoch": 5.027294117647059, "grad_norm": 1.3949198048337208, "learning_rate": 2.6544758526875193e-05, "loss": 1.0703, "step": 10683 }, { "epoch": 5.027764705882353, "grad_norm": 1.5445293979215264, "learning_rate": 2.651986525664455e-05, "loss": 1.0547, "step": 10684 }, { "epoch": 5.028235294117647, "grad_norm": 0.9412622880360565, "learning_rate": 2.649498253222307e-05, "loss": 0.5938, "step": 10685 }, { "epoch": 5.028705882352941, "grad_norm": 1.1991259520383064, "learning_rate": 2.647011035573588e-05, "loss": 0.9727, "step": 10686 }, { "epoch": 5.029176470588236, "grad_norm": 1.195103796367829, "learning_rate": 2.6445248729307165e-05, "loss": 0.7695, "step": 10687 }, { "epoch": 5.02964705882353, "grad_norm": 1.4299504151494096, "learning_rate": 2.6420397655060282e-05, "loss": 1.0781, "step": 10688 }, { "epoch": 5.030117647058823, "grad_norm": 1.3952189888432815, "learning_rate": 2.639555713511756e-05, "loss": 1.6875, "step": 10689 }, { "epoch": 5.030588235294117, "grad_norm": 1.1594521772788031, "learning_rate": 2.637072717160056e-05, "loss": 1.1641, "step": 10690 }, { "epoch": 5.031058823529412, "grad_norm": 1.0796574011212494, "learning_rate": 2.6345907766629843e-05, "loss": 0.7578, "step": 10691 }, { "epoch": 5.031529411764706, "grad_norm": 1.2825447883821135, "learning_rate": 2.6321098922325102e-05, "loss": 1.6797, "step": 10692 }, { "epoch": 5.032, "grad_norm": 0.7795950810695599, "learning_rate": 2.6296300640805206e-05, "loss": 0.5312, "step": 10693 }, { "epoch": 5.032470588235294, "grad_norm": 0.0984842306613324, "learning_rate": 2.627151292418795e-05, "loss": 0.0247, "step": 10694 }, { "epoch": 5.0329411764705885, "grad_norm": 1.6335818451409876, "learning_rate": 2.6246735774590383e-05, "loss": 1.1484, "step": 10695 }, { "epoch": 5.033411764705883, "grad_norm": 1.1569527407512206, "learning_rate": 2.6221969194128617e-05, "loss": 0.4863, "step": 10696 }, { "epoch": 5.033882352941176, "grad_norm": 1.1146542429379456, "learning_rate": 2.6197213184917775e-05, "loss": 0.6914, "step": 10697 }, { "epoch": 5.03435294117647, "grad_norm": 1.1497562123475604, "learning_rate": 2.6172467749072185e-05, "loss": 0.4824, "step": 10698 }, { "epoch": 5.0348235294117645, "grad_norm": 1.440684870185467, "learning_rate": 2.6147732888705235e-05, "loss": 1.5391, "step": 10699 }, { "epoch": 5.035294117647059, "grad_norm": 9.401431542760035, "learning_rate": 2.6123008605929375e-05, "loss": 1.1328, "step": 10700 }, { "epoch": 5.035764705882353, "grad_norm": 1.1075181513991665, "learning_rate": 2.6098294902856177e-05, "loss": 0.8086, "step": 10701 }, { "epoch": 5.036235294117647, "grad_norm": 1.357585820327283, "learning_rate": 2.607359178159637e-05, "loss": 1.0234, "step": 10702 }, { "epoch": 5.036705882352941, "grad_norm": 1.5051029677037269, "learning_rate": 2.604889924425965e-05, "loss": 1.2422, "step": 10703 }, { "epoch": 5.037176470588236, "grad_norm": 0.9875656725605203, "learning_rate": 2.6024217292954917e-05, "loss": 0.6719, "step": 10704 }, { "epoch": 5.03764705882353, "grad_norm": 1.1760049738729594, "learning_rate": 2.599954592979015e-05, "loss": 0.5664, "step": 10705 }, { "epoch": 5.038117647058823, "grad_norm": 1.2849688514726183, "learning_rate": 2.5974885156872345e-05, "loss": 1.4453, "step": 10706 }, { "epoch": 5.038588235294117, "grad_norm": 0.9530647562368499, "learning_rate": 2.5950234976307682e-05, "loss": 0.4961, "step": 10707 }, { "epoch": 5.039058823529412, "grad_norm": 0.8695982956867101, "learning_rate": 2.5925595390201466e-05, "loss": 0.4863, "step": 10708 }, { "epoch": 5.039529411764706, "grad_norm": 0.7958234638795214, "learning_rate": 2.5900966400657923e-05, "loss": 0.5312, "step": 10709 }, { "epoch": 5.04, "grad_norm": 0.8622445685060739, "learning_rate": 2.5876348009780564e-05, "loss": 0.2344, "step": 10710 }, { "epoch": 5.040470588235294, "grad_norm": 0.6935351760679032, "learning_rate": 2.5851740219671945e-05, "loss": 0.5664, "step": 10711 }, { "epoch": 5.0409411764705885, "grad_norm": 1.09619943711229, "learning_rate": 2.5827143032433627e-05, "loss": 0.7734, "step": 10712 }, { "epoch": 5.041411764705883, "grad_norm": 0.99835744867248, "learning_rate": 2.5802556450166396e-05, "loss": 0.5117, "step": 10713 }, { "epoch": 5.041882352941176, "grad_norm": 1.022642685981477, "learning_rate": 2.5777980474969983e-05, "loss": 0.9492, "step": 10714 }, { "epoch": 5.04235294117647, "grad_norm": 0.7144961231914058, "learning_rate": 2.5753415108943332e-05, "loss": 0.2305, "step": 10715 }, { "epoch": 5.0428235294117645, "grad_norm": 1.096802012348611, "learning_rate": 2.572886035418451e-05, "loss": 0.7266, "step": 10716 }, { "epoch": 5.043294117647059, "grad_norm": 2.0864642797409028, "learning_rate": 2.5704316212790527e-05, "loss": 1.0312, "step": 10717 }, { "epoch": 5.043764705882353, "grad_norm": 0.731141836765188, "learning_rate": 2.5679782686857586e-05, "loss": 0.1699, "step": 10718 }, { "epoch": 5.044235294117647, "grad_norm": 0.7808733083796535, "learning_rate": 2.5655259778481045e-05, "loss": 0.5703, "step": 10719 }, { "epoch": 5.044705882352941, "grad_norm": 1.124691033265802, "learning_rate": 2.5630747489755182e-05, "loss": 0.9531, "step": 10720 }, { "epoch": 5.045176470588236, "grad_norm": 0.8423481793424069, "learning_rate": 2.560624582277352e-05, "loss": 0.5, "step": 10721 }, { "epoch": 5.04564705882353, "grad_norm": 0.9884720580349803, "learning_rate": 2.558175477962865e-05, "loss": 0.5469, "step": 10722 }, { "epoch": 5.046117647058823, "grad_norm": 1.1923548762632836, "learning_rate": 2.5557274362412132e-05, "loss": 0.8477, "step": 10723 }, { "epoch": 5.046588235294117, "grad_norm": 0.6159764302416755, "learning_rate": 2.55328045732148e-05, "loss": 0.1719, "step": 10724 }, { "epoch": 5.047058823529412, "grad_norm": 0.7714428650852447, "learning_rate": 2.5508345414126474e-05, "loss": 0.168, "step": 10725 }, { "epoch": 5.047529411764706, "grad_norm": 0.5743219900954437, "learning_rate": 2.5483896887236056e-05, "loss": 0.1006, "step": 10726 }, { "epoch": 5.048, "grad_norm": 0.8976407821707255, "learning_rate": 2.5459458994631604e-05, "loss": 0.457, "step": 10727 }, { "epoch": 5.048470588235294, "grad_norm": 0.6086839106073446, "learning_rate": 2.5435031738400234e-05, "loss": 0.1357, "step": 10728 }, { "epoch": 5.0489411764705885, "grad_norm": 0.8105045418149568, "learning_rate": 2.541061512062808e-05, "loss": 0.5781, "step": 10729 }, { "epoch": 5.049411764705883, "grad_norm": 1.0138863955783781, "learning_rate": 2.538620914340057e-05, "loss": 0.3809, "step": 10730 }, { "epoch": 5.049882352941176, "grad_norm": 1.127123670143384, "learning_rate": 2.5361813808802007e-05, "loss": 0.293, "step": 10731 }, { "epoch": 5.05035294117647, "grad_norm": 0.5991385133575041, "learning_rate": 2.533742911891587e-05, "loss": 0.4531, "step": 10732 }, { "epoch": 5.0508235294117645, "grad_norm": 0.7968581622713726, "learning_rate": 2.531305507582481e-05, "loss": 0.1064, "step": 10733 }, { "epoch": 5.051294117647059, "grad_norm": 1.7570623560667809, "learning_rate": 2.5288691681610383e-05, "loss": 1.0703, "step": 10734 }, { "epoch": 5.051764705882353, "grad_norm": 0.4709723594252069, "learning_rate": 2.526433893835339e-05, "loss": 0.0698, "step": 10735 }, { "epoch": 5.052235294117647, "grad_norm": 0.8860732589007263, "learning_rate": 2.5239996848133715e-05, "loss": 0.4941, "step": 10736 }, { "epoch": 5.052705882352941, "grad_norm": 0.9447653448344426, "learning_rate": 2.5215665413030213e-05, "loss": 0.6367, "step": 10737 }, { "epoch": 5.053176470588236, "grad_norm": 0.6138319089098702, "learning_rate": 2.519134463512095e-05, "loss": 0.1699, "step": 10738 }, { "epoch": 5.05364705882353, "grad_norm": 1.2020486710236247, "learning_rate": 2.516703451648309e-05, "loss": 0.832, "step": 10739 }, { "epoch": 5.054117647058823, "grad_norm": 0.46213489498405164, "learning_rate": 2.5142735059192737e-05, "loss": 0.0469, "step": 10740 }, { "epoch": 5.054588235294117, "grad_norm": 1.1133821297624118, "learning_rate": 2.511844626532526e-05, "loss": 1.2422, "step": 10741 }, { "epoch": 5.055058823529412, "grad_norm": 1.2923166891446227, "learning_rate": 2.5094168136954985e-05, "loss": 1.0703, "step": 10742 }, { "epoch": 5.055529411764706, "grad_norm": 1.029475391657278, "learning_rate": 2.506990067615542e-05, "loss": 0.9844, "step": 10743 }, { "epoch": 5.056, "grad_norm": 0.7565829386333349, "learning_rate": 2.504564388499914e-05, "loss": 0.4648, "step": 10744 }, { "epoch": 5.056470588235294, "grad_norm": 1.0228144268826378, "learning_rate": 2.502139776555772e-05, "loss": 0.4219, "step": 10745 }, { "epoch": 5.0569411764705885, "grad_norm": 1.277132499628327, "learning_rate": 2.4997162319901977e-05, "loss": 0.9062, "step": 10746 }, { "epoch": 5.057411764705883, "grad_norm": 0.9868777672286547, "learning_rate": 2.497293755010173e-05, "loss": 0.6484, "step": 10747 }, { "epoch": 5.057882352941176, "grad_norm": 1.406800858928939, "learning_rate": 2.4948723458225827e-05, "loss": 0.6172, "step": 10748 }, { "epoch": 5.05835294117647, "grad_norm": 1.089868287141011, "learning_rate": 2.4924520046342317e-05, "loss": 0.8398, "step": 10749 }, { "epoch": 5.0588235294117645, "grad_norm": 1.0482338882441473, "learning_rate": 2.4900327316518326e-05, "loss": 0.8359, "step": 10750 }, { "epoch": 5.059294117647059, "grad_norm": 1.18276348244352, "learning_rate": 2.4876145270819954e-05, "loss": 1.1406, "step": 10751 }, { "epoch": 5.059764705882353, "grad_norm": 1.9005188600366851, "learning_rate": 2.485197391131251e-05, "loss": 1.3203, "step": 10752 }, { "epoch": 5.060235294117647, "grad_norm": 0.866982629197556, "learning_rate": 2.4827813240060367e-05, "loss": 0.2773, "step": 10753 }, { "epoch": 5.060705882352941, "grad_norm": 0.8666856021609599, "learning_rate": 2.480366325912692e-05, "loss": 0.3496, "step": 10754 }, { "epoch": 5.061176470588236, "grad_norm": 1.07100991831641, "learning_rate": 2.4779523970574703e-05, "loss": 0.7188, "step": 10755 }, { "epoch": 5.06164705882353, "grad_norm": 0.9728108075353493, "learning_rate": 2.4755395376465386e-05, "loss": 1.3281, "step": 10756 }, { "epoch": 5.062117647058823, "grad_norm": 0.619177417008957, "learning_rate": 2.4731277478859586e-05, "loss": 0.3984, "step": 10757 }, { "epoch": 5.062588235294117, "grad_norm": 1.0524726021952717, "learning_rate": 2.4707170279817118e-05, "loss": 1.2891, "step": 10758 }, { "epoch": 5.063058823529412, "grad_norm": 1.3607648745811405, "learning_rate": 2.4683073781396924e-05, "loss": 0.8281, "step": 10759 }, { "epoch": 5.063529411764706, "grad_norm": 1.4094274093225687, "learning_rate": 2.465898798565685e-05, "loss": 1.4375, "step": 10760 }, { "epoch": 5.064, "grad_norm": 0.9951085367001553, "learning_rate": 2.4634912894653996e-05, "loss": 0.832, "step": 10761 }, { "epoch": 5.064470588235294, "grad_norm": 1.1151474243736956, "learning_rate": 2.461084851044454e-05, "loss": 0.7539, "step": 10762 }, { "epoch": 5.0649411764705885, "grad_norm": 1.0067145657486236, "learning_rate": 2.4586794835083613e-05, "loss": 0.3203, "step": 10763 }, { "epoch": 5.065411764705883, "grad_norm": 0.7191914254215125, "learning_rate": 2.4562751870625597e-05, "loss": 0.1826, "step": 10764 }, { "epoch": 5.065882352941176, "grad_norm": 0.9191290144833306, "learning_rate": 2.4538719619123786e-05, "loss": 0.25, "step": 10765 }, { "epoch": 5.06635294117647, "grad_norm": 0.7963126643399376, "learning_rate": 2.4514698082630706e-05, "loss": 0.3105, "step": 10766 }, { "epoch": 5.0668235294117645, "grad_norm": 1.1754918799773584, "learning_rate": 2.449068726319795e-05, "loss": 0.5352, "step": 10767 }, { "epoch": 5.067294117647059, "grad_norm": 1.1830768599376238, "learning_rate": 2.4466687162876093e-05, "loss": 0.8359, "step": 10768 }, { "epoch": 5.067764705882353, "grad_norm": 1.7306334319807561, "learning_rate": 2.4442697783714894e-05, "loss": 0.7578, "step": 10769 }, { "epoch": 5.068235294117647, "grad_norm": 1.192665658911271, "learning_rate": 2.4418719127763183e-05, "loss": 0.6289, "step": 10770 }, { "epoch": 5.068705882352941, "grad_norm": 0.9829762467312492, "learning_rate": 2.43947511970688e-05, "loss": 0.5742, "step": 10771 }, { "epoch": 5.069176470588236, "grad_norm": 1.1280944062693716, "learning_rate": 2.437079399367875e-05, "loss": 0.9844, "step": 10772 }, { "epoch": 5.06964705882353, "grad_norm": 1.0738661600442014, "learning_rate": 2.434684751963914e-05, "loss": 0.4043, "step": 10773 }, { "epoch": 5.070117647058823, "grad_norm": 0.8298475991123505, "learning_rate": 2.4322911776995002e-05, "loss": 0.2168, "step": 10774 }, { "epoch": 5.070588235294117, "grad_norm": 0.588139477074786, "learning_rate": 2.429898676779072e-05, "loss": 0.0981, "step": 10775 }, { "epoch": 5.071058823529412, "grad_norm": 1.250588546556498, "learning_rate": 2.4275072494069487e-05, "loss": 0.8828, "step": 10776 }, { "epoch": 5.071529411764706, "grad_norm": 1.222131944020969, "learning_rate": 2.425116895787374e-05, "loss": 0.9961, "step": 10777 }, { "epoch": 5.072, "grad_norm": 1.0085533296666513, "learning_rate": 2.4227276161244995e-05, "loss": 0.4512, "step": 10778 }, { "epoch": 5.072470588235294, "grad_norm": 0.6457842272537946, "learning_rate": 2.4203394106223755e-05, "loss": 0.1191, "step": 10779 }, { "epoch": 5.0729411764705885, "grad_norm": 1.283564816365072, "learning_rate": 2.417952279484967e-05, "loss": 0.9102, "step": 10780 }, { "epoch": 5.073411764705883, "grad_norm": 0.7284176691070812, "learning_rate": 2.4155662229161527e-05, "loss": 0.7891, "step": 10781 }, { "epoch": 5.073882352941176, "grad_norm": 1.2409160380045816, "learning_rate": 2.4131812411197045e-05, "loss": 0.9961, "step": 10782 }, { "epoch": 5.07435294117647, "grad_norm": 1.1840391151726957, "learning_rate": 2.4107973342993174e-05, "loss": 0.9258, "step": 10783 }, { "epoch": 5.0748235294117645, "grad_norm": 0.9451537921235106, "learning_rate": 2.408414502658591e-05, "loss": 0.6719, "step": 10784 }, { "epoch": 5.075294117647059, "grad_norm": 1.127841990876044, "learning_rate": 2.4060327464010254e-05, "loss": 1.2656, "step": 10785 }, { "epoch": 5.075764705882353, "grad_norm": 0.7240549415150053, "learning_rate": 2.4036520657300346e-05, "loss": 0.1719, "step": 10786 }, { "epoch": 5.076235294117647, "grad_norm": 1.6313102938459767, "learning_rate": 2.401272460848946e-05, "loss": 1.3594, "step": 10787 }, { "epoch": 5.076705882352941, "grad_norm": 1.069855095997984, "learning_rate": 2.3988939319609817e-05, "loss": 0.6758, "step": 10788 }, { "epoch": 5.077176470588236, "grad_norm": 1.0721242749172597, "learning_rate": 2.3965164792692833e-05, "loss": 1.1797, "step": 10789 }, { "epoch": 5.07764705882353, "grad_norm": 1.1138936760819835, "learning_rate": 2.3941401029769024e-05, "loss": 0.5977, "step": 10790 }, { "epoch": 5.078117647058823, "grad_norm": 0.9538343588212059, "learning_rate": 2.3917648032867844e-05, "loss": 0.3223, "step": 10791 }, { "epoch": 5.078588235294117, "grad_norm": 0.9964579723131118, "learning_rate": 2.3893905804017977e-05, "loss": 0.2539, "step": 10792 }, { "epoch": 5.079058823529412, "grad_norm": 0.9270688011478676, "learning_rate": 2.387017434524707e-05, "loss": 0.3438, "step": 10793 }, { "epoch": 5.079529411764706, "grad_norm": 1.037964102838087, "learning_rate": 2.3846453658581927e-05, "loss": 0.8789, "step": 10794 }, { "epoch": 5.08, "grad_norm": 0.9541779496487187, "learning_rate": 2.382274374604847e-05, "loss": 0.3789, "step": 10795 }, { "epoch": 5.080470588235294, "grad_norm": 0.5380277579956572, "learning_rate": 2.3799044609671558e-05, "loss": 0.0742, "step": 10796 }, { "epoch": 5.0809411764705885, "grad_norm": 0.6434002772127883, "learning_rate": 2.3775356251475248e-05, "loss": 0.1504, "step": 10797 }, { "epoch": 5.081411764705883, "grad_norm": 0.6362400555630285, "learning_rate": 2.375167867348268e-05, "loss": 0.3828, "step": 10798 }, { "epoch": 5.081882352941176, "grad_norm": 1.2606819243527947, "learning_rate": 2.3728011877715947e-05, "loss": 0.793, "step": 10799 }, { "epoch": 5.08235294117647, "grad_norm": 0.8541483134112935, "learning_rate": 2.3704355866196373e-05, "loss": 0.8711, "step": 10800 }, { "epoch": 5.0828235294117645, "grad_norm": 0.9711353240047101, "learning_rate": 2.3680710640944328e-05, "loss": 0.4922, "step": 10801 }, { "epoch": 5.083294117647059, "grad_norm": 1.2155817952520978, "learning_rate": 2.3657076203979132e-05, "loss": 0.5273, "step": 10802 }, { "epoch": 5.083764705882353, "grad_norm": 1.2916054627860194, "learning_rate": 2.363345255731935e-05, "loss": 1.0938, "step": 10803 }, { "epoch": 5.084235294117647, "grad_norm": 1.2136145314141857, "learning_rate": 2.360983970298257e-05, "loss": 0.4238, "step": 10804 }, { "epoch": 5.084705882352941, "grad_norm": 1.0475378023782826, "learning_rate": 2.3586237642985396e-05, "loss": 0.5469, "step": 10805 }, { "epoch": 5.085176470588236, "grad_norm": 0.8579546606194209, "learning_rate": 2.356264637934358e-05, "loss": 0.2354, "step": 10806 }, { "epoch": 5.085647058823529, "grad_norm": 0.5746024643792302, "learning_rate": 2.3539065914071963e-05, "loss": 0.1099, "step": 10807 }, { "epoch": 5.086117647058823, "grad_norm": 1.196193732252822, "learning_rate": 2.351549624918437e-05, "loss": 1.0781, "step": 10808 }, { "epoch": 5.086588235294117, "grad_norm": 0.5157179566906617, "learning_rate": 2.34919373866938e-05, "loss": 0.1055, "step": 10809 }, { "epoch": 5.087058823529412, "grad_norm": 1.3021457773604839, "learning_rate": 2.3468389328612348e-05, "loss": 1.3125, "step": 10810 }, { "epoch": 5.087529411764706, "grad_norm": 0.8196664199685865, "learning_rate": 2.344485207695102e-05, "loss": 0.3027, "step": 10811 }, { "epoch": 5.088, "grad_norm": 0.8043790556225708, "learning_rate": 2.3421325633720094e-05, "loss": 0.2168, "step": 10812 }, { "epoch": 5.088470588235294, "grad_norm": 0.8696020788606041, "learning_rate": 2.3397810000928845e-05, "loss": 0.3301, "step": 10813 }, { "epoch": 5.0889411764705885, "grad_norm": 1.08296379956349, "learning_rate": 2.337430518058557e-05, "loss": 0.668, "step": 10814 }, { "epoch": 5.089411764705883, "grad_norm": 1.1040076093751983, "learning_rate": 2.335081117469777e-05, "loss": 0.8828, "step": 10815 }, { "epoch": 5.089882352941176, "grad_norm": 1.0640113411637888, "learning_rate": 2.3327327985271865e-05, "loss": 0.8984, "step": 10816 }, { "epoch": 5.09035294117647, "grad_norm": 0.7972793871566456, "learning_rate": 2.3303855614313483e-05, "loss": 0.2168, "step": 10817 }, { "epoch": 5.0908235294117645, "grad_norm": 0.8123709115104245, "learning_rate": 2.328039406382727e-05, "loss": 0.2852, "step": 10818 }, { "epoch": 5.091294117647059, "grad_norm": 0.7440435542373283, "learning_rate": 2.325694333581695e-05, "loss": 0.1631, "step": 10819 }, { "epoch": 5.091764705882353, "grad_norm": 1.3253555082283244, "learning_rate": 2.32335034322854e-05, "loss": 1.3359, "step": 10820 }, { "epoch": 5.092235294117647, "grad_norm": 1.1103573824711825, "learning_rate": 2.3210074355234392e-05, "loss": 0.9727, "step": 10821 }, { "epoch": 5.092705882352941, "grad_norm": 1.1479517515608346, "learning_rate": 2.318665610666493e-05, "loss": 0.9648, "step": 10822 }, { "epoch": 5.093176470588236, "grad_norm": 1.1467408890790747, "learning_rate": 2.3163248688577098e-05, "loss": 0.4082, "step": 10823 }, { "epoch": 5.09364705882353, "grad_norm": 0.94955539066638, "learning_rate": 2.3139852102969906e-05, "loss": 0.3066, "step": 10824 }, { "epoch": 5.094117647058823, "grad_norm": 0.9823098749210268, "learning_rate": 2.31164663518416e-05, "loss": 0.3008, "step": 10825 }, { "epoch": 5.094588235294117, "grad_norm": 1.0187842246802208, "learning_rate": 2.3093091437189476e-05, "loss": 0.4492, "step": 10826 }, { "epoch": 5.095058823529412, "grad_norm": 0.7732848718197107, "learning_rate": 2.306972736100977e-05, "loss": 0.2051, "step": 10827 }, { "epoch": 5.095529411764706, "grad_norm": 1.4362830641051203, "learning_rate": 2.304637412529794e-05, "loss": 1.6406, "step": 10828 }, { "epoch": 5.096, "grad_norm": 0.85575001097886, "learning_rate": 2.3023031732048504e-05, "loss": 0.5938, "step": 10829 }, { "epoch": 5.096470588235294, "grad_norm": 1.0063306927866977, "learning_rate": 2.2999700183254934e-05, "loss": 0.543, "step": 10830 }, { "epoch": 5.0969411764705885, "grad_norm": 0.97946769292146, "learning_rate": 2.29763794809099e-05, "loss": 0.4297, "step": 10831 }, { "epoch": 5.097411764705883, "grad_norm": 0.6334547101147953, "learning_rate": 2.2953069627005132e-05, "loss": 0.1699, "step": 10832 }, { "epoch": 5.097882352941176, "grad_norm": 0.8394089801398978, "learning_rate": 2.2929770623531368e-05, "loss": 0.5195, "step": 10833 }, { "epoch": 5.09835294117647, "grad_norm": 1.280800286723497, "learning_rate": 2.290648247247845e-05, "loss": 0.8867, "step": 10834 }, { "epoch": 5.0988235294117645, "grad_norm": 1.0925024942886092, "learning_rate": 2.2883205175835368e-05, "loss": 0.625, "step": 10835 }, { "epoch": 5.099294117647059, "grad_norm": 0.9535161631883764, "learning_rate": 2.2859938735590017e-05, "loss": 0.5195, "step": 10836 }, { "epoch": 5.099764705882353, "grad_norm": 0.8845362247950458, "learning_rate": 2.2836683153729522e-05, "loss": 0.7656, "step": 10837 }, { "epoch": 5.100235294117647, "grad_norm": 1.1124003204415576, "learning_rate": 2.2813438432240067e-05, "loss": 0.543, "step": 10838 }, { "epoch": 5.100705882352941, "grad_norm": 2.6789604038834853, "learning_rate": 2.2790204573106756e-05, "loss": 0.7891, "step": 10839 }, { "epoch": 5.101176470588236, "grad_norm": 1.2580213380874672, "learning_rate": 2.276698157831396e-05, "loss": 0.5859, "step": 10840 }, { "epoch": 5.101647058823529, "grad_norm": 1.258021390732933, "learning_rate": 2.2743769449845035e-05, "loss": 0.4961, "step": 10841 }, { "epoch": 5.102117647058823, "grad_norm": 0.9096144443918152, "learning_rate": 2.272056818968235e-05, "loss": 0.6484, "step": 10842 }, { "epoch": 5.102588235294117, "grad_norm": 0.8740426059931083, "learning_rate": 2.269737779980748e-05, "loss": 0.707, "step": 10843 }, { "epoch": 5.103058823529412, "grad_norm": 0.5518014218078322, "learning_rate": 2.2674198282200934e-05, "loss": 0.1426, "step": 10844 }, { "epoch": 5.103529411764706, "grad_norm": 1.0678842472624437, "learning_rate": 2.2651029638842372e-05, "loss": 0.7617, "step": 10845 }, { "epoch": 5.104, "grad_norm": 1.1290417680414369, "learning_rate": 2.262787187171058e-05, "loss": 0.4473, "step": 10846 }, { "epoch": 5.104470588235294, "grad_norm": 1.2702617290970446, "learning_rate": 2.2604724982783224e-05, "loss": 1.1328, "step": 10847 }, { "epoch": 5.1049411764705885, "grad_norm": 1.3244584219322046, "learning_rate": 2.258158897403724e-05, "loss": 1.0703, "step": 10848 }, { "epoch": 5.105411764705883, "grad_norm": 1.0954043459933303, "learning_rate": 2.2558463847448566e-05, "loss": 0.9414, "step": 10849 }, { "epoch": 5.105882352941176, "grad_norm": 0.7937467469916174, "learning_rate": 2.2535349604992153e-05, "loss": 0.1611, "step": 10850 }, { "epoch": 5.10635294117647, "grad_norm": 1.442592422543826, "learning_rate": 2.251224624864209e-05, "loss": 1.0625, "step": 10851 }, { "epoch": 5.1068235294117645, "grad_norm": 1.3841263817667993, "learning_rate": 2.2489153780371566e-05, "loss": 1.1016, "step": 10852 }, { "epoch": 5.107294117647059, "grad_norm": 1.4307609189386656, "learning_rate": 2.2466072202152702e-05, "loss": 1.3281, "step": 10853 }, { "epoch": 5.107764705882353, "grad_norm": 0.8332089629460621, "learning_rate": 2.2443001515956823e-05, "loss": 0.2275, "step": 10854 }, { "epoch": 5.108235294117647, "grad_norm": 1.7644384911703375, "learning_rate": 2.24199417237543e-05, "loss": 0.8555, "step": 10855 }, { "epoch": 5.108705882352941, "grad_norm": 1.6124434072966392, "learning_rate": 2.2396892827514507e-05, "loss": 1.3984, "step": 10856 }, { "epoch": 5.109176470588236, "grad_norm": 0.7736938423110546, "learning_rate": 2.2373854829205957e-05, "loss": 0.1904, "step": 10857 }, { "epoch": 5.109647058823529, "grad_norm": 1.1233751359776645, "learning_rate": 2.235082773079624e-05, "loss": 0.9062, "step": 10858 }, { "epoch": 5.110117647058823, "grad_norm": 1.1937383446498047, "learning_rate": 2.2327811534251933e-05, "loss": 1.0625, "step": 10859 }, { "epoch": 5.110588235294117, "grad_norm": 0.9540744042631839, "learning_rate": 2.230480624153873e-05, "loss": 0.3359, "step": 10860 }, { "epoch": 5.111058823529412, "grad_norm": 0.910075986638712, "learning_rate": 2.2281811854621445e-05, "loss": 0.5078, "step": 10861 }, { "epoch": 5.111529411764706, "grad_norm": 1.309860002082637, "learning_rate": 2.2258828375463855e-05, "loss": 1.8125, "step": 10862 }, { "epoch": 5.112, "grad_norm": 0.611245081348772, "learning_rate": 2.223585580602889e-05, "loss": 0.4512, "step": 10863 }, { "epoch": 5.112470588235294, "grad_norm": 0.8909237276877409, "learning_rate": 2.221289414827853e-05, "loss": 0.332, "step": 10864 }, { "epoch": 5.1129411764705885, "grad_norm": 1.044266919455117, "learning_rate": 2.218994340417378e-05, "loss": 0.6289, "step": 10865 }, { "epoch": 5.113411764705883, "grad_norm": 1.3639857225779048, "learning_rate": 2.2167003575674824e-05, "loss": 0.8281, "step": 10866 }, { "epoch": 5.113882352941176, "grad_norm": 1.1066883508436471, "learning_rate": 2.2144074664740723e-05, "loss": 0.3809, "step": 10867 }, { "epoch": 5.11435294117647, "grad_norm": 1.4254536706674215, "learning_rate": 2.2121156673329794e-05, "loss": 0.5977, "step": 10868 }, { "epoch": 5.1148235294117645, "grad_norm": 1.0765998022788208, "learning_rate": 2.2098249603399348e-05, "loss": 0.9609, "step": 10869 }, { "epoch": 5.115294117647059, "grad_norm": 1.1815077346318466, "learning_rate": 2.2075353456905727e-05, "loss": 0.7891, "step": 10870 }, { "epoch": 5.115764705882353, "grad_norm": 1.0073266138811368, "learning_rate": 2.2052468235804393e-05, "loss": 0.1826, "step": 10871 }, { "epoch": 5.116235294117647, "grad_norm": 1.5738185048630486, "learning_rate": 2.2029593942049838e-05, "loss": 0.957, "step": 10872 }, { "epoch": 5.116705882352941, "grad_norm": 1.3645409106765203, "learning_rate": 2.200673057759565e-05, "loss": 1.1016, "step": 10873 }, { "epoch": 5.117176470588236, "grad_norm": 1.0422839801117185, "learning_rate": 2.1983878144394497e-05, "loss": 0.9766, "step": 10874 }, { "epoch": 5.117647058823529, "grad_norm": 0.2326890778460096, "learning_rate": 2.1961036644398035e-05, "loss": 0.0114, "step": 10875 }, { "epoch": 5.118117647058823, "grad_norm": 1.0418097922108074, "learning_rate": 2.193820607955707e-05, "loss": 0.4082, "step": 10876 }, { "epoch": 5.118588235294117, "grad_norm": 1.556832867059655, "learning_rate": 2.1915386451821494e-05, "loss": 0.5273, "step": 10877 }, { "epoch": 5.119058823529412, "grad_norm": 0.7131421805532243, "learning_rate": 2.1892577763140112e-05, "loss": 0.1748, "step": 10878 }, { "epoch": 5.119529411764706, "grad_norm": 0.6810312837922293, "learning_rate": 2.186978001546097e-05, "loss": 0.4258, "step": 10879 }, { "epoch": 5.12, "grad_norm": 1.0483224743079378, "learning_rate": 2.1846993210731108e-05, "loss": 0.457, "step": 10880 }, { "epoch": 5.120470588235294, "grad_norm": 1.2361199255961124, "learning_rate": 2.1824217350896585e-05, "loss": 0.7969, "step": 10881 }, { "epoch": 5.1209411764705886, "grad_norm": 0.9161860505396231, "learning_rate": 2.180145243790261e-05, "loss": 0.2949, "step": 10882 }, { "epoch": 5.121411764705883, "grad_norm": 1.3358221092007503, "learning_rate": 2.1778698473693435e-05, "loss": 0.8633, "step": 10883 }, { "epoch": 5.121882352941176, "grad_norm": 1.317236633942489, "learning_rate": 2.1755955460212282e-05, "loss": 0.6875, "step": 10884 }, { "epoch": 5.12235294117647, "grad_norm": 0.8187837439106268, "learning_rate": 2.1733223399401585e-05, "loss": 0.3906, "step": 10885 }, { "epoch": 5.1228235294117646, "grad_norm": 1.1694367095182734, "learning_rate": 2.1710502293202786e-05, "loss": 1.0156, "step": 10886 }, { "epoch": 5.123294117647059, "grad_norm": 0.7374836814356257, "learning_rate": 2.1687792143556303e-05, "loss": 0.1787, "step": 10887 }, { "epoch": 5.123764705882353, "grad_norm": 1.288109545781589, "learning_rate": 2.1665092952401754e-05, "loss": 1.3359, "step": 10888 }, { "epoch": 5.124235294117647, "grad_norm": 1.475751378808634, "learning_rate": 2.1642404721677765e-05, "loss": 0.8125, "step": 10889 }, { "epoch": 5.124705882352941, "grad_norm": 1.113983310793565, "learning_rate": 2.1619727453321974e-05, "loss": 0.4766, "step": 10890 }, { "epoch": 5.125176470588236, "grad_norm": 0.4893495220138916, "learning_rate": 2.159706114927116e-05, "loss": 0.0649, "step": 10891 }, { "epoch": 5.125647058823529, "grad_norm": 0.9355398342006725, "learning_rate": 2.157440581146118e-05, "loss": 0.8242, "step": 10892 }, { "epoch": 5.126117647058823, "grad_norm": 0.9813956876116976, "learning_rate": 2.155176144182682e-05, "loss": 0.7695, "step": 10893 }, { "epoch": 5.1265882352941174, "grad_norm": 1.414810038346964, "learning_rate": 2.1529128042302113e-05, "loss": 0.6992, "step": 10894 }, { "epoch": 5.127058823529412, "grad_norm": 1.4913105482843523, "learning_rate": 2.150650561481998e-05, "loss": 1.0078, "step": 10895 }, { "epoch": 5.127529411764706, "grad_norm": 1.3597208819684614, "learning_rate": 2.1483894161312525e-05, "loss": 1.1953, "step": 10896 }, { "epoch": 5.128, "grad_norm": 0.877116925260309, "learning_rate": 2.1461293683710905e-05, "loss": 0.6055, "step": 10897 }, { "epoch": 5.128470588235294, "grad_norm": 0.6305387732288694, "learning_rate": 2.143870418394526e-05, "loss": 0.1484, "step": 10898 }, { "epoch": 5.128941176470589, "grad_norm": 1.2856415817462616, "learning_rate": 2.141612566394486e-05, "loss": 0.5781, "step": 10899 }, { "epoch": 5.129411764705883, "grad_norm": 1.0049964464320231, "learning_rate": 2.1393558125638066e-05, "loss": 0.9492, "step": 10900 }, { "epoch": 5.129882352941176, "grad_norm": 1.478321886712285, "learning_rate": 2.1371001570952186e-05, "loss": 0.9414, "step": 10901 }, { "epoch": 5.13035294117647, "grad_norm": 1.3132524499441445, "learning_rate": 2.1348456001813674e-05, "loss": 0.7305, "step": 10902 }, { "epoch": 5.130823529411765, "grad_norm": 1.229009721153195, "learning_rate": 2.132592142014811e-05, "loss": 1.0234, "step": 10903 }, { "epoch": 5.131294117647059, "grad_norm": 1.261862717229986, "learning_rate": 2.130339782787996e-05, "loss": 0.8359, "step": 10904 }, { "epoch": 5.131764705882353, "grad_norm": 1.2749775413804503, "learning_rate": 2.1280885226932877e-05, "loss": 0.75, "step": 10905 }, { "epoch": 5.132235294117647, "grad_norm": 0.6792812351668153, "learning_rate": 2.125838361922961e-05, "loss": 0.1777, "step": 10906 }, { "epoch": 5.1327058823529415, "grad_norm": 0.9178285181523912, "learning_rate": 2.1235893006691786e-05, "loss": 0.3047, "step": 10907 }, { "epoch": 5.133176470588236, "grad_norm": 0.824235003268363, "learning_rate": 2.1213413391240335e-05, "loss": 0.1641, "step": 10908 }, { "epoch": 5.133647058823529, "grad_norm": 1.2273468495807665, "learning_rate": 2.119094477479506e-05, "loss": 1.1719, "step": 10909 }, { "epoch": 5.134117647058823, "grad_norm": 1.2587387348582013, "learning_rate": 2.116848715927489e-05, "loss": 1.7344, "step": 10910 }, { "epoch": 5.1345882352941175, "grad_norm": 1.1215801260772083, "learning_rate": 2.114604054659787e-05, "loss": 0.6523, "step": 10911 }, { "epoch": 5.135058823529412, "grad_norm": 1.2424738825094668, "learning_rate": 2.1123604938680987e-05, "loss": 0.8516, "step": 10912 }, { "epoch": 5.135529411764706, "grad_norm": 1.1134143955564346, "learning_rate": 2.1101180337440378e-05, "loss": 0.707, "step": 10913 }, { "epoch": 5.136, "grad_norm": 1.0343732508900105, "learning_rate": 2.1078766744791238e-05, "loss": 0.4141, "step": 10914 }, { "epoch": 5.136470588235294, "grad_norm": 1.0650862391919307, "learning_rate": 2.1056364162647736e-05, "loss": 1.2812, "step": 10915 }, { "epoch": 5.136941176470589, "grad_norm": 1.2161592137238906, "learning_rate": 2.1033972592923214e-05, "loss": 1.1641, "step": 10916 }, { "epoch": 5.137411764705883, "grad_norm": 1.2099807375990408, "learning_rate": 2.1011592037530023e-05, "loss": 0.7422, "step": 10917 }, { "epoch": 5.137882352941176, "grad_norm": 0.9923156626760891, "learning_rate": 2.098922249837953e-05, "loss": 0.2871, "step": 10918 }, { "epoch": 5.13835294117647, "grad_norm": 0.9511440161558181, "learning_rate": 2.096686397738223e-05, "loss": 0.6328, "step": 10919 }, { "epoch": 5.138823529411765, "grad_norm": 1.304425563612744, "learning_rate": 2.094451647644769e-05, "loss": 0.6484, "step": 10920 }, { "epoch": 5.139294117647059, "grad_norm": 1.0457318863185259, "learning_rate": 2.0922179997484424e-05, "loss": 0.4551, "step": 10921 }, { "epoch": 5.139764705882353, "grad_norm": 0.6943662396725554, "learning_rate": 2.0899854542400147e-05, "loss": 0.2617, "step": 10922 }, { "epoch": 5.140235294117647, "grad_norm": 1.0264851680785598, "learning_rate": 2.087754011310147e-05, "loss": 0.3066, "step": 10923 }, { "epoch": 5.1407058823529415, "grad_norm": 1.225018932473765, "learning_rate": 2.0855236711494223e-05, "loss": 0.7109, "step": 10924 }, { "epoch": 5.141176470588236, "grad_norm": 1.122908646972321, "learning_rate": 2.083294433948324e-05, "loss": 0.875, "step": 10925 }, { "epoch": 5.141647058823529, "grad_norm": 1.1326761799464018, "learning_rate": 2.0810662998972345e-05, "loss": 1.3047, "step": 10926 }, { "epoch": 5.142117647058823, "grad_norm": 0.9734745098945606, "learning_rate": 2.0788392691864504e-05, "loss": 0.3555, "step": 10927 }, { "epoch": 5.1425882352941175, "grad_norm": 1.154424432001397, "learning_rate": 2.0766133420061732e-05, "loss": 0.7695, "step": 10928 }, { "epoch": 5.143058823529412, "grad_norm": 0.6658890337769972, "learning_rate": 2.0743885185465025e-05, "loss": 0.1279, "step": 10929 }, { "epoch": 5.143529411764706, "grad_norm": 1.3667041304507583, "learning_rate": 2.0721647989974514e-05, "loss": 0.9258, "step": 10930 }, { "epoch": 5.144, "grad_norm": 1.1459934791977218, "learning_rate": 2.0699421835489417e-05, "loss": 0.7852, "step": 10931 }, { "epoch": 5.144470588235294, "grad_norm": 1.5474247341212815, "learning_rate": 2.0677206723907857e-05, "loss": 1.1484, "step": 10932 }, { "epoch": 5.144941176470589, "grad_norm": 1.3172305711080041, "learning_rate": 2.0655002657127188e-05, "loss": 0.6602, "step": 10933 }, { "epoch": 5.145411764705882, "grad_norm": 1.0064408517852104, "learning_rate": 2.0632809637043747e-05, "loss": 0.4102, "step": 10934 }, { "epoch": 5.145882352941176, "grad_norm": 1.1354944843252575, "learning_rate": 2.0610627665552875e-05, "loss": 0.5469, "step": 10935 }, { "epoch": 5.14635294117647, "grad_norm": 1.1953577544625582, "learning_rate": 2.058845674454905e-05, "loss": 0.6172, "step": 10936 }, { "epoch": 5.146823529411765, "grad_norm": 1.0948321808009596, "learning_rate": 2.0566296875925813e-05, "loss": 0.9688, "step": 10937 }, { "epoch": 5.147294117647059, "grad_norm": 0.6328746549634011, "learning_rate": 2.054414806157566e-05, "loss": 0.5586, "step": 10938 }, { "epoch": 5.147764705882353, "grad_norm": 0.8388446353125855, "learning_rate": 2.052201030339024e-05, "loss": 0.3711, "step": 10939 }, { "epoch": 5.148235294117647, "grad_norm": 0.9477086748246825, "learning_rate": 2.0499883603260274e-05, "loss": 0.3105, "step": 10940 }, { "epoch": 5.1487058823529415, "grad_norm": 0.4538554069972298, "learning_rate": 2.0477767963075386e-05, "loss": 0.1523, "step": 10941 }, { "epoch": 5.149176470588236, "grad_norm": 1.4149807906553153, "learning_rate": 2.045566338472443e-05, "loss": 0.8672, "step": 10942 }, { "epoch": 5.149647058823529, "grad_norm": 0.8944484754513439, "learning_rate": 2.043356987009526e-05, "loss": 0.5195, "step": 10943 }, { "epoch": 5.150117647058823, "grad_norm": 1.0758008900947966, "learning_rate": 2.041148742107471e-05, "loss": 0.4961, "step": 10944 }, { "epoch": 5.1505882352941175, "grad_norm": 0.8597772467247207, "learning_rate": 2.038941603954879e-05, "loss": 0.5859, "step": 10945 }, { "epoch": 5.151058823529412, "grad_norm": 0.8467519062187918, "learning_rate": 2.036735572740245e-05, "loss": 0.2432, "step": 10946 }, { "epoch": 5.151529411764706, "grad_norm": 0.7327464524917674, "learning_rate": 2.0345306486519787e-05, "loss": 0.3398, "step": 10947 }, { "epoch": 5.152, "grad_norm": 1.5801715418519255, "learning_rate": 2.032326831878392e-05, "loss": 1.75, "step": 10948 }, { "epoch": 5.152470588235294, "grad_norm": 1.3823815775306125, "learning_rate": 2.0301241226076963e-05, "loss": 1.1094, "step": 10949 }, { "epoch": 5.152941176470589, "grad_norm": 0.917484614806025, "learning_rate": 2.027922521028018e-05, "loss": 0.2598, "step": 10950 }, { "epoch": 5.153411764705883, "grad_norm": 1.0025360162410495, "learning_rate": 2.025722027327385e-05, "loss": 0.8281, "step": 10951 }, { "epoch": 5.153882352941176, "grad_norm": 1.0028477855426643, "learning_rate": 2.0235226416937284e-05, "loss": 0.4023, "step": 10952 }, { "epoch": 5.15435294117647, "grad_norm": 1.488960580448064, "learning_rate": 2.0213243643148914e-05, "loss": 0.8867, "step": 10953 }, { "epoch": 5.154823529411765, "grad_norm": 1.0664684530639283, "learning_rate": 2.0191271953786104e-05, "loss": 0.5625, "step": 10954 }, { "epoch": 5.155294117647059, "grad_norm": 1.2078124383533178, "learning_rate": 2.016931135072537e-05, "loss": 0.8906, "step": 10955 }, { "epoch": 5.155764705882353, "grad_norm": 1.322895074114253, "learning_rate": 2.0147361835842296e-05, "loss": 1.0391, "step": 10956 }, { "epoch": 5.156235294117647, "grad_norm": 0.7425730015775861, "learning_rate": 2.0125423411011416e-05, "loss": 0.5586, "step": 10957 }, { "epoch": 5.1567058823529415, "grad_norm": 0.7328476167233259, "learning_rate": 2.0103496078106396e-05, "loss": 0.4688, "step": 10958 }, { "epoch": 5.157176470588236, "grad_norm": 1.1720276359923052, "learning_rate": 2.0081579839000005e-05, "loss": 0.6641, "step": 10959 }, { "epoch": 5.157647058823529, "grad_norm": 0.5306246297710958, "learning_rate": 2.00596746955639e-05, "loss": 0.0874, "step": 10960 }, { "epoch": 5.158117647058823, "grad_norm": 1.3461208642394862, "learning_rate": 2.0037780649668912e-05, "loss": 0.7148, "step": 10961 }, { "epoch": 5.1585882352941175, "grad_norm": 1.0601885211248268, "learning_rate": 2.001589770318498e-05, "loss": 0.4297, "step": 10962 }, { "epoch": 5.159058823529412, "grad_norm": 0.9803805796861731, "learning_rate": 1.9994025857980896e-05, "loss": 1.2734, "step": 10963 }, { "epoch": 5.159529411764706, "grad_norm": 0.7875903229487635, "learning_rate": 1.9972165115924695e-05, "loss": 0.8047, "step": 10964 }, { "epoch": 5.16, "grad_norm": 1.2098576537537, "learning_rate": 1.99503154788834e-05, "loss": 0.8438, "step": 10965 }, { "epoch": 5.160470588235294, "grad_norm": 1.3355096072643757, "learning_rate": 1.9928476948723022e-05, "loss": 0.7031, "step": 10966 }, { "epoch": 5.160941176470589, "grad_norm": 1.2137845523140114, "learning_rate": 1.99066495273087e-05, "loss": 1.0312, "step": 10967 }, { "epoch": 5.161411764705882, "grad_norm": 0.9557012837154532, "learning_rate": 1.9884833216504657e-05, "loss": 0.9922, "step": 10968 }, { "epoch": 5.161882352941176, "grad_norm": 0.5211044083041486, "learning_rate": 1.9863028018174043e-05, "loss": 0.1289, "step": 10969 }, { "epoch": 5.16235294117647, "grad_norm": 0.8764177645419071, "learning_rate": 1.9841233934179145e-05, "loss": 0.2695, "step": 10970 }, { "epoch": 5.162823529411765, "grad_norm": 1.0541791626325492, "learning_rate": 1.9819450966381328e-05, "loss": 1.1875, "step": 10971 }, { "epoch": 5.163294117647059, "grad_norm": 0.9786059845459968, "learning_rate": 1.9797679116640903e-05, "loss": 0.3516, "step": 10972 }, { "epoch": 5.163764705882353, "grad_norm": 0.7277740121814004, "learning_rate": 1.9775918386817346e-05, "loss": 0.1533, "step": 10973 }, { "epoch": 5.164235294117647, "grad_norm": 0.5653433929493971, "learning_rate": 1.9754168778769058e-05, "loss": 0.1504, "step": 10974 }, { "epoch": 5.1647058823529415, "grad_norm": 0.8043480063490224, "learning_rate": 1.9732430294353636e-05, "loss": 0.2754, "step": 10975 }, { "epoch": 5.165176470588236, "grad_norm": 1.039651967906187, "learning_rate": 1.9710702935427642e-05, "loss": 0.3652, "step": 10976 }, { "epoch": 5.165647058823529, "grad_norm": 0.8958490660492542, "learning_rate": 1.9688986703846658e-05, "loss": 0.6094, "step": 10977 }, { "epoch": 5.166117647058823, "grad_norm": 0.987181081669325, "learning_rate": 1.9667281601465372e-05, "loss": 0.2871, "step": 10978 }, { "epoch": 5.1665882352941175, "grad_norm": 0.6235214532420948, "learning_rate": 1.964558763013755e-05, "loss": 0.165, "step": 10979 }, { "epoch": 5.167058823529412, "grad_norm": 0.5763534645038166, "learning_rate": 1.9623904791715907e-05, "loss": 0.1338, "step": 10980 }, { "epoch": 5.167529411764706, "grad_norm": 1.0700473075556278, "learning_rate": 1.9602233088052277e-05, "loss": 0.332, "step": 10981 }, { "epoch": 5.168, "grad_norm": 0.8788152100182485, "learning_rate": 1.9580572520997566e-05, "loss": 0.3145, "step": 10982 }, { "epoch": 5.168470588235294, "grad_norm": 1.3063173358883915, "learning_rate": 1.9558923092401645e-05, "loss": 1.2656, "step": 10983 }, { "epoch": 5.168941176470589, "grad_norm": 1.0563515276515172, "learning_rate": 1.95372848041135e-05, "loss": 0.5625, "step": 10984 }, { "epoch": 5.169411764705882, "grad_norm": 1.2503210491522125, "learning_rate": 1.951565765798117e-05, "loss": 0.5117, "step": 10985 }, { "epoch": 5.169882352941176, "grad_norm": 0.9961351582960185, "learning_rate": 1.9494041655851677e-05, "loss": 0.5898, "step": 10986 }, { "epoch": 5.17035294117647, "grad_norm": 1.0161800946670014, "learning_rate": 1.9472436799571145e-05, "loss": 0.6172, "step": 10987 }, { "epoch": 5.170823529411765, "grad_norm": 1.1587133313518414, "learning_rate": 1.9450843090984797e-05, "loss": 0.8086, "step": 10988 }, { "epoch": 5.171294117647059, "grad_norm": 1.4027847360206895, "learning_rate": 1.9429260531936743e-05, "loss": 0.9609, "step": 10989 }, { "epoch": 5.171764705882353, "grad_norm": 1.2906458257531068, "learning_rate": 1.940768912427027e-05, "loss": 1.4219, "step": 10990 }, { "epoch": 5.172235294117647, "grad_norm": 1.4190477897161966, "learning_rate": 1.9386128869827743e-05, "loss": 0.6172, "step": 10991 }, { "epoch": 5.1727058823529415, "grad_norm": 1.0028163460241726, "learning_rate": 1.936457977045044e-05, "loss": 0.4102, "step": 10992 }, { "epoch": 5.173176470588236, "grad_norm": 0.9714034203005473, "learning_rate": 1.9343041827978785e-05, "loss": 0.2617, "step": 10993 }, { "epoch": 5.173647058823529, "grad_norm": 0.9829290302281622, "learning_rate": 1.9321515044252273e-05, "loss": 0.7852, "step": 10994 }, { "epoch": 5.174117647058823, "grad_norm": 1.422865992047852, "learning_rate": 1.9299999421109303e-05, "loss": 0.9961, "step": 10995 }, { "epoch": 5.1745882352941175, "grad_norm": 0.8598951689906448, "learning_rate": 1.9278494960387452e-05, "loss": 0.2812, "step": 10996 }, { "epoch": 5.175058823529412, "grad_norm": 1.0146516706637136, "learning_rate": 1.9257001663923338e-05, "loss": 0.4062, "step": 10997 }, { "epoch": 5.175529411764706, "grad_norm": 1.417721243163452, "learning_rate": 1.9235519533552556e-05, "loss": 1.0234, "step": 10998 }, { "epoch": 5.176, "grad_norm": 1.274436228515537, "learning_rate": 1.921404857110984e-05, "loss": 0.5625, "step": 10999 }, { "epoch": 5.176470588235294, "grad_norm": 0.679770053376418, "learning_rate": 1.9192588778428842e-05, "loss": 0.5508, "step": 11000 }, { "epoch": 5.176941176470589, "grad_norm": 0.6226892765646168, "learning_rate": 1.9171140157342407e-05, "loss": 0.1562, "step": 11001 }, { "epoch": 5.177411764705882, "grad_norm": 0.9375148807396186, "learning_rate": 1.9149702709682292e-05, "loss": 0.5039, "step": 11002 }, { "epoch": 5.177882352941176, "grad_norm": 1.2568082607934066, "learning_rate": 1.9128276437279362e-05, "loss": 1.0078, "step": 11003 }, { "epoch": 5.17835294117647, "grad_norm": 0.935980668567683, "learning_rate": 1.9106861341963587e-05, "loss": 0.5078, "step": 11004 }, { "epoch": 5.178823529411765, "grad_norm": 1.0130848343508354, "learning_rate": 1.9085457425563855e-05, "loss": 0.5312, "step": 11005 }, { "epoch": 5.179294117647059, "grad_norm": 1.1893232315131648, "learning_rate": 1.9064064689908187e-05, "loss": 0.8789, "step": 11006 }, { "epoch": 5.179764705882353, "grad_norm": 0.7210196368230855, "learning_rate": 1.9042683136823654e-05, "loss": 0.207, "step": 11007 }, { "epoch": 5.180235294117647, "grad_norm": 1.1962390572531185, "learning_rate": 1.9021312768136314e-05, "loss": 1.1406, "step": 11008 }, { "epoch": 5.1807058823529415, "grad_norm": 1.4323116318178615, "learning_rate": 1.899995358567129e-05, "loss": 0.8281, "step": 11009 }, { "epoch": 5.181176470588236, "grad_norm": 0.7746514837441028, "learning_rate": 1.897860559125282e-05, "loss": 0.2344, "step": 11010 }, { "epoch": 5.181647058823529, "grad_norm": 1.1544359584581119, "learning_rate": 1.895726878670405e-05, "loss": 0.8242, "step": 11011 }, { "epoch": 5.182117647058823, "grad_norm": 0.9809155190422711, "learning_rate": 1.8935943173847302e-05, "loss": 0.4258, "step": 11012 }, { "epoch": 5.1825882352941175, "grad_norm": 0.5257157186429202, "learning_rate": 1.8914628754503885e-05, "loss": 0.1299, "step": 11013 }, { "epoch": 5.183058823529412, "grad_norm": 0.8350257363236205, "learning_rate": 1.889332553049413e-05, "loss": 0.668, "step": 11014 }, { "epoch": 5.183529411764706, "grad_norm": 1.10376043037415, "learning_rate": 1.887203350363743e-05, "loss": 0.4258, "step": 11015 }, { "epoch": 5.184, "grad_norm": 0.9037174919595335, "learning_rate": 1.8850752675752295e-05, "loss": 0.2754, "step": 11016 }, { "epoch": 5.184470588235294, "grad_norm": 0.9826788409046088, "learning_rate": 1.8829483048656124e-05, "loss": 0.8125, "step": 11017 }, { "epoch": 5.184941176470589, "grad_norm": 0.7704010860005533, "learning_rate": 1.8808224624165497e-05, "loss": 0.1768, "step": 11018 }, { "epoch": 5.185411764705882, "grad_norm": 0.7944944976356207, "learning_rate": 1.878697740409601e-05, "loss": 0.2949, "step": 11019 }, { "epoch": 5.185882352941176, "grad_norm": 0.9784298938033106, "learning_rate": 1.8765741390262212e-05, "loss": 0.668, "step": 11020 }, { "epoch": 5.18635294117647, "grad_norm": 0.6802207069973027, "learning_rate": 1.8744516584477803e-05, "loss": 0.1787, "step": 11021 }, { "epoch": 5.186823529411765, "grad_norm": 1.1715505202949061, "learning_rate": 1.87233029885555e-05, "loss": 0.5898, "step": 11022 }, { "epoch": 5.187294117647059, "grad_norm": 0.608814517740927, "learning_rate": 1.8702100604307018e-05, "loss": 0.377, "step": 11023 }, { "epoch": 5.187764705882353, "grad_norm": 1.0522944194627035, "learning_rate": 1.868090943354318e-05, "loss": 0.3398, "step": 11024 }, { "epoch": 5.188235294117647, "grad_norm": 1.2188520306891233, "learning_rate": 1.8659729478073764e-05, "loss": 1.1016, "step": 11025 }, { "epoch": 5.1887058823529415, "grad_norm": 1.0397091392641125, "learning_rate": 1.8638560739707664e-05, "loss": 1.0391, "step": 11026 }, { "epoch": 5.189176470588236, "grad_norm": 0.7359470503503159, "learning_rate": 1.8617403220252846e-05, "loss": 0.1904, "step": 11027 }, { "epoch": 5.189647058823529, "grad_norm": 0.6743386889710178, "learning_rate": 1.8596256921516166e-05, "loss": 0.1523, "step": 11028 }, { "epoch": 5.190117647058823, "grad_norm": 0.9620574462405764, "learning_rate": 1.8575121845303698e-05, "loss": 1.1172, "step": 11029 }, { "epoch": 5.1905882352941175, "grad_norm": 0.8273562841936988, "learning_rate": 1.8553997993420495e-05, "loss": 0.1973, "step": 11030 }, { "epoch": 5.191058823529412, "grad_norm": 1.1571778436072626, "learning_rate": 1.8532885367670547e-05, "loss": 0.9258, "step": 11031 }, { "epoch": 5.191529411764706, "grad_norm": 1.1097713370834847, "learning_rate": 1.8511783969857046e-05, "loss": 0.4531, "step": 11032 }, { "epoch": 5.192, "grad_norm": 1.2778635332074737, "learning_rate": 1.849069380178217e-05, "loss": 1.125, "step": 11033 }, { "epoch": 5.192470588235294, "grad_norm": 0.8422892702218713, "learning_rate": 1.8469614865247075e-05, "loss": 0.2197, "step": 11034 }, { "epoch": 5.192941176470589, "grad_norm": 1.0330764239514911, "learning_rate": 1.8448547162052002e-05, "loss": 0.3516, "step": 11035 }, { "epoch": 5.193411764705882, "grad_norm": 1.0373628569632047, "learning_rate": 1.8427490693996298e-05, "loss": 0.7891, "step": 11036 }, { "epoch": 5.193882352941176, "grad_norm": 0.9520854712347354, "learning_rate": 1.840644546287821e-05, "loss": 0.5352, "step": 11037 }, { "epoch": 5.19435294117647, "grad_norm": 1.046853796935797, "learning_rate": 1.838541147049515e-05, "loss": 0.6797, "step": 11038 }, { "epoch": 5.194823529411765, "grad_norm": 0.9750642737101631, "learning_rate": 1.8364388718643542e-05, "loss": 0.4707, "step": 11039 }, { "epoch": 5.195294117647059, "grad_norm": 1.1235482595055328, "learning_rate": 1.8343377209118755e-05, "loss": 0.5508, "step": 11040 }, { "epoch": 5.195764705882353, "grad_norm": 0.48617244632947426, "learning_rate": 1.8322376943715382e-05, "loss": 0.1768, "step": 11041 }, { "epoch": 5.196235294117647, "grad_norm": 2.0927395362282977, "learning_rate": 1.8301387924226863e-05, "loss": 0.5508, "step": 11042 }, { "epoch": 5.1967058823529415, "grad_norm": 0.9278901083074753, "learning_rate": 1.828041015244579e-05, "loss": 0.3652, "step": 11043 }, { "epoch": 5.197176470588236, "grad_norm": 1.1504573726858474, "learning_rate": 1.82594436301638e-05, "loss": 0.9219, "step": 11044 }, { "epoch": 5.197647058823529, "grad_norm": 1.0870324947977446, "learning_rate": 1.8238488359171476e-05, "loss": 0.3926, "step": 11045 }, { "epoch": 5.198117647058823, "grad_norm": 0.8098590992866813, "learning_rate": 1.821754434125854e-05, "loss": 0.2275, "step": 11046 }, { "epoch": 5.1985882352941175, "grad_norm": 1.4292162067459768, "learning_rate": 1.8196611578213738e-05, "loss": 0.8438, "step": 11047 }, { "epoch": 5.199058823529412, "grad_norm": 1.0165051694117657, "learning_rate": 1.817569007182476e-05, "loss": 0.5156, "step": 11048 }, { "epoch": 5.199529411764706, "grad_norm": 1.0365952735479742, "learning_rate": 1.815477982387846e-05, "loss": 0.4414, "step": 11049 }, { "epoch": 5.2, "grad_norm": 1.097416207678917, "learning_rate": 1.813388083616068e-05, "loss": 0.4648, "step": 11050 }, { "epoch": 5.200470588235294, "grad_norm": 0.646827065927911, "learning_rate": 1.8112993110456235e-05, "loss": 0.1758, "step": 11051 }, { "epoch": 5.200941176470589, "grad_norm": 1.420708510397792, "learning_rate": 1.8092116648549136e-05, "loss": 1.0234, "step": 11052 }, { "epoch": 5.201411764705882, "grad_norm": 1.2312876793549334, "learning_rate": 1.8071251452222224e-05, "loss": 1.2266, "step": 11053 }, { "epoch": 5.201882352941176, "grad_norm": 1.0518693848522824, "learning_rate": 1.8050397523257555e-05, "loss": 0.7148, "step": 11054 }, { "epoch": 5.20235294117647, "grad_norm": 1.2334379220918437, "learning_rate": 1.8029554863436186e-05, "loss": 0.5391, "step": 11055 }, { "epoch": 5.202823529411765, "grad_norm": 0.8161422374535726, "learning_rate": 1.8008723474538095e-05, "loss": 0.2559, "step": 11056 }, { "epoch": 5.203294117647059, "grad_norm": 0.6879390329075865, "learning_rate": 1.798790335834244e-05, "loss": 0.1348, "step": 11057 }, { "epoch": 5.203764705882353, "grad_norm": 1.28664513121827, "learning_rate": 1.796709451662739e-05, "loss": 0.8828, "step": 11058 }, { "epoch": 5.204235294117647, "grad_norm": 1.106127033421241, "learning_rate": 1.7946296951170065e-05, "loss": 0.8516, "step": 11059 }, { "epoch": 5.2047058823529415, "grad_norm": 1.4150965426455526, "learning_rate": 1.792551066374668e-05, "loss": 0.5156, "step": 11060 }, { "epoch": 5.205176470588235, "grad_norm": 1.074222678613171, "learning_rate": 1.7904735656132545e-05, "loss": 0.6797, "step": 11061 }, { "epoch": 5.205647058823529, "grad_norm": 1.0574834959554813, "learning_rate": 1.7883971930101883e-05, "loss": 0.9102, "step": 11062 }, { "epoch": 5.206117647058823, "grad_norm": 1.037051227446744, "learning_rate": 1.786321948742803e-05, "loss": 0.3379, "step": 11063 }, { "epoch": 5.2065882352941175, "grad_norm": 0.6547621208485788, "learning_rate": 1.78424783298834e-05, "loss": 0.4785, "step": 11064 }, { "epoch": 5.207058823529412, "grad_norm": 1.164046415358965, "learning_rate": 1.782174845923931e-05, "loss": 0.7148, "step": 11065 }, { "epoch": 5.207529411764706, "grad_norm": 0.8759925686826876, "learning_rate": 1.7801029877266237e-05, "loss": 0.5, "step": 11066 }, { "epoch": 5.208, "grad_norm": 1.3257398457976608, "learning_rate": 1.778032258573368e-05, "loss": 0.9023, "step": 11067 }, { "epoch": 5.208470588235294, "grad_norm": 1.3162636352124881, "learning_rate": 1.7759626586410074e-05, "loss": 1.1406, "step": 11068 }, { "epoch": 5.208941176470589, "grad_norm": 1.642517872860155, "learning_rate": 1.7738941881063007e-05, "loss": 0.7852, "step": 11069 }, { "epoch": 5.209411764705882, "grad_norm": 1.4654394377119182, "learning_rate": 1.7718268471459045e-05, "loss": 0.8398, "step": 11070 }, { "epoch": 5.209882352941176, "grad_norm": 1.1168904029635982, "learning_rate": 1.769760635936378e-05, "loss": 0.3594, "step": 11071 }, { "epoch": 5.21035294117647, "grad_norm": 1.2948745589324258, "learning_rate": 1.7676955546541867e-05, "loss": 1.3516, "step": 11072 }, { "epoch": 5.210823529411765, "grad_norm": 1.0277358536002092, "learning_rate": 1.7656316034757024e-05, "loss": 0.5742, "step": 11073 }, { "epoch": 5.211294117647059, "grad_norm": 0.9233778379172858, "learning_rate": 1.7635687825771894e-05, "loss": 0.3008, "step": 11074 }, { "epoch": 5.211764705882353, "grad_norm": 1.1821575034273297, "learning_rate": 1.7615070921348316e-05, "loss": 0.5781, "step": 11075 }, { "epoch": 5.212235294117647, "grad_norm": 1.0597794917949848, "learning_rate": 1.7594465323246987e-05, "loss": 0.7109, "step": 11076 }, { "epoch": 5.2127058823529415, "grad_norm": 1.0718113773769713, "learning_rate": 1.757387103322776e-05, "loss": 1.125, "step": 11077 }, { "epoch": 5.213176470588236, "grad_norm": 1.3641624956553429, "learning_rate": 1.7553288053049515e-05, "loss": 0.9883, "step": 11078 }, { "epoch": 5.213647058823529, "grad_norm": 1.3837145721372257, "learning_rate": 1.7532716384470108e-05, "loss": 1.2188, "step": 11079 }, { "epoch": 5.214117647058823, "grad_norm": 1.3228699595758193, "learning_rate": 1.7512156029246453e-05, "loss": 1.25, "step": 11080 }, { "epoch": 5.2145882352941175, "grad_norm": 1.360377200398905, "learning_rate": 1.7491606989134566e-05, "loss": 0.668, "step": 11081 }, { "epoch": 5.215058823529412, "grad_norm": 0.3422914274537903, "learning_rate": 1.7471069265889353e-05, "loss": 0.2969, "step": 11082 }, { "epoch": 5.215529411764706, "grad_norm": 0.910617258721075, "learning_rate": 1.7450542861264883e-05, "loss": 0.4219, "step": 11083 }, { "epoch": 5.216, "grad_norm": 0.8356445825854067, "learning_rate": 1.7430027777014217e-05, "loss": 0.3027, "step": 11084 }, { "epoch": 5.216470588235294, "grad_norm": 1.1674310689249174, "learning_rate": 1.7409524014889388e-05, "loss": 0.4922, "step": 11085 }, { "epoch": 5.216941176470589, "grad_norm": 0.6357390573856087, "learning_rate": 1.738903157664161e-05, "loss": 0.3867, "step": 11086 }, { "epoch": 5.217411764705882, "grad_norm": 0.9766603802528385, "learning_rate": 1.736855046402097e-05, "loss": 0.377, "step": 11087 }, { "epoch": 5.217882352941176, "grad_norm": 1.1976837834342933, "learning_rate": 1.7348080678776665e-05, "loss": 0.5703, "step": 11088 }, { "epoch": 5.21835294117647, "grad_norm": 1.242653079167708, "learning_rate": 1.732762222265695e-05, "loss": 0.793, "step": 11089 }, { "epoch": 5.218823529411765, "grad_norm": 0.7851196105873545, "learning_rate": 1.7307175097409044e-05, "loss": 0.7148, "step": 11090 }, { "epoch": 5.219294117647059, "grad_norm": 1.0266648799804476, "learning_rate": 1.7286739304779214e-05, "loss": 0.6484, "step": 11091 }, { "epoch": 5.219764705882353, "grad_norm": 1.1956467672499278, "learning_rate": 1.7266314846512848e-05, "loss": 1.0781, "step": 11092 }, { "epoch": 5.220235294117647, "grad_norm": 0.9766666320817629, "learning_rate": 1.7245901724354218e-05, "loss": 0.3477, "step": 11093 }, { "epoch": 5.2207058823529415, "grad_norm": 0.928589375776236, "learning_rate": 1.7225499940046726e-05, "loss": 0.416, "step": 11094 }, { "epoch": 5.221176470588235, "grad_norm": 0.542364386370703, "learning_rate": 1.720510949533282e-05, "loss": 0.1099, "step": 11095 }, { "epoch": 5.221647058823529, "grad_norm": 1.03745030196547, "learning_rate": 1.71847303919539e-05, "loss": 0.6914, "step": 11096 }, { "epoch": 5.222117647058823, "grad_norm": 0.8554877237381096, "learning_rate": 1.7164362631650457e-05, "loss": 0.2422, "step": 11097 }, { "epoch": 5.2225882352941175, "grad_norm": 0.9155743749851325, "learning_rate": 1.7144006216162036e-05, "loss": 0.832, "step": 11098 }, { "epoch": 5.223058823529412, "grad_norm": 1.0705202269292802, "learning_rate": 1.7123661147227096e-05, "loss": 0.3672, "step": 11099 }, { "epoch": 5.223529411764706, "grad_norm": 0.6185966777610363, "learning_rate": 1.7103327426583265e-05, "loss": 0.1758, "step": 11100 }, { "epoch": 5.224, "grad_norm": 0.7006198188190187, "learning_rate": 1.7083005055967146e-05, "loss": 0.3828, "step": 11101 }, { "epoch": 5.224470588235294, "grad_norm": 1.2294478160840443, "learning_rate": 1.7062694037114326e-05, "loss": 0.5703, "step": 11102 }, { "epoch": 5.224941176470589, "grad_norm": 0.97585912170314, "learning_rate": 1.7042394371759516e-05, "loss": 0.7656, "step": 11103 }, { "epoch": 5.225411764705882, "grad_norm": 0.8198669324371072, "learning_rate": 1.7022106061636342e-05, "loss": 0.5352, "step": 11104 }, { "epoch": 5.225882352941176, "grad_norm": 1.1119553431326044, "learning_rate": 1.700182910847757e-05, "loss": 0.6211, "step": 11105 }, { "epoch": 5.22635294117647, "grad_norm": 0.3238935417757951, "learning_rate": 1.6981563514014967e-05, "loss": 0.022, "step": 11106 }, { "epoch": 5.226823529411765, "grad_norm": 1.1493953685933616, "learning_rate": 1.6961309279979278e-05, "loss": 0.6328, "step": 11107 }, { "epoch": 5.227294117647059, "grad_norm": 1.151528778210476, "learning_rate": 1.6941066408100308e-05, "loss": 0.5078, "step": 11108 }, { "epoch": 5.227764705882353, "grad_norm": 0.9195191529646447, "learning_rate": 1.6920834900106967e-05, "loss": 1.0703, "step": 11109 }, { "epoch": 5.228235294117647, "grad_norm": 1.5424263902014168, "learning_rate": 1.690061475772703e-05, "loss": 1.0234, "step": 11110 }, { "epoch": 5.2287058823529415, "grad_norm": 1.1541971856967934, "learning_rate": 1.6880405982687455e-05, "loss": 0.832, "step": 11111 }, { "epoch": 5.229176470588235, "grad_norm": 0.8898495815999556, "learning_rate": 1.6860208576714173e-05, "loss": 0.832, "step": 11112 }, { "epoch": 5.229647058823529, "grad_norm": 0.9114433916542026, "learning_rate": 1.68400225415321e-05, "loss": 1.1484, "step": 11113 }, { "epoch": 5.230117647058823, "grad_norm": 1.3179858473295902, "learning_rate": 1.6819847878865255e-05, "loss": 1.4375, "step": 11114 }, { "epoch": 5.2305882352941175, "grad_norm": 1.051901209030145, "learning_rate": 1.6799684590436663e-05, "loss": 0.707, "step": 11115 }, { "epoch": 5.231058823529412, "grad_norm": 1.202197797503544, "learning_rate": 1.6779532677968327e-05, "loss": 0.9727, "step": 11116 }, { "epoch": 5.231529411764706, "grad_norm": 1.2107059302996943, "learning_rate": 1.6759392143181354e-05, "loss": 0.9297, "step": 11117 }, { "epoch": 5.232, "grad_norm": 1.0389256653706453, "learning_rate": 1.6739262987795854e-05, "loss": 0.3145, "step": 11118 }, { "epoch": 5.232470588235294, "grad_norm": 1.0564370914419576, "learning_rate": 1.6719145213530898e-05, "loss": 0.3828, "step": 11119 }, { "epoch": 5.232941176470589, "grad_norm": 1.3556782912639396, "learning_rate": 1.669903882210468e-05, "loss": 1.3125, "step": 11120 }, { "epoch": 5.233411764705882, "grad_norm": 0.8476931212998516, "learning_rate": 1.667894381523443e-05, "loss": 0.1641, "step": 11121 }, { "epoch": 5.233882352941176, "grad_norm": 1.218533509571915, "learning_rate": 1.665886019463627e-05, "loss": 1.1875, "step": 11122 }, { "epoch": 5.23435294117647, "grad_norm": 0.9118249876033389, "learning_rate": 1.6638787962025495e-05, "loss": 0.3574, "step": 11123 }, { "epoch": 5.234823529411765, "grad_norm": 0.9047501686298327, "learning_rate": 1.6618727119116388e-05, "loss": 0.3223, "step": 11124 }, { "epoch": 5.235294117647059, "grad_norm": 1.719554960718313, "learning_rate": 1.6598677667622175e-05, "loss": 1.6562, "step": 11125 }, { "epoch": 5.235764705882353, "grad_norm": 0.9022384714052962, "learning_rate": 1.657863960925525e-05, "loss": 0.2871, "step": 11126 }, { "epoch": 5.236235294117647, "grad_norm": 0.9349638891073807, "learning_rate": 1.655861294572692e-05, "loss": 0.2451, "step": 11127 }, { "epoch": 5.2367058823529415, "grad_norm": 1.1452029096322671, "learning_rate": 1.6538597678747555e-05, "loss": 0.9883, "step": 11128 }, { "epoch": 5.237176470588235, "grad_norm": 0.9483409108816108, "learning_rate": 1.651859381002658e-05, "loss": 0.7461, "step": 11129 }, { "epoch": 5.237647058823529, "grad_norm": 1.0139499446262719, "learning_rate": 1.649860134127241e-05, "loss": 0.5, "step": 11130 }, { "epoch": 5.238117647058823, "grad_norm": 1.16292999633491, "learning_rate": 1.647862027419251e-05, "loss": 0.7031, "step": 11131 }, { "epoch": 5.2385882352941175, "grad_norm": 1.1137987421532622, "learning_rate": 1.6458650610493385e-05, "loss": 0.7617, "step": 11132 }, { "epoch": 5.239058823529412, "grad_norm": 0.7321162443242756, "learning_rate": 1.6438692351880482e-05, "loss": 0.1709, "step": 11133 }, { "epoch": 5.239529411764706, "grad_norm": 1.1757581164299375, "learning_rate": 1.6418745500058395e-05, "loss": 0.5078, "step": 11134 }, { "epoch": 5.24, "grad_norm": 0.7818337790699362, "learning_rate": 1.639881005673064e-05, "loss": 0.2559, "step": 11135 }, { "epoch": 5.240470588235294, "grad_norm": 1.3003760012271224, "learning_rate": 1.6378886023599803e-05, "loss": 0.5039, "step": 11136 }, { "epoch": 5.240941176470589, "grad_norm": 1.2911689041983077, "learning_rate": 1.6358973402367524e-05, "loss": 1.2266, "step": 11137 }, { "epoch": 5.241411764705882, "grad_norm": 1.0770687169483508, "learning_rate": 1.6339072194734414e-05, "loss": 0.9336, "step": 11138 }, { "epoch": 5.241882352941176, "grad_norm": 1.0816795544672857, "learning_rate": 1.6319182402400144e-05, "loss": 0.7891, "step": 11139 }, { "epoch": 5.24235294117647, "grad_norm": 1.0451789990520972, "learning_rate": 1.629930402706342e-05, "loss": 0.3477, "step": 11140 }, { "epoch": 5.242823529411765, "grad_norm": 1.0831137889865459, "learning_rate": 1.6279437070421885e-05, "loss": 0.3496, "step": 11141 }, { "epoch": 5.243294117647059, "grad_norm": 1.242839045834151, "learning_rate": 1.625958153417235e-05, "loss": 0.5781, "step": 11142 }, { "epoch": 5.243764705882353, "grad_norm": 1.5351949430334548, "learning_rate": 1.6239737420010558e-05, "loss": 0.7148, "step": 11143 }, { "epoch": 5.244235294117647, "grad_norm": 1.3611191234651772, "learning_rate": 1.621990472963125e-05, "loss": 1.4062, "step": 11144 }, { "epoch": 5.2447058823529416, "grad_norm": 0.7816309125438453, "learning_rate": 1.620008346472826e-05, "loss": 0.4102, "step": 11145 }, { "epoch": 5.245176470588235, "grad_norm": 1.192059158877473, "learning_rate": 1.618027362699446e-05, "loss": 0.2412, "step": 11146 }, { "epoch": 5.245647058823529, "grad_norm": 1.1017486922988489, "learning_rate": 1.6160475218121654e-05, "loss": 0.5234, "step": 11147 }, { "epoch": 5.246117647058823, "grad_norm": 1.1475159382331652, "learning_rate": 1.6140688239800736e-05, "loss": 0.6641, "step": 11148 }, { "epoch": 5.2465882352941176, "grad_norm": 0.6782499069571956, "learning_rate": 1.6120912693721654e-05, "loss": 0.1357, "step": 11149 }, { "epoch": 5.247058823529412, "grad_norm": 0.6265514948955101, "learning_rate": 1.6101148581573274e-05, "loss": 0.1494, "step": 11150 }, { "epoch": 5.247529411764706, "grad_norm": 1.1935886525676718, "learning_rate": 1.6081395905043565e-05, "loss": 1.2969, "step": 11151 }, { "epoch": 5.248, "grad_norm": 1.2190342847715847, "learning_rate": 1.606165466581954e-05, "loss": 0.4102, "step": 11152 }, { "epoch": 5.248470588235294, "grad_norm": 1.2432315824215452, "learning_rate": 1.6041924865587152e-05, "loss": 0.5117, "step": 11153 }, { "epoch": 5.248941176470589, "grad_norm": 1.2944949373922123, "learning_rate": 1.602220650603147e-05, "loss": 0.6602, "step": 11154 }, { "epoch": 5.249411764705882, "grad_norm": 1.0538880574455975, "learning_rate": 1.600249958883648e-05, "loss": 0.8164, "step": 11155 }, { "epoch": 5.249882352941176, "grad_norm": 1.043203756626182, "learning_rate": 1.5982804115685273e-05, "loss": 0.3594, "step": 11156 }, { "epoch": 5.2503529411764704, "grad_norm": 0.7479008704363647, "learning_rate": 1.5963120088259977e-05, "loss": 0.1748, "step": 11157 }, { "epoch": 5.250823529411765, "grad_norm": 1.3044652448882617, "learning_rate": 1.594344750824164e-05, "loss": 0.8281, "step": 11158 }, { "epoch": 5.251294117647059, "grad_norm": 1.0336906208260879, "learning_rate": 1.5923786377310433e-05, "loss": 0.3652, "step": 11159 }, { "epoch": 5.251764705882353, "grad_norm": 1.5139078166498827, "learning_rate": 1.5904136697145542e-05, "loss": 1.2891, "step": 11160 }, { "epoch": 5.252235294117647, "grad_norm": 0.7478436438293702, "learning_rate": 1.5884498469425078e-05, "loss": 0.1875, "step": 11161 }, { "epoch": 5.252705882352942, "grad_norm": 1.404119293164912, "learning_rate": 1.5864871695826282e-05, "loss": 0.5273, "step": 11162 }, { "epoch": 5.253176470588235, "grad_norm": 1.0105363998869608, "learning_rate": 1.5845256378025395e-05, "loss": 0.9883, "step": 11163 }, { "epoch": 5.253647058823529, "grad_norm": 1.2048511645303879, "learning_rate": 1.582565251769761e-05, "loss": 0.3848, "step": 11164 }, { "epoch": 5.254117647058823, "grad_norm": 1.0030331156434036, "learning_rate": 1.5806060116517227e-05, "loss": 0.2949, "step": 11165 }, { "epoch": 5.254588235294118, "grad_norm": 1.5314255535681465, "learning_rate": 1.578647917615757e-05, "loss": 0.8516, "step": 11166 }, { "epoch": 5.255058823529412, "grad_norm": 1.3212021529654079, "learning_rate": 1.5766909698290863e-05, "loss": 0.8047, "step": 11167 }, { "epoch": 5.255529411764706, "grad_norm": 1.3882756492667112, "learning_rate": 1.574735168458846e-05, "loss": 0.918, "step": 11168 }, { "epoch": 5.256, "grad_norm": 1.324233145692118, "learning_rate": 1.572780513672079e-05, "loss": 0.668, "step": 11169 }, { "epoch": 5.2564705882352944, "grad_norm": 1.0106927846261051, "learning_rate": 1.5708270056357115e-05, "loss": 0.459, "step": 11170 }, { "epoch": 5.256941176470589, "grad_norm": 1.2377627544267347, "learning_rate": 1.5688746445165867e-05, "loss": 0.8711, "step": 11171 }, { "epoch": 5.257411764705882, "grad_norm": 1.1402934836418837, "learning_rate": 1.5669234304814515e-05, "loss": 0.6797, "step": 11172 }, { "epoch": 5.257882352941176, "grad_norm": 0.6052070368878327, "learning_rate": 1.564973363696939e-05, "loss": 0.0996, "step": 11173 }, { "epoch": 5.2583529411764705, "grad_norm": 1.0106088064314935, "learning_rate": 1.5630244443296013e-05, "loss": 0.9492, "step": 11174 }, { "epoch": 5.258823529411765, "grad_norm": 1.4475981848169777, "learning_rate": 1.5610766725458834e-05, "loss": 0.9336, "step": 11175 }, { "epoch": 5.259294117647059, "grad_norm": 0.9432465193243065, "learning_rate": 1.5591300485121354e-05, "loss": 1.1719, "step": 11176 }, { "epoch": 5.259764705882353, "grad_norm": 0.9632929787330708, "learning_rate": 1.5571845723946113e-05, "loss": 0.5938, "step": 11177 }, { "epoch": 5.260235294117647, "grad_norm": 1.3749705391454405, "learning_rate": 1.555240244359458e-05, "loss": 0.8867, "step": 11178 }, { "epoch": 5.260705882352942, "grad_norm": 1.0802756142363865, "learning_rate": 1.5532970645727344e-05, "loss": 1.0156, "step": 11179 }, { "epoch": 5.261176470588235, "grad_norm": 1.1870448134766645, "learning_rate": 1.551355033200401e-05, "loss": 1.1406, "step": 11180 }, { "epoch": 5.261647058823529, "grad_norm": 0.8443066680564489, "learning_rate": 1.5494141504083087e-05, "loss": 0.1846, "step": 11181 }, { "epoch": 5.262117647058823, "grad_norm": 0.5214570595669693, "learning_rate": 1.547474416362226e-05, "loss": 0.0889, "step": 11182 }, { "epoch": 5.262588235294118, "grad_norm": 1.0587888719405916, "learning_rate": 1.5455358312278093e-05, "loss": 0.3594, "step": 11183 }, { "epoch": 5.263058823529412, "grad_norm": 0.6013529975149198, "learning_rate": 1.5435983951706276e-05, "loss": 0.1982, "step": 11184 }, { "epoch": 5.263529411764706, "grad_norm": 1.0294862305763952, "learning_rate": 1.5416621083561486e-05, "loss": 0.4785, "step": 11185 }, { "epoch": 5.264, "grad_norm": 1.007428474034362, "learning_rate": 1.5397269709497352e-05, "loss": 0.5156, "step": 11186 }, { "epoch": 5.2644705882352945, "grad_norm": 1.3731750815248063, "learning_rate": 1.5377929831166597e-05, "loss": 1.2656, "step": 11187 }, { "epoch": 5.264941176470588, "grad_norm": 1.00909812100006, "learning_rate": 1.5358601450220988e-05, "loss": 0.7344, "step": 11188 }, { "epoch": 5.265411764705882, "grad_norm": 0.774421442031066, "learning_rate": 1.53392845683112e-05, "loss": 0.1758, "step": 11189 }, { "epoch": 5.265882352941176, "grad_norm": 0.8529447740400276, "learning_rate": 1.5319979187087018e-05, "loss": 0.2119, "step": 11190 }, { "epoch": 5.2663529411764705, "grad_norm": 0.7783360343321226, "learning_rate": 1.5300685308197254e-05, "loss": 0.457, "step": 11191 }, { "epoch": 5.266823529411765, "grad_norm": 1.0792231294266834, "learning_rate": 1.528140293328962e-05, "loss": 0.4199, "step": 11192 }, { "epoch": 5.267294117647059, "grad_norm": 1.2294556870275428, "learning_rate": 1.5262132064010956e-05, "loss": 0.9492, "step": 11193 }, { "epoch": 5.267764705882353, "grad_norm": 0.9959742181070481, "learning_rate": 1.5242872702007142e-05, "loss": 0.4863, "step": 11194 }, { "epoch": 5.268235294117647, "grad_norm": 0.8323717068493609, "learning_rate": 1.522362484892296e-05, "loss": 0.7305, "step": 11195 }, { "epoch": 5.268705882352942, "grad_norm": 1.09017896876934, "learning_rate": 1.5204388506402281e-05, "loss": 0.4688, "step": 11196 }, { "epoch": 5.269176470588235, "grad_norm": 1.2920118874520807, "learning_rate": 1.5185163676088007e-05, "loss": 0.7578, "step": 11197 }, { "epoch": 5.269647058823529, "grad_norm": 1.3451112580017044, "learning_rate": 1.5165950359622009e-05, "loss": 1.2969, "step": 11198 }, { "epoch": 5.270117647058823, "grad_norm": 1.491800898518122, "learning_rate": 1.5146748558645205e-05, "loss": 1.0156, "step": 11199 }, { "epoch": 5.270588235294118, "grad_norm": 0.5778557648086754, "learning_rate": 1.5127558274797535e-05, "loss": 0.1187, "step": 11200 }, { "epoch": 5.271058823529412, "grad_norm": 1.202915476734743, "learning_rate": 1.5108379509717922e-05, "loss": 0.582, "step": 11201 }, { "epoch": 5.271529411764706, "grad_norm": 1.1746179774148808, "learning_rate": 1.5089212265044336e-05, "loss": 0.707, "step": 11202 }, { "epoch": 5.272, "grad_norm": 0.9269180311608092, "learning_rate": 1.5070056542413788e-05, "loss": 0.459, "step": 11203 }, { "epoch": 5.2724705882352945, "grad_norm": 1.1510142170703623, "learning_rate": 1.5050912343462202e-05, "loss": 0.8555, "step": 11204 }, { "epoch": 5.272941176470589, "grad_norm": 1.005966515148958, "learning_rate": 1.5031779669824668e-05, "loss": 0.6758, "step": 11205 }, { "epoch": 5.273411764705882, "grad_norm": 1.1535559997221014, "learning_rate": 1.5012658523135113e-05, "loss": 0.7617, "step": 11206 }, { "epoch": 5.273882352941176, "grad_norm": 0.9687908344622813, "learning_rate": 1.4993548905026648e-05, "loss": 0.2852, "step": 11207 }, { "epoch": 5.2743529411764705, "grad_norm": 1.0247420968671284, "learning_rate": 1.4974450817131334e-05, "loss": 0.3906, "step": 11208 }, { "epoch": 5.274823529411765, "grad_norm": 1.2435548234528788, "learning_rate": 1.4955364261080183e-05, "loss": 0.7852, "step": 11209 }, { "epoch": 5.275294117647059, "grad_norm": 1.5055576653196086, "learning_rate": 1.4936289238503324e-05, "loss": 1.6875, "step": 11210 }, { "epoch": 5.275764705882353, "grad_norm": 1.3796953277507433, "learning_rate": 1.491722575102987e-05, "loss": 0.5, "step": 11211 }, { "epoch": 5.276235294117647, "grad_norm": 1.121458727379057, "learning_rate": 1.4898173800287888e-05, "loss": 0.9727, "step": 11212 }, { "epoch": 5.276705882352942, "grad_norm": 0.8366476173052767, "learning_rate": 1.4879133387904523e-05, "loss": 0.2461, "step": 11213 }, { "epoch": 5.277176470588235, "grad_norm": 1.0807297082066605, "learning_rate": 1.4860104515505977e-05, "loss": 0.9883, "step": 11214 }, { "epoch": 5.277647058823529, "grad_norm": 0.7262623365511158, "learning_rate": 1.4841087184717315e-05, "loss": 0.1143, "step": 11215 }, { "epoch": 5.278117647058823, "grad_norm": 1.393986914085657, "learning_rate": 1.4822081397162771e-05, "loss": 0.9688, "step": 11216 }, { "epoch": 5.278588235294118, "grad_norm": 1.1087365163081537, "learning_rate": 1.4803087154465548e-05, "loss": 0.6211, "step": 11217 }, { "epoch": 5.279058823529412, "grad_norm": 0.6640614535447418, "learning_rate": 1.4784104458247764e-05, "loss": 0.3809, "step": 11218 }, { "epoch": 5.279529411764706, "grad_norm": 1.2425265426869012, "learning_rate": 1.476513331013074e-05, "loss": 0.8945, "step": 11219 }, { "epoch": 5.28, "grad_norm": 1.0144908969342843, "learning_rate": 1.4746173711734627e-05, "loss": 0.5469, "step": 11220 }, { "epoch": 5.2804705882352945, "grad_norm": 0.7114545374449014, "learning_rate": 1.4727225664678699e-05, "loss": 0.2256, "step": 11221 }, { "epoch": 5.280941176470588, "grad_norm": 1.089530345028301, "learning_rate": 1.4708289170581227e-05, "loss": 0.7148, "step": 11222 }, { "epoch": 5.281411764705882, "grad_norm": 0.9619587885069133, "learning_rate": 1.4689364231059436e-05, "loss": 0.2432, "step": 11223 }, { "epoch": 5.281882352941176, "grad_norm": 1.1238609606034289, "learning_rate": 1.4670450847729648e-05, "loss": 0.4316, "step": 11224 }, { "epoch": 5.2823529411764705, "grad_norm": 1.1247482367462538, "learning_rate": 1.465154902220717e-05, "loss": 1.1875, "step": 11225 }, { "epoch": 5.282823529411765, "grad_norm": 1.567003379998774, "learning_rate": 1.4632658756106247e-05, "loss": 1.5625, "step": 11226 }, { "epoch": 5.283294117647059, "grad_norm": 0.5637171741058414, "learning_rate": 1.4613780051040236e-05, "loss": 0.1865, "step": 11227 }, { "epoch": 5.283764705882353, "grad_norm": 1.1990678505981889, "learning_rate": 1.4594912908621515e-05, "loss": 0.707, "step": 11228 }, { "epoch": 5.284235294117647, "grad_norm": 0.7587147099361193, "learning_rate": 1.4576057330461361e-05, "loss": 0.2295, "step": 11229 }, { "epoch": 5.284705882352942, "grad_norm": 1.5204889368793184, "learning_rate": 1.4557213318170152e-05, "loss": 1.6016, "step": 11230 }, { "epoch": 5.285176470588235, "grad_norm": 1.2001103767968033, "learning_rate": 1.453838087335732e-05, "loss": 0.7305, "step": 11231 }, { "epoch": 5.285647058823529, "grad_norm": 0.7675181248250748, "learning_rate": 1.4519559997631142e-05, "loss": 0.3008, "step": 11232 }, { "epoch": 5.286117647058823, "grad_norm": 0.6979187310880114, "learning_rate": 1.45007506925991e-05, "loss": 0.168, "step": 11233 }, { "epoch": 5.286588235294118, "grad_norm": 1.1072579874258741, "learning_rate": 1.448195295986756e-05, "loss": 0.7852, "step": 11234 }, { "epoch": 5.287058823529412, "grad_norm": 0.597337286889817, "learning_rate": 1.4463166801041936e-05, "loss": 0.4727, "step": 11235 }, { "epoch": 5.287529411764706, "grad_norm": 0.6928119468396267, "learning_rate": 1.4444392217726697e-05, "loss": 0.1338, "step": 11236 }, { "epoch": 5.288, "grad_norm": 0.9781634390574953, "learning_rate": 1.442562921152524e-05, "loss": 0.457, "step": 11237 }, { "epoch": 5.2884705882352945, "grad_norm": 1.1002648316876085, "learning_rate": 1.4406877784040021e-05, "loss": 0.6367, "step": 11238 }, { "epoch": 5.288941176470589, "grad_norm": 0.8102956840414292, "learning_rate": 1.4388137936872556e-05, "loss": 0.4336, "step": 11239 }, { "epoch": 5.289411764705882, "grad_norm": 1.234760727269053, "learning_rate": 1.4369409671623267e-05, "loss": 1.1172, "step": 11240 }, { "epoch": 5.289882352941176, "grad_norm": 0.8440592751690617, "learning_rate": 1.4350692989891638e-05, "loss": 0.1797, "step": 11241 }, { "epoch": 5.2903529411764705, "grad_norm": 1.0905224332874224, "learning_rate": 1.4331987893276225e-05, "loss": 0.377, "step": 11242 }, { "epoch": 5.290823529411765, "grad_norm": 0.6880873981368071, "learning_rate": 1.4313294383374468e-05, "loss": 0.1592, "step": 11243 }, { "epoch": 5.291294117647059, "grad_norm": 1.2027615081148881, "learning_rate": 1.429461246178289e-05, "loss": 1.1797, "step": 11244 }, { "epoch": 5.291764705882353, "grad_norm": 1.4309598345313972, "learning_rate": 1.4275942130097096e-05, "loss": 0.8555, "step": 11245 }, { "epoch": 5.292235294117647, "grad_norm": 1.017444862775915, "learning_rate": 1.4257283389911512e-05, "loss": 0.7266, "step": 11246 }, { "epoch": 5.292705882352941, "grad_norm": 1.1292333051082235, "learning_rate": 1.4238636242819761e-05, "loss": 0.8828, "step": 11247 }, { "epoch": 5.293176470588235, "grad_norm": 0.47481919828914587, "learning_rate": 1.422000069041439e-05, "loss": 0.1865, "step": 11248 }, { "epoch": 5.293647058823529, "grad_norm": 1.0483856859926224, "learning_rate": 1.420137673428694e-05, "loss": 0.4043, "step": 11249 }, { "epoch": 5.294117647058823, "grad_norm": 0.9404780285735285, "learning_rate": 1.4182764376028006e-05, "loss": 1.0469, "step": 11250 }, { "epoch": 5.294588235294118, "grad_norm": 0.8333321693417419, "learning_rate": 1.4164163617227197e-05, "loss": 0.2295, "step": 11251 }, { "epoch": 5.295058823529412, "grad_norm": 1.4374331642922336, "learning_rate": 1.4145574459473064e-05, "loss": 0.3105, "step": 11252 }, { "epoch": 5.295529411764706, "grad_norm": 1.4747317232951382, "learning_rate": 1.4126996904353233e-05, "loss": 0.9844, "step": 11253 }, { "epoch": 5.296, "grad_norm": 0.9393637140492335, "learning_rate": 1.4108430953454352e-05, "loss": 0.4531, "step": 11254 }, { "epoch": 5.2964705882352945, "grad_norm": 1.0574702204894906, "learning_rate": 1.4089876608361971e-05, "loss": 0.7891, "step": 11255 }, { "epoch": 5.296941176470588, "grad_norm": 1.1915381058452075, "learning_rate": 1.4071333870660807e-05, "loss": 1.2109, "step": 11256 }, { "epoch": 5.297411764705882, "grad_norm": 0.9241045891892152, "learning_rate": 1.4052802741934404e-05, "loss": 0.4512, "step": 11257 }, { "epoch": 5.297882352941176, "grad_norm": 0.6479284589364269, "learning_rate": 1.4034283223765485e-05, "loss": 0.1328, "step": 11258 }, { "epoch": 5.2983529411764705, "grad_norm": 0.47804714837475926, "learning_rate": 1.4015775317735717e-05, "loss": 0.1533, "step": 11259 }, { "epoch": 5.298823529411765, "grad_norm": 1.1291726346969027, "learning_rate": 1.3997279025425701e-05, "loss": 0.9922, "step": 11260 }, { "epoch": 5.299294117647059, "grad_norm": 1.215799666902867, "learning_rate": 1.397879434841514e-05, "loss": 0.582, "step": 11261 }, { "epoch": 5.299764705882353, "grad_norm": 1.2081408335068626, "learning_rate": 1.3960321288282722e-05, "loss": 1.3281, "step": 11262 }, { "epoch": 5.300235294117647, "grad_norm": 0.72871642353846, "learning_rate": 1.3941859846606152e-05, "loss": 0.498, "step": 11263 }, { "epoch": 5.300705882352942, "grad_norm": 1.0128373338452938, "learning_rate": 1.3923410024962117e-05, "loss": 0.9648, "step": 11264 }, { "epoch": 5.301176470588235, "grad_norm": 0.8514305065880968, "learning_rate": 1.390497182492631e-05, "loss": 0.2852, "step": 11265 }, { "epoch": 5.301647058823529, "grad_norm": 0.9327672747395868, "learning_rate": 1.388654524807345e-05, "loss": 0.7969, "step": 11266 }, { "epoch": 5.302117647058823, "grad_norm": 1.2281077374408753, "learning_rate": 1.3868130295977282e-05, "loss": 0.5312, "step": 11267 }, { "epoch": 5.302588235294118, "grad_norm": 0.5575975135845752, "learning_rate": 1.3849726970210479e-05, "loss": 0.1104, "step": 11268 }, { "epoch": 5.303058823529412, "grad_norm": 0.8769817403199124, "learning_rate": 1.3831335272344801e-05, "loss": 0.3613, "step": 11269 }, { "epoch": 5.303529411764706, "grad_norm": 1.0773457067021528, "learning_rate": 1.3812955203951026e-05, "loss": 0.6758, "step": 11270 }, { "epoch": 5.304, "grad_norm": 0.9008445977136725, "learning_rate": 1.3794586766598848e-05, "loss": 0.2412, "step": 11271 }, { "epoch": 5.3044705882352945, "grad_norm": 0.94628925251436, "learning_rate": 1.3776229961857043e-05, "loss": 0.4922, "step": 11272 }, { "epoch": 5.304941176470588, "grad_norm": 0.8253450719462193, "learning_rate": 1.3757884791293394e-05, "loss": 0.332, "step": 11273 }, { "epoch": 5.305411764705882, "grad_norm": 0.802905149120976, "learning_rate": 1.373955125647463e-05, "loss": 0.2754, "step": 11274 }, { "epoch": 5.305882352941176, "grad_norm": 1.2400051099259877, "learning_rate": 1.3721229358966545e-05, "loss": 0.377, "step": 11275 }, { "epoch": 5.3063529411764705, "grad_norm": 1.097906996510531, "learning_rate": 1.3702919100333925e-05, "loss": 0.4121, "step": 11276 }, { "epoch": 5.306823529411765, "grad_norm": 1.0412686083946356, "learning_rate": 1.3684620482140534e-05, "loss": 0.668, "step": 11277 }, { "epoch": 5.307294117647059, "grad_norm": 1.4614380674504126, "learning_rate": 1.3666333505949173e-05, "loss": 0.8516, "step": 11278 }, { "epoch": 5.307764705882353, "grad_norm": 1.179640978445835, "learning_rate": 1.3648058173321675e-05, "loss": 0.7461, "step": 11279 }, { "epoch": 5.308235294117647, "grad_norm": 1.135889686935062, "learning_rate": 1.3629794485818791e-05, "loss": 0.8672, "step": 11280 }, { "epoch": 5.308705882352941, "grad_norm": 0.8391607668897936, "learning_rate": 1.361154244500034e-05, "loss": 0.25, "step": 11281 }, { "epoch": 5.309176470588235, "grad_norm": 0.8441184435648638, "learning_rate": 1.3593302052425159e-05, "loss": 0.6797, "step": 11282 }, { "epoch": 5.309647058823529, "grad_norm": 0.9191414937729696, "learning_rate": 1.3575073309651052e-05, "loss": 0.2852, "step": 11283 }, { "epoch": 5.310117647058823, "grad_norm": 1.1485861706785783, "learning_rate": 1.3556856218234858e-05, "loss": 0.9375, "step": 11284 }, { "epoch": 5.310588235294118, "grad_norm": 0.5959103403182396, "learning_rate": 1.3538650779732362e-05, "loss": 0.4707, "step": 11285 }, { "epoch": 5.311058823529412, "grad_norm": 0.6518943802297991, "learning_rate": 1.3520456995698443e-05, "loss": 0.1226, "step": 11286 }, { "epoch": 5.311529411764706, "grad_norm": 0.7027781788056574, "learning_rate": 1.3502274867686936e-05, "loss": 0.2031, "step": 11287 }, { "epoch": 5.312, "grad_norm": 1.5477616071520897, "learning_rate": 1.348410439725065e-05, "loss": 1.2656, "step": 11288 }, { "epoch": 5.3124705882352945, "grad_norm": 0.9192706650414582, "learning_rate": 1.3465945585941462e-05, "loss": 0.4434, "step": 11289 }, { "epoch": 5.312941176470588, "grad_norm": 0.8957822839803452, "learning_rate": 1.3447798435310247e-05, "loss": 0.3203, "step": 11290 }, { "epoch": 5.313411764705882, "grad_norm": 0.9811288046776877, "learning_rate": 1.3429662946906799e-05, "loss": 0.3711, "step": 11291 }, { "epoch": 5.313882352941176, "grad_norm": 0.8039591232891697, "learning_rate": 1.3411539122279996e-05, "loss": 0.2656, "step": 11292 }, { "epoch": 5.3143529411764705, "grad_norm": 1.1192858680098765, "learning_rate": 1.3393426962977765e-05, "loss": 0.6953, "step": 11293 }, { "epoch": 5.314823529411765, "grad_norm": 1.0128689439741134, "learning_rate": 1.3375326470546872e-05, "loss": 0.5234, "step": 11294 }, { "epoch": 5.315294117647059, "grad_norm": 1.139795110831558, "learning_rate": 1.3357237646533258e-05, "loss": 0.6562, "step": 11295 }, { "epoch": 5.315764705882353, "grad_norm": 1.0475414385932342, "learning_rate": 1.3339160492481793e-05, "loss": 0.832, "step": 11296 }, { "epoch": 5.316235294117647, "grad_norm": 0.736003978575548, "learning_rate": 1.3321095009936321e-05, "loss": 0.7109, "step": 11297 }, { "epoch": 5.316705882352942, "grad_norm": 0.8022869214178866, "learning_rate": 1.3303041200439729e-05, "loss": 0.2002, "step": 11298 }, { "epoch": 5.317176470588235, "grad_norm": 1.2518427130307581, "learning_rate": 1.3284999065533946e-05, "loss": 0.7461, "step": 11299 }, { "epoch": 5.317647058823529, "grad_norm": 1.042706405789828, "learning_rate": 1.326696860675981e-05, "loss": 0.625, "step": 11300 }, { "epoch": 5.318117647058823, "grad_norm": 1.366903375427497, "learning_rate": 1.3248949825657217e-05, "loss": 0.9023, "step": 11301 }, { "epoch": 5.318588235294118, "grad_norm": 0.3390692002916147, "learning_rate": 1.3230942723765092e-05, "loss": 0.0145, "step": 11302 }, { "epoch": 5.319058823529412, "grad_norm": 1.1804737208400076, "learning_rate": 1.3212947302621284e-05, "loss": 1.1328, "step": 11303 }, { "epoch": 5.319529411764706, "grad_norm": 1.0523610863248183, "learning_rate": 1.3194963563762718e-05, "loss": 0.5625, "step": 11304 }, { "epoch": 5.32, "grad_norm": 1.0708938548728446, "learning_rate": 1.317699150872531e-05, "loss": 0.6641, "step": 11305 }, { "epoch": 5.3204705882352945, "grad_norm": 0.9831928245006845, "learning_rate": 1.3159031139043918e-05, "loss": 0.8594, "step": 11306 }, { "epoch": 5.320941176470588, "grad_norm": 1.0293838921085587, "learning_rate": 1.3141082456252467e-05, "loss": 0.5039, "step": 11307 }, { "epoch": 5.321411764705882, "grad_norm": 0.9916984849434146, "learning_rate": 1.312314546188386e-05, "loss": 0.3926, "step": 11308 }, { "epoch": 5.321882352941176, "grad_norm": 0.9607548121353299, "learning_rate": 1.3105220157470009e-05, "loss": 0.3223, "step": 11309 }, { "epoch": 5.3223529411764705, "grad_norm": 0.969742452810394, "learning_rate": 1.3087306544541836e-05, "loss": 0.8828, "step": 11310 }, { "epoch": 5.322823529411765, "grad_norm": 1.0105982141007497, "learning_rate": 1.306940462462922e-05, "loss": 1.0156, "step": 11311 }, { "epoch": 5.323294117647059, "grad_norm": 1.0345515803198482, "learning_rate": 1.3051514399261104e-05, "loss": 0.3691, "step": 11312 }, { "epoch": 5.323764705882353, "grad_norm": 1.2082556442175734, "learning_rate": 1.3033635869965398e-05, "loss": 0.6289, "step": 11313 }, { "epoch": 5.324235294117647, "grad_norm": 1.2520704431622052, "learning_rate": 1.3015769038268981e-05, "loss": 0.8086, "step": 11314 }, { "epoch": 5.324705882352941, "grad_norm": 1.2087274624400788, "learning_rate": 1.2997913905697816e-05, "loss": 1.3125, "step": 11315 }, { "epoch": 5.325176470588235, "grad_norm": 1.1444959923038738, "learning_rate": 1.2980070473776782e-05, "loss": 0.8828, "step": 11316 }, { "epoch": 5.325647058823529, "grad_norm": 1.2096320715286353, "learning_rate": 1.2962238744029795e-05, "loss": 1.1172, "step": 11317 }, { "epoch": 5.326117647058823, "grad_norm": 0.6202006834117569, "learning_rate": 1.2944418717979804e-05, "loss": 0.2539, "step": 11318 }, { "epoch": 5.326588235294118, "grad_norm": 1.1751699862683491, "learning_rate": 1.2926610397148707e-05, "loss": 0.7109, "step": 11319 }, { "epoch": 5.327058823529412, "grad_norm": 0.8027694547895787, "learning_rate": 1.2908813783057404e-05, "loss": 0.6328, "step": 11320 }, { "epoch": 5.327529411764706, "grad_norm": 1.439527147457344, "learning_rate": 1.2891028877225863e-05, "loss": 0.7969, "step": 11321 }, { "epoch": 5.328, "grad_norm": 1.1270870253724221, "learning_rate": 1.287325568117295e-05, "loss": 0.6094, "step": 11322 }, { "epoch": 5.3284705882352945, "grad_norm": 1.2266916259157066, "learning_rate": 1.2855494196416604e-05, "loss": 0.6758, "step": 11323 }, { "epoch": 5.328941176470588, "grad_norm": 0.8967685129195183, "learning_rate": 1.2837744424473761e-05, "loss": 0.7539, "step": 11324 }, { "epoch": 5.329411764705882, "grad_norm": 1.0326401419469893, "learning_rate": 1.2820006366860292e-05, "loss": 0.6172, "step": 11325 }, { "epoch": 5.329882352941176, "grad_norm": 0.9829372317618827, "learning_rate": 1.2802280025091165e-05, "loss": 0.3926, "step": 11326 }, { "epoch": 5.3303529411764705, "grad_norm": 0.9895180037525875, "learning_rate": 1.2784565400680291e-05, "loss": 0.3574, "step": 11327 }, { "epoch": 5.330823529411765, "grad_norm": 0.9667701359628553, "learning_rate": 1.276686249514054e-05, "loss": 0.3184, "step": 11328 }, { "epoch": 5.331294117647059, "grad_norm": 0.9134879144353131, "learning_rate": 1.2749171309983852e-05, "loss": 0.2949, "step": 11329 }, { "epoch": 5.331764705882353, "grad_norm": 1.096857777456168, "learning_rate": 1.273149184672117e-05, "loss": 0.7383, "step": 11330 }, { "epoch": 5.332235294117647, "grad_norm": 1.407221253057911, "learning_rate": 1.2713824106862369e-05, "loss": 0.8281, "step": 11331 }, { "epoch": 5.332705882352942, "grad_norm": 0.9813967282793992, "learning_rate": 1.2696168091916376e-05, "loss": 0.6055, "step": 11332 }, { "epoch": 5.333176470588235, "grad_norm": 0.9111981241625107, "learning_rate": 1.2678523803391116e-05, "loss": 0.4082, "step": 11333 }, { "epoch": 5.333647058823529, "grad_norm": 0.6469914692757884, "learning_rate": 1.2660891242793453e-05, "loss": 0.1475, "step": 11334 }, { "epoch": 5.334117647058823, "grad_norm": 1.0568231246634476, "learning_rate": 1.2643270411629346e-05, "loss": 0.5312, "step": 11335 }, { "epoch": 5.334588235294118, "grad_norm": 1.386197300677654, "learning_rate": 1.2625661311403662e-05, "loss": 0.9141, "step": 11336 }, { "epoch": 5.335058823529412, "grad_norm": 1.4042511500026293, "learning_rate": 1.2608063943620312e-05, "loss": 0.8086, "step": 11337 }, { "epoch": 5.335529411764706, "grad_norm": 1.2751396425497585, "learning_rate": 1.2590478309782243e-05, "loss": 0.5312, "step": 11338 }, { "epoch": 5.336, "grad_norm": 1.0474996642420136, "learning_rate": 1.2572904411391272e-05, "loss": 0.5156, "step": 11339 }, { "epoch": 5.3364705882352945, "grad_norm": 1.170879914577282, "learning_rate": 1.2555342249948347e-05, "loss": 1.3828, "step": 11340 }, { "epoch": 5.336941176470588, "grad_norm": 1.0221379596178712, "learning_rate": 1.2537791826953387e-05, "loss": 0.6758, "step": 11341 }, { "epoch": 5.337411764705882, "grad_norm": 0.6400105852052307, "learning_rate": 1.2520253143905223e-05, "loss": 0.1104, "step": 11342 }, { "epoch": 5.337882352941176, "grad_norm": 0.9056754035574206, "learning_rate": 1.2502726202301777e-05, "loss": 0.8125, "step": 11343 }, { "epoch": 5.3383529411764705, "grad_norm": 1.1065155885909153, "learning_rate": 1.2485211003639967e-05, "loss": 0.8242, "step": 11344 }, { "epoch": 5.338823529411765, "grad_norm": 0.8718039178796765, "learning_rate": 1.2467707549415612e-05, "loss": 0.3516, "step": 11345 }, { "epoch": 5.339294117647059, "grad_norm": 1.3411803253904622, "learning_rate": 1.2450215841123618e-05, "loss": 0.7891, "step": 11346 }, { "epoch": 5.339764705882353, "grad_norm": 1.221282783001632, "learning_rate": 1.2432735880257904e-05, "loss": 0.9102, "step": 11347 }, { "epoch": 5.340235294117647, "grad_norm": 1.0537516597757932, "learning_rate": 1.2415267668311263e-05, "loss": 1.3125, "step": 11348 }, { "epoch": 5.340705882352941, "grad_norm": 1.2835035262431989, "learning_rate": 1.2397811206775632e-05, "loss": 0.7383, "step": 11349 }, { "epoch": 5.341176470588235, "grad_norm": 0.8726746544192322, "learning_rate": 1.2380366497141886e-05, "loss": 1.3281, "step": 11350 }, { "epoch": 5.341647058823529, "grad_norm": 1.3061412653964537, "learning_rate": 1.23629335408998e-05, "loss": 1.1406, "step": 11351 }, { "epoch": 5.342117647058823, "grad_norm": 1.1444086688720316, "learning_rate": 1.234551233953835e-05, "loss": 0.7656, "step": 11352 }, { "epoch": 5.342588235294118, "grad_norm": 0.8424806433286552, "learning_rate": 1.232810289454531e-05, "loss": 0.2715, "step": 11353 }, { "epoch": 5.343058823529412, "grad_norm": 0.9384299331936888, "learning_rate": 1.2310705207407561e-05, "loss": 0.668, "step": 11354 }, { "epoch": 5.343529411764706, "grad_norm": 1.105325352070729, "learning_rate": 1.229331927961098e-05, "loss": 0.9102, "step": 11355 }, { "epoch": 5.344, "grad_norm": 1.1620694685778257, "learning_rate": 1.2275945112640363e-05, "loss": 0.4688, "step": 11356 }, { "epoch": 5.3444705882352945, "grad_norm": 1.0587752845590956, "learning_rate": 1.2258582707979554e-05, "loss": 0.8008, "step": 11357 }, { "epoch": 5.344941176470588, "grad_norm": 1.129751014193178, "learning_rate": 1.2241232067111418e-05, "loss": 0.6523, "step": 11358 }, { "epoch": 5.345411764705882, "grad_norm": 1.1893827906815089, "learning_rate": 1.2223893191517753e-05, "loss": 0.3594, "step": 11359 }, { "epoch": 5.345882352941176, "grad_norm": 0.9030164302322693, "learning_rate": 1.2206566082679392e-05, "loss": 0.7148, "step": 11360 }, { "epoch": 5.3463529411764705, "grad_norm": 1.2234785042349308, "learning_rate": 1.2189250742076184e-05, "loss": 1.0859, "step": 11361 }, { "epoch": 5.346823529411765, "grad_norm": 1.3528081862473065, "learning_rate": 1.2171947171186897e-05, "loss": 0.6992, "step": 11362 }, { "epoch": 5.347294117647059, "grad_norm": 1.9085945498320327, "learning_rate": 1.2154655371489397e-05, "loss": 0.5039, "step": 11363 }, { "epoch": 5.347764705882353, "grad_norm": 0.9984203916320984, "learning_rate": 1.213737534446042e-05, "loss": 0.6719, "step": 11364 }, { "epoch": 5.348235294117647, "grad_norm": 1.0376158455044895, "learning_rate": 1.2120107091575803e-05, "loss": 1.0938, "step": 11365 }, { "epoch": 5.348705882352942, "grad_norm": 1.155437684841398, "learning_rate": 1.2102850614310366e-05, "loss": 0.4414, "step": 11366 }, { "epoch": 5.349176470588235, "grad_norm": 0.5290812327961145, "learning_rate": 1.2085605914137847e-05, "loss": 0.0698, "step": 11367 }, { "epoch": 5.349647058823529, "grad_norm": 1.3035995795200466, "learning_rate": 1.206837299253105e-05, "loss": 1.1016, "step": 11368 }, { "epoch": 5.350117647058823, "grad_norm": 1.2182941490434644, "learning_rate": 1.2051151850961766e-05, "loss": 0.707, "step": 11369 }, { "epoch": 5.350588235294118, "grad_norm": 1.4431205121955046, "learning_rate": 1.2033942490900716e-05, "loss": 0.8477, "step": 11370 }, { "epoch": 5.351058823529412, "grad_norm": 0.6315503744999622, "learning_rate": 1.201674491381771e-05, "loss": 0.4434, "step": 11371 }, { "epoch": 5.351529411764706, "grad_norm": 0.8344134601989034, "learning_rate": 1.1999559121181523e-05, "loss": 0.2188, "step": 11372 }, { "epoch": 5.352, "grad_norm": 1.8021572470543952, "learning_rate": 1.1982385114459846e-05, "loss": 0.793, "step": 11373 }, { "epoch": 5.352470588235294, "grad_norm": 0.655527880735825, "learning_rate": 1.1965222895119442e-05, "loss": 0.1235, "step": 11374 }, { "epoch": 5.352941176470588, "grad_norm": 0.9704880885679558, "learning_rate": 1.1948072464626101e-05, "loss": 0.5469, "step": 11375 }, { "epoch": 5.353411764705882, "grad_norm": 0.4920547945513902, "learning_rate": 1.1930933824444472e-05, "loss": 0.0698, "step": 11376 }, { "epoch": 5.353882352941176, "grad_norm": 0.8414596622270396, "learning_rate": 1.1913806976038331e-05, "loss": 0.2871, "step": 11377 }, { "epoch": 5.3543529411764705, "grad_norm": 0.8068315697163831, "learning_rate": 1.1896691920870394e-05, "loss": 0.2236, "step": 11378 }, { "epoch": 5.354823529411765, "grad_norm": 1.3477660385277117, "learning_rate": 1.1879588660402356e-05, "loss": 1.2812, "step": 11379 }, { "epoch": 5.355294117647059, "grad_norm": 0.6471184683153594, "learning_rate": 1.18624971960949e-05, "loss": 0.1514, "step": 11380 }, { "epoch": 5.355764705882353, "grad_norm": 1.100823063509663, "learning_rate": 1.1845417529407791e-05, "loss": 1.0703, "step": 11381 }, { "epoch": 5.356235294117647, "grad_norm": 0.9996251127792084, "learning_rate": 1.1828349661799629e-05, "loss": 0.4766, "step": 11382 }, { "epoch": 5.356705882352941, "grad_norm": 0.9371311047562385, "learning_rate": 1.1811293594728133e-05, "loss": 0.3105, "step": 11383 }, { "epoch": 5.357176470588235, "grad_norm": 0.974859022206983, "learning_rate": 1.179424932965002e-05, "loss": 0.3867, "step": 11384 }, { "epoch": 5.357647058823529, "grad_norm": 1.2463636124709323, "learning_rate": 1.1777216868020872e-05, "loss": 0.8398, "step": 11385 }, { "epoch": 5.358117647058823, "grad_norm": 0.6821733472515895, "learning_rate": 1.1760196211295398e-05, "loss": 0.5, "step": 11386 }, { "epoch": 5.358588235294118, "grad_norm": 1.3340112665677597, "learning_rate": 1.1743187360927214e-05, "loss": 0.6484, "step": 11387 }, { "epoch": 5.359058823529412, "grad_norm": 1.2189293719063305, "learning_rate": 1.1726190318368977e-05, "loss": 0.7148, "step": 11388 }, { "epoch": 5.359529411764706, "grad_norm": 1.16860721466992, "learning_rate": 1.1709205085072344e-05, "loss": 0.8477, "step": 11389 }, { "epoch": 5.36, "grad_norm": 1.0703724766856435, "learning_rate": 1.1692231662487866e-05, "loss": 0.8945, "step": 11390 }, { "epoch": 5.3604705882352945, "grad_norm": 0.8551359548707049, "learning_rate": 1.1675270052065222e-05, "loss": 0.3398, "step": 11391 }, { "epoch": 5.360941176470588, "grad_norm": 1.0832814170525924, "learning_rate": 1.1658320255253e-05, "loss": 0.7656, "step": 11392 }, { "epoch": 5.361411764705882, "grad_norm": 1.221444896411831, "learning_rate": 1.1641382273498777e-05, "loss": 0.6133, "step": 11393 }, { "epoch": 5.361882352941176, "grad_norm": 1.1599262775751937, "learning_rate": 1.1624456108249164e-05, "loss": 1.0625, "step": 11394 }, { "epoch": 5.3623529411764705, "grad_norm": 1.4642550195538815, "learning_rate": 1.1607541760949735e-05, "loss": 0.4375, "step": 11395 }, { "epoch": 5.362823529411765, "grad_norm": 1.2399013731171633, "learning_rate": 1.1590639233045002e-05, "loss": 0.5742, "step": 11396 }, { "epoch": 5.363294117647059, "grad_norm": 1.1058362745139492, "learning_rate": 1.1573748525978626e-05, "loss": 0.6602, "step": 11397 }, { "epoch": 5.363764705882353, "grad_norm": 1.0109230565807208, "learning_rate": 1.155686964119309e-05, "loss": 1.2109, "step": 11398 }, { "epoch": 5.364235294117647, "grad_norm": 1.2299238803286112, "learning_rate": 1.154000258012992e-05, "loss": 0.7891, "step": 11399 }, { "epoch": 5.364705882352941, "grad_norm": 1.1260159746755098, "learning_rate": 1.1523147344229716e-05, "loss": 0.7773, "step": 11400 }, { "epoch": 5.365176470588235, "grad_norm": 1.1951561827365194, "learning_rate": 1.1506303934931926e-05, "loss": 0.7305, "step": 11401 }, { "epoch": 5.365647058823529, "grad_norm": 1.2184741249440159, "learning_rate": 1.1489472353675083e-05, "loss": 0.7656, "step": 11402 }, { "epoch": 5.366117647058823, "grad_norm": 1.200413786829217, "learning_rate": 1.147265260189672e-05, "loss": 1.1719, "step": 11403 }, { "epoch": 5.366588235294118, "grad_norm": 0.7910897891010211, "learning_rate": 1.1455844681033272e-05, "loss": 0.7188, "step": 11404 }, { "epoch": 5.367058823529412, "grad_norm": 1.1154776788554845, "learning_rate": 1.1439048592520256e-05, "loss": 0.8359, "step": 11405 }, { "epoch": 5.367529411764706, "grad_norm": 1.0518971912576343, "learning_rate": 1.1422264337792125e-05, "loss": 0.8008, "step": 11406 }, { "epoch": 5.368, "grad_norm": 1.1600452733137925, "learning_rate": 1.1405491918282333e-05, "loss": 0.4902, "step": 11407 }, { "epoch": 5.368470588235294, "grad_norm": 0.8969604818627903, "learning_rate": 1.1388731335423334e-05, "loss": 0.7812, "step": 11408 }, { "epoch": 5.368941176470588, "grad_norm": 0.765492389269505, "learning_rate": 1.1371982590646584e-05, "loss": 0.3711, "step": 11409 }, { "epoch": 5.369411764705882, "grad_norm": 0.8463906029975384, "learning_rate": 1.135524568538247e-05, "loss": 0.6094, "step": 11410 }, { "epoch": 5.369882352941176, "grad_norm": 0.6725823989810557, "learning_rate": 1.1338520621060415e-05, "loss": 0.0898, "step": 11411 }, { "epoch": 5.3703529411764706, "grad_norm": 0.9303799491223379, "learning_rate": 1.1321807399108845e-05, "loss": 1.1016, "step": 11412 }, { "epoch": 5.370823529411765, "grad_norm": 1.3349655410280714, "learning_rate": 1.1305106020955135e-05, "loss": 1.0234, "step": 11413 }, { "epoch": 5.371294117647059, "grad_norm": 1.037641060291771, "learning_rate": 1.1288416488025675e-05, "loss": 0.3535, "step": 11414 }, { "epoch": 5.371764705882353, "grad_norm": 0.6546555820180296, "learning_rate": 1.1271738801745795e-05, "loss": 0.6875, "step": 11415 }, { "epoch": 5.372235294117647, "grad_norm": 1.287269664178865, "learning_rate": 1.1255072963539885e-05, "loss": 0.75, "step": 11416 }, { "epoch": 5.372705882352941, "grad_norm": 0.9162871238138846, "learning_rate": 1.123841897483131e-05, "loss": 0.3633, "step": 11417 }, { "epoch": 5.373176470588235, "grad_norm": 1.3537400503998387, "learning_rate": 1.1221776837042345e-05, "loss": 0.625, "step": 11418 }, { "epoch": 5.373647058823529, "grad_norm": 0.8308293374980833, "learning_rate": 1.1205146551594357e-05, "loss": 0.7656, "step": 11419 }, { "epoch": 5.374117647058823, "grad_norm": 1.0792620272857931, "learning_rate": 1.1188528119907642e-05, "loss": 0.334, "step": 11420 }, { "epoch": 5.374588235294118, "grad_norm": 0.8399138063187098, "learning_rate": 1.117192154340148e-05, "loss": 0.2041, "step": 11421 }, { "epoch": 5.375058823529412, "grad_norm": 0.809153178804655, "learning_rate": 1.115532682349417e-05, "loss": 0.8633, "step": 11422 }, { "epoch": 5.375529411764706, "grad_norm": 0.9311100923303427, "learning_rate": 1.1138743961602998e-05, "loss": 0.3457, "step": 11423 }, { "epoch": 5.376, "grad_norm": 0.23064688732199198, "learning_rate": 1.1122172959144177e-05, "loss": 0.0105, "step": 11424 }, { "epoch": 5.376470588235295, "grad_norm": 0.8119185642163237, "learning_rate": 1.1105613817532976e-05, "loss": 0.1973, "step": 11425 }, { "epoch": 5.376941176470588, "grad_norm": 1.009024134117757, "learning_rate": 1.108906653818365e-05, "loss": 0.8945, "step": 11426 }, { "epoch": 5.377411764705882, "grad_norm": 1.2747199678546763, "learning_rate": 1.1072531122509381e-05, "loss": 0.8945, "step": 11427 }, { "epoch": 5.377882352941176, "grad_norm": 1.0265640953231407, "learning_rate": 1.1056007571922377e-05, "loss": 0.7617, "step": 11428 }, { "epoch": 5.378352941176471, "grad_norm": 0.9103984380285836, "learning_rate": 1.1039495887833872e-05, "loss": 0.8789, "step": 11429 }, { "epoch": 5.378823529411765, "grad_norm": 1.0293418547736233, "learning_rate": 1.1022996071653988e-05, "loss": 0.7461, "step": 11430 }, { "epoch": 5.379294117647059, "grad_norm": 0.7454542740212045, "learning_rate": 1.1006508124791935e-05, "loss": 0.1768, "step": 11431 }, { "epoch": 5.379764705882353, "grad_norm": 0.8743850375999608, "learning_rate": 1.099003204865585e-05, "loss": 0.7891, "step": 11432 }, { "epoch": 5.3802352941176474, "grad_norm": 1.052646367956787, "learning_rate": 1.0973567844652858e-05, "loss": 0.8242, "step": 11433 }, { "epoch": 5.380705882352941, "grad_norm": 0.8033609403586298, "learning_rate": 1.0957115514189086e-05, "loss": 0.2715, "step": 11434 }, { "epoch": 5.381176470588235, "grad_norm": 1.5835297433544289, "learning_rate": 1.0940675058669673e-05, "loss": 0.8867, "step": 11435 }, { "epoch": 5.381647058823529, "grad_norm": 0.9677785926607759, "learning_rate": 1.0924246479498682e-05, "loss": 0.2539, "step": 11436 }, { "epoch": 5.3821176470588235, "grad_norm": 1.0637036696739801, "learning_rate": 1.0907829778079207e-05, "loss": 1.1016, "step": 11437 }, { "epoch": 5.382588235294118, "grad_norm": 1.0042138112981662, "learning_rate": 1.089142495581331e-05, "loss": 1.4375, "step": 11438 }, { "epoch": 5.383058823529412, "grad_norm": 1.0662865048467924, "learning_rate": 1.0875032014102036e-05, "loss": 0.7305, "step": 11439 }, { "epoch": 5.383529411764706, "grad_norm": 1.564558413644165, "learning_rate": 1.0858650954345433e-05, "loss": 1.3359, "step": 11440 }, { "epoch": 5.384, "grad_norm": 1.1696935535672943, "learning_rate": 1.0842281777942514e-05, "loss": 0.6602, "step": 11441 }, { "epoch": 5.384470588235294, "grad_norm": 0.8856441001878486, "learning_rate": 1.082592448629131e-05, "loss": 0.3184, "step": 11442 }, { "epoch": 5.384941176470588, "grad_norm": 1.2128572826407167, "learning_rate": 1.0809579080788805e-05, "loss": 0.8359, "step": 11443 }, { "epoch": 5.385411764705882, "grad_norm": 0.6119710145874971, "learning_rate": 1.0793245562830966e-05, "loss": 0.1338, "step": 11444 }, { "epoch": 5.385882352941176, "grad_norm": 1.0377987471988779, "learning_rate": 1.0776923933812775e-05, "loss": 0.3398, "step": 11445 }, { "epoch": 5.386352941176471, "grad_norm": 0.8348395994251414, "learning_rate": 1.0760614195128153e-05, "loss": 0.3164, "step": 11446 }, { "epoch": 5.386823529411765, "grad_norm": 2.0377730430589245, "learning_rate": 1.0744316348170035e-05, "loss": 0.7344, "step": 11447 }, { "epoch": 5.387294117647059, "grad_norm": 1.2800939878935926, "learning_rate": 1.0728030394330373e-05, "loss": 0.5312, "step": 11448 }, { "epoch": 5.387764705882353, "grad_norm": 0.9370442369581478, "learning_rate": 1.0711756335000005e-05, "loss": 0.543, "step": 11449 }, { "epoch": 5.3882352941176475, "grad_norm": 1.1514073739059978, "learning_rate": 1.069549417156887e-05, "loss": 0.5547, "step": 11450 }, { "epoch": 5.388705882352941, "grad_norm": 1.1142455089333803, "learning_rate": 1.0679243905425822e-05, "loss": 0.832, "step": 11451 }, { "epoch": 5.389176470588235, "grad_norm": 1.2594325304735658, "learning_rate": 1.0663005537958702e-05, "loss": 0.7188, "step": 11452 }, { "epoch": 5.389647058823529, "grad_norm": 1.13619055390359, "learning_rate": 1.0646779070554335e-05, "loss": 0.9922, "step": 11453 }, { "epoch": 5.3901176470588235, "grad_norm": 1.0309862809809995, "learning_rate": 1.0630564504598593e-05, "loss": 0.6406, "step": 11454 }, { "epoch": 5.390588235294118, "grad_norm": 1.0395655810324984, "learning_rate": 1.0614361841476238e-05, "loss": 0.7188, "step": 11455 }, { "epoch": 5.391058823529412, "grad_norm": 1.2424854096466396, "learning_rate": 1.0598171082571044e-05, "loss": 0.6016, "step": 11456 }, { "epoch": 5.391529411764706, "grad_norm": 0.9089584592431734, "learning_rate": 1.0581992229265823e-05, "loss": 0.625, "step": 11457 }, { "epoch": 5.392, "grad_norm": 1.1454713300106425, "learning_rate": 1.0565825282942303e-05, "loss": 0.9609, "step": 11458 }, { "epoch": 5.392470588235295, "grad_norm": 1.1456421482478487, "learning_rate": 1.0549670244981211e-05, "loss": 0.377, "step": 11459 }, { "epoch": 5.392941176470588, "grad_norm": 1.2683257246581428, "learning_rate": 1.0533527116762296e-05, "loss": 1.2109, "step": 11460 }, { "epoch": 5.393411764705882, "grad_norm": 1.1134843852403538, "learning_rate": 1.0517395899664239e-05, "loss": 0.6641, "step": 11461 }, { "epoch": 5.393882352941176, "grad_norm": 0.8619686834466359, "learning_rate": 1.0501276595064716e-05, "loss": 0.1631, "step": 11462 }, { "epoch": 5.394352941176471, "grad_norm": 0.8930475812986277, "learning_rate": 1.0485169204340432e-05, "loss": 0.5469, "step": 11463 }, { "epoch": 5.394823529411765, "grad_norm": 1.0522812039995633, "learning_rate": 1.0469073728867e-05, "loss": 0.6641, "step": 11464 }, { "epoch": 5.395294117647059, "grad_norm": 0.34885550411573835, "learning_rate": 1.0452990170019088e-05, "loss": 0.0466, "step": 11465 }, { "epoch": 5.395764705882353, "grad_norm": 0.9894085241790758, "learning_rate": 1.0436918529170264e-05, "loss": 0.6602, "step": 11466 }, { "epoch": 5.3962352941176475, "grad_norm": 0.7284950957452478, "learning_rate": 1.0420858807693162e-05, "loss": 0.2871, "step": 11467 }, { "epoch": 5.396705882352941, "grad_norm": 0.8144921060160493, "learning_rate": 1.0404811006959352e-05, "loss": 0.4141, "step": 11468 }, { "epoch": 5.397176470588235, "grad_norm": 1.1342303845190558, "learning_rate": 1.0388775128339389e-05, "loss": 0.9531, "step": 11469 }, { "epoch": 5.397647058823529, "grad_norm": 0.8306599269449723, "learning_rate": 1.037275117320281e-05, "loss": 0.1914, "step": 11470 }, { "epoch": 5.3981176470588235, "grad_norm": 1.2473957616956062, "learning_rate": 1.0356739142918169e-05, "loss": 0.5977, "step": 11471 }, { "epoch": 5.398588235294118, "grad_norm": 1.1267334161216656, "learning_rate": 1.0340739038852924e-05, "loss": 0.875, "step": 11472 }, { "epoch": 5.399058823529412, "grad_norm": 0.49604571813858483, "learning_rate": 1.0324750862373598e-05, "loss": 0.0781, "step": 11473 }, { "epoch": 5.399529411764706, "grad_norm": 1.006314233209326, "learning_rate": 1.0308774614845683e-05, "loss": 0.7852, "step": 11474 }, { "epoch": 5.4, "grad_norm": 0.9654799916986101, "learning_rate": 1.0292810297633569e-05, "loss": 0.3848, "step": 11475 }, { "epoch": 5.400470588235294, "grad_norm": 0.3921060925591494, "learning_rate": 1.02768579121007e-05, "loss": 0.0767, "step": 11476 }, { "epoch": 5.400941176470588, "grad_norm": 1.1956312754522767, "learning_rate": 1.0260917459609535e-05, "loss": 1.0078, "step": 11477 }, { "epoch": 5.401411764705882, "grad_norm": 1.7106767310073956, "learning_rate": 1.024498894152142e-05, "loss": 1.5156, "step": 11478 }, { "epoch": 5.401882352941176, "grad_norm": 1.0580049778063496, "learning_rate": 1.0229072359196749e-05, "loss": 0.6172, "step": 11479 }, { "epoch": 5.402352941176471, "grad_norm": 0.5188440235798321, "learning_rate": 1.0213167713994885e-05, "loss": 0.0986, "step": 11480 }, { "epoch": 5.402823529411765, "grad_norm": 0.9096632698375691, "learning_rate": 1.0197275007274126e-05, "loss": 0.2969, "step": 11481 }, { "epoch": 5.403294117647059, "grad_norm": 0.8276370295573724, "learning_rate": 1.0181394240391798e-05, "loss": 0.2051, "step": 11482 }, { "epoch": 5.403764705882353, "grad_norm": 1.1315449035341527, "learning_rate": 1.016552541470424e-05, "loss": 0.5234, "step": 11483 }, { "epoch": 5.4042352941176475, "grad_norm": 1.093773720518058, "learning_rate": 1.0149668531566662e-05, "loss": 1.0, "step": 11484 }, { "epoch": 5.404705882352941, "grad_norm": 1.1318190208479049, "learning_rate": 1.0133823592333385e-05, "loss": 0.7539, "step": 11485 }, { "epoch": 5.405176470588235, "grad_norm": 1.3782756247954688, "learning_rate": 1.0117990598357589e-05, "loss": 0.2422, "step": 11486 }, { "epoch": 5.405647058823529, "grad_norm": 1.1512925496224715, "learning_rate": 1.0102169550991529e-05, "loss": 0.5586, "step": 11487 }, { "epoch": 5.4061176470588235, "grad_norm": 1.1209907714624314, "learning_rate": 1.008636045158639e-05, "loss": 0.6211, "step": 11488 }, { "epoch": 5.406588235294118, "grad_norm": 0.7514051569009542, "learning_rate": 1.0070563301492323e-05, "loss": 0.5352, "step": 11489 }, { "epoch": 5.407058823529412, "grad_norm": 0.9828410726813687, "learning_rate": 1.0054778102058502e-05, "loss": 0.5625, "step": 11490 }, { "epoch": 5.407529411764706, "grad_norm": 1.2475142881978185, "learning_rate": 1.0039004854633082e-05, "loss": 1.1406, "step": 11491 }, { "epoch": 5.408, "grad_norm": 0.9292306222346189, "learning_rate": 1.0023243560563132e-05, "loss": 0.2715, "step": 11492 }, { "epoch": 5.408470588235295, "grad_norm": 1.4903382095607087, "learning_rate": 1.0007494221194762e-05, "loss": 1.0703, "step": 11493 }, { "epoch": 5.408941176470588, "grad_norm": 0.9303993720668682, "learning_rate": 9.991756837873061e-06, "loss": 0.4863, "step": 11494 }, { "epoch": 5.409411764705882, "grad_norm": 1.1186525002292602, "learning_rate": 9.976031411942054e-06, "loss": 0.4863, "step": 11495 }, { "epoch": 5.409882352941176, "grad_norm": 0.9158557062970839, "learning_rate": 9.960317944744783e-06, "loss": 0.6523, "step": 11496 }, { "epoch": 5.410352941176471, "grad_norm": 1.46886266809845, "learning_rate": 9.944616437623243e-06, "loss": 0.7266, "step": 11497 }, { "epoch": 5.410823529411765, "grad_norm": 1.0526722284610432, "learning_rate": 9.928926891918426e-06, "loss": 0.5586, "step": 11498 }, { "epoch": 5.411294117647059, "grad_norm": 0.7478210702749628, "learning_rate": 9.913249308970312e-06, "loss": 0.1777, "step": 11499 }, { "epoch": 5.411764705882353, "grad_norm": 1.320153196409516, "learning_rate": 9.89758369011781e-06, "loss": 1.2656, "step": 11500 }, { "epoch": 5.412235294117647, "grad_norm": 1.1484962671861554, "learning_rate": 9.881930036698854e-06, "loss": 0.5977, "step": 11501 }, { "epoch": 5.412705882352941, "grad_norm": 1.4704009349159235, "learning_rate": 9.866288350050372e-06, "loss": 1.5391, "step": 11502 }, { "epoch": 5.413176470588235, "grad_norm": 1.106668067259666, "learning_rate": 9.850658631508196e-06, "loss": 0.8555, "step": 11503 }, { "epoch": 5.413647058823529, "grad_norm": 1.3922081371409116, "learning_rate": 9.835040882407207e-06, "loss": 1.0391, "step": 11504 }, { "epoch": 5.4141176470588235, "grad_norm": 1.0770560205626116, "learning_rate": 9.819435104081242e-06, "loss": 0.7734, "step": 11505 }, { "epoch": 5.414588235294118, "grad_norm": 0.8435853536379461, "learning_rate": 9.803841297863096e-06, "loss": 0.3691, "step": 11506 }, { "epoch": 5.415058823529412, "grad_norm": 1.0048637073610307, "learning_rate": 9.788259465084541e-06, "loss": 0.3398, "step": 11507 }, { "epoch": 5.415529411764706, "grad_norm": 1.0710689708104417, "learning_rate": 9.772689607076394e-06, "loss": 0.5156, "step": 11508 }, { "epoch": 5.416, "grad_norm": 1.4153218189636232, "learning_rate": 9.757131725168343e-06, "loss": 0.9727, "step": 11509 }, { "epoch": 5.416470588235294, "grad_norm": 1.3407592989618013, "learning_rate": 9.74158582068914e-06, "loss": 0.5781, "step": 11510 }, { "epoch": 5.416941176470588, "grad_norm": 0.6347170776480977, "learning_rate": 9.726051894966492e-06, "loss": 0.4863, "step": 11511 }, { "epoch": 5.417411764705882, "grad_norm": 1.2393538753415332, "learning_rate": 9.710529949327018e-06, "loss": 1.0234, "step": 11512 }, { "epoch": 5.417882352941176, "grad_norm": 1.2008575661677465, "learning_rate": 9.695019985096409e-06, "loss": 1.0391, "step": 11513 }, { "epoch": 5.418352941176471, "grad_norm": 0.969356135700851, "learning_rate": 9.679522003599305e-06, "loss": 0.7031, "step": 11514 }, { "epoch": 5.418823529411765, "grad_norm": 1.2420992201980054, "learning_rate": 9.664036006159266e-06, "loss": 0.9961, "step": 11515 }, { "epoch": 5.419294117647059, "grad_norm": 1.1387122503116454, "learning_rate": 9.648561994098913e-06, "loss": 0.9219, "step": 11516 }, { "epoch": 5.419764705882353, "grad_norm": 1.0686379548131046, "learning_rate": 9.633099968739777e-06, "loss": 0.5273, "step": 11517 }, { "epoch": 5.4202352941176475, "grad_norm": 0.9783177083949106, "learning_rate": 9.617649931402399e-06, "loss": 0.4844, "step": 11518 }, { "epoch": 5.420705882352941, "grad_norm": 1.1810194463717125, "learning_rate": 9.602211883406308e-06, "loss": 0.6406, "step": 11519 }, { "epoch": 5.421176470588235, "grad_norm": 0.9984341908124544, "learning_rate": 9.586785826069932e-06, "loss": 0.8359, "step": 11520 }, { "epoch": 5.421647058823529, "grad_norm": 0.9724011349046328, "learning_rate": 9.571371760710783e-06, "loss": 0.5586, "step": 11521 }, { "epoch": 5.4221176470588235, "grad_norm": 1.3533698916176424, "learning_rate": 9.55596968864531e-06, "loss": 1.1641, "step": 11522 }, { "epoch": 5.422588235294118, "grad_norm": 0.8172341480623064, "learning_rate": 9.540579611188875e-06, "loss": 0.2734, "step": 11523 }, { "epoch": 5.423058823529412, "grad_norm": 1.1087367151699394, "learning_rate": 9.52520152965588e-06, "loss": 0.6523, "step": 11524 }, { "epoch": 5.423529411764706, "grad_norm": 0.658582816626752, "learning_rate": 9.509835445359736e-06, "loss": 0.1934, "step": 11525 }, { "epoch": 5.424, "grad_norm": 1.168014920500004, "learning_rate": 9.494481359612715e-06, "loss": 0.3691, "step": 11526 }, { "epoch": 5.424470588235294, "grad_norm": 1.1957055357440267, "learning_rate": 9.479139273726166e-06, "loss": 1.0703, "step": 11527 }, { "epoch": 5.424941176470588, "grad_norm": 1.1771206764450188, "learning_rate": 9.46380918901039e-06, "loss": 1.1406, "step": 11528 }, { "epoch": 5.425411764705882, "grad_norm": 1.1040333257297408, "learning_rate": 9.448491106774591e-06, "loss": 0.5859, "step": 11529 }, { "epoch": 5.425882352941176, "grad_norm": 1.6120873608655306, "learning_rate": 9.433185028327106e-06, "loss": 0.5547, "step": 11530 }, { "epoch": 5.426352941176471, "grad_norm": 1.0926824951131222, "learning_rate": 9.417890954975071e-06, "loss": 0.8242, "step": 11531 }, { "epoch": 5.426823529411765, "grad_norm": 1.2777930746268265, "learning_rate": 9.402608888024692e-06, "loss": 1.1094, "step": 11532 }, { "epoch": 5.427294117647059, "grad_norm": 1.2864710888516897, "learning_rate": 9.38733882878116e-06, "loss": 0.9688, "step": 11533 }, { "epoch": 5.427764705882353, "grad_norm": 0.9254170951237299, "learning_rate": 9.37208077854858e-06, "loss": 0.4688, "step": 11534 }, { "epoch": 5.428235294117647, "grad_norm": 1.1095189889991146, "learning_rate": 9.356834738630075e-06, "loss": 1.0391, "step": 11535 }, { "epoch": 5.428705882352941, "grad_norm": 1.0582295921946854, "learning_rate": 9.341600710327773e-06, "loss": 0.6094, "step": 11536 }, { "epoch": 5.429176470588235, "grad_norm": 0.9431034487171807, "learning_rate": 9.32637869494265e-06, "loss": 0.5547, "step": 11537 }, { "epoch": 5.429647058823529, "grad_norm": 1.0453343843530096, "learning_rate": 9.311168693774796e-06, "loss": 0.3496, "step": 11538 }, { "epoch": 5.4301176470588235, "grad_norm": 1.5362574643493896, "learning_rate": 9.295970708123223e-06, "loss": 0.7305, "step": 11539 }, { "epoch": 5.430588235294118, "grad_norm": 1.2280550642129147, "learning_rate": 9.280784739285896e-06, "loss": 0.9453, "step": 11540 }, { "epoch": 5.431058823529412, "grad_norm": 1.0364319933910058, "learning_rate": 9.265610788559758e-06, "loss": 0.6406, "step": 11541 }, { "epoch": 5.431529411764706, "grad_norm": 1.047766227590988, "learning_rate": 9.250448857240789e-06, "loss": 0.668, "step": 11542 }, { "epoch": 5.432, "grad_norm": 1.406660605835763, "learning_rate": 9.235298946623838e-06, "loss": 1.0547, "step": 11543 }, { "epoch": 5.432470588235294, "grad_norm": 0.4157418286930365, "learning_rate": 9.220161058002834e-06, "loss": 0.0767, "step": 11544 }, { "epoch": 5.432941176470588, "grad_norm": 1.1101574671832395, "learning_rate": 9.205035192670563e-06, "loss": 0.8438, "step": 11545 }, { "epoch": 5.433411764705882, "grad_norm": 0.7504883875551848, "learning_rate": 9.189921351918888e-06, "loss": 0.2217, "step": 11546 }, { "epoch": 5.433882352941176, "grad_norm": 1.199140585540056, "learning_rate": 9.174819537038614e-06, "loss": 0.5078, "step": 11547 }, { "epoch": 5.434352941176471, "grad_norm": 1.2990477527179662, "learning_rate": 9.159729749319473e-06, "loss": 0.9883, "step": 11548 }, { "epoch": 5.434823529411765, "grad_norm": 1.0396171204509883, "learning_rate": 9.144651990050234e-06, "loss": 0.8086, "step": 11549 }, { "epoch": 5.435294117647059, "grad_norm": 1.0864290777809584, "learning_rate": 9.129586260518634e-06, "loss": 0.8438, "step": 11550 }, { "epoch": 5.435764705882353, "grad_norm": 1.0370461762028125, "learning_rate": 9.114532562011312e-06, "loss": 0.5078, "step": 11551 }, { "epoch": 5.4362352941176475, "grad_norm": 0.814165642128028, "learning_rate": 9.099490895813938e-06, "loss": 0.2412, "step": 11552 }, { "epoch": 5.436705882352941, "grad_norm": 1.213261602337837, "learning_rate": 9.084461263211168e-06, "loss": 0.75, "step": 11553 }, { "epoch": 5.437176470588235, "grad_norm": 1.0233965444044302, "learning_rate": 9.069443665486593e-06, "loss": 0.8125, "step": 11554 }, { "epoch": 5.437647058823529, "grad_norm": 0.7765235699671498, "learning_rate": 9.054438103922767e-06, "loss": 0.3027, "step": 11555 }, { "epoch": 5.4381176470588235, "grad_norm": 0.9414419230898761, "learning_rate": 9.039444579801302e-06, "loss": 0.6719, "step": 11556 }, { "epoch": 5.438588235294118, "grad_norm": 0.9061635850766585, "learning_rate": 9.024463094402656e-06, "loss": 0.5352, "step": 11557 }, { "epoch": 5.439058823529412, "grad_norm": 1.0316697045059935, "learning_rate": 9.009493649006322e-06, "loss": 0.6172, "step": 11558 }, { "epoch": 5.439529411764706, "grad_norm": 1.2165447870606447, "learning_rate": 8.994536244890827e-06, "loss": 1.0781, "step": 11559 }, { "epoch": 5.44, "grad_norm": 1.4665397817666566, "learning_rate": 8.979590883333532e-06, "loss": 0.75, "step": 11560 }, { "epoch": 5.440470588235294, "grad_norm": 0.7815140504589932, "learning_rate": 8.964657565610884e-06, "loss": 0.5234, "step": 11561 }, { "epoch": 5.440941176470588, "grad_norm": 0.7602776881498439, "learning_rate": 8.949736292998277e-06, "loss": 0.1719, "step": 11562 }, { "epoch": 5.441411764705882, "grad_norm": 0.8666389930206737, "learning_rate": 8.934827066770012e-06, "loss": 0.2451, "step": 11563 }, { "epoch": 5.441882352941176, "grad_norm": 1.059560498958582, "learning_rate": 8.919929888199435e-06, "loss": 0.8594, "step": 11564 }, { "epoch": 5.442352941176471, "grad_norm": 1.2447941160666882, "learning_rate": 8.90504475855886e-06, "loss": 0.4922, "step": 11565 }, { "epoch": 5.442823529411765, "grad_norm": 0.454670305754251, "learning_rate": 8.890171679119506e-06, "loss": 0.0635, "step": 11566 }, { "epoch": 5.443294117647059, "grad_norm": 1.064892488284783, "learning_rate": 8.875310651151657e-06, "loss": 0.5742, "step": 11567 }, { "epoch": 5.443764705882353, "grad_norm": 1.3539717926903825, "learning_rate": 8.86046167592448e-06, "loss": 1.4219, "step": 11568 }, { "epoch": 5.444235294117647, "grad_norm": 0.9624563032124496, "learning_rate": 8.845624754706144e-06, "loss": 0.5273, "step": 11569 }, { "epoch": 5.444705882352941, "grad_norm": 1.7069648341202608, "learning_rate": 8.830799888763834e-06, "loss": 1.0859, "step": 11570 }, { "epoch": 5.445176470588235, "grad_norm": 0.5852183477102447, "learning_rate": 8.81598707936364e-06, "loss": 0.085, "step": 11571 }, { "epoch": 5.445647058823529, "grad_norm": 1.1821041262021492, "learning_rate": 8.801186327770632e-06, "loss": 0.793, "step": 11572 }, { "epoch": 5.4461176470588235, "grad_norm": 1.0603684824420576, "learning_rate": 8.7863976352489e-06, "loss": 1.4922, "step": 11573 }, { "epoch": 5.446588235294118, "grad_norm": 1.0123579961136033, "learning_rate": 8.771621003061463e-06, "loss": 0.8477, "step": 11574 }, { "epoch": 5.447058823529412, "grad_norm": 1.1219724451837307, "learning_rate": 8.756856432470315e-06, "loss": 0.5859, "step": 11575 }, { "epoch": 5.447529411764706, "grad_norm": 0.43235374181443287, "learning_rate": 8.742103924736416e-06, "loss": 0.3398, "step": 11576 }, { "epoch": 5.448, "grad_norm": 0.7372008809158627, "learning_rate": 8.727363481119704e-06, "loss": 0.1875, "step": 11577 }, { "epoch": 5.448470588235294, "grad_norm": 1.2080213312172123, "learning_rate": 8.712635102879124e-06, "loss": 1.5469, "step": 11578 }, { "epoch": 5.448941176470588, "grad_norm": 0.7836982188808195, "learning_rate": 8.697918791272485e-06, "loss": 0.2168, "step": 11579 }, { "epoch": 5.449411764705882, "grad_norm": 1.3122661939958375, "learning_rate": 8.68321454755665e-06, "loss": 0.7773, "step": 11580 }, { "epoch": 5.449882352941176, "grad_norm": 0.9475992833422758, "learning_rate": 8.668522372987479e-06, "loss": 0.3281, "step": 11581 }, { "epoch": 5.450352941176471, "grad_norm": 0.9780991386634188, "learning_rate": 8.653842268819723e-06, "loss": 0.6094, "step": 11582 }, { "epoch": 5.450823529411765, "grad_norm": 1.1363885731376537, "learning_rate": 8.639174236307106e-06, "loss": 0.582, "step": 11583 }, { "epoch": 5.451294117647059, "grad_norm": 1.0157906120866034, "learning_rate": 8.624518276702414e-06, "loss": 0.6758, "step": 11584 }, { "epoch": 5.451764705882353, "grad_norm": 1.0506029447457819, "learning_rate": 8.609874391257293e-06, "loss": 0.7266, "step": 11585 }, { "epoch": 5.4522352941176475, "grad_norm": 0.8435018614322549, "learning_rate": 8.595242581222394e-06, "loss": 0.2754, "step": 11586 }, { "epoch": 5.452705882352941, "grad_norm": 0.9349507474434202, "learning_rate": 8.580622847847402e-06, "loss": 0.3105, "step": 11587 }, { "epoch": 5.453176470588235, "grad_norm": 0.9577947135522659, "learning_rate": 8.566015192380849e-06, "loss": 0.3281, "step": 11588 }, { "epoch": 5.453647058823529, "grad_norm": 0.7689186655980701, "learning_rate": 8.551419616070322e-06, "loss": 1.1562, "step": 11589 }, { "epoch": 5.4541176470588235, "grad_norm": 1.524708982740994, "learning_rate": 8.536836120162389e-06, "loss": 0.7969, "step": 11590 }, { "epoch": 5.454588235294118, "grad_norm": 1.0443415836212584, "learning_rate": 8.522264705902504e-06, "loss": 0.793, "step": 11591 }, { "epoch": 5.455058823529412, "grad_norm": 1.0737368521290154, "learning_rate": 8.507705374535157e-06, "loss": 0.9141, "step": 11592 }, { "epoch": 5.455529411764706, "grad_norm": 1.0022729167910658, "learning_rate": 8.4931581273038e-06, "loss": 0.7305, "step": 11593 }, { "epoch": 5.456, "grad_norm": 1.0922988090233103, "learning_rate": 8.47862296545081e-06, "loss": 0.918, "step": 11594 }, { "epoch": 5.456470588235294, "grad_norm": 1.0912815427330926, "learning_rate": 8.464099890217607e-06, "loss": 0.918, "step": 11595 }, { "epoch": 5.456941176470588, "grad_norm": 0.7231541629857987, "learning_rate": 8.449588902844484e-06, "loss": 0.2539, "step": 11596 }, { "epoch": 5.457411764705882, "grad_norm": 0.8530965877566611, "learning_rate": 8.43509000457075e-06, "loss": 0.3047, "step": 11597 }, { "epoch": 5.457882352941176, "grad_norm": 1.0633008955509666, "learning_rate": 8.420603196634745e-06, "loss": 0.4102, "step": 11598 }, { "epoch": 5.458352941176471, "grad_norm": 1.8041216303460135, "learning_rate": 8.406128480273633e-06, "loss": 1.7812, "step": 11599 }, { "epoch": 5.458823529411765, "grad_norm": 0.7171315616774326, "learning_rate": 8.391665856723655e-06, "loss": 0.2559, "step": 11600 }, { "epoch": 5.459294117647059, "grad_norm": 1.3685986212135217, "learning_rate": 8.377215327220043e-06, "loss": 0.5352, "step": 11601 }, { "epoch": 5.459764705882353, "grad_norm": 0.9337858699523297, "learning_rate": 8.362776892996858e-06, "loss": 0.8047, "step": 11602 }, { "epoch": 5.460235294117647, "grad_norm": 1.1080111221551172, "learning_rate": 8.348350555287264e-06, "loss": 0.543, "step": 11603 }, { "epoch": 5.460705882352941, "grad_norm": 0.5523464798732701, "learning_rate": 8.333936315323343e-06, "loss": 0.1147, "step": 11604 }, { "epoch": 5.461176470588235, "grad_norm": 0.7374288405007496, "learning_rate": 8.319534174336106e-06, "loss": 0.2002, "step": 11605 }, { "epoch": 5.461647058823529, "grad_norm": 0.7897831887069506, "learning_rate": 8.305144133555608e-06, "loss": 1.0625, "step": 11606 }, { "epoch": 5.4621176470588235, "grad_norm": 0.9441892409016647, "learning_rate": 8.290766194210813e-06, "loss": 0.6172, "step": 11607 }, { "epoch": 5.462588235294118, "grad_norm": 1.2282944644147769, "learning_rate": 8.276400357529655e-06, "loss": 0.5664, "step": 11608 }, { "epoch": 5.463058823529412, "grad_norm": 0.7145658951428809, "learning_rate": 8.262046624739038e-06, "loss": 0.4355, "step": 11609 }, { "epoch": 5.463529411764706, "grad_norm": 0.8990016575153635, "learning_rate": 8.247704997064896e-06, "loss": 0.5078, "step": 11610 }, { "epoch": 5.464, "grad_norm": 1.0894438491188516, "learning_rate": 8.233375475732018e-06, "loss": 0.9844, "step": 11611 }, { "epoch": 5.464470588235294, "grad_norm": 0.9657781167331639, "learning_rate": 8.21905806196424e-06, "loss": 1.1562, "step": 11612 }, { "epoch": 5.464941176470588, "grad_norm": 1.0810443432267436, "learning_rate": 8.204752756984351e-06, "loss": 0.582, "step": 11613 }, { "epoch": 5.465411764705882, "grad_norm": 1.1944489886160377, "learning_rate": 8.19045956201404e-06, "loss": 1.1172, "step": 11614 }, { "epoch": 5.465882352941176, "grad_norm": 1.1507631321349807, "learning_rate": 8.176178478274082e-06, "loss": 0.5117, "step": 11615 }, { "epoch": 5.466352941176471, "grad_norm": 0.9125904109695661, "learning_rate": 8.161909506984132e-06, "loss": 0.3945, "step": 11616 }, { "epoch": 5.466823529411765, "grad_norm": 0.9622977556164469, "learning_rate": 8.147652649362802e-06, "loss": 0.416, "step": 11617 }, { "epoch": 5.467294117647059, "grad_norm": 0.9055989191395077, "learning_rate": 8.133407906627715e-06, "loss": 0.1807, "step": 11618 }, { "epoch": 5.467764705882353, "grad_norm": 1.4210554611308228, "learning_rate": 8.119175279995448e-06, "loss": 0.9883, "step": 11619 }, { "epoch": 5.4682352941176475, "grad_norm": 1.2045319947848252, "learning_rate": 8.104954770681532e-06, "loss": 0.8711, "step": 11620 }, { "epoch": 5.468705882352941, "grad_norm": 0.7262572729889267, "learning_rate": 8.090746379900475e-06, "loss": 0.332, "step": 11621 }, { "epoch": 5.469176470588235, "grad_norm": 0.7910006985290665, "learning_rate": 8.076550108865743e-06, "loss": 0.2891, "step": 11622 }, { "epoch": 5.469647058823529, "grad_norm": 1.2125170650723402, "learning_rate": 8.062365958789746e-06, "loss": 0.7656, "step": 11623 }, { "epoch": 5.4701176470588235, "grad_norm": 0.9356083500210474, "learning_rate": 8.048193930883933e-06, "loss": 0.9766, "step": 11624 }, { "epoch": 5.470588235294118, "grad_norm": 2.4247063731660243, "learning_rate": 8.034034026358587e-06, "loss": 0.9492, "step": 11625 }, { "epoch": 5.471058823529412, "grad_norm": 1.1400139011901695, "learning_rate": 8.019886246423107e-06, "loss": 0.6172, "step": 11626 }, { "epoch": 5.471529411764706, "grad_norm": 1.579880132161384, "learning_rate": 8.005750592285725e-06, "loss": 1.8828, "step": 11627 }, { "epoch": 5.4719999999999995, "grad_norm": 1.182791385867576, "learning_rate": 7.991627065153727e-06, "loss": 0.875, "step": 11628 }, { "epoch": 5.472470588235294, "grad_norm": 1.4209906340411467, "learning_rate": 7.977515666233347e-06, "loss": 0.4824, "step": 11629 }, { "epoch": 5.472941176470588, "grad_norm": 0.9631042476171182, "learning_rate": 7.963416396729722e-06, "loss": 0.9844, "step": 11630 }, { "epoch": 5.473411764705882, "grad_norm": 0.7448879385312126, "learning_rate": 7.94932925784702e-06, "loss": 0.0996, "step": 11631 }, { "epoch": 5.473882352941176, "grad_norm": 0.8128615924696151, "learning_rate": 7.935254250788364e-06, "loss": 0.3438, "step": 11632 }, { "epoch": 5.474352941176471, "grad_norm": 0.8147664232641263, "learning_rate": 7.921191376755808e-06, "loss": 0.249, "step": 11633 }, { "epoch": 5.474823529411765, "grad_norm": 0.8205553594072555, "learning_rate": 7.90714063695041e-06, "loss": 0.8203, "step": 11634 }, { "epoch": 5.475294117647059, "grad_norm": 1.084500468930776, "learning_rate": 7.893102032572174e-06, "loss": 0.7969, "step": 11635 }, { "epoch": 5.475764705882353, "grad_norm": 1.1191532992981805, "learning_rate": 7.879075564820025e-06, "loss": 0.4961, "step": 11636 }, { "epoch": 5.476235294117647, "grad_norm": 1.106451615814787, "learning_rate": 7.86506123489194e-06, "loss": 0.8438, "step": 11637 }, { "epoch": 5.476705882352941, "grad_norm": 1.0222575372258822, "learning_rate": 7.85105904398481e-06, "loss": 0.7383, "step": 11638 }, { "epoch": 5.477176470588235, "grad_norm": 0.9185671629445913, "learning_rate": 7.837068993294443e-06, "loss": 0.6016, "step": 11639 }, { "epoch": 5.477647058823529, "grad_norm": 1.0874527122654754, "learning_rate": 7.823091084015687e-06, "loss": 0.4512, "step": 11640 }, { "epoch": 5.4781176470588235, "grad_norm": 1.2485664258295954, "learning_rate": 7.809125317342369e-06, "loss": 0.6875, "step": 11641 }, { "epoch": 5.478588235294118, "grad_norm": 1.3660743350612656, "learning_rate": 7.79517169446715e-06, "loss": 1.3281, "step": 11642 }, { "epoch": 5.479058823529412, "grad_norm": 0.5730243394978586, "learning_rate": 7.781230216581779e-06, "loss": 0.2539, "step": 11643 }, { "epoch": 5.479529411764706, "grad_norm": 0.9509788247834486, "learning_rate": 7.767300884876971e-06, "loss": 0.2988, "step": 11644 }, { "epoch": 5.48, "grad_norm": 1.3857952509410576, "learning_rate": 7.753383700542271e-06, "loss": 1.5547, "step": 11645 }, { "epoch": 5.480470588235294, "grad_norm": 0.8687615610055092, "learning_rate": 7.739478664766347e-06, "loss": 0.5352, "step": 11646 }, { "epoch": 5.480941176470588, "grad_norm": 1.3512801830116055, "learning_rate": 7.725585778736698e-06, "loss": 1.0859, "step": 11647 }, { "epoch": 5.481411764705882, "grad_norm": 0.8984346998610049, "learning_rate": 7.711705043639893e-06, "loss": 0.377, "step": 11648 }, { "epoch": 5.481882352941176, "grad_norm": 0.9857460776025038, "learning_rate": 7.697836460661399e-06, "loss": 0.7344, "step": 11649 }, { "epoch": 5.482352941176471, "grad_norm": 0.8165214028392251, "learning_rate": 7.683980030985654e-06, "loss": 0.2754, "step": 11650 }, { "epoch": 5.482823529411765, "grad_norm": 1.0077507022767382, "learning_rate": 7.670135755796074e-06, "loss": 0.7266, "step": 11651 }, { "epoch": 5.483294117647059, "grad_norm": 1.0144986333347505, "learning_rate": 7.656303636275036e-06, "loss": 0.832, "step": 11652 }, { "epoch": 5.483764705882353, "grad_norm": 0.9681764919801074, "learning_rate": 7.64248367360384e-06, "loss": 0.9453, "step": 11653 }, { "epoch": 5.484235294117647, "grad_norm": 1.2573660123491273, "learning_rate": 7.6286758689627915e-06, "loss": 1.0859, "step": 11654 }, { "epoch": 5.484705882352941, "grad_norm": 1.1701866013439737, "learning_rate": 7.614880223531183e-06, "loss": 0.7461, "step": 11655 }, { "epoch": 5.485176470588235, "grad_norm": 1.0193093258200803, "learning_rate": 7.6010967384871725e-06, "loss": 0.6172, "step": 11656 }, { "epoch": 5.485647058823529, "grad_norm": 0.9160315605471492, "learning_rate": 7.587325415007967e-06, "loss": 0.291, "step": 11657 }, { "epoch": 5.4861176470588235, "grad_norm": 0.8482623016978219, "learning_rate": 7.573566254269708e-06, "loss": 0.4688, "step": 11658 }, { "epoch": 5.486588235294118, "grad_norm": 1.2292708458624166, "learning_rate": 7.559819257447492e-06, "loss": 0.8633, "step": 11659 }, { "epoch": 5.487058823529412, "grad_norm": 1.1237551118015197, "learning_rate": 7.546084425715359e-06, "loss": 0.582, "step": 11660 }, { "epoch": 5.487529411764706, "grad_norm": 1.1165372783155738, "learning_rate": 7.532361760246375e-06, "loss": 0.9102, "step": 11661 }, { "epoch": 5.4879999999999995, "grad_norm": 1.1885343453738377, "learning_rate": 7.5186512622124495e-06, "loss": 0.8594, "step": 11662 }, { "epoch": 5.488470588235294, "grad_norm": 0.9616134500320915, "learning_rate": 7.5049529327846135e-06, "loss": 0.4629, "step": 11663 }, { "epoch": 5.488941176470588, "grad_norm": 1.5448593111396143, "learning_rate": 7.491266773132715e-06, "loss": 0.3223, "step": 11664 }, { "epoch": 5.489411764705882, "grad_norm": 1.3797268510100371, "learning_rate": 7.477592784425634e-06, "loss": 1.3594, "step": 11665 }, { "epoch": 5.489882352941176, "grad_norm": 1.6638013080160894, "learning_rate": 7.463930967831222e-06, "loss": 0.875, "step": 11666 }, { "epoch": 5.490352941176471, "grad_norm": 1.0603641245731765, "learning_rate": 7.450281324516228e-06, "loss": 0.4336, "step": 11667 }, { "epoch": 5.490823529411765, "grad_norm": 0.7726854873980152, "learning_rate": 7.436643855646402e-06, "loss": 0.2324, "step": 11668 }, { "epoch": 5.491294117647059, "grad_norm": 1.1979875994498728, "learning_rate": 7.423018562386479e-06, "loss": 0.8789, "step": 11669 }, { "epoch": 5.491764705882353, "grad_norm": 1.3093050422494803, "learning_rate": 7.40940544590008e-06, "loss": 0.4688, "step": 11670 }, { "epoch": 5.492235294117647, "grad_norm": 1.5797675819126944, "learning_rate": 7.395804507349873e-06, "loss": 1.2344, "step": 11671 }, { "epoch": 5.492705882352941, "grad_norm": 1.2044961023219432, "learning_rate": 7.3822157478974465e-06, "loss": 0.5586, "step": 11672 }, { "epoch": 5.493176470588235, "grad_norm": 0.6920675035445674, "learning_rate": 7.368639168703306e-06, "loss": 0.1133, "step": 11673 }, { "epoch": 5.493647058823529, "grad_norm": 0.9055732644512251, "learning_rate": 7.3550747709269885e-06, "loss": 0.2539, "step": 11674 }, { "epoch": 5.4941176470588236, "grad_norm": 1.4752031717438783, "learning_rate": 7.34152255572697e-06, "loss": 0.9141, "step": 11675 }, { "epoch": 5.494588235294118, "grad_norm": 0.8487215848248865, "learning_rate": 7.327982524260639e-06, "loss": 0.1177, "step": 11676 }, { "epoch": 5.495058823529412, "grad_norm": 1.1138497530686005, "learning_rate": 7.314454677684439e-06, "loss": 0.5508, "step": 11677 }, { "epoch": 5.495529411764706, "grad_norm": 1.176888559016679, "learning_rate": 7.300939017153645e-06, "loss": 1.0547, "step": 11678 }, { "epoch": 5.496, "grad_norm": 1.5523550271795512, "learning_rate": 7.287435543822584e-06, "loss": 1.4375, "step": 11679 }, { "epoch": 5.496470588235294, "grad_norm": 1.0213493558284978, "learning_rate": 7.2739442588445666e-06, "loss": 0.7656, "step": 11680 }, { "epoch": 5.496941176470588, "grad_norm": 0.8012049675915024, "learning_rate": 7.260465163371754e-06, "loss": 0.2236, "step": 11681 }, { "epoch": 5.497411764705882, "grad_norm": 1.467363073255128, "learning_rate": 7.246998258555342e-06, "loss": 1.4844, "step": 11682 }, { "epoch": 5.497882352941176, "grad_norm": 1.0815126247291103, "learning_rate": 7.233543545545511e-06, "loss": 0.5859, "step": 11683 }, { "epoch": 5.498352941176471, "grad_norm": 1.0797408515941354, "learning_rate": 7.220101025491293e-06, "loss": 1.0078, "step": 11684 }, { "epoch": 5.498823529411765, "grad_norm": 1.3197681710511373, "learning_rate": 7.2066706995407845e-06, "loss": 1.0781, "step": 11685 }, { "epoch": 5.499294117647059, "grad_norm": 0.842257958462761, "learning_rate": 7.193252568841018e-06, "loss": 1.3047, "step": 11686 }, { "epoch": 5.499764705882353, "grad_norm": 0.8970160166633134, "learning_rate": 7.1798466345379115e-06, "loss": 0.2637, "step": 11687 }, { "epoch": 5.5002352941176476, "grad_norm": 1.0276081165103985, "learning_rate": 7.166452897776448e-06, "loss": 0.6328, "step": 11688 }, { "epoch": 5.500705882352941, "grad_norm": 0.8639301222683479, "learning_rate": 7.153071359700513e-06, "loss": 0.3691, "step": 11689 }, { "epoch": 5.501176470588235, "grad_norm": 1.438239014862573, "learning_rate": 7.1397020214529255e-06, "loss": 1.0859, "step": 11690 }, { "epoch": 5.501647058823529, "grad_norm": 1.0174957753585248, "learning_rate": 7.126344884175522e-06, "loss": 0.6172, "step": 11691 }, { "epoch": 5.502117647058824, "grad_norm": 0.8582804295704474, "learning_rate": 7.1129999490090735e-06, "loss": 0.3906, "step": 11692 }, { "epoch": 5.502588235294118, "grad_norm": 0.9400397008657483, "learning_rate": 7.099667217093252e-06, "loss": 0.4473, "step": 11693 }, { "epoch": 5.503058823529412, "grad_norm": 0.6643777312165108, "learning_rate": 7.08634668956678e-06, "loss": 0.126, "step": 11694 }, { "epoch": 5.503529411764706, "grad_norm": 1.31016471401536, "learning_rate": 7.073038367567313e-06, "loss": 0.7773, "step": 11695 }, { "epoch": 5.504, "grad_norm": 1.0481623786821823, "learning_rate": 7.059742252231393e-06, "loss": 0.6836, "step": 11696 }, { "epoch": 5.504470588235294, "grad_norm": 0.7916402741802571, "learning_rate": 7.046458344694628e-06, "loss": 0.2314, "step": 11697 }, { "epoch": 5.504941176470588, "grad_norm": 0.9135855072492554, "learning_rate": 7.033186646091477e-06, "loss": 0.7109, "step": 11698 }, { "epoch": 5.505411764705882, "grad_norm": 1.237469438384039, "learning_rate": 7.019927157555449e-06, "loss": 0.5703, "step": 11699 }, { "epoch": 5.5058823529411764, "grad_norm": 0.46228162596602207, "learning_rate": 7.006679880218974e-06, "loss": 0.1289, "step": 11700 }, { "epoch": 5.506352941176471, "grad_norm": 1.58097614249117, "learning_rate": 6.993444815213378e-06, "loss": 1.2422, "step": 11701 }, { "epoch": 5.506823529411765, "grad_norm": 1.0377775963071183, "learning_rate": 6.9802219636690585e-06, "loss": 0.8789, "step": 11702 }, { "epoch": 5.507294117647059, "grad_norm": 0.5154380016154892, "learning_rate": 6.967011326715294e-06, "loss": 0.0859, "step": 11703 }, { "epoch": 5.507764705882353, "grad_norm": 0.7877069105719343, "learning_rate": 6.953812905480316e-06, "loss": 0.1758, "step": 11704 }, { "epoch": 5.508235294117647, "grad_norm": 0.8645520831504376, "learning_rate": 6.940626701091356e-06, "loss": 0.2412, "step": 11705 }, { "epoch": 5.508705882352941, "grad_norm": 1.3726001062270299, "learning_rate": 6.927452714674614e-06, "loss": 0.7773, "step": 11706 }, { "epoch": 5.509176470588235, "grad_norm": 1.0750068043278258, "learning_rate": 6.914290947355122e-06, "loss": 0.6992, "step": 11707 }, { "epoch": 5.509647058823529, "grad_norm": 0.8516976862930772, "learning_rate": 6.901141400257048e-06, "loss": 0.2363, "step": 11708 }, { "epoch": 5.510117647058824, "grad_norm": 0.8681163136665727, "learning_rate": 6.888004074503395e-06, "loss": 0.3887, "step": 11709 }, { "epoch": 5.510588235294118, "grad_norm": 1.1829611502775945, "learning_rate": 6.87487897121613e-06, "loss": 0.4883, "step": 11710 }, { "epoch": 5.511058823529412, "grad_norm": 1.0164078378113757, "learning_rate": 6.861766091516274e-06, "loss": 0.8555, "step": 11711 }, { "epoch": 5.511529411764706, "grad_norm": 0.7413055310358646, "learning_rate": 6.848665436523631e-06, "loss": 0.2988, "step": 11712 }, { "epoch": 5.5120000000000005, "grad_norm": 0.8588732534800458, "learning_rate": 6.83557700735714e-06, "loss": 0.3535, "step": 11713 }, { "epoch": 5.512470588235294, "grad_norm": 1.2545107190124256, "learning_rate": 6.82250080513459e-06, "loss": 0.9062, "step": 11714 }, { "epoch": 5.512941176470588, "grad_norm": 0.615071871215332, "learning_rate": 6.809436830972753e-06, "loss": 0.1338, "step": 11715 }, { "epoch": 5.513411764705882, "grad_norm": 0.8905419715868509, "learning_rate": 6.796385085987338e-06, "loss": 0.5703, "step": 11716 }, { "epoch": 5.5138823529411765, "grad_norm": 0.42620801669649516, "learning_rate": 6.783345571293069e-06, "loss": 0.1738, "step": 11717 }, { "epoch": 5.514352941176471, "grad_norm": 1.3791176553131081, "learning_rate": 6.770318288003556e-06, "loss": 1.5156, "step": 11718 }, { "epoch": 5.514823529411765, "grad_norm": 1.3075599068282426, "learning_rate": 6.757303237231393e-06, "loss": 0.7031, "step": 11719 }, { "epoch": 5.515294117647059, "grad_norm": 0.963946149802394, "learning_rate": 6.744300420088156e-06, "loss": 0.3242, "step": 11720 }, { "epoch": 5.5157647058823525, "grad_norm": 1.3152659987078839, "learning_rate": 6.73130983768429e-06, "loss": 0.918, "step": 11721 }, { "epoch": 5.516235294117647, "grad_norm": 1.0351813787280835, "learning_rate": 6.718331491129309e-06, "loss": 0.5156, "step": 11722 }, { "epoch": 5.516705882352941, "grad_norm": 1.3207710896244362, "learning_rate": 6.705365381531641e-06, "loss": 0.9492, "step": 11723 }, { "epoch": 5.517176470588235, "grad_norm": 0.800210597988851, "learning_rate": 6.6924115099985856e-06, "loss": 0.2168, "step": 11724 }, { "epoch": 5.517647058823529, "grad_norm": 0.41801062715503845, "learning_rate": 6.679469877636523e-06, "loss": 0.0986, "step": 11725 }, { "epoch": 5.518117647058824, "grad_norm": 1.0251303799615101, "learning_rate": 6.6665404855507195e-06, "loss": 0.4434, "step": 11726 }, { "epoch": 5.518588235294118, "grad_norm": 1.1270820060750604, "learning_rate": 6.6536233348453925e-06, "loss": 0.7969, "step": 11727 }, { "epoch": 5.519058823529412, "grad_norm": 1.259449819958065, "learning_rate": 6.640718426623759e-06, "loss": 0.5938, "step": 11728 }, { "epoch": 5.519529411764706, "grad_norm": 0.5430523207353318, "learning_rate": 6.627825761987937e-06, "loss": 0.1055, "step": 11729 }, { "epoch": 5.52, "grad_norm": 1.4085996831173273, "learning_rate": 6.61494534203903e-06, "loss": 1.0234, "step": 11730 }, { "epoch": 5.520470588235294, "grad_norm": 1.247765502223723, "learning_rate": 6.6020771678771066e-06, "loss": 1.3359, "step": 11731 }, { "epoch": 5.520941176470588, "grad_norm": 0.9526321344052657, "learning_rate": 6.589221240601123e-06, "loss": 0.3223, "step": 11732 }, { "epoch": 5.521411764705882, "grad_norm": 1.031951651354125, "learning_rate": 6.576377561309082e-06, "loss": 0.5508, "step": 11733 }, { "epoch": 5.5218823529411765, "grad_norm": 0.8091285900625516, "learning_rate": 6.563546131097907e-06, "loss": 0.2148, "step": 11734 }, { "epoch": 5.522352941176471, "grad_norm": 0.8079391966588486, "learning_rate": 6.550726951063423e-06, "loss": 0.6484, "step": 11735 }, { "epoch": 5.522823529411765, "grad_norm": 0.7986719037329604, "learning_rate": 6.5379200223004855e-06, "loss": 0.4043, "step": 11736 }, { "epoch": 5.523294117647059, "grad_norm": 1.0431913248041533, "learning_rate": 6.5251253459028545e-06, "loss": 0.5391, "step": 11737 }, { "epoch": 5.523764705882353, "grad_norm": 1.5392135498325032, "learning_rate": 6.512342922963254e-06, "loss": 1.4688, "step": 11738 }, { "epoch": 5.524235294117647, "grad_norm": 0.6858386311590727, "learning_rate": 6.499572754573362e-06, "loss": 0.1992, "step": 11739 }, { "epoch": 5.524705882352941, "grad_norm": 1.1423824619320673, "learning_rate": 6.48681484182384e-06, "loss": 0.918, "step": 11740 }, { "epoch": 5.525176470588235, "grad_norm": 0.8382551846037606, "learning_rate": 6.474069185804248e-06, "loss": 0.2373, "step": 11741 }, { "epoch": 5.525647058823529, "grad_norm": 1.2914422870996762, "learning_rate": 6.461335787603134e-06, "loss": 0.4648, "step": 11742 }, { "epoch": 5.526117647058824, "grad_norm": 1.0235383283049764, "learning_rate": 6.448614648308009e-06, "loss": 0.4004, "step": 11743 }, { "epoch": 5.526588235294118, "grad_norm": 0.9536636462080695, "learning_rate": 6.435905769005306e-06, "loss": 0.3145, "step": 11744 }, { "epoch": 5.527058823529412, "grad_norm": 1.0411985992262647, "learning_rate": 6.423209150780423e-06, "loss": 0.6211, "step": 11745 }, { "epoch": 5.527529411764706, "grad_norm": 1.023953471896566, "learning_rate": 6.4105247947177256e-06, "loss": 0.4023, "step": 11746 }, { "epoch": 5.5280000000000005, "grad_norm": 0.643184259177599, "learning_rate": 6.3978527019004964e-06, "loss": 0.1758, "step": 11747 }, { "epoch": 5.528470588235294, "grad_norm": 1.2008187833235318, "learning_rate": 6.385192873411038e-06, "loss": 0.7461, "step": 11748 }, { "epoch": 5.528941176470588, "grad_norm": 1.156183710122074, "learning_rate": 6.372545310330518e-06, "loss": 0.5703, "step": 11749 }, { "epoch": 5.529411764705882, "grad_norm": 1.0897132419876192, "learning_rate": 6.359910013739122e-06, "loss": 1.1094, "step": 11750 }, { "epoch": 5.5298823529411765, "grad_norm": 0.9077465167591221, "learning_rate": 6.347286984715938e-06, "loss": 0.3906, "step": 11751 }, { "epoch": 5.530352941176471, "grad_norm": 1.0868078886640302, "learning_rate": 6.334676224339086e-06, "loss": 0.5898, "step": 11752 }, { "epoch": 5.530823529411765, "grad_norm": 1.2878710398938265, "learning_rate": 6.322077733685538e-06, "loss": 0.707, "step": 11753 }, { "epoch": 5.531294117647059, "grad_norm": 1.1509129248962406, "learning_rate": 6.309491513831317e-06, "loss": 0.9609, "step": 11754 }, { "epoch": 5.5317647058823525, "grad_norm": 0.7329372827964636, "learning_rate": 6.296917565851295e-06, "loss": 0.3516, "step": 11755 }, { "epoch": 5.532235294117647, "grad_norm": 1.4243223108110048, "learning_rate": 6.284355890819398e-06, "loss": 1.3203, "step": 11756 }, { "epoch": 5.532705882352941, "grad_norm": 1.0734078174316786, "learning_rate": 6.271806489808401e-06, "loss": 0.4453, "step": 11757 }, { "epoch": 5.533176470588235, "grad_norm": 1.2223849212684692, "learning_rate": 6.2592693638901125e-06, "loss": 1.3281, "step": 11758 }, { "epoch": 5.533647058823529, "grad_norm": 0.9902933927273981, "learning_rate": 6.246744514135293e-06, "loss": 0.6758, "step": 11759 }, { "epoch": 5.534117647058824, "grad_norm": 1.1649008499878144, "learning_rate": 6.234231941613571e-06, "loss": 1.1875, "step": 11760 }, { "epoch": 5.534588235294118, "grad_norm": 1.0455043008214437, "learning_rate": 6.221731647393607e-06, "loss": 0.793, "step": 11761 }, { "epoch": 5.535058823529412, "grad_norm": 1.1426641709235894, "learning_rate": 6.209243632543015e-06, "loss": 0.7773, "step": 11762 }, { "epoch": 5.535529411764706, "grad_norm": 0.9630782451871973, "learning_rate": 6.196767898128291e-06, "loss": 0.5156, "step": 11763 }, { "epoch": 5.536, "grad_norm": 0.6822519093313458, "learning_rate": 6.184304445214932e-06, "loss": 0.2637, "step": 11764 }, { "epoch": 5.536470588235294, "grad_norm": 1.173038416132842, "learning_rate": 6.171853274867406e-06, "loss": 0.6172, "step": 11765 }, { "epoch": 5.536941176470588, "grad_norm": 0.9194093901181658, "learning_rate": 6.159414388149059e-06, "loss": 0.5352, "step": 11766 }, { "epoch": 5.537411764705882, "grad_norm": 1.6451706004141033, "learning_rate": 6.146987786122259e-06, "loss": 1.1875, "step": 11767 }, { "epoch": 5.5378823529411765, "grad_norm": 0.8942141515148772, "learning_rate": 6.134573469848325e-06, "loss": 0.334, "step": 11768 }, { "epoch": 5.538352941176471, "grad_norm": 1.1403272940113895, "learning_rate": 6.1221714403874415e-06, "loss": 0.5195, "step": 11769 }, { "epoch": 5.538823529411765, "grad_norm": 0.40642581613483125, "learning_rate": 6.109781698798843e-06, "loss": 0.0527, "step": 11770 }, { "epoch": 5.539294117647059, "grad_norm": 1.2549412980080994, "learning_rate": 6.097404246140669e-06, "loss": 1.1016, "step": 11771 }, { "epoch": 5.539764705882353, "grad_norm": 1.267028354481295, "learning_rate": 6.085039083470006e-06, "loss": 0.8867, "step": 11772 }, { "epoch": 5.540235294117647, "grad_norm": 1.4442795230700693, "learning_rate": 6.072686211842909e-06, "loss": 1.125, "step": 11773 }, { "epoch": 5.540705882352941, "grad_norm": 0.49681577380207204, "learning_rate": 6.060345632314384e-06, "loss": 0.2832, "step": 11774 }, { "epoch": 5.541176470588235, "grad_norm": 0.9277361627465305, "learning_rate": 6.048017345938339e-06, "loss": 0.4941, "step": 11775 }, { "epoch": 5.541647058823529, "grad_norm": 1.2528203344394433, "learning_rate": 6.035701353767713e-06, "loss": 0.707, "step": 11776 }, { "epoch": 5.542117647058824, "grad_norm": 1.0291405864788357, "learning_rate": 6.023397656854301e-06, "loss": 0.5195, "step": 11777 }, { "epoch": 5.542588235294118, "grad_norm": 1.3268461904395716, "learning_rate": 6.011106256248943e-06, "loss": 0.9375, "step": 11778 }, { "epoch": 5.543058823529412, "grad_norm": 1.279984305145315, "learning_rate": 5.998827153001401e-06, "loss": 0.8906, "step": 11779 }, { "epoch": 5.543529411764706, "grad_norm": 1.115287180011099, "learning_rate": 5.986560348160302e-06, "loss": 0.5469, "step": 11780 }, { "epoch": 5.5440000000000005, "grad_norm": 0.662688618233323, "learning_rate": 5.9743058427733414e-06, "loss": 0.1855, "step": 11781 }, { "epoch": 5.544470588235294, "grad_norm": 1.1761903424753248, "learning_rate": 5.9620636378871325e-06, "loss": 1.2656, "step": 11782 }, { "epoch": 5.544941176470588, "grad_norm": 0.8604073019330815, "learning_rate": 5.9498337345471544e-06, "loss": 0.5312, "step": 11783 }, { "epoch": 5.545411764705882, "grad_norm": 1.2229752791797026, "learning_rate": 5.93761613379794e-06, "loss": 0.8906, "step": 11784 }, { "epoch": 5.5458823529411765, "grad_norm": 0.9104249286439778, "learning_rate": 5.925410836682953e-06, "loss": 0.3242, "step": 11785 }, { "epoch": 5.546352941176471, "grad_norm": 0.9700215843182893, "learning_rate": 5.913217844244545e-06, "loss": 0.8125, "step": 11786 }, { "epoch": 5.546823529411765, "grad_norm": 0.8883318055043848, "learning_rate": 5.90103715752408e-06, "loss": 0.3164, "step": 11787 }, { "epoch": 5.547294117647059, "grad_norm": 1.1352561895694044, "learning_rate": 5.888868777561862e-06, "loss": 0.9688, "step": 11788 }, { "epoch": 5.5477647058823525, "grad_norm": 1.0683755364362069, "learning_rate": 5.8767127053970745e-06, "loss": 0.8125, "step": 11789 }, { "epoch": 5.548235294117647, "grad_norm": 1.088440030632635, "learning_rate": 5.864568942067954e-06, "loss": 0.5586, "step": 11790 }, { "epoch": 5.548705882352941, "grad_norm": 1.1742818575945884, "learning_rate": 5.852437488611656e-06, "loss": 0.7578, "step": 11791 }, { "epoch": 5.549176470588235, "grad_norm": 0.3873441045004298, "learning_rate": 5.840318346064199e-06, "loss": 0.0525, "step": 11792 }, { "epoch": 5.549647058823529, "grad_norm": 0.6833848475134413, "learning_rate": 5.828211515460673e-06, "loss": 0.3047, "step": 11793 }, { "epoch": 5.550117647058824, "grad_norm": 1.0622204542159537, "learning_rate": 5.816116997835052e-06, "loss": 0.8203, "step": 11794 }, { "epoch": 5.550588235294118, "grad_norm": 0.8077613559886823, "learning_rate": 5.804034794220208e-06, "loss": 0.4805, "step": 11795 }, { "epoch": 5.551058823529412, "grad_norm": 1.098594107384731, "learning_rate": 5.791964905648116e-06, "loss": 0.668, "step": 11796 }, { "epoch": 5.551529411764706, "grad_norm": 0.7384257743922535, "learning_rate": 5.779907333149553e-06, "loss": 0.3945, "step": 11797 }, { "epoch": 5.552, "grad_norm": 0.9032884620745555, "learning_rate": 5.767862077754276e-06, "loss": 0.2852, "step": 11798 }, { "epoch": 5.552470588235294, "grad_norm": 0.7963591344606676, "learning_rate": 5.755829140491064e-06, "loss": 0.6406, "step": 11799 }, { "epoch": 5.552941176470588, "grad_norm": 1.062692242134238, "learning_rate": 5.743808522387544e-06, "loss": 1.0312, "step": 11800 }, { "epoch": 5.553411764705882, "grad_norm": 1.0960599622411733, "learning_rate": 5.7318002244703285e-06, "loss": 0.6562, "step": 11801 }, { "epoch": 5.5538823529411765, "grad_norm": 0.5700756885127352, "learning_rate": 5.7198042477650475e-06, "loss": 0.0957, "step": 11802 }, { "epoch": 5.554352941176471, "grad_norm": 1.0099108160940247, "learning_rate": 5.7078205932961476e-06, "loss": 0.3301, "step": 11803 }, { "epoch": 5.554823529411765, "grad_norm": 1.2952803926167935, "learning_rate": 5.69584926208711e-06, "loss": 1.4688, "step": 11804 }, { "epoch": 5.555294117647059, "grad_norm": 1.0517641489402296, "learning_rate": 5.683890255160367e-06, "loss": 0.668, "step": 11805 }, { "epoch": 5.555764705882353, "grad_norm": 1.242020387792222, "learning_rate": 5.671943573537269e-06, "loss": 0.8047, "step": 11806 }, { "epoch": 5.556235294117647, "grad_norm": 0.9552084174349847, "learning_rate": 5.6600092182381156e-06, "loss": 0.5039, "step": 11807 }, { "epoch": 5.556705882352941, "grad_norm": 1.1728595774437864, "learning_rate": 5.648087190282141e-06, "loss": 0.7891, "step": 11808 }, { "epoch": 5.557176470588235, "grad_norm": 1.2163970177620556, "learning_rate": 5.636177490687565e-06, "loss": 0.2793, "step": 11809 }, { "epoch": 5.557647058823529, "grad_norm": 0.7514953576685599, "learning_rate": 5.624280120471542e-06, "loss": 0.2969, "step": 11810 }, { "epoch": 5.558117647058824, "grad_norm": 0.7783810748165026, "learning_rate": 5.612395080650156e-06, "loss": 0.2246, "step": 11811 }, { "epoch": 5.558588235294118, "grad_norm": 0.2539389494336572, "learning_rate": 5.600522372238431e-06, "loss": 0.0232, "step": 11812 }, { "epoch": 5.559058823529412, "grad_norm": 1.2092174980500294, "learning_rate": 5.588661996250387e-06, "loss": 1.3438, "step": 11813 }, { "epoch": 5.559529411764705, "grad_norm": 0.7773697275713791, "learning_rate": 5.576813953698933e-06, "loss": 0.2734, "step": 11814 }, { "epoch": 5.5600000000000005, "grad_norm": 1.3617284830142509, "learning_rate": 5.564978245595958e-06, "loss": 0.6445, "step": 11815 }, { "epoch": 5.560470588235294, "grad_norm": 1.0374644489195273, "learning_rate": 5.553154872952303e-06, "loss": 0.3691, "step": 11816 }, { "epoch": 5.560941176470588, "grad_norm": 1.0942818472520035, "learning_rate": 5.541343836777729e-06, "loss": 0.3672, "step": 11817 }, { "epoch": 5.561411764705882, "grad_norm": 1.202335773201332, "learning_rate": 5.529545138080943e-06, "loss": 0.9102, "step": 11818 }, { "epoch": 5.5618823529411765, "grad_norm": 1.026070510427272, "learning_rate": 5.517758777869641e-06, "loss": 0.7383, "step": 11819 }, { "epoch": 5.562352941176471, "grad_norm": 1.2456604465431254, "learning_rate": 5.5059847571504e-06, "loss": 1.0156, "step": 11820 }, { "epoch": 5.562823529411765, "grad_norm": 0.9416231150397585, "learning_rate": 5.494223076928817e-06, "loss": 0.418, "step": 11821 }, { "epoch": 5.563294117647059, "grad_norm": 1.8620608440792137, "learning_rate": 5.482473738209386e-06, "loss": 1.6562, "step": 11822 }, { "epoch": 5.5637647058823525, "grad_norm": 1.0912794768868437, "learning_rate": 5.470736741995541e-06, "loss": 0.6289, "step": 11823 }, { "epoch": 5.564235294117647, "grad_norm": 1.0068169007314733, "learning_rate": 5.4590120892896775e-06, "loss": 0.4297, "step": 11824 }, { "epoch": 5.564705882352941, "grad_norm": 1.0996113371404306, "learning_rate": 5.447299781093178e-06, "loss": 0.3984, "step": 11825 }, { "epoch": 5.565176470588235, "grad_norm": 1.1777403684025287, "learning_rate": 5.435599818406278e-06, "loss": 0.6797, "step": 11826 }, { "epoch": 5.565647058823529, "grad_norm": 1.0701912627212935, "learning_rate": 5.423912202228275e-06, "loss": 0.9727, "step": 11827 }, { "epoch": 5.566117647058824, "grad_norm": 1.3101413543138571, "learning_rate": 5.412236933557274e-06, "loss": 1.0938, "step": 11828 }, { "epoch": 5.566588235294118, "grad_norm": 1.1513149631227104, "learning_rate": 5.400574013390441e-06, "loss": 0.6016, "step": 11829 }, { "epoch": 5.567058823529412, "grad_norm": 1.3673164751553408, "learning_rate": 5.388923442723864e-06, "loss": 0.6992, "step": 11830 }, { "epoch": 5.567529411764706, "grad_norm": 1.3936214746466193, "learning_rate": 5.37728522255253e-06, "loss": 1.3594, "step": 11831 }, { "epoch": 5.568, "grad_norm": 1.5038111534493646, "learning_rate": 5.365659353870394e-06, "loss": 1.0312, "step": 11832 }, { "epoch": 5.568470588235294, "grad_norm": 1.0076213188059888, "learning_rate": 5.354045837670412e-06, "loss": 0.6484, "step": 11833 }, { "epoch": 5.568941176470588, "grad_norm": 0.9889979417232205, "learning_rate": 5.342444674944374e-06, "loss": 0.377, "step": 11834 }, { "epoch": 5.569411764705882, "grad_norm": 0.8756179141875189, "learning_rate": 5.330855866683087e-06, "loss": 0.5703, "step": 11835 }, { "epoch": 5.5698823529411765, "grad_norm": 1.0813058664850648, "learning_rate": 5.319279413876343e-06, "loss": 0.9062, "step": 11836 }, { "epoch": 5.570352941176471, "grad_norm": 0.8486465534717792, "learning_rate": 5.307715317512767e-06, "loss": 0.2637, "step": 11837 }, { "epoch": 5.570823529411765, "grad_norm": 1.1103656164995102, "learning_rate": 5.29616357858002e-06, "loss": 0.5859, "step": 11838 }, { "epoch": 5.571294117647059, "grad_norm": 0.9623699664173413, "learning_rate": 5.2846241980646795e-06, "loss": 0.5234, "step": 11839 }, { "epoch": 5.571764705882353, "grad_norm": 0.9472750178923849, "learning_rate": 5.273097176952224e-06, "loss": 0.248, "step": 11840 }, { "epoch": 5.572235294117647, "grad_norm": 1.0560696960458655, "learning_rate": 5.261582516227197e-06, "loss": 0.8047, "step": 11841 }, { "epoch": 5.572705882352941, "grad_norm": 1.2106264612692401, "learning_rate": 5.250080216872932e-06, "loss": 0.4121, "step": 11842 }, { "epoch": 5.573176470588235, "grad_norm": 0.9449022578258743, "learning_rate": 5.238590279871824e-06, "loss": 0.9844, "step": 11843 }, { "epoch": 5.573647058823529, "grad_norm": 1.0211672776542462, "learning_rate": 5.227112706205155e-06, "loss": 0.8789, "step": 11844 }, { "epoch": 5.574117647058824, "grad_norm": 1.5165392540926, "learning_rate": 5.2156474968531595e-06, "loss": 0.9531, "step": 11845 }, { "epoch": 5.574588235294118, "grad_norm": 1.1202334597588557, "learning_rate": 5.204194652795035e-06, "loss": 0.8477, "step": 11846 }, { "epoch": 5.575058823529412, "grad_norm": 0.8426578936446573, "learning_rate": 5.1927541750089175e-06, "loss": 0.6758, "step": 11847 }, { "epoch": 5.575529411764705, "grad_norm": 1.4179588593897268, "learning_rate": 5.181326064471858e-06, "loss": 0.8477, "step": 11848 }, { "epoch": 5.576, "grad_norm": 1.1411687193838933, "learning_rate": 5.169910322159876e-06, "loss": 0.3633, "step": 11849 }, { "epoch": 5.576470588235294, "grad_norm": 1.6758910171950296, "learning_rate": 5.158506949047975e-06, "loss": 0.4961, "step": 11850 }, { "epoch": 5.576941176470588, "grad_norm": 1.0837884333594912, "learning_rate": 5.147115946109992e-06, "loss": 1.1172, "step": 11851 }, { "epoch": 5.577411764705882, "grad_norm": 1.055627997593996, "learning_rate": 5.1357373143188e-06, "loss": 0.6133, "step": 11852 }, { "epoch": 5.5778823529411765, "grad_norm": 0.9766766270377878, "learning_rate": 5.124371054646237e-06, "loss": 0.3398, "step": 11853 }, { "epoch": 5.578352941176471, "grad_norm": 1.1584627726088383, "learning_rate": 5.11301716806296e-06, "loss": 0.5078, "step": 11854 }, { "epoch": 5.578823529411765, "grad_norm": 1.116765143382932, "learning_rate": 5.101675655538695e-06, "loss": 0.6758, "step": 11855 }, { "epoch": 5.579294117647059, "grad_norm": 1.0538638594799261, "learning_rate": 5.090346518042049e-06, "loss": 0.4199, "step": 11856 }, { "epoch": 5.5797647058823525, "grad_norm": 1.274275260104198, "learning_rate": 5.079029756540598e-06, "loss": 1.2891, "step": 11857 }, { "epoch": 5.580235294117647, "grad_norm": 1.4822998711131519, "learning_rate": 5.067725372000836e-06, "loss": 1.2031, "step": 11858 }, { "epoch": 5.580705882352941, "grad_norm": 0.8920345826176076, "learning_rate": 5.056433365388207e-06, "loss": 0.3125, "step": 11859 }, { "epoch": 5.581176470588235, "grad_norm": 1.3103920579475474, "learning_rate": 5.045153737667107e-06, "loss": 0.8047, "step": 11860 }, { "epoch": 5.581647058823529, "grad_norm": 1.0753184397680198, "learning_rate": 5.033886489800898e-06, "loss": 0.5586, "step": 11861 }, { "epoch": 5.582117647058824, "grad_norm": 1.155838611806683, "learning_rate": 5.022631622751827e-06, "loss": 0.6172, "step": 11862 }, { "epoch": 5.582588235294118, "grad_norm": 0.9267301677259542, "learning_rate": 5.011389137481109e-06, "loss": 0.543, "step": 11863 }, { "epoch": 5.583058823529412, "grad_norm": 0.8681912270462637, "learning_rate": 5.000159034948958e-06, "loss": 0.7188, "step": 11864 }, { "epoch": 5.583529411764706, "grad_norm": 1.0650246906037204, "learning_rate": 4.988941316114425e-06, "loss": 0.5625, "step": 11865 }, { "epoch": 5.584, "grad_norm": 0.8706548235383748, "learning_rate": 4.9777359819355765e-06, "loss": 0.3789, "step": 11866 }, { "epoch": 5.584470588235294, "grad_norm": 1.245080417317674, "learning_rate": 4.9665430333694135e-06, "loss": 0.4316, "step": 11867 }, { "epoch": 5.584941176470588, "grad_norm": 0.8132429269390368, "learning_rate": 4.955362471371854e-06, "loss": 0.209, "step": 11868 }, { "epoch": 5.585411764705882, "grad_norm": 1.523021728946803, "learning_rate": 4.944194296897785e-06, "loss": 1.0391, "step": 11869 }, { "epoch": 5.5858823529411765, "grad_norm": 1.3859409893518975, "learning_rate": 4.933038510901027e-06, "loss": 0.8203, "step": 11870 }, { "epoch": 5.586352941176471, "grad_norm": 1.4800095406989693, "learning_rate": 4.921895114334301e-06, "loss": 0.9844, "step": 11871 }, { "epoch": 5.586823529411765, "grad_norm": 1.0030677594832758, "learning_rate": 4.910764108149346e-06, "loss": 0.3945, "step": 11872 }, { "epoch": 5.587294117647059, "grad_norm": 0.9080672059630137, "learning_rate": 4.899645493296817e-06, "loss": 0.293, "step": 11873 }, { "epoch": 5.587764705882353, "grad_norm": 1.0949925976604316, "learning_rate": 4.888539270726255e-06, "loss": 0.4238, "step": 11874 }, { "epoch": 5.588235294117647, "grad_norm": 1.3220571100498735, "learning_rate": 4.877445441386218e-06, "loss": 0.4863, "step": 11875 }, { "epoch": 5.588705882352941, "grad_norm": 0.7974898364253564, "learning_rate": 4.866364006224166e-06, "loss": 0.4688, "step": 11876 }, { "epoch": 5.589176470588235, "grad_norm": 1.2177607434377287, "learning_rate": 4.85529496618649e-06, "loss": 0.5469, "step": 11877 }, { "epoch": 5.589647058823529, "grad_norm": 0.7721925204108209, "learning_rate": 4.844238322218568e-06, "loss": 0.3262, "step": 11878 }, { "epoch": 5.590117647058824, "grad_norm": 0.8695056237616718, "learning_rate": 4.833194075264663e-06, "loss": 0.2295, "step": 11879 }, { "epoch": 5.590588235294118, "grad_norm": 1.4636430202715036, "learning_rate": 4.8221622262680336e-06, "loss": 0.5938, "step": 11880 }, { "epoch": 5.591058823529412, "grad_norm": 0.7050332727833674, "learning_rate": 4.811142776170845e-06, "loss": 0.207, "step": 11881 }, { "epoch": 5.591529411764705, "grad_norm": 1.6600706019893217, "learning_rate": 4.800135725914195e-06, "loss": 0.9297, "step": 11882 }, { "epoch": 5.592, "grad_norm": 1.1619704622238562, "learning_rate": 4.789141076438163e-06, "loss": 0.6797, "step": 11883 }, { "epoch": 5.592470588235294, "grad_norm": 1.1666020546139448, "learning_rate": 4.778158828681733e-06, "loss": 0.582, "step": 11884 }, { "epoch": 5.592941176470588, "grad_norm": 1.133212381225842, "learning_rate": 4.767188983582854e-06, "loss": 0.9258, "step": 11885 }, { "epoch": 5.593411764705882, "grad_norm": 1.6429251573843233, "learning_rate": 4.75623154207841e-06, "loss": 0.8086, "step": 11886 }, { "epoch": 5.5938823529411765, "grad_norm": 0.9360144929549814, "learning_rate": 4.745286505104201e-06, "loss": 0.418, "step": 11887 }, { "epoch": 5.594352941176471, "grad_norm": 1.1078382150695134, "learning_rate": 4.734353873594998e-06, "loss": 0.6172, "step": 11888 }, { "epoch": 5.594823529411765, "grad_norm": 1.3136099304883306, "learning_rate": 4.723433648484503e-06, "loss": 0.793, "step": 11889 }, { "epoch": 5.595294117647059, "grad_norm": 1.0343761317217286, "learning_rate": 4.712525830705338e-06, "loss": 0.8008, "step": 11890 }, { "epoch": 5.5957647058823525, "grad_norm": 1.348685261971096, "learning_rate": 4.70163042118909e-06, "loss": 0.7305, "step": 11891 }, { "epoch": 5.596235294117647, "grad_norm": 0.9898986651957279, "learning_rate": 4.690747420866314e-06, "loss": 0.4043, "step": 11892 }, { "epoch": 5.596705882352941, "grad_norm": 1.0261013115802446, "learning_rate": 4.679876830666434e-06, "loss": 0.3945, "step": 11893 }, { "epoch": 5.597176470588235, "grad_norm": 0.8288265610835638, "learning_rate": 4.669018651517859e-06, "loss": 0.1934, "step": 11894 }, { "epoch": 5.597647058823529, "grad_norm": 1.1707822075415526, "learning_rate": 4.658172884347943e-06, "loss": 0.4688, "step": 11895 }, { "epoch": 5.598117647058824, "grad_norm": 1.1729922978091192, "learning_rate": 4.6473395300829495e-06, "loss": 1.1328, "step": 11896 }, { "epoch": 5.598588235294118, "grad_norm": 0.9451835678609549, "learning_rate": 4.636518589648119e-06, "loss": 0.582, "step": 11897 }, { "epoch": 5.599058823529412, "grad_norm": 1.025337242143275, "learning_rate": 4.625710063967613e-06, "loss": 0.8906, "step": 11898 }, { "epoch": 5.599529411764706, "grad_norm": 1.2471646935210012, "learning_rate": 4.614913953964511e-06, "loss": 1.2812, "step": 11899 }, { "epoch": 5.6, "grad_norm": 0.9489220681043935, "learning_rate": 4.604130260560873e-06, "loss": 0.2617, "step": 11900 }, { "epoch": 5.600470588235294, "grad_norm": 0.8209671411825487, "learning_rate": 4.593358984677681e-06, "loss": 0.2119, "step": 11901 }, { "epoch": 5.600941176470588, "grad_norm": 1.305912369135822, "learning_rate": 4.5826001272348315e-06, "loss": 0.8555, "step": 11902 }, { "epoch": 5.601411764705882, "grad_norm": 0.6041089085651946, "learning_rate": 4.571853689151206e-06, "loss": 0.1533, "step": 11903 }, { "epoch": 5.6018823529411765, "grad_norm": 0.5095704114027447, "learning_rate": 4.5611196713446045e-06, "loss": 0.0693, "step": 11904 }, { "epoch": 5.602352941176471, "grad_norm": 1.0285653434545898, "learning_rate": 4.5503980747317425e-06, "loss": 0.5586, "step": 11905 }, { "epoch": 5.602823529411765, "grad_norm": 0.40388715128230906, "learning_rate": 4.539688900228322e-06, "loss": 0.1245, "step": 11906 }, { "epoch": 5.603294117647059, "grad_norm": 1.2233389210642702, "learning_rate": 4.528992148748944e-06, "loss": 0.6445, "step": 11907 }, { "epoch": 5.603764705882353, "grad_norm": 0.7546923872699373, "learning_rate": 4.518307821207162e-06, "loss": 0.2412, "step": 11908 }, { "epoch": 5.604235294117647, "grad_norm": 1.6773126313271145, "learning_rate": 4.507635918515496e-06, "loss": 1.0547, "step": 11909 }, { "epoch": 5.604705882352941, "grad_norm": 0.7608665631290631, "learning_rate": 4.496976441585348e-06, "loss": 0.9688, "step": 11910 }, { "epoch": 5.605176470588235, "grad_norm": 1.1231039741536575, "learning_rate": 4.486329391327093e-06, "loss": 1.0156, "step": 11911 }, { "epoch": 5.605647058823529, "grad_norm": 1.1436290537820344, "learning_rate": 4.475694768650051e-06, "loss": 0.5039, "step": 11912 }, { "epoch": 5.606117647058824, "grad_norm": 1.198501006169764, "learning_rate": 4.4650725744624635e-06, "loss": 0.6133, "step": 11913 }, { "epoch": 5.606588235294118, "grad_norm": 0.8786865062323865, "learning_rate": 4.454462809671538e-06, "loss": 0.3516, "step": 11914 }, { "epoch": 5.607058823529412, "grad_norm": 0.678517121433099, "learning_rate": 4.443865475183384e-06, "loss": 0.1455, "step": 11915 }, { "epoch": 5.607529411764705, "grad_norm": 1.7957038908020229, "learning_rate": 4.4332805719030586e-06, "loss": 0.7656, "step": 11916 }, { "epoch": 5.608, "grad_norm": 0.9930080141244052, "learning_rate": 4.4227081007345565e-06, "loss": 0.4219, "step": 11917 }, { "epoch": 5.608470588235294, "grad_norm": 1.0414721384529841, "learning_rate": 4.412148062580873e-06, "loss": 0.7773, "step": 11918 }, { "epoch": 5.608941176470588, "grad_norm": 1.2347642248174822, "learning_rate": 4.401600458343818e-06, "loss": 1.6016, "step": 11919 }, { "epoch": 5.609411764705882, "grad_norm": 0.8962402152978148, "learning_rate": 4.39106528892424e-06, "loss": 0.4707, "step": 11920 }, { "epoch": 5.6098823529411765, "grad_norm": 0.4936614599850078, "learning_rate": 4.380542555221917e-06, "loss": 0.3359, "step": 11921 }, { "epoch": 5.610352941176471, "grad_norm": 0.9479446398557795, "learning_rate": 4.370032258135514e-06, "loss": 0.2441, "step": 11922 }, { "epoch": 5.610823529411765, "grad_norm": 1.2626904967870285, "learning_rate": 4.3595343985626476e-06, "loss": 0.7461, "step": 11923 }, { "epoch": 5.611294117647059, "grad_norm": 1.1219793289582853, "learning_rate": 4.349048977399933e-06, "loss": 0.543, "step": 11924 }, { "epoch": 5.6117647058823525, "grad_norm": 1.1929742011809799, "learning_rate": 4.33857599554282e-06, "loss": 1.2656, "step": 11925 }, { "epoch": 5.612235294117647, "grad_norm": 0.7465132990088371, "learning_rate": 4.328115453885794e-06, "loss": 0.248, "step": 11926 }, { "epoch": 5.612705882352941, "grad_norm": 0.5853574561133825, "learning_rate": 4.317667353322241e-06, "loss": 0.0884, "step": 11927 }, { "epoch": 5.613176470588235, "grad_norm": 0.8045189810637929, "learning_rate": 4.307231694744445e-06, "loss": 0.3477, "step": 11928 }, { "epoch": 5.613647058823529, "grad_norm": 1.2528387770738962, "learning_rate": 4.296808479043678e-06, "loss": 1.2578, "step": 11929 }, { "epoch": 5.614117647058824, "grad_norm": 0.7593842505454291, "learning_rate": 4.286397707110145e-06, "loss": 0.1572, "step": 11930 }, { "epoch": 5.614588235294118, "grad_norm": 0.8073578602836213, "learning_rate": 4.275999379832967e-06, "loss": 0.2676, "step": 11931 }, { "epoch": 5.615058823529412, "grad_norm": 1.0543305492110338, "learning_rate": 4.265613498100234e-06, "loss": 0.5977, "step": 11932 }, { "epoch": 5.615529411764706, "grad_norm": 1.07269263645614, "learning_rate": 4.255240062798904e-06, "loss": 0.3965, "step": 11933 }, { "epoch": 5.616, "grad_norm": 0.9344953262719026, "learning_rate": 4.24487907481495e-06, "loss": 0.3398, "step": 11934 }, { "epoch": 5.616470588235294, "grad_norm": 0.808591411669064, "learning_rate": 4.234530535033265e-06, "loss": 0.2422, "step": 11935 }, { "epoch": 5.616941176470588, "grad_norm": 1.1826251174272753, "learning_rate": 4.224194444337642e-06, "loss": 0.9141, "step": 11936 }, { "epoch": 5.617411764705882, "grad_norm": 1.1589827763343044, "learning_rate": 4.213870803610842e-06, "loss": 1.3984, "step": 11937 }, { "epoch": 5.6178823529411765, "grad_norm": 0.8575723997096742, "learning_rate": 4.203559613734542e-06, "loss": 0.2598, "step": 11938 }, { "epoch": 5.618352941176471, "grad_norm": 0.6837224822153019, "learning_rate": 4.193260875589371e-06, "loss": 0.2695, "step": 11939 }, { "epoch": 5.618823529411765, "grad_norm": 0.4538257031343197, "learning_rate": 4.1829745900549275e-06, "loss": 0.0679, "step": 11940 }, { "epoch": 5.619294117647058, "grad_norm": 0.9169548882802104, "learning_rate": 4.1727007580096565e-06, "loss": 0.3398, "step": 11941 }, { "epoch": 5.619764705882353, "grad_norm": 1.2953561365797603, "learning_rate": 4.162439380331023e-06, "loss": 0.7656, "step": 11942 }, { "epoch": 5.620235294117647, "grad_norm": 1.0411393720247224, "learning_rate": 4.15219045789541e-06, "loss": 0.3184, "step": 11943 }, { "epoch": 5.620705882352941, "grad_norm": 1.176968396790456, "learning_rate": 4.141953991578101e-06, "loss": 0.4258, "step": 11944 }, { "epoch": 5.621176470588235, "grad_norm": 1.2260385095128201, "learning_rate": 4.131729982253345e-06, "loss": 0.5938, "step": 11945 }, { "epoch": 5.621647058823529, "grad_norm": 1.0628618133485692, "learning_rate": 4.121518430794346e-06, "loss": 0.6836, "step": 11946 }, { "epoch": 5.622117647058824, "grad_norm": 0.9425214861837184, "learning_rate": 4.111319338073188e-06, "loss": 1.0391, "step": 11947 }, { "epoch": 5.622588235294118, "grad_norm": 1.224849152234669, "learning_rate": 4.101132704960941e-06, "loss": 0.918, "step": 11948 }, { "epoch": 5.623058823529412, "grad_norm": 0.7719513016426474, "learning_rate": 4.09095853232761e-06, "loss": 0.2324, "step": 11949 }, { "epoch": 5.623529411764705, "grad_norm": 1.5103432372383239, "learning_rate": 4.080796821042082e-06, "loss": 0.9609, "step": 11950 }, { "epoch": 5.624, "grad_norm": 1.1256854979482436, "learning_rate": 4.07064757197223e-06, "loss": 0.4434, "step": 11951 }, { "epoch": 5.624470588235294, "grad_norm": 0.9642965592147542, "learning_rate": 4.0605107859848604e-06, "loss": 0.6875, "step": 11952 }, { "epoch": 5.624941176470588, "grad_norm": 1.233755628896581, "learning_rate": 4.050386463945699e-06, "loss": 0.6836, "step": 11953 }, { "epoch": 5.625411764705882, "grad_norm": 1.026328553278169, "learning_rate": 4.040274606719418e-06, "loss": 0.4258, "step": 11954 }, { "epoch": 5.625882352941177, "grad_norm": 1.2847723229169479, "learning_rate": 4.030175215169612e-06, "loss": 0.8125, "step": 11955 }, { "epoch": 5.626352941176471, "grad_norm": 1.3639990492896774, "learning_rate": 4.020088290158807e-06, "loss": 0.875, "step": 11956 }, { "epoch": 5.626823529411765, "grad_norm": 1.131347362631778, "learning_rate": 4.010013832548514e-06, "loss": 0.8789, "step": 11957 }, { "epoch": 5.627294117647059, "grad_norm": 1.4611819638742718, "learning_rate": 3.999951843199095e-06, "loss": 1.4062, "step": 11958 }, { "epoch": 5.627764705882353, "grad_norm": 1.4041091379769635, "learning_rate": 3.989902322969929e-06, "loss": 0.6445, "step": 11959 }, { "epoch": 5.628235294117647, "grad_norm": 1.2493654029349617, "learning_rate": 3.979865272719279e-06, "loss": 1.2812, "step": 11960 }, { "epoch": 5.628705882352941, "grad_norm": 1.1168627307304664, "learning_rate": 3.969840693304344e-06, "loss": 1.0703, "step": 11961 }, { "epoch": 5.629176470588235, "grad_norm": 1.0616634684364383, "learning_rate": 3.959828585581287e-06, "loss": 0.3984, "step": 11962 }, { "epoch": 5.6296470588235294, "grad_norm": 0.8072229997255481, "learning_rate": 3.949828950405226e-06, "loss": 0.3594, "step": 11963 }, { "epoch": 5.630117647058824, "grad_norm": 1.3603933571572013, "learning_rate": 3.939841788630111e-06, "loss": 0.8516, "step": 11964 }, { "epoch": 5.630588235294118, "grad_norm": 1.2365239928220433, "learning_rate": 3.9298671011089245e-06, "loss": 0.7031, "step": 11965 }, { "epoch": 5.631058823529412, "grad_norm": 1.1826042618326293, "learning_rate": 3.9199048886935865e-06, "loss": 0.9102, "step": 11966 }, { "epoch": 5.631529411764706, "grad_norm": 1.3966906566650266, "learning_rate": 3.909955152234867e-06, "loss": 1.1016, "step": 11967 }, { "epoch": 5.632, "grad_norm": 0.9637574925359785, "learning_rate": 3.900017892582552e-06, "loss": 0.8906, "step": 11968 }, { "epoch": 5.632470588235294, "grad_norm": 0.7233681762620938, "learning_rate": 3.8900931105853294e-06, "loss": 0.1689, "step": 11969 }, { "epoch": 5.632941176470588, "grad_norm": 1.1858042192261868, "learning_rate": 3.8801808070908215e-06, "loss": 1.0, "step": 11970 }, { "epoch": 5.633411764705882, "grad_norm": 1.1249378280290752, "learning_rate": 3.870280982945567e-06, "loss": 1.0625, "step": 11971 }, { "epoch": 5.633882352941177, "grad_norm": 1.1242614040816379, "learning_rate": 3.860393638995107e-06, "loss": 0.5391, "step": 11972 }, { "epoch": 5.634352941176471, "grad_norm": 1.1532483598696004, "learning_rate": 3.850518776083817e-06, "loss": 0.7617, "step": 11973 }, { "epoch": 5.634823529411765, "grad_norm": 1.1307405993530655, "learning_rate": 3.840656395055103e-06, "loss": 0.5195, "step": 11974 }, { "epoch": 5.635294117647058, "grad_norm": 1.1165440075160755, "learning_rate": 3.8308064967512095e-06, "loss": 0.4961, "step": 11975 }, { "epoch": 5.635764705882353, "grad_norm": 1.253848960514257, "learning_rate": 3.820969082013414e-06, "loss": 0.7344, "step": 11976 }, { "epoch": 5.636235294117647, "grad_norm": 0.29204300917040854, "learning_rate": 3.8111441516818776e-06, "loss": 0.0143, "step": 11977 }, { "epoch": 5.636705882352941, "grad_norm": 1.3547007203756865, "learning_rate": 3.8013317065956784e-06, "loss": 1.5625, "step": 11978 }, { "epoch": 5.637176470588235, "grad_norm": 1.1893704763726036, "learning_rate": 3.7915317475928296e-06, "loss": 0.543, "step": 11979 }, { "epoch": 5.6376470588235295, "grad_norm": 0.6568350134946002, "learning_rate": 3.7817442755103457e-06, "loss": 0.0859, "step": 11980 }, { "epoch": 5.638117647058824, "grad_norm": 1.161826380154852, "learning_rate": 3.7719692911840905e-06, "loss": 0.7773, "step": 11981 }, { "epoch": 5.638588235294118, "grad_norm": 0.8955627785519508, "learning_rate": 3.76220679544888e-06, "loss": 0.7734, "step": 11982 }, { "epoch": 5.639058823529412, "grad_norm": 1.1504746208004792, "learning_rate": 3.7524567891385305e-06, "loss": 0.6445, "step": 11983 }, { "epoch": 5.6395294117647055, "grad_norm": 1.0714502691832488, "learning_rate": 3.7427192730856927e-06, "loss": 0.6172, "step": 11984 }, { "epoch": 5.64, "grad_norm": 0.8042622342717592, "learning_rate": 3.7329942481220176e-06, "loss": 1.0781, "step": 11985 }, { "epoch": 5.640470588235294, "grad_norm": 0.7723978730141896, "learning_rate": 3.723281715078091e-06, "loss": 0.2158, "step": 11986 }, { "epoch": 5.640941176470588, "grad_norm": 1.047188928726572, "learning_rate": 3.7135816747833667e-06, "loss": 0.8789, "step": 11987 }, { "epoch": 5.641411764705882, "grad_norm": 0.9947704536987061, "learning_rate": 3.7038941280663147e-06, "loss": 0.4102, "step": 11988 }, { "epoch": 5.641882352941177, "grad_norm": 1.1538722216760873, "learning_rate": 3.6942190757542735e-06, "loss": 1.1172, "step": 11989 }, { "epoch": 5.642352941176471, "grad_norm": 1.3027574260478625, "learning_rate": 3.684556518673548e-06, "loss": 0.8242, "step": 11990 }, { "epoch": 5.642823529411765, "grad_norm": 1.18315062662193, "learning_rate": 3.6749064576493793e-06, "loss": 0.6328, "step": 11991 }, { "epoch": 5.643294117647059, "grad_norm": 0.9011625657836118, "learning_rate": 3.6652688935058904e-06, "loss": 0.4824, "step": 11992 }, { "epoch": 5.643764705882353, "grad_norm": 0.49294172462827757, "learning_rate": 3.6556438270662236e-06, "loss": 0.1934, "step": 11993 }, { "epoch": 5.644235294117647, "grad_norm": 1.1397569461745332, "learning_rate": 3.6460312591524044e-06, "loss": 0.8906, "step": 11994 }, { "epoch": 5.644705882352941, "grad_norm": 1.1126199248114836, "learning_rate": 3.6364311905853594e-06, "loss": 0.2676, "step": 11995 }, { "epoch": 5.645176470588235, "grad_norm": 1.3072972896904769, "learning_rate": 3.6268436221849994e-06, "loss": 0.9023, "step": 11996 }, { "epoch": 5.6456470588235295, "grad_norm": 1.1632219828925174, "learning_rate": 3.6172685547701695e-06, "loss": 0.6523, "step": 11997 }, { "epoch": 5.646117647058824, "grad_norm": 1.3545386713140903, "learning_rate": 3.6077059891585826e-06, "loss": 0.4688, "step": 11998 }, { "epoch": 5.646588235294118, "grad_norm": 1.0060530471037534, "learning_rate": 3.598155926166968e-06, "loss": 0.6875, "step": 11999 }, { "epoch": 5.647058823529412, "grad_norm": 0.9645179049781611, "learning_rate": 3.5886183666109405e-06, "loss": 1.3516, "step": 12000 }, { "epoch": 5.647529411764706, "grad_norm": 0.9968505112884197, "learning_rate": 3.579093311305031e-06, "loss": 0.3008, "step": 12001 }, { "epoch": 5.648, "grad_norm": 1.0314830507309434, "learning_rate": 3.5695807610627402e-06, "loss": 0.9297, "step": 12002 }, { "epoch": 5.648470588235294, "grad_norm": 1.091567927872185, "learning_rate": 3.5600807166965173e-06, "loss": 0.5469, "step": 12003 }, { "epoch": 5.648941176470588, "grad_norm": 1.0560710526317856, "learning_rate": 3.5505931790176635e-06, "loss": 0.7773, "step": 12004 }, { "epoch": 5.649411764705882, "grad_norm": 1.251406702843279, "learning_rate": 3.5411181488364802e-06, "loss": 0.5312, "step": 12005 }, { "epoch": 5.649882352941177, "grad_norm": 1.1086528156116748, "learning_rate": 3.531655626962204e-06, "loss": 0.4258, "step": 12006 }, { "epoch": 5.650352941176471, "grad_norm": 0.9216400956734015, "learning_rate": 3.5222056142029387e-06, "loss": 0.2988, "step": 12007 }, { "epoch": 5.650823529411765, "grad_norm": 0.7583875648367432, "learning_rate": 3.5127681113658044e-06, "loss": 0.2793, "step": 12008 }, { "epoch": 5.651294117647058, "grad_norm": 1.3383461898833466, "learning_rate": 3.503343119256774e-06, "loss": 1.2188, "step": 12009 }, { "epoch": 5.651764705882353, "grad_norm": 1.0823483169289536, "learning_rate": 3.493930638680803e-06, "loss": 0.6406, "step": 12010 }, { "epoch": 5.652235294117647, "grad_norm": 1.3782915961762907, "learning_rate": 3.484530670441782e-06, "loss": 1.4609, "step": 12011 }, { "epoch": 5.652705882352941, "grad_norm": 0.954843377068807, "learning_rate": 3.4751432153424687e-06, "loss": 0.9922, "step": 12012 }, { "epoch": 5.653176470588235, "grad_norm": 0.7883191285416619, "learning_rate": 3.465768274184638e-06, "loss": 0.4258, "step": 12013 }, { "epoch": 5.6536470588235295, "grad_norm": 1.3679733730965047, "learning_rate": 3.4564058477689503e-06, "loss": 0.9922, "step": 12014 }, { "epoch": 5.654117647058824, "grad_norm": 1.123232422264737, "learning_rate": 3.4470559368949825e-06, "loss": 1.1484, "step": 12015 }, { "epoch": 5.654588235294118, "grad_norm": 1.341750213392269, "learning_rate": 3.4377185423612785e-06, "loss": 0.4434, "step": 12016 }, { "epoch": 5.655058823529412, "grad_norm": 0.7434194374940136, "learning_rate": 3.428393664965301e-06, "loss": 0.1738, "step": 12017 }, { "epoch": 5.6555294117647055, "grad_norm": 1.4973800077005543, "learning_rate": 3.4190813055034127e-06, "loss": 0.75, "step": 12018 }, { "epoch": 5.656, "grad_norm": 0.873909792655999, "learning_rate": 3.4097814647709775e-06, "loss": 0.2012, "step": 12019 }, { "epoch": 5.656470588235294, "grad_norm": 0.9635876912191158, "learning_rate": 3.400494143562227e-06, "loss": 0.334, "step": 12020 }, { "epoch": 5.656941176470588, "grad_norm": 1.2063229141459773, "learning_rate": 3.391219342670326e-06, "loss": 0.5742, "step": 12021 }, { "epoch": 5.657411764705882, "grad_norm": 0.6319045480675026, "learning_rate": 3.3819570628874414e-06, "loss": 0.1206, "step": 12022 }, { "epoch": 5.657882352941177, "grad_norm": 1.106013018273515, "learning_rate": 3.3727073050045406e-06, "loss": 0.6719, "step": 12023 }, { "epoch": 5.658352941176471, "grad_norm": 0.9570393371154181, "learning_rate": 3.3634700698116577e-06, "loss": 0.3887, "step": 12024 }, { "epoch": 5.658823529411765, "grad_norm": 0.9756709003050197, "learning_rate": 3.354245358097679e-06, "loss": 0.7031, "step": 12025 }, { "epoch": 5.659294117647059, "grad_norm": 1.147945821225784, "learning_rate": 3.345033170650441e-06, "loss": 0.5312, "step": 12026 }, { "epoch": 5.659764705882353, "grad_norm": 1.0443223113415072, "learning_rate": 3.3358335082566966e-06, "loss": 0.7305, "step": 12027 }, { "epoch": 5.660235294117647, "grad_norm": 1.1153692742597578, "learning_rate": 3.326646371702185e-06, "loss": 0.5469, "step": 12028 }, { "epoch": 5.660705882352941, "grad_norm": 1.1339156084247655, "learning_rate": 3.317471761771462e-06, "loss": 0.582, "step": 12029 }, { "epoch": 5.661176470588235, "grad_norm": 1.2848443515511956, "learning_rate": 3.30830967924815e-06, "loss": 0.4902, "step": 12030 }, { "epoch": 5.6616470588235295, "grad_norm": 0.6571399435035254, "learning_rate": 3.299160124914707e-06, "loss": 0.2246, "step": 12031 }, { "epoch": 5.662117647058824, "grad_norm": 0.7994471044807587, "learning_rate": 3.290023099552541e-06, "loss": 0.1748, "step": 12032 }, { "epoch": 5.662588235294118, "grad_norm": 1.1121990032572129, "learning_rate": 3.2808986039420115e-06, "loss": 0.5898, "step": 12033 }, { "epoch": 5.663058823529412, "grad_norm": 0.9644640264451503, "learning_rate": 3.2717866388624114e-06, "loss": 0.7188, "step": 12034 }, { "epoch": 5.663529411764706, "grad_norm": 0.9016172098884911, "learning_rate": 3.2626872050919185e-06, "loss": 0.7461, "step": 12035 }, { "epoch": 5.664, "grad_norm": 0.9586295845125884, "learning_rate": 3.2536003034076774e-06, "loss": 1.1406, "step": 12036 }, { "epoch": 5.664470588235294, "grad_norm": 1.1045096150712799, "learning_rate": 3.2445259345857676e-06, "loss": 0.5508, "step": 12037 }, { "epoch": 5.664941176470588, "grad_norm": 1.145224893503774, "learning_rate": 3.2354640994011693e-06, "loss": 1.4219, "step": 12038 }, { "epoch": 5.665411764705882, "grad_norm": 1.28856324395081, "learning_rate": 3.22641479862783e-06, "loss": 1.1719, "step": 12039 }, { "epoch": 5.665882352941177, "grad_norm": 2.2289865250034624, "learning_rate": 3.2173780330385647e-06, "loss": 1.7422, "step": 12040 }, { "epoch": 5.666352941176471, "grad_norm": 1.3582813874632587, "learning_rate": 3.2083538034051726e-06, "loss": 0.4023, "step": 12041 }, { "epoch": 5.666823529411765, "grad_norm": 1.0374337910143632, "learning_rate": 3.199342110498404e-06, "loss": 0.3574, "step": 12042 }, { "epoch": 5.667294117647058, "grad_norm": 1.0873229074640924, "learning_rate": 3.1903429550878435e-06, "loss": 0.4023, "step": 12043 }, { "epoch": 5.667764705882353, "grad_norm": 0.7669490422219365, "learning_rate": 3.181356337942109e-06, "loss": 0.2227, "step": 12044 }, { "epoch": 5.668235294117647, "grad_norm": 1.3077558414047532, "learning_rate": 3.1723822598286875e-06, "loss": 1.1016, "step": 12045 }, { "epoch": 5.668705882352941, "grad_norm": 0.770761838854835, "learning_rate": 3.163420721513982e-06, "loss": 0.2246, "step": 12046 }, { "epoch": 5.669176470588235, "grad_norm": 1.3648857835497774, "learning_rate": 3.1544717237633965e-06, "loss": 0.9883, "step": 12047 }, { "epoch": 5.6696470588235295, "grad_norm": 1.5169245198514218, "learning_rate": 3.145535267341204e-06, "loss": 0.2373, "step": 12048 }, { "epoch": 5.670117647058824, "grad_norm": 0.9326340826873778, "learning_rate": 3.1366113530106097e-06, "loss": 0.3867, "step": 12049 }, { "epoch": 5.670588235294118, "grad_norm": 0.6908122936171582, "learning_rate": 3.1276999815337544e-06, "loss": 0.1641, "step": 12050 }, { "epoch": 5.671058823529412, "grad_norm": 0.8621233210541912, "learning_rate": 3.118801153671746e-06, "loss": 0.4629, "step": 12051 }, { "epoch": 5.6715294117647055, "grad_norm": 1.0334035716541117, "learning_rate": 3.1099148701845435e-06, "loss": 0.6602, "step": 12052 }, { "epoch": 5.672, "grad_norm": 1.1453576678326847, "learning_rate": 3.1010411318311226e-06, "loss": 0.4453, "step": 12053 }, { "epoch": 5.672470588235294, "grad_norm": 0.9637034042425155, "learning_rate": 3.092179939369327e-06, "loss": 0.4004, "step": 12054 }, { "epoch": 5.672941176470588, "grad_norm": 1.021190134754215, "learning_rate": 3.083331293555935e-06, "loss": 1.3828, "step": 12055 }, { "epoch": 5.673411764705882, "grad_norm": 0.9523369510437649, "learning_rate": 3.074495195146659e-06, "loss": 0.7812, "step": 12056 }, { "epoch": 5.673882352941177, "grad_norm": 1.3945265898864605, "learning_rate": 3.0656716448961784e-06, "loss": 0.793, "step": 12057 }, { "epoch": 5.674352941176471, "grad_norm": 1.030078103711338, "learning_rate": 3.05686064355804e-06, "loss": 0.3027, "step": 12058 }, { "epoch": 5.674823529411765, "grad_norm": 1.35326054941843, "learning_rate": 3.048062191884759e-06, "loss": 0.8984, "step": 12059 }, { "epoch": 5.675294117647059, "grad_norm": 1.0364326730963687, "learning_rate": 3.039276290627768e-06, "loss": 0.9375, "step": 12060 }, { "epoch": 5.675764705882353, "grad_norm": 1.2091545036373015, "learning_rate": 3.0305029405373993e-06, "loss": 0.5078, "step": 12061 }, { "epoch": 5.676235294117647, "grad_norm": 1.0674842435348553, "learning_rate": 3.0217421423629706e-06, "loss": 0.8164, "step": 12062 }, { "epoch": 5.676705882352941, "grad_norm": 1.221395627644229, "learning_rate": 3.0129938968526836e-06, "loss": 0.875, "step": 12063 }, { "epoch": 5.677176470588235, "grad_norm": 0.874255438412743, "learning_rate": 3.00425820475369e-06, "loss": 0.6328, "step": 12064 }, { "epoch": 5.6776470588235295, "grad_norm": 1.4056387840907316, "learning_rate": 2.9955350668120605e-06, "loss": 0.9219, "step": 12065 }, { "epoch": 5.678117647058824, "grad_norm": 0.9124494985653944, "learning_rate": 2.9868244837727815e-06, "loss": 0.3594, "step": 12066 }, { "epoch": 5.678588235294118, "grad_norm": 1.1049996087501495, "learning_rate": 2.978126456379809e-06, "loss": 0.5586, "step": 12067 }, { "epoch": 5.679058823529411, "grad_norm": 0.9470923777338025, "learning_rate": 2.969440985375948e-06, "loss": 0.3047, "step": 12068 }, { "epoch": 5.679529411764706, "grad_norm": 1.3443146420134469, "learning_rate": 2.960768071503006e-06, "loss": 1.4141, "step": 12069 }, { "epoch": 5.68, "grad_norm": 0.8691431510914759, "learning_rate": 2.9521077155017224e-06, "loss": 0.2656, "step": 12070 }, { "epoch": 5.680470588235294, "grad_norm": 1.1547035068668756, "learning_rate": 2.9434599181116736e-06, "loss": 0.7031, "step": 12071 }, { "epoch": 5.680941176470588, "grad_norm": 1.1124736380967706, "learning_rate": 2.934824680071468e-06, "loss": 0.8438, "step": 12072 }, { "epoch": 5.681411764705882, "grad_norm": 1.1793613057353833, "learning_rate": 2.9262020021185993e-06, "loss": 1.0469, "step": 12073 }, { "epoch": 5.681882352941177, "grad_norm": 1.384088534760186, "learning_rate": 2.917591884989462e-06, "loss": 1.6328, "step": 12074 }, { "epoch": 5.682352941176471, "grad_norm": 1.1481312541579853, "learning_rate": 2.908994329419401e-06, "loss": 0.4453, "step": 12075 }, { "epoch": 5.682823529411765, "grad_norm": 0.9625216273415202, "learning_rate": 2.9004093361427282e-06, "loss": 0.9336, "step": 12076 }, { "epoch": 5.683294117647058, "grad_norm": 1.100687902577334, "learning_rate": 2.891836905892592e-06, "loss": 1.0, "step": 12077 }, { "epoch": 5.683764705882353, "grad_norm": 0.9464815298426336, "learning_rate": 2.883277039401155e-06, "loss": 0.8086, "step": 12078 }, { "epoch": 5.684235294117647, "grad_norm": 1.0714781935419764, "learning_rate": 2.874729737399467e-06, "loss": 0.7031, "step": 12079 }, { "epoch": 5.684705882352941, "grad_norm": 1.083235936077726, "learning_rate": 2.866195000617494e-06, "loss": 0.373, "step": 12080 }, { "epoch": 5.685176470588235, "grad_norm": 1.2466957429417025, "learning_rate": 2.857672829784169e-06, "loss": 1.0859, "step": 12081 }, { "epoch": 5.6856470588235295, "grad_norm": 1.5187539169285336, "learning_rate": 2.84916322562731e-06, "loss": 1.2812, "step": 12082 }, { "epoch": 5.686117647058824, "grad_norm": 0.8377318715332558, "learning_rate": 2.840666188873669e-06, "loss": 0.1973, "step": 12083 }, { "epoch": 5.686588235294118, "grad_norm": 1.4824109526231026, "learning_rate": 2.8321817202489493e-06, "loss": 1.0859, "step": 12084 }, { "epoch": 5.687058823529412, "grad_norm": 1.033382571631857, "learning_rate": 2.8237098204777874e-06, "loss": 0.9727, "step": 12085 }, { "epoch": 5.6875294117647055, "grad_norm": 0.9131780710145869, "learning_rate": 2.815250490283688e-06, "loss": 0.6562, "step": 12086 }, { "epoch": 5.688, "grad_norm": 1.2056469702864463, "learning_rate": 2.8068037303891233e-06, "loss": 0.4863, "step": 12087 }, { "epoch": 5.688470588235294, "grad_norm": 1.1937181902436103, "learning_rate": 2.7983695415155162e-06, "loss": 0.8516, "step": 12088 }, { "epoch": 5.688941176470588, "grad_norm": 1.463283390518046, "learning_rate": 2.789947924383157e-06, "loss": 0.8086, "step": 12089 }, { "epoch": 5.689411764705882, "grad_norm": 1.3656773349163307, "learning_rate": 2.78153887971132e-06, "loss": 1.2422, "step": 12090 }, { "epoch": 5.689882352941177, "grad_norm": 1.4022946357478425, "learning_rate": 2.773142408218165e-06, "loss": 1.0234, "step": 12091 }, { "epoch": 5.690352941176471, "grad_norm": 0.7717962197571867, "learning_rate": 2.7647585106207682e-06, "loss": 0.291, "step": 12092 }, { "epoch": 5.690823529411765, "grad_norm": 1.0046284904617486, "learning_rate": 2.7563871876352063e-06, "loss": 0.6367, "step": 12093 }, { "epoch": 5.691294117647059, "grad_norm": 1.309219611075757, "learning_rate": 2.7480284399763907e-06, "loss": 1.375, "step": 12094 }, { "epoch": 5.691764705882353, "grad_norm": 1.735014327538808, "learning_rate": 2.7396822683582177e-06, "loss": 0.793, "step": 12095 }, { "epoch": 5.692235294117647, "grad_norm": 1.2220653238299257, "learning_rate": 2.731348673493483e-06, "loss": 0.9297, "step": 12096 }, { "epoch": 5.692705882352941, "grad_norm": 1.3954744192703612, "learning_rate": 2.723027656093918e-06, "loss": 0.543, "step": 12097 }, { "epoch": 5.693176470588235, "grad_norm": 0.9673240307694205, "learning_rate": 2.714719216870187e-06, "loss": 0.3223, "step": 12098 }, { "epoch": 5.6936470588235295, "grad_norm": 1.134904774826432, "learning_rate": 2.706423356531873e-06, "loss": 0.875, "step": 12099 }, { "epoch": 5.694117647058824, "grad_norm": 0.9491087603735884, "learning_rate": 2.6981400757874584e-06, "loss": 0.3711, "step": 12100 }, { "epoch": 5.694588235294118, "grad_norm": 0.947237921937353, "learning_rate": 2.6898693753443945e-06, "loss": 0.9219, "step": 12101 }, { "epoch": 5.695058823529411, "grad_norm": 1.3012484871322971, "learning_rate": 2.68161125590905e-06, "loss": 1.4062, "step": 12102 }, { "epoch": 5.6955294117647055, "grad_norm": 1.21931310703802, "learning_rate": 2.673365718186693e-06, "loss": 0.7422, "step": 12103 }, { "epoch": 5.696, "grad_norm": 0.9050736495665637, "learning_rate": 2.6651327628815277e-06, "loss": 0.2422, "step": 12104 }, { "epoch": 5.696470588235294, "grad_norm": 0.6990670813090097, "learning_rate": 2.656912390696708e-06, "loss": 0.5273, "step": 12105 }, { "epoch": 5.696941176470588, "grad_norm": 0.5830617999274719, "learning_rate": 2.6487046023342563e-06, "loss": 0.2734, "step": 12106 }, { "epoch": 5.697411764705882, "grad_norm": 1.2471545394301828, "learning_rate": 2.6405093984952275e-06, "loss": 0.7305, "step": 12107 }, { "epoch": 5.697882352941177, "grad_norm": 0.81918464102133, "learning_rate": 2.6323267798794625e-06, "loss": 0.4375, "step": 12108 }, { "epoch": 5.698352941176471, "grad_norm": 0.767018283053609, "learning_rate": 2.624156747185835e-06, "loss": 0.4844, "step": 12109 }, { "epoch": 5.698823529411765, "grad_norm": 0.9228359430263875, "learning_rate": 2.615999301112104e-06, "loss": 0.8008, "step": 12110 }, { "epoch": 5.699294117647058, "grad_norm": 1.6167299474544319, "learning_rate": 2.6078544423549286e-06, "loss": 1.0312, "step": 12111 }, { "epoch": 5.699764705882353, "grad_norm": 1.0714976190307852, "learning_rate": 2.599722171609936e-06, "loss": 0.3828, "step": 12112 }, { "epoch": 5.700235294117647, "grad_norm": 0.7420430152308746, "learning_rate": 2.591602489571687e-06, "loss": 0.1777, "step": 12113 }, { "epoch": 5.700705882352941, "grad_norm": 0.9715470975775481, "learning_rate": 2.5834953969336105e-06, "loss": 0.3145, "step": 12114 }, { "epoch": 5.701176470588235, "grad_norm": 1.0297486231007802, "learning_rate": 2.5754008943880855e-06, "loss": 0.4961, "step": 12115 }, { "epoch": 5.7016470588235295, "grad_norm": 1.448113142462278, "learning_rate": 2.5673189826264596e-06, "loss": 1.3359, "step": 12116 }, { "epoch": 5.702117647058824, "grad_norm": 0.9680470529708973, "learning_rate": 2.559249662338947e-06, "loss": 0.4141, "step": 12117 }, { "epoch": 5.702588235294118, "grad_norm": 1.0403285954058146, "learning_rate": 2.5511929342146964e-06, "loss": 0.6719, "step": 12118 }, { "epoch": 5.703058823529412, "grad_norm": 1.2873773609689738, "learning_rate": 2.543148798941791e-06, "loss": 1.2266, "step": 12119 }, { "epoch": 5.7035294117647055, "grad_norm": 1.0330202654882075, "learning_rate": 2.535117257207264e-06, "loss": 0.8438, "step": 12120 }, { "epoch": 5.704, "grad_norm": 0.6957172165085878, "learning_rate": 2.5270983096970344e-06, "loss": 0.5703, "step": 12121 }, { "epoch": 5.704470588235294, "grad_norm": 0.7933194098329873, "learning_rate": 2.5190919570959533e-06, "loss": 0.2295, "step": 12122 }, { "epoch": 5.704941176470588, "grad_norm": 2.0926663185812493, "learning_rate": 2.511098200087808e-06, "loss": 1.2109, "step": 12123 }, { "epoch": 5.705411764705882, "grad_norm": 0.5197054429976234, "learning_rate": 2.503117039355318e-06, "loss": 0.0767, "step": 12124 }, { "epoch": 5.705882352941177, "grad_norm": 1.3081077589209598, "learning_rate": 2.4951484755800886e-06, "loss": 1.0078, "step": 12125 }, { "epoch": 5.706352941176471, "grad_norm": 0.9806579318543852, "learning_rate": 2.487192509442676e-06, "loss": 0.5977, "step": 12126 }, { "epoch": 5.706823529411765, "grad_norm": 0.8868180172679467, "learning_rate": 2.4792491416226024e-06, "loss": 0.334, "step": 12127 }, { "epoch": 5.707294117647059, "grad_norm": 1.0276807436764883, "learning_rate": 2.471318372798209e-06, "loss": 1.0312, "step": 12128 }, { "epoch": 5.707764705882353, "grad_norm": 0.8126958483731461, "learning_rate": 2.4634002036468536e-06, "loss": 0.3086, "step": 12129 }, { "epoch": 5.708235294117647, "grad_norm": 0.6923360119044691, "learning_rate": 2.4554946348447958e-06, "loss": 0.2949, "step": 12130 }, { "epoch": 5.708705882352941, "grad_norm": 1.0062684906328396, "learning_rate": 2.447601667067195e-06, "loss": 1.0703, "step": 12131 }, { "epoch": 5.709176470588235, "grad_norm": 1.1692186506122006, "learning_rate": 2.4397213009881456e-06, "loss": 0.8906, "step": 12132 }, { "epoch": 5.7096470588235295, "grad_norm": 1.0671990101141287, "learning_rate": 2.4318535372806757e-06, "loss": 1.2188, "step": 12133 }, { "epoch": 5.710117647058824, "grad_norm": 1.5730700404922904, "learning_rate": 2.4239983766167315e-06, "loss": 0.8594, "step": 12134 }, { "epoch": 5.710588235294118, "grad_norm": 1.6659000139751632, "learning_rate": 2.416155819667193e-06, "loss": 0.6953, "step": 12135 }, { "epoch": 5.711058823529411, "grad_norm": 1.166464415074176, "learning_rate": 2.4083258671018403e-06, "loss": 0.9219, "step": 12136 }, { "epoch": 5.7115294117647055, "grad_norm": 1.074740680804947, "learning_rate": 2.4005085195893725e-06, "loss": 0.9141, "step": 12137 }, { "epoch": 5.712, "grad_norm": 0.30712872585442574, "learning_rate": 2.3927037777974722e-06, "loss": 0.0176, "step": 12138 }, { "epoch": 5.712470588235294, "grad_norm": 0.9458849449624341, "learning_rate": 2.384911642392673e-06, "loss": 0.6484, "step": 12139 }, { "epoch": 5.712941176470588, "grad_norm": 1.0693703621250565, "learning_rate": 2.377132114040459e-06, "loss": 0.4277, "step": 12140 }, { "epoch": 5.713411764705882, "grad_norm": 0.7296051353835612, "learning_rate": 2.369365193405248e-06, "loss": 0.2012, "step": 12141 }, { "epoch": 5.713882352941177, "grad_norm": 0.9251484212360317, "learning_rate": 2.3616108811503775e-06, "loss": 0.6562, "step": 12142 }, { "epoch": 5.714352941176471, "grad_norm": 0.84340496346154, "learning_rate": 2.3538691779381003e-06, "loss": 0.6406, "step": 12143 }, { "epoch": 5.714823529411765, "grad_norm": 1.110811930265371, "learning_rate": 2.346140084429604e-06, "loss": 1.3594, "step": 12144 }, { "epoch": 5.715294117647058, "grad_norm": 1.1595064491424092, "learning_rate": 2.3384236012849778e-06, "loss": 0.9609, "step": 12145 }, { "epoch": 5.715764705882353, "grad_norm": 1.0663413151648136, "learning_rate": 2.330719729163244e-06, "loss": 0.7656, "step": 12146 }, { "epoch": 5.716235294117647, "grad_norm": 1.0816502057916049, "learning_rate": 2.32302846872236e-06, "loss": 0.6641, "step": 12147 }, { "epoch": 5.716705882352941, "grad_norm": 1.2518831279043325, "learning_rate": 2.3153498206192e-06, "loss": 0.7578, "step": 12148 }, { "epoch": 5.717176470588235, "grad_norm": 1.345227008133301, "learning_rate": 2.3076837855095563e-06, "loss": 0.9531, "step": 12149 }, { "epoch": 5.7176470588235295, "grad_norm": 1.1125001206838312, "learning_rate": 2.3000303640481386e-06, "loss": 0.4238, "step": 12150 }, { "epoch": 5.718117647058824, "grad_norm": 1.0611411187375903, "learning_rate": 2.292389556888591e-06, "loss": 0.918, "step": 12151 }, { "epoch": 5.718588235294118, "grad_norm": 0.9245710653869674, "learning_rate": 2.2847613646834905e-06, "loss": 0.2734, "step": 12152 }, { "epoch": 5.719058823529412, "grad_norm": 0.9560964184159293, "learning_rate": 2.2771457880842834e-06, "loss": 0.3984, "step": 12153 }, { "epoch": 5.7195294117647055, "grad_norm": 1.1120794625390324, "learning_rate": 2.2695428277414165e-06, "loss": 0.7109, "step": 12154 }, { "epoch": 5.72, "grad_norm": 1.0421566780177567, "learning_rate": 2.261952484304219e-06, "loss": 0.4609, "step": 12155 }, { "epoch": 5.720470588235294, "grad_norm": 0.88800041633949, "learning_rate": 2.254374758420924e-06, "loss": 0.2812, "step": 12156 }, { "epoch": 5.720941176470588, "grad_norm": 1.281309746032342, "learning_rate": 2.2468096507386956e-06, "loss": 0.5234, "step": 12157 }, { "epoch": 5.721411764705882, "grad_norm": 1.1292960385127528, "learning_rate": 2.239257161903668e-06, "loss": 0.4004, "step": 12158 }, { "epoch": 5.721882352941177, "grad_norm": 0.8834714956688648, "learning_rate": 2.231717292560825e-06, "loss": 0.8438, "step": 12159 }, { "epoch": 5.722352941176471, "grad_norm": 0.9994002713855005, "learning_rate": 2.224190043354135e-06, "loss": 0.3867, "step": 12160 }, { "epoch": 5.722823529411765, "grad_norm": 0.5986204867478534, "learning_rate": 2.2166754149264667e-06, "loss": 0.1021, "step": 12161 }, { "epoch": 5.723294117647059, "grad_norm": 0.8044149991033088, "learning_rate": 2.2091734079195565e-06, "loss": 0.2812, "step": 12162 }, { "epoch": 5.723764705882353, "grad_norm": 1.3318128202442876, "learning_rate": 2.2016840229741593e-06, "loss": 1.1562, "step": 12163 }, { "epoch": 5.724235294117647, "grad_norm": 0.919237705035914, "learning_rate": 2.1942072607299135e-06, "loss": 0.9258, "step": 12164 }, { "epoch": 5.724705882352941, "grad_norm": 0.7147217877366082, "learning_rate": 2.1867431218253406e-06, "loss": 0.1836, "step": 12165 }, { "epoch": 5.725176470588235, "grad_norm": 0.6927406624545843, "learning_rate": 2.179291606897915e-06, "loss": 0.5039, "step": 12166 }, { "epoch": 5.7256470588235295, "grad_norm": 1.284157293892044, "learning_rate": 2.171852716584044e-06, "loss": 0.707, "step": 12167 }, { "epoch": 5.726117647058824, "grad_norm": 1.4046927720758506, "learning_rate": 2.1644264515190524e-06, "loss": 1.2969, "step": 12168 }, { "epoch": 5.726588235294118, "grad_norm": 1.0352969109713988, "learning_rate": 2.1570128123371667e-06, "loss": 0.625, "step": 12169 }, { "epoch": 5.727058823529411, "grad_norm": 1.2114883785156108, "learning_rate": 2.1496117996715636e-06, "loss": 0.5039, "step": 12170 }, { "epoch": 5.7275294117647055, "grad_norm": 0.7204722117915011, "learning_rate": 2.142223414154287e-06, "loss": 0.4668, "step": 12171 }, { "epoch": 5.728, "grad_norm": 1.4437547192204545, "learning_rate": 2.1348476564163997e-06, "loss": 1.3281, "step": 12172 }, { "epoch": 5.728470588235294, "grad_norm": 1.1846147933163722, "learning_rate": 2.1274845270877804e-06, "loss": 0.8516, "step": 12173 }, { "epoch": 5.728941176470588, "grad_norm": 1.0157395813319197, "learning_rate": 2.1201340267972933e-06, "loss": 1.0547, "step": 12174 }, { "epoch": 5.729411764705882, "grad_norm": 1.184811562078706, "learning_rate": 2.1127961561727193e-06, "loss": 0.7891, "step": 12175 }, { "epoch": 5.729882352941177, "grad_norm": 0.8325210096397164, "learning_rate": 2.1054709158407235e-06, "loss": 0.3203, "step": 12176 }, { "epoch": 5.730352941176471, "grad_norm": 1.119850155385451, "learning_rate": 2.0981583064269226e-06, "loss": 0.9531, "step": 12177 }, { "epoch": 5.730823529411765, "grad_norm": 0.7118853652969863, "learning_rate": 2.0908583285558834e-06, "loss": 0.416, "step": 12178 }, { "epoch": 5.731294117647058, "grad_norm": 0.7054368683872155, "learning_rate": 2.083570982851007e-06, "loss": 0.2119, "step": 12179 }, { "epoch": 5.731764705882353, "grad_norm": 0.8370425160353906, "learning_rate": 2.0762962699346965e-06, "loss": 0.3555, "step": 12180 }, { "epoch": 5.732235294117647, "grad_norm": 1.1975515068439626, "learning_rate": 2.06903419042827e-06, "loss": 0.7773, "step": 12181 }, { "epoch": 5.732705882352941, "grad_norm": 1.4404408902500927, "learning_rate": 2.0617847449519e-06, "loss": 0.7344, "step": 12182 }, { "epoch": 5.733176470588235, "grad_norm": 0.9869018884559851, "learning_rate": 2.054547934124756e-06, "loss": 1.0469, "step": 12183 }, { "epoch": 5.7336470588235295, "grad_norm": 0.7143528913993614, "learning_rate": 2.047323758564878e-06, "loss": 0.2168, "step": 12184 }, { "epoch": 5.734117647058824, "grad_norm": 1.0562561716336234, "learning_rate": 2.040112218889256e-06, "loss": 0.6172, "step": 12185 }, { "epoch": 5.734588235294118, "grad_norm": 1.104370411745543, "learning_rate": 2.0329133157137965e-06, "loss": 0.6641, "step": 12186 }, { "epoch": 5.735058823529412, "grad_norm": 1.0107013215362382, "learning_rate": 2.0257270496533086e-06, "loss": 0.8086, "step": 12187 }, { "epoch": 5.7355294117647055, "grad_norm": 1.0897809590177092, "learning_rate": 2.0185534213215506e-06, "loss": 0.7734, "step": 12188 }, { "epoch": 5.736, "grad_norm": 1.0936719579939103, "learning_rate": 2.011392431331166e-06, "loss": 0.3262, "step": 12189 }, { "epoch": 5.736470588235294, "grad_norm": 1.3176872499318066, "learning_rate": 2.0042440802937485e-06, "loss": 0.543, "step": 12190 }, { "epoch": 5.736941176470588, "grad_norm": 0.991942587380315, "learning_rate": 1.9971083688197943e-06, "loss": 0.6562, "step": 12191 }, { "epoch": 5.737411764705882, "grad_norm": 0.6893804902269677, "learning_rate": 1.989985297518748e-06, "loss": 0.4336, "step": 12192 }, { "epoch": 5.737882352941177, "grad_norm": 1.5572014155742748, "learning_rate": 1.9828748669989405e-06, "loss": 0.7969, "step": 12193 }, { "epoch": 5.738352941176471, "grad_norm": 1.1870108013133631, "learning_rate": 1.9757770778676183e-06, "loss": 0.4395, "step": 12194 }, { "epoch": 5.738823529411764, "grad_norm": 1.1041440083586105, "learning_rate": 1.9686919307310145e-06, "loss": 0.5781, "step": 12195 }, { "epoch": 5.739294117647058, "grad_norm": 0.7702342907701758, "learning_rate": 1.961619426194194e-06, "loss": 0.4922, "step": 12196 }, { "epoch": 5.739764705882353, "grad_norm": 1.092762703133432, "learning_rate": 1.9545595648612078e-06, "loss": 0.7617, "step": 12197 }, { "epoch": 5.740235294117647, "grad_norm": 0.8504134842289703, "learning_rate": 1.9475123473349897e-06, "loss": 0.5977, "step": 12198 }, { "epoch": 5.740705882352941, "grad_norm": 1.0386045936827937, "learning_rate": 1.9404777742174083e-06, "loss": 0.3652, "step": 12199 }, { "epoch": 5.741176470588235, "grad_norm": 0.6415586682666453, "learning_rate": 1.9334558461092663e-06, "loss": 0.1147, "step": 12200 }, { "epoch": 5.7416470588235295, "grad_norm": 0.8552966154788111, "learning_rate": 1.9264465636102342e-06, "loss": 0.4492, "step": 12201 }, { "epoch": 5.742117647058824, "grad_norm": 1.355131099501951, "learning_rate": 1.9194499273189834e-06, "loss": 0.5469, "step": 12202 }, { "epoch": 5.742588235294118, "grad_norm": 1.288208003624945, "learning_rate": 1.912465937833035e-06, "loss": 0.6367, "step": 12203 }, { "epoch": 5.743058823529411, "grad_norm": 0.8758281273939957, "learning_rate": 1.9054945957488466e-06, "loss": 0.7422, "step": 12204 }, { "epoch": 5.7435294117647056, "grad_norm": 1.1202738955114095, "learning_rate": 1.898535901661824e-06, "loss": 0.4141, "step": 12205 }, { "epoch": 5.744, "grad_norm": 0.8429049245797677, "learning_rate": 1.8915898561662923e-06, "loss": 0.3691, "step": 12206 }, { "epoch": 5.744470588235294, "grad_norm": 0.5315217546042964, "learning_rate": 1.884656459855427e-06, "loss": 0.2246, "step": 12207 }, { "epoch": 5.744941176470588, "grad_norm": 1.3910569964671204, "learning_rate": 1.8777357133214044e-06, "loss": 0.6992, "step": 12208 }, { "epoch": 5.745411764705882, "grad_norm": 0.9550379250999931, "learning_rate": 1.8708276171553017e-06, "loss": 1.2422, "step": 12209 }, { "epoch": 5.745882352941177, "grad_norm": 1.1676632352813614, "learning_rate": 1.86393217194708e-06, "loss": 0.4805, "step": 12210 }, { "epoch": 5.746352941176471, "grad_norm": 1.0716253526603896, "learning_rate": 1.857049378285652e-06, "loss": 0.4746, "step": 12211 }, { "epoch": 5.746823529411765, "grad_norm": 1.0234627036399186, "learning_rate": 1.8501792367588642e-06, "loss": 0.6992, "step": 12212 }, { "epoch": 5.747294117647058, "grad_norm": 0.9550992748385798, "learning_rate": 1.84332174795343e-06, "loss": 0.8359, "step": 12213 }, { "epoch": 5.747764705882353, "grad_norm": 0.854698188261507, "learning_rate": 1.8364769124550149e-06, "loss": 0.2969, "step": 12214 }, { "epoch": 5.748235294117647, "grad_norm": 1.209434433295653, "learning_rate": 1.8296447308482176e-06, "loss": 0.6602, "step": 12215 }, { "epoch": 5.748705882352941, "grad_norm": 0.955991039681084, "learning_rate": 1.8228252037165215e-06, "loss": 0.8867, "step": 12216 }, { "epoch": 5.749176470588235, "grad_norm": 0.6132118216592523, "learning_rate": 1.8160183316423605e-06, "loss": 0.1553, "step": 12217 }, { "epoch": 5.7496470588235296, "grad_norm": 1.079423972476024, "learning_rate": 1.8092241152070696e-06, "loss": 0.6211, "step": 12218 }, { "epoch": 5.750117647058824, "grad_norm": 0.9641332881312005, "learning_rate": 1.8024425549909183e-06, "loss": 0.4492, "step": 12219 }, { "epoch": 5.750588235294118, "grad_norm": 1.3767825597435186, "learning_rate": 1.795673651573093e-06, "loss": 0.9922, "step": 12220 }, { "epoch": 5.751058823529412, "grad_norm": 1.0217958735980808, "learning_rate": 1.788917405531648e-06, "loss": 0.457, "step": 12221 }, { "epoch": 5.751529411764706, "grad_norm": 1.1981448810280981, "learning_rate": 1.7821738174436384e-06, "loss": 1.3359, "step": 12222 }, { "epoch": 5.752, "grad_norm": 1.0816510193912443, "learning_rate": 1.7754428878849868e-06, "loss": 0.8789, "step": 12223 }, { "epoch": 5.752470588235294, "grad_norm": 1.4230241221069562, "learning_rate": 1.7687246174305502e-06, "loss": 1.2188, "step": 12224 }, { "epoch": 5.752941176470588, "grad_norm": 1.2932446279922976, "learning_rate": 1.762019006654103e-06, "loss": 0.7461, "step": 12225 }, { "epoch": 5.7534117647058824, "grad_norm": 1.0789261952586673, "learning_rate": 1.7553260561283533e-06, "loss": 0.7266, "step": 12226 }, { "epoch": 5.753882352941177, "grad_norm": 1.440055721207568, "learning_rate": 1.748645766424861e-06, "loss": 0.707, "step": 12227 }, { "epoch": 5.754352941176471, "grad_norm": 0.763352421136889, "learning_rate": 1.7419781381142028e-06, "loss": 0.3477, "step": 12228 }, { "epoch": 5.754823529411764, "grad_norm": 1.1338562717937783, "learning_rate": 1.7353231717658233e-06, "loss": 0.5391, "step": 12229 }, { "epoch": 5.7552941176470584, "grad_norm": 1.240515295424528, "learning_rate": 1.728680867948068e-06, "loss": 0.7734, "step": 12230 }, { "epoch": 5.755764705882353, "grad_norm": 1.5383085334206543, "learning_rate": 1.7220512272282328e-06, "loss": 1.5156, "step": 12231 }, { "epoch": 5.756235294117647, "grad_norm": 0.9675515814596426, "learning_rate": 1.7154342501725316e-06, "loss": 0.4805, "step": 12232 }, { "epoch": 5.756705882352941, "grad_norm": 1.0772551149809366, "learning_rate": 1.708829937346079e-06, "loss": 0.9609, "step": 12233 }, { "epoch": 5.757176470588235, "grad_norm": 0.7247109323006043, "learning_rate": 1.7022382893129072e-06, "loss": 0.1602, "step": 12234 }, { "epoch": 5.75764705882353, "grad_norm": 0.7450470247044502, "learning_rate": 1.6956593066359825e-06, "loss": 0.2266, "step": 12235 }, { "epoch": 5.758117647058824, "grad_norm": 0.9951777376824925, "learning_rate": 1.6890929898771722e-06, "loss": 0.2041, "step": 12236 }, { "epoch": 5.758588235294118, "grad_norm": 0.9165965245644038, "learning_rate": 1.6825393395972942e-06, "loss": 0.3047, "step": 12237 }, { "epoch": 5.759058823529411, "grad_norm": 0.7398589159478354, "learning_rate": 1.6759983563560508e-06, "loss": 0.5391, "step": 12238 }, { "epoch": 5.759529411764706, "grad_norm": 1.0224773728613874, "learning_rate": 1.669470040712062e-06, "loss": 0.8867, "step": 12239 }, { "epoch": 5.76, "grad_norm": 1.102994498607953, "learning_rate": 1.6629543932228817e-06, "loss": 0.5078, "step": 12240 }, { "epoch": 5.760470588235294, "grad_norm": 0.9913177424023552, "learning_rate": 1.6564514144449982e-06, "loss": 0.332, "step": 12241 }, { "epoch": 5.760941176470588, "grad_norm": 1.4685121464414161, "learning_rate": 1.6499611049337835e-06, "loss": 0.8555, "step": 12242 }, { "epoch": 5.7614117647058825, "grad_norm": 0.6834444695149473, "learning_rate": 1.6434834652435447e-06, "loss": 0.1055, "step": 12243 }, { "epoch": 5.761882352941177, "grad_norm": 1.016748948172931, "learning_rate": 1.6370184959275058e-06, "loss": 0.957, "step": 12244 }, { "epoch": 5.762352941176471, "grad_norm": 1.1244511069830312, "learning_rate": 1.6305661975377916e-06, "loss": 0.9102, "step": 12245 }, { "epoch": 5.762823529411765, "grad_norm": 1.2271090136200535, "learning_rate": 1.6241265706254947e-06, "loss": 0.6797, "step": 12246 }, { "epoch": 5.7632941176470585, "grad_norm": 0.811273336212705, "learning_rate": 1.617699615740542e-06, "loss": 0.1816, "step": 12247 }, { "epoch": 5.763764705882353, "grad_norm": 1.4573559411870374, "learning_rate": 1.6112853334318775e-06, "loss": 0.6914, "step": 12248 }, { "epoch": 5.764235294117647, "grad_norm": 0.8236639696661195, "learning_rate": 1.6048837242472634e-06, "loss": 0.4746, "step": 12249 }, { "epoch": 5.764705882352941, "grad_norm": 0.909249545819014, "learning_rate": 1.598494788733462e-06, "loss": 0.3125, "step": 12250 }, { "epoch": 5.765176470588235, "grad_norm": 0.9431920606546391, "learning_rate": 1.5921185274361203e-06, "loss": 1.2109, "step": 12251 }, { "epoch": 5.76564705882353, "grad_norm": 0.9190381747952744, "learning_rate": 1.585754940899786e-06, "loss": 0.4531, "step": 12252 }, { "epoch": 5.766117647058824, "grad_norm": 1.1676901923804879, "learning_rate": 1.5794040296679411e-06, "loss": 0.8555, "step": 12253 }, { "epoch": 5.766588235294118, "grad_norm": 0.9911786038000842, "learning_rate": 1.5730657942830183e-06, "loss": 0.5117, "step": 12254 }, { "epoch": 5.767058823529412, "grad_norm": 1.039116393449271, "learning_rate": 1.5667402352862846e-06, "loss": 0.8555, "step": 12255 }, { "epoch": 5.767529411764706, "grad_norm": 1.2172102293651632, "learning_rate": 1.5604273532179913e-06, "loss": 0.8516, "step": 12256 }, { "epoch": 5.768, "grad_norm": 1.4551307538515974, "learning_rate": 1.554127148617307e-06, "loss": 0.5859, "step": 12257 }, { "epoch": 5.768470588235294, "grad_norm": 1.0590360690775544, "learning_rate": 1.5478396220222678e-06, "loss": 0.7891, "step": 12258 }, { "epoch": 5.768941176470588, "grad_norm": 1.0193798011934494, "learning_rate": 1.5415647739698944e-06, "loss": 0.3848, "step": 12259 }, { "epoch": 5.7694117647058825, "grad_norm": 1.3296319391449265, "learning_rate": 1.5353026049960748e-06, "loss": 1.2266, "step": 12260 }, { "epoch": 5.769882352941177, "grad_norm": 0.7325090929694896, "learning_rate": 1.5290531156356146e-06, "loss": 0.2676, "step": 12261 }, { "epoch": 5.770352941176471, "grad_norm": 1.2526128481290906, "learning_rate": 1.5228163064222698e-06, "loss": 0.707, "step": 12262 }, { "epoch": 5.770823529411764, "grad_norm": 1.0103366472887825, "learning_rate": 1.516592177888698e-06, "loss": 0.9375, "step": 12263 }, { "epoch": 5.7712941176470585, "grad_norm": 0.7492392707231127, "learning_rate": 1.5103807305664573e-06, "loss": 0.6289, "step": 12264 }, { "epoch": 5.771764705882353, "grad_norm": 1.0432060184808534, "learning_rate": 1.5041819649860399e-06, "loss": 0.377, "step": 12265 }, { "epoch": 5.772235294117647, "grad_norm": 1.3649409980460137, "learning_rate": 1.4979958816768556e-06, "loss": 1.1016, "step": 12266 }, { "epoch": 5.772705882352941, "grad_norm": 1.1850687520777161, "learning_rate": 1.4918224811672153e-06, "loss": 0.8359, "step": 12267 }, { "epoch": 5.773176470588235, "grad_norm": 1.266661825717207, "learning_rate": 1.485661763984347e-06, "loss": 0.8633, "step": 12268 }, { "epoch": 5.77364705882353, "grad_norm": 0.9333093206130209, "learning_rate": 1.479513730654447e-06, "loss": 0.5273, "step": 12269 }, { "epoch": 5.774117647058824, "grad_norm": 0.8871279316596862, "learning_rate": 1.4733783817025613e-06, "loss": 0.3066, "step": 12270 }, { "epoch": 5.774588235294118, "grad_norm": 0.8955686601969907, "learning_rate": 1.4672557176526712e-06, "loss": 0.2988, "step": 12271 }, { "epoch": 5.775058823529411, "grad_norm": 1.0978385201515406, "learning_rate": 1.4611457390276915e-06, "loss": 1.0312, "step": 12272 }, { "epoch": 5.775529411764706, "grad_norm": 0.9799025328908355, "learning_rate": 1.4550484463494384e-06, "loss": 0.3086, "step": 12273 }, { "epoch": 5.776, "grad_norm": 0.9279351830044891, "learning_rate": 1.4489638401386783e-06, "loss": 0.6797, "step": 12274 }, { "epoch": 5.776470588235294, "grad_norm": 1.0438865292891981, "learning_rate": 1.442891920915029e-06, "loss": 0.3066, "step": 12275 }, { "epoch": 5.776941176470588, "grad_norm": 1.1296403678866669, "learning_rate": 1.436832689197076e-06, "loss": 0.8281, "step": 12276 }, { "epoch": 5.7774117647058825, "grad_norm": 1.1117321753089593, "learning_rate": 1.4307861455023218e-06, "loss": 0.7812, "step": 12277 }, { "epoch": 5.777882352941177, "grad_norm": 0.9523962746342728, "learning_rate": 1.4247522903471531e-06, "loss": 0.3379, "step": 12278 }, { "epoch": 5.778352941176471, "grad_norm": 1.0057040083358033, "learning_rate": 1.4187311242468914e-06, "loss": 0.3633, "step": 12279 }, { "epoch": 5.778823529411765, "grad_norm": 1.1119769066060363, "learning_rate": 1.4127226477157915e-06, "loss": 0.9062, "step": 12280 }, { "epoch": 5.7792941176470585, "grad_norm": 0.6559705307470307, "learning_rate": 1.4067268612669769e-06, "loss": 0.5547, "step": 12281 }, { "epoch": 5.779764705882353, "grad_norm": 0.2688205979899552, "learning_rate": 1.4007437654125375e-06, "loss": 0.0762, "step": 12282 }, { "epoch": 5.780235294117647, "grad_norm": 1.154908281801955, "learning_rate": 1.3947733606634647e-06, "loss": 0.7891, "step": 12283 }, { "epoch": 5.780705882352941, "grad_norm": 1.1215278604930847, "learning_rate": 1.388815647529634e-06, "loss": 0.8906, "step": 12284 }, { "epoch": 5.781176470588235, "grad_norm": 1.1118420076650504, "learning_rate": 1.3828706265199051e-06, "loss": 0.7422, "step": 12285 }, { "epoch": 5.78164705882353, "grad_norm": 1.3639391145805986, "learning_rate": 1.376938298141972e-06, "loss": 0.7812, "step": 12286 }, { "epoch": 5.782117647058824, "grad_norm": 1.03538697008873, "learning_rate": 1.3710186629025121e-06, "loss": 0.8281, "step": 12287 }, { "epoch": 5.782588235294118, "grad_norm": 0.5501620220611713, "learning_rate": 1.3651117213070716e-06, "loss": 0.124, "step": 12288 }, { "epoch": 5.783058823529412, "grad_norm": 1.4588633795646009, "learning_rate": 1.3592174738601302e-06, "loss": 0.543, "step": 12289 }, { "epoch": 5.783529411764706, "grad_norm": 1.0071232708592013, "learning_rate": 1.3533359210651184e-06, "loss": 0.3281, "step": 12290 }, { "epoch": 5.784, "grad_norm": 0.9853109202996806, "learning_rate": 1.3474670634243178e-06, "loss": 0.7539, "step": 12291 }, { "epoch": 5.784470588235294, "grad_norm": 1.0509597601639586, "learning_rate": 1.341610901438961e-06, "loss": 1.0391, "step": 12292 }, { "epoch": 5.784941176470588, "grad_norm": 0.7593287335024618, "learning_rate": 1.3357674356091974e-06, "loss": 0.1953, "step": 12293 }, { "epoch": 5.7854117647058825, "grad_norm": 0.7927875149681161, "learning_rate": 1.3299366664340948e-06, "loss": 0.5977, "step": 12294 }, { "epoch": 5.785882352941177, "grad_norm": 0.48606309736924563, "learning_rate": 1.3241185944116217e-06, "loss": 0.1123, "step": 12295 }, { "epoch": 5.786352941176471, "grad_norm": 0.9231696721329451, "learning_rate": 1.3183132200386637e-06, "loss": 0.2432, "step": 12296 }, { "epoch": 5.786823529411764, "grad_norm": 0.648425187441983, "learning_rate": 1.3125205438110408e-06, "loss": 0.1035, "step": 12297 }, { "epoch": 5.7872941176470585, "grad_norm": 1.0813376393603238, "learning_rate": 1.3067405662234743e-06, "loss": 0.3223, "step": 12298 }, { "epoch": 5.787764705882353, "grad_norm": 1.2472272069400185, "learning_rate": 1.3009732877696022e-06, "loss": 1.4062, "step": 12299 }, { "epoch": 5.788235294117647, "grad_norm": 1.0074008121557132, "learning_rate": 1.2952187089419642e-06, "loss": 0.4629, "step": 12300 }, { "epoch": 5.788705882352941, "grad_norm": 0.8471856103042722, "learning_rate": 1.28947683023205e-06, "loss": 0.25, "step": 12301 }, { "epoch": 5.789176470588235, "grad_norm": 1.0180559995044927, "learning_rate": 1.2837476521302514e-06, "loss": 0.3633, "step": 12302 }, { "epoch": 5.78964705882353, "grad_norm": 1.4918207442846119, "learning_rate": 1.2780311751258265e-06, "loss": 0.8203, "step": 12303 }, { "epoch": 5.790117647058824, "grad_norm": 0.8621668847553874, "learning_rate": 1.2723273997070182e-06, "loss": 0.1963, "step": 12304 }, { "epoch": 5.790588235294118, "grad_norm": 1.2016933154758564, "learning_rate": 1.2666363263609702e-06, "loss": 0.5898, "step": 12305 }, { "epoch": 5.791058823529411, "grad_norm": 1.180218435648326, "learning_rate": 1.2609579555737104e-06, "loss": 0.8398, "step": 12306 }, { "epoch": 5.791529411764706, "grad_norm": 0.9572038584204733, "learning_rate": 1.2552922878301841e-06, "loss": 0.3984, "step": 12307 }, { "epoch": 5.792, "grad_norm": 1.3326981815252168, "learning_rate": 1.2496393236143044e-06, "loss": 0.8477, "step": 12308 }, { "epoch": 5.792470588235294, "grad_norm": 1.1142751141233425, "learning_rate": 1.2439990634088348e-06, "loss": 0.6328, "step": 12309 }, { "epoch": 5.792941176470588, "grad_norm": 0.9831036306293361, "learning_rate": 1.2383715076954902e-06, "loss": 0.5547, "step": 12310 }, { "epoch": 5.7934117647058825, "grad_norm": 1.3412304009154496, "learning_rate": 1.232756656954903e-06, "loss": 1.2188, "step": 12311 }, { "epoch": 5.793882352941177, "grad_norm": 0.6441701024676372, "learning_rate": 1.2271545116665727e-06, "loss": 0.209, "step": 12312 }, { "epoch": 5.794352941176471, "grad_norm": 0.9951784599179027, "learning_rate": 1.2215650723089832e-06, "loss": 0.2715, "step": 12313 }, { "epoch": 5.794823529411765, "grad_norm": 1.2374376653927557, "learning_rate": 1.2159883393595026e-06, "loss": 0.4434, "step": 12314 }, { "epoch": 5.7952941176470585, "grad_norm": 0.8998390977747442, "learning_rate": 1.2104243132943836e-06, "loss": 0.4629, "step": 12315 }, { "epoch": 5.795764705882353, "grad_norm": 0.9750832758687615, "learning_rate": 1.2048729945888458e-06, "loss": 0.5508, "step": 12316 }, { "epoch": 5.796235294117647, "grad_norm": 0.5722072826204087, "learning_rate": 1.1993343837169933e-06, "loss": 0.334, "step": 12317 }, { "epoch": 5.796705882352941, "grad_norm": 1.179320908830275, "learning_rate": 1.1938084811518311e-06, "loss": 1.3203, "step": 12318 }, { "epoch": 5.797176470588235, "grad_norm": 1.262962169038733, "learning_rate": 1.1882952873653317e-06, "loss": 0.832, "step": 12319 }, { "epoch": 5.79764705882353, "grad_norm": 0.6948749650799436, "learning_rate": 1.1827948028283352e-06, "loss": 0.1992, "step": 12320 }, { "epoch": 5.798117647058824, "grad_norm": 0.6532981884854178, "learning_rate": 1.1773070280105822e-06, "loss": 0.1138, "step": 12321 }, { "epoch": 5.798588235294117, "grad_norm": 0.5834548573252147, "learning_rate": 1.171831963380815e-06, "loss": 0.0767, "step": 12322 }, { "epoch": 5.799058823529411, "grad_norm": 1.4429882754183454, "learning_rate": 1.1663696094065755e-06, "loss": 0.9922, "step": 12323 }, { "epoch": 5.799529411764706, "grad_norm": 0.8728974835600996, "learning_rate": 1.160919966554391e-06, "loss": 0.3672, "step": 12324 }, { "epoch": 5.8, "grad_norm": 0.7467168784053232, "learning_rate": 1.155483035289706e-06, "loss": 0.2412, "step": 12325 }, { "epoch": 5.800470588235294, "grad_norm": 1.1665805789768144, "learning_rate": 1.150058816076832e-06, "loss": 0.4609, "step": 12326 }, { "epoch": 5.800941176470588, "grad_norm": 0.835279676652282, "learning_rate": 1.1446473093790487e-06, "loss": 0.2275, "step": 12327 }, { "epoch": 5.8014117647058825, "grad_norm": 1.0858341210592348, "learning_rate": 1.1392485156585196e-06, "loss": 0.832, "step": 12328 }, { "epoch": 5.801882352941177, "grad_norm": 1.6266158731685492, "learning_rate": 1.1338624353763093e-06, "loss": 0.5781, "step": 12329 }, { "epoch": 5.802352941176471, "grad_norm": 0.9046490225542362, "learning_rate": 1.1284890689924497e-06, "loss": 0.6719, "step": 12330 }, { "epoch": 5.802823529411764, "grad_norm": 0.977664847154124, "learning_rate": 1.1231284169658239e-06, "loss": 0.8398, "step": 12331 }, { "epoch": 5.8032941176470585, "grad_norm": 1.1029312664439073, "learning_rate": 1.1177804797542821e-06, "loss": 1.0234, "step": 12332 }, { "epoch": 5.803764705882353, "grad_norm": 1.2477674657721596, "learning_rate": 1.1124452578145425e-06, "loss": 1.5859, "step": 12333 }, { "epoch": 5.804235294117647, "grad_norm": 0.7486617189773486, "learning_rate": 1.1071227516022574e-06, "loss": 0.1865, "step": 12334 }, { "epoch": 5.804705882352941, "grad_norm": 1.0859522896578884, "learning_rate": 1.1018129615720129e-06, "loss": 0.5156, "step": 12335 }, { "epoch": 5.805176470588235, "grad_norm": 0.9492032424925366, "learning_rate": 1.0965158881772961e-06, "loss": 0.3066, "step": 12336 }, { "epoch": 5.80564705882353, "grad_norm": 1.1110450991179222, "learning_rate": 1.091231531870479e-06, "loss": 1.1094, "step": 12337 }, { "epoch": 5.806117647058824, "grad_norm": 1.2663068003259372, "learning_rate": 1.0859598931028668e-06, "loss": 1.3906, "step": 12338 }, { "epoch": 5.806588235294118, "grad_norm": 1.2441382356064354, "learning_rate": 1.0807009723247328e-06, "loss": 0.4629, "step": 12339 }, { "epoch": 5.807058823529411, "grad_norm": 0.9516401284960135, "learning_rate": 1.0754547699851513e-06, "loss": 0.2637, "step": 12340 }, { "epoch": 5.807529411764706, "grad_norm": 1.5095934443487176, "learning_rate": 1.0702212865322135e-06, "loss": 1.7188, "step": 12341 }, { "epoch": 5.808, "grad_norm": 1.27491991030451, "learning_rate": 1.0650005224128788e-06, "loss": 0.6133, "step": 12342 }, { "epoch": 5.808470588235294, "grad_norm": 0.7958694466741588, "learning_rate": 1.059792478073007e-06, "loss": 0.75, "step": 12343 }, { "epoch": 5.808941176470588, "grad_norm": 1.3419268520005978, "learning_rate": 1.0545971539574094e-06, "loss": 1.0469, "step": 12344 }, { "epoch": 5.8094117647058825, "grad_norm": 1.377270224611854, "learning_rate": 1.0494145505097973e-06, "loss": 0.9883, "step": 12345 }, { "epoch": 5.809882352941177, "grad_norm": 0.7249923219514237, "learning_rate": 1.0442446681727834e-06, "loss": 0.1631, "step": 12346 }, { "epoch": 5.810352941176471, "grad_norm": 0.7710780228119831, "learning_rate": 1.0390875073878812e-06, "loss": 0.166, "step": 12347 }, { "epoch": 5.810823529411765, "grad_norm": 1.1015161448774637, "learning_rate": 1.0339430685955719e-06, "loss": 0.3184, "step": 12348 }, { "epoch": 5.8112941176470585, "grad_norm": 0.904660641785306, "learning_rate": 1.0288113522352038e-06, "loss": 0.4004, "step": 12349 }, { "epoch": 5.811764705882353, "grad_norm": 1.4450782334484489, "learning_rate": 1.0236923587450263e-06, "loss": 1.1797, "step": 12350 }, { "epoch": 5.812235294117647, "grad_norm": 1.385933675163474, "learning_rate": 1.0185860885622565e-06, "loss": 0.8086, "step": 12351 }, { "epoch": 5.812705882352941, "grad_norm": 1.181070698359397, "learning_rate": 1.013492542122979e-06, "loss": 0.5312, "step": 12352 }, { "epoch": 5.813176470588235, "grad_norm": 1.0161501111446718, "learning_rate": 1.0084117198622287e-06, "loss": 0.707, "step": 12353 }, { "epoch": 5.81364705882353, "grad_norm": 1.0756441013575033, "learning_rate": 1.0033436222138924e-06, "loss": 0.8867, "step": 12354 }, { "epoch": 5.814117647058824, "grad_norm": 0.7572960648608515, "learning_rate": 9.98288249610857e-07, "loss": 0.1631, "step": 12355 }, { "epoch": 5.814588235294117, "grad_norm": 0.6900095539625268, "learning_rate": 9.932456024848435e-07, "loss": 0.3457, "step": 12356 }, { "epoch": 5.815058823529411, "grad_norm": 1.0867279520514768, "learning_rate": 9.882156812665244e-07, "loss": 0.5234, "step": 12357 }, { "epoch": 5.815529411764706, "grad_norm": 0.3187125235405837, "learning_rate": 9.831984863854891e-07, "loss": 0.0684, "step": 12358 }, { "epoch": 5.816, "grad_norm": 0.9510183305582459, "learning_rate": 9.781940182702451e-07, "loss": 0.5664, "step": 12359 }, { "epoch": 5.816470588235294, "grad_norm": 1.0575714267857719, "learning_rate": 9.732022773481506e-07, "loss": 0.2324, "step": 12360 }, { "epoch": 5.816941176470588, "grad_norm": 0.9693178366886861, "learning_rate": 9.682232640455644e-07, "loss": 0.8164, "step": 12361 }, { "epoch": 5.8174117647058825, "grad_norm": 1.5398929822739842, "learning_rate": 9.632569787877298e-07, "loss": 0.7891, "step": 12362 }, { "epoch": 5.817882352941177, "grad_norm": 0.6921075865485296, "learning_rate": 9.583034219987406e-07, "loss": 0.5195, "step": 12363 }, { "epoch": 5.818352941176471, "grad_norm": 1.1355856257840737, "learning_rate": 9.53362594101692e-07, "loss": 0.4512, "step": 12364 }, { "epoch": 5.818823529411764, "grad_norm": 1.006413799118567, "learning_rate": 9.48434495518563e-07, "loss": 0.2754, "step": 12365 }, { "epoch": 5.8192941176470585, "grad_norm": 0.8987384497958666, "learning_rate": 9.435191266702003e-07, "loss": 0.3086, "step": 12366 }, { "epoch": 5.819764705882353, "grad_norm": 0.6642007840544867, "learning_rate": 9.386164879764346e-07, "loss": 0.2021, "step": 12367 }, { "epoch": 5.820235294117647, "grad_norm": 0.742814201967705, "learning_rate": 9.337265798559645e-07, "loss": 0.2256, "step": 12368 }, { "epoch": 5.820705882352941, "grad_norm": 0.4345975403392816, "learning_rate": 9.28849402726406e-07, "loss": 0.0391, "step": 12369 }, { "epoch": 5.821176470588235, "grad_norm": 1.293160119253435, "learning_rate": 9.239849570042924e-07, "loss": 0.832, "step": 12370 }, { "epoch": 5.82164705882353, "grad_norm": 1.0492388595575606, "learning_rate": 9.191332431050747e-07, "loss": 1.0781, "step": 12371 }, { "epoch": 5.822117647058824, "grad_norm": 0.8445641054910697, "learning_rate": 9.142942614431215e-07, "loss": 0.4883, "step": 12372 }, { "epoch": 5.822588235294118, "grad_norm": 1.4931631203900302, "learning_rate": 9.094680124317022e-07, "loss": 1.0625, "step": 12373 }, { "epoch": 5.823058823529411, "grad_norm": 1.0536735357851705, "learning_rate": 9.046544964830038e-07, "loss": 0.5742, "step": 12374 }, { "epoch": 5.823529411764706, "grad_norm": 0.6400511626149281, "learning_rate": 8.998537140081141e-07, "loss": 0.1855, "step": 12375 }, { "epoch": 5.824, "grad_norm": 1.0715748660254094, "learning_rate": 8.950656654170552e-07, "loss": 0.6328, "step": 12376 }, { "epoch": 5.824470588235294, "grad_norm": 0.8841630945729585, "learning_rate": 8.902903511187332e-07, "loss": 0.4766, "step": 12377 }, { "epoch": 5.824941176470588, "grad_norm": 1.2750377768778678, "learning_rate": 8.855277715210052e-07, "loss": 0.5312, "step": 12378 }, { "epoch": 5.8254117647058825, "grad_norm": 0.90633752001827, "learning_rate": 8.807779270305959e-07, "loss": 0.6523, "step": 12379 }, { "epoch": 5.825882352941177, "grad_norm": 1.0215114608097624, "learning_rate": 8.760408180531809e-07, "loss": 0.3516, "step": 12380 }, { "epoch": 5.826352941176471, "grad_norm": 0.7579120364005434, "learning_rate": 8.713164449933364e-07, "loss": 0.4043, "step": 12381 }, { "epoch": 5.826823529411765, "grad_norm": 1.3695646140137592, "learning_rate": 8.666048082545396e-07, "loss": 0.7422, "step": 12382 }, { "epoch": 5.8272941176470585, "grad_norm": 1.3008117454978128, "learning_rate": 8.619059082391689e-07, "loss": 1.1641, "step": 12383 }, { "epoch": 5.827764705882353, "grad_norm": 1.3112965765390214, "learning_rate": 8.572197453485696e-07, "loss": 0.9102, "step": 12384 }, { "epoch": 5.828235294117647, "grad_norm": 1.3675283740081532, "learning_rate": 8.525463199829385e-07, "loss": 0.8086, "step": 12385 }, { "epoch": 5.828705882352941, "grad_norm": 0.848418175189584, "learning_rate": 8.47885632541423e-07, "loss": 0.2402, "step": 12386 }, { "epoch": 5.829176470588235, "grad_norm": 0.8770247135249007, "learning_rate": 8.432376834220544e-07, "loss": 0.2988, "step": 12387 }, { "epoch": 5.82964705882353, "grad_norm": 1.0716764158360281, "learning_rate": 8.386024730217988e-07, "loss": 0.3379, "step": 12388 }, { "epoch": 5.830117647058824, "grad_norm": 0.867977726304289, "learning_rate": 8.339800017365228e-07, "loss": 0.2656, "step": 12389 }, { "epoch": 5.830588235294117, "grad_norm": 0.9961759317092775, "learning_rate": 8.293702699610272e-07, "loss": 0.8398, "step": 12390 }, { "epoch": 5.831058823529411, "grad_norm": 1.1257684540074175, "learning_rate": 8.247732780889637e-07, "loss": 0.3848, "step": 12391 }, { "epoch": 5.831529411764706, "grad_norm": 1.7137418731729026, "learning_rate": 8.201890265129684e-07, "loss": 1.3984, "step": 12392 }, { "epoch": 5.832, "grad_norm": 0.9774095693048602, "learning_rate": 8.156175156245614e-07, "loss": 0.9023, "step": 12393 }, { "epoch": 5.832470588235294, "grad_norm": 0.47232832461567864, "learning_rate": 8.110587458141637e-07, "loss": 0.0688, "step": 12394 }, { "epoch": 5.832941176470588, "grad_norm": 0.5791648715699781, "learning_rate": 8.065127174711139e-07, "loss": 0.1143, "step": 12395 }, { "epoch": 5.8334117647058825, "grad_norm": 1.1332675849018978, "learning_rate": 8.01979430983668e-07, "loss": 0.6602, "step": 12396 }, { "epoch": 5.833882352941177, "grad_norm": 0.7872672747204923, "learning_rate": 7.97458886738983e-07, "loss": 0.373, "step": 12397 }, { "epoch": 5.834352941176471, "grad_norm": 0.8662217780621999, "learning_rate": 7.929510851231502e-07, "loss": 0.8633, "step": 12398 }, { "epoch": 5.834823529411764, "grad_norm": 0.63670155563851, "learning_rate": 7.884560265211615e-07, "loss": 0.2383, "step": 12399 }, { "epoch": 5.8352941176470585, "grad_norm": 1.2025458445320454, "learning_rate": 7.839737113168931e-07, "loss": 1.1406, "step": 12400 }, { "epoch": 5.835764705882353, "grad_norm": 0.7614516219127359, "learning_rate": 7.795041398931723e-07, "loss": 0.1875, "step": 12401 }, { "epoch": 5.836235294117647, "grad_norm": 1.11764562959999, "learning_rate": 7.750473126317269e-07, "loss": 0.75, "step": 12402 }, { "epoch": 5.836705882352941, "grad_norm": 0.40862150217204374, "learning_rate": 7.706032299131692e-07, "loss": 0.052, "step": 12403 }, { "epoch": 5.837176470588235, "grad_norm": 0.8909296593512284, "learning_rate": 7.66171892117079e-07, "loss": 0.8359, "step": 12404 }, { "epoch": 5.83764705882353, "grad_norm": 1.4974300166934151, "learning_rate": 7.617532996218867e-07, "loss": 0.8359, "step": 12405 }, { "epoch": 5.838117647058824, "grad_norm": 1.0475262256619824, "learning_rate": 7.573474528049739e-07, "loss": 0.3359, "step": 12406 }, { "epoch": 5.838588235294118, "grad_norm": 0.8638153217686367, "learning_rate": 7.529543520426228e-07, "loss": 0.3906, "step": 12407 }, { "epoch": 5.839058823529411, "grad_norm": 0.6097409095263111, "learning_rate": 7.485739977100336e-07, "loss": 0.1133, "step": 12408 }, { "epoch": 5.839529411764706, "grad_norm": 1.319509903967754, "learning_rate": 7.442063901812734e-07, "loss": 0.5156, "step": 12409 }, { "epoch": 5.84, "grad_norm": 1.1632043556885086, "learning_rate": 7.398515298294106e-07, "loss": 0.8672, "step": 12410 }, { "epoch": 5.840470588235294, "grad_norm": 1.0229526937382232, "learning_rate": 7.355094170263143e-07, "loss": 0.4102, "step": 12411 }, { "epoch": 5.840941176470588, "grad_norm": 0.9369411319744819, "learning_rate": 7.311800521428712e-07, "loss": 0.377, "step": 12412 }, { "epoch": 5.8414117647058825, "grad_norm": 1.1365595389001535, "learning_rate": 7.26863435548819e-07, "loss": 0.7148, "step": 12413 }, { "epoch": 5.841882352941177, "grad_norm": 0.773336454412613, "learning_rate": 7.22559567612796e-07, "loss": 0.2305, "step": 12414 }, { "epoch": 5.842352941176471, "grad_norm": 0.7859612371254601, "learning_rate": 7.182684487023748e-07, "loss": 0.2285, "step": 12415 }, { "epoch": 5.842823529411765, "grad_norm": 1.1701507820194927, "learning_rate": 7.139900791840792e-07, "loss": 0.6992, "step": 12416 }, { "epoch": 5.8432941176470585, "grad_norm": 1.1615114527952197, "learning_rate": 7.097244594232498e-07, "loss": 0.4434, "step": 12417 }, { "epoch": 5.843764705882353, "grad_norm": 1.5250362186483952, "learning_rate": 7.054715897842289e-07, "loss": 1.2734, "step": 12418 }, { "epoch": 5.844235294117647, "grad_norm": 1.13029981971601, "learning_rate": 7.012314706302257e-07, "loss": 0.5977, "step": 12419 }, { "epoch": 5.844705882352941, "grad_norm": 1.3420765086712092, "learning_rate": 6.970041023233508e-07, "loss": 0.9883, "step": 12420 }, { "epoch": 5.845176470588235, "grad_norm": 0.7669101583189795, "learning_rate": 6.927894852246485e-07, "loss": 0.1426, "step": 12421 }, { "epoch": 5.84564705882353, "grad_norm": 1.235535439021394, "learning_rate": 6.88587619694081e-07, "loss": 0.7031, "step": 12422 }, { "epoch": 5.846117647058824, "grad_norm": 1.7152796718761558, "learning_rate": 6.843985060904944e-07, "loss": 1.1562, "step": 12423 }, { "epoch": 5.846588235294117, "grad_norm": 0.8652343196442204, "learning_rate": 6.802221447716694e-07, "loss": 0.1807, "step": 12424 }, { "epoch": 5.847058823529411, "grad_norm": 1.0701920495341104, "learning_rate": 6.760585360942872e-07, "loss": 0.5195, "step": 12425 }, { "epoch": 5.847529411764706, "grad_norm": 1.4423596692044152, "learning_rate": 6.719076804139134e-07, "loss": 0.7305, "step": 12426 }, { "epoch": 5.848, "grad_norm": 1.0016033390314514, "learning_rate": 6.677695780850978e-07, "loss": 1.0547, "step": 12427 }, { "epoch": 5.848470588235294, "grad_norm": 1.0324530673811936, "learning_rate": 6.636442294612243e-07, "loss": 0.7734, "step": 12428 }, { "epoch": 5.848941176470588, "grad_norm": 1.6537853158150724, "learning_rate": 6.595316348946278e-07, "loss": 1.5, "step": 12429 }, { "epoch": 5.8494117647058825, "grad_norm": 1.343901693927667, "learning_rate": 6.554317947365273e-07, "loss": 1.3516, "step": 12430 }, { "epoch": 5.849882352941177, "grad_norm": 1.2099963895126535, "learning_rate": 6.513447093370927e-07, "loss": 0.8281, "step": 12431 }, { "epoch": 5.850352941176471, "grad_norm": 1.1174952695428084, "learning_rate": 6.472703790453781e-07, "loss": 0.4199, "step": 12432 }, { "epoch": 5.850823529411764, "grad_norm": 0.5849395113628965, "learning_rate": 6.43208804209322e-07, "loss": 0.124, "step": 12433 }, { "epoch": 5.8512941176470585, "grad_norm": 0.8227127590802388, "learning_rate": 6.391599851758466e-07, "loss": 0.2363, "step": 12434 }, { "epoch": 5.851764705882353, "grad_norm": 1.1207201650027414, "learning_rate": 6.351239222907089e-07, "loss": 0.6094, "step": 12435 }, { "epoch": 5.852235294117647, "grad_norm": 1.2131551256203648, "learning_rate": 6.311006158986165e-07, "loss": 1.1797, "step": 12436 }, { "epoch": 5.852705882352941, "grad_norm": 1.575698732404555, "learning_rate": 6.270900663431777e-07, "loss": 1.4141, "step": 12437 }, { "epoch": 5.853176470588235, "grad_norm": 0.8440963884198912, "learning_rate": 6.230922739669353e-07, "loss": 0.4258, "step": 12438 }, { "epoch": 5.85364705882353, "grad_norm": 1.5121755265778876, "learning_rate": 6.191072391112828e-07, "loss": 1.5078, "step": 12439 }, { "epoch": 5.854117647058824, "grad_norm": 1.2314273829376832, "learning_rate": 6.151349621165979e-07, "loss": 1.0234, "step": 12440 }, { "epoch": 5.854588235294118, "grad_norm": 1.2154935807858935, "learning_rate": 6.111754433221094e-07, "loss": 0.4336, "step": 12441 }, { "epoch": 5.855058823529411, "grad_norm": 0.9982718644381084, "learning_rate": 6.0722868306598e-07, "loss": 0.6797, "step": 12442 }, { "epoch": 5.855529411764706, "grad_norm": 0.8396964589615541, "learning_rate": 6.032946816852902e-07, "loss": 0.2344, "step": 12443 }, { "epoch": 5.856, "grad_norm": 1.4570376693001712, "learning_rate": 5.99373439516021e-07, "loss": 1.2266, "step": 12444 }, { "epoch": 5.856470588235294, "grad_norm": 1.0719696013965867, "learning_rate": 5.954649568930714e-07, "loss": 0.7227, "step": 12445 }, { "epoch": 5.856941176470588, "grad_norm": 0.9484911555283628, "learning_rate": 5.91569234150241e-07, "loss": 0.6562, "step": 12446 }, { "epoch": 5.8574117647058825, "grad_norm": 1.270251523960922, "learning_rate": 5.876862716202302e-07, "loss": 0.9414, "step": 12447 }, { "epoch": 5.857882352941177, "grad_norm": 0.9223132115222976, "learning_rate": 5.838160696346906e-07, "loss": 0.8633, "step": 12448 }, { "epoch": 5.85835294117647, "grad_norm": 0.7511428704738587, "learning_rate": 5.799586285241242e-07, "loss": 0.1377, "step": 12449 }, { "epoch": 5.858823529411764, "grad_norm": 0.9186510663393935, "learning_rate": 5.761139486180178e-07, "loss": 0.5664, "step": 12450 }, { "epoch": 5.8592941176470585, "grad_norm": 0.919651574359357, "learning_rate": 5.722820302446751e-07, "loss": 0.3379, "step": 12451 }, { "epoch": 5.859764705882353, "grad_norm": 1.2575848718736924, "learning_rate": 5.684628737314012e-07, "loss": 0.5, "step": 12452 }, { "epoch": 5.860235294117647, "grad_norm": 0.9849447015034137, "learning_rate": 5.646564794043518e-07, "loss": 0.3027, "step": 12453 }, { "epoch": 5.860705882352941, "grad_norm": 1.0215247856298124, "learning_rate": 5.608628475886001e-07, "loss": 0.3672, "step": 12454 }, { "epoch": 5.861176470588235, "grad_norm": 0.725202063614507, "learning_rate": 5.570819786081704e-07, "loss": 0.2617, "step": 12455 }, { "epoch": 5.86164705882353, "grad_norm": 0.7852998425557122, "learning_rate": 5.533138727859543e-07, "loss": 0.4434, "step": 12456 }, { "epoch": 5.862117647058824, "grad_norm": 0.9053909857079712, "learning_rate": 5.495585304437611e-07, "loss": 0.3281, "step": 12457 }, { "epoch": 5.862588235294117, "grad_norm": 0.7440082448811502, "learning_rate": 5.458159519023175e-07, "loss": 0.167, "step": 12458 }, { "epoch": 5.863058823529411, "grad_norm": 0.8928019274366081, "learning_rate": 5.42086137481268e-07, "loss": 0.2295, "step": 12459 }, { "epoch": 5.863529411764706, "grad_norm": 1.1803759270620526, "learning_rate": 5.383690874991408e-07, "loss": 0.4199, "step": 12460 }, { "epoch": 5.864, "grad_norm": 0.8866752527529576, "learning_rate": 5.34664802273399e-07, "loss": 1.0078, "step": 12461 }, { "epoch": 5.864470588235294, "grad_norm": 1.2288466412238548, "learning_rate": 5.30973282120406e-07, "loss": 0.9609, "step": 12462 }, { "epoch": 5.864941176470588, "grad_norm": 0.9474050964988481, "learning_rate": 5.272945273554429e-07, "loss": 0.6641, "step": 12463 }, { "epoch": 5.8654117647058825, "grad_norm": 0.8210192085459059, "learning_rate": 5.236285382926752e-07, "loss": 0.1797, "step": 12464 }, { "epoch": 5.865882352941177, "grad_norm": 1.0611411295607838, "learning_rate": 5.19975315245219e-07, "loss": 0.8711, "step": 12465 }, { "epoch": 5.866352941176471, "grad_norm": 1.1067556244216097, "learning_rate": 5.163348585250748e-07, "loss": 0.6719, "step": 12466 }, { "epoch": 5.866823529411764, "grad_norm": 0.8008474306577431, "learning_rate": 5.127071684431272e-07, "loss": 0.3359, "step": 12467 }, { "epoch": 5.8672941176470585, "grad_norm": 0.9572781385974852, "learning_rate": 5.090922453092283e-07, "loss": 0.5977, "step": 12468 }, { "epoch": 5.867764705882353, "grad_norm": 0.8081526972261048, "learning_rate": 5.054900894320979e-07, "loss": 0.293, "step": 12469 }, { "epoch": 5.868235294117647, "grad_norm": 1.1341397754160085, "learning_rate": 5.019007011193732e-07, "loss": 0.4766, "step": 12470 }, { "epoch": 5.868705882352941, "grad_norm": 0.9186469902224245, "learning_rate": 4.983240806776257e-07, "loss": 0.3203, "step": 12471 }, { "epoch": 5.869176470588235, "grad_norm": 1.1648672120263615, "learning_rate": 4.947602284122942e-07, "loss": 0.8398, "step": 12472 }, { "epoch": 5.86964705882353, "grad_norm": 1.3934116669944312, "learning_rate": 4.912091446277688e-07, "loss": 0.8125, "step": 12473 }, { "epoch": 5.870117647058824, "grad_norm": 0.8903173721679928, "learning_rate": 4.876708296273068e-07, "loss": 0.2754, "step": 12474 }, { "epoch": 5.870588235294118, "grad_norm": 1.3085781402646075, "learning_rate": 4.841452837131166e-07, "loss": 1.0, "step": 12475 }, { "epoch": 5.871058823529411, "grad_norm": 0.6920229630774906, "learning_rate": 4.806325071862905e-07, "loss": 0.1738, "step": 12476 }, { "epoch": 5.871529411764706, "grad_norm": 1.032939592648904, "learning_rate": 4.771325003468385e-07, "loss": 0.9609, "step": 12477 }, { "epoch": 5.872, "grad_norm": 1.4304411520813343, "learning_rate": 4.736452634936716e-07, "loss": 1.3438, "step": 12478 }, { "epoch": 5.872470588235294, "grad_norm": 1.4380732211538905, "learning_rate": 4.701707969246182e-07, "loss": 0.9141, "step": 12479 }, { "epoch": 5.872941176470588, "grad_norm": 1.0884332515197794, "learning_rate": 4.6670910093642435e-07, "loss": 0.9062, "step": 12480 }, { "epoch": 5.8734117647058826, "grad_norm": 1.0495487600364493, "learning_rate": 4.6326017582473674e-07, "loss": 0.7734, "step": 12481 }, { "epoch": 5.873882352941177, "grad_norm": 1.0293422200481812, "learning_rate": 4.5982402188408653e-07, "loss": 0.3672, "step": 12482 }, { "epoch": 5.87435294117647, "grad_norm": 0.860793106161992, "learning_rate": 4.5640063940797223e-07, "loss": 0.4258, "step": 12483 }, { "epoch": 5.874823529411764, "grad_norm": 1.116073739297348, "learning_rate": 4.529900286887433e-07, "loss": 1.0078, "step": 12484 }, { "epoch": 5.875294117647059, "grad_norm": 0.6467794676792307, "learning_rate": 4.4959219001768355e-07, "loss": 0.1406, "step": 12485 }, { "epoch": 5.875764705882353, "grad_norm": 0.5434361832850211, "learning_rate": 4.4620712368501065e-07, "loss": 0.0664, "step": 12486 }, { "epoch": 5.876235294117647, "grad_norm": 1.5138485937144681, "learning_rate": 4.428348299797768e-07, "loss": 1.75, "step": 12487 }, { "epoch": 5.876705882352941, "grad_norm": 1.3057118276269497, "learning_rate": 4.394753091900516e-07, "loss": 0.6758, "step": 12488 }, { "epoch": 5.8771764705882354, "grad_norm": 0.46437105625758923, "learning_rate": 4.361285616027055e-07, "loss": 0.1138, "step": 12489 }, { "epoch": 5.87764705882353, "grad_norm": 1.2052742074470666, "learning_rate": 4.327945875035932e-07, "loss": 1.2891, "step": 12490 }, { "epoch": 5.878117647058824, "grad_norm": 1.179628278647903, "learning_rate": 4.2947338717743694e-07, "loss": 0.5781, "step": 12491 }, { "epoch": 5.878588235294117, "grad_norm": 1.015673910244789, "learning_rate": 4.2616496090790983e-07, "loss": 0.3359, "step": 12492 }, { "epoch": 5.8790588235294114, "grad_norm": 0.5614679745508293, "learning_rate": 4.228693089775526e-07, "loss": 0.1162, "step": 12493 }, { "epoch": 5.879529411764706, "grad_norm": 0.7960090139010156, "learning_rate": 4.195864316678066e-07, "loss": 0.2363, "step": 12494 }, { "epoch": 5.88, "grad_norm": 0.8375594363774639, "learning_rate": 4.16316329259081e-07, "loss": 0.5664, "step": 12495 }, { "epoch": 5.880470588235294, "grad_norm": 0.845038547882751, "learning_rate": 4.130590020306523e-07, "loss": 0.3594, "step": 12496 }, { "epoch": 5.880941176470588, "grad_norm": 1.0838697160751938, "learning_rate": 4.0981445026068147e-07, "loss": 0.8359, "step": 12497 }, { "epoch": 5.881411764705883, "grad_norm": 1.6569232360388317, "learning_rate": 4.0658267422631343e-07, "loss": 1.1328, "step": 12498 }, { "epoch": 5.881882352941177, "grad_norm": 1.1126649102376465, "learning_rate": 4.033636742035273e-07, "loss": 1.3438, "step": 12499 }, { "epoch": 5.882352941176471, "grad_norm": 0.7255614601305371, "learning_rate": 4.0015745046725336e-07, "loss": 0.1885, "step": 12500 }, { "epoch": 5.882823529411764, "grad_norm": 0.7936450050264034, "learning_rate": 3.9696400329132236e-07, "loss": 0.5156, "step": 12501 }, { "epoch": 5.883294117647059, "grad_norm": 1.1483693780040711, "learning_rate": 3.9378333294844967e-07, "loss": 0.7031, "step": 12502 }, { "epoch": 5.883764705882353, "grad_norm": 0.9026767852331654, "learning_rate": 3.906154397103012e-07, "loss": 1.0781, "step": 12503 }, { "epoch": 5.884235294117647, "grad_norm": 1.278060760001669, "learning_rate": 3.874603238474272e-07, "loss": 0.6016, "step": 12504 }, { "epoch": 5.884705882352941, "grad_norm": 0.834835704938219, "learning_rate": 3.8431798562927887e-07, "loss": 0.2578, "step": 12505 }, { "epoch": 5.8851764705882355, "grad_norm": 0.8552917243456923, "learning_rate": 3.8118842532424145e-07, "loss": 0.4961, "step": 12506 }, { "epoch": 5.88564705882353, "grad_norm": 1.1474362006123682, "learning_rate": 3.780716431995845e-07, "loss": 0.5312, "step": 12507 }, { "epoch": 5.886117647058824, "grad_norm": 1.0213740213413571, "learning_rate": 3.7496763952151177e-07, "loss": 0.8008, "step": 12508 }, { "epoch": 5.886588235294118, "grad_norm": 0.785105389189194, "learning_rate": 3.7187641455509455e-07, "loss": 0.5, "step": 12509 }, { "epoch": 5.8870588235294115, "grad_norm": 1.426105364302714, "learning_rate": 3.687979685643549e-07, "loss": 0.498, "step": 12510 }, { "epoch": 5.887529411764706, "grad_norm": 1.4092486672086684, "learning_rate": 3.6573230181221604e-07, "loss": 0.9375, "step": 12511 }, { "epoch": 5.888, "grad_norm": 1.2329487866173312, "learning_rate": 3.62679414560485e-07, "loss": 0.9492, "step": 12512 }, { "epoch": 5.888470588235294, "grad_norm": 1.3154012452077306, "learning_rate": 3.596393070699033e-07, "loss": 0.707, "step": 12513 }, { "epoch": 5.888941176470588, "grad_norm": 1.068692426529055, "learning_rate": 3.566119796000966e-07, "loss": 0.7344, "step": 12514 }, { "epoch": 5.889411764705883, "grad_norm": 0.8607495920382907, "learning_rate": 3.535974324096413e-07, "loss": 0.4043, "step": 12515 }, { "epoch": 5.889882352941177, "grad_norm": 1.1801177661588929, "learning_rate": 3.5059566575596497e-07, "loss": 1.0156, "step": 12516 }, { "epoch": 5.89035294117647, "grad_norm": 1.0423823047241818, "learning_rate": 3.4760667989544575e-07, "loss": 0.8828, "step": 12517 }, { "epoch": 5.890823529411764, "grad_norm": 1.2778835430614064, "learning_rate": 3.446304750833462e-07, "loss": 0.6797, "step": 12518 }, { "epoch": 5.891294117647059, "grad_norm": 1.126753244229261, "learning_rate": 3.416670515738629e-07, "loss": 0.3691, "step": 12519 }, { "epoch": 5.891764705882353, "grad_norm": 1.3236885987288816, "learning_rate": 3.387164096200934e-07, "loss": 1.3516, "step": 12520 }, { "epoch": 5.892235294117647, "grad_norm": 0.8926576753958749, "learning_rate": 3.3577854947403615e-07, "loss": 0.2432, "step": 12521 }, { "epoch": 5.892705882352941, "grad_norm": 0.886582758432534, "learning_rate": 3.3285347138657383e-07, "loss": 0.4277, "step": 12522 }, { "epoch": 5.8931764705882355, "grad_norm": 1.0983676130220368, "learning_rate": 3.2994117560753986e-07, "loss": 0.3438, "step": 12523 }, { "epoch": 5.89364705882353, "grad_norm": 0.6192988500768778, "learning_rate": 3.2704166238565197e-07, "loss": 0.2471, "step": 12524 }, { "epoch": 5.894117647058824, "grad_norm": 1.265715829734998, "learning_rate": 3.241549319685621e-07, "loss": 1.0703, "step": 12525 }, { "epoch": 5.894588235294117, "grad_norm": 1.0830168875453938, "learning_rate": 3.212809846027897e-07, "loss": 0.457, "step": 12526 }, { "epoch": 5.8950588235294115, "grad_norm": 1.0519829186302436, "learning_rate": 3.1841982053378846e-07, "loss": 0.377, "step": 12527 }, { "epoch": 5.895529411764706, "grad_norm": 1.2976102258934235, "learning_rate": 3.155714400059128e-07, "loss": 0.4531, "step": 12528 }, { "epoch": 5.896, "grad_norm": 0.8436463777239436, "learning_rate": 3.127358432624516e-07, "loss": 0.9297, "step": 12529 }, { "epoch": 5.896470588235294, "grad_norm": 1.3363384122045356, "learning_rate": 3.099130305455444e-07, "loss": 0.5273, "step": 12530 }, { "epoch": 5.896941176470588, "grad_norm": 0.978513507681428, "learning_rate": 3.0710300209629836e-07, "loss": 0.8672, "step": 12531 }, { "epoch": 5.897411764705883, "grad_norm": 0.9911866591573888, "learning_rate": 3.043057581547048e-07, "loss": 0.3672, "step": 12532 }, { "epoch": 5.897882352941177, "grad_norm": 1.234515589619992, "learning_rate": 3.0152129895963937e-07, "loss": 0.7695, "step": 12533 }, { "epoch": 5.898352941176471, "grad_norm": 0.8376201108148009, "learning_rate": 2.987496247489285e-07, "loss": 0.2773, "step": 12534 }, { "epoch": 5.898823529411764, "grad_norm": 0.7655397205553349, "learning_rate": 2.959907357592661e-07, "loss": 0.1445, "step": 12535 }, { "epoch": 5.899294117647059, "grad_norm": 1.1012244171010561, "learning_rate": 2.9324463222629715e-07, "loss": 0.9141, "step": 12536 }, { "epoch": 5.899764705882353, "grad_norm": 0.5772790157755094, "learning_rate": 2.9051131438455055e-07, "loss": 0.1196, "step": 12537 }, { "epoch": 5.900235294117647, "grad_norm": 1.1557439022016402, "learning_rate": 2.8779078246745635e-07, "loss": 0.7969, "step": 12538 }, { "epoch": 5.900705882352941, "grad_norm": 1.0539546976451577, "learning_rate": 2.85083036707362e-07, "loss": 0.9727, "step": 12539 }, { "epoch": 5.9011764705882355, "grad_norm": 1.0935482698412466, "learning_rate": 2.8238807733551583e-07, "loss": 0.3203, "step": 12540 }, { "epoch": 5.90164705882353, "grad_norm": 0.5586015299415228, "learning_rate": 2.797059045821004e-07, "loss": 0.2432, "step": 12541 }, { "epoch": 5.902117647058824, "grad_norm": 0.664843945784017, "learning_rate": 2.7703651867616583e-07, "loss": 0.1494, "step": 12542 }, { "epoch": 5.902588235294118, "grad_norm": 0.9243865729450026, "learning_rate": 2.743799198457131e-07, "loss": 0.543, "step": 12543 }, { "epoch": 5.9030588235294115, "grad_norm": 0.8543185108660857, "learning_rate": 2.71736108317594e-07, "loss": 0.3867, "step": 12544 }, { "epoch": 5.903529411764706, "grad_norm": 1.0414491347480643, "learning_rate": 2.6910508431764457e-07, "loss": 0.6016, "step": 12545 }, { "epoch": 5.904, "grad_norm": 1.4577566429693907, "learning_rate": 2.6648684807053513e-07, "loss": 1.0625, "step": 12546 }, { "epoch": 5.904470588235294, "grad_norm": 0.955206009231257, "learning_rate": 2.6388139979988677e-07, "loss": 0.3496, "step": 12547 }, { "epoch": 5.904941176470588, "grad_norm": 1.2757680469355859, "learning_rate": 2.612887397282215e-07, "loss": 0.9688, "step": 12548 }, { "epoch": 5.905411764705883, "grad_norm": 1.4963537266041442, "learning_rate": 2.5870886807696225e-07, "loss": 0.7812, "step": 12549 }, { "epoch": 5.905882352941177, "grad_norm": 0.9856698083313467, "learning_rate": 2.5614178506644934e-07, "loss": 0.6836, "step": 12550 }, { "epoch": 5.90635294117647, "grad_norm": 1.2055631396864348, "learning_rate": 2.5358749091590746e-07, "loss": 0.793, "step": 12551 }, { "epoch": 5.906823529411764, "grad_norm": 0.7140008594018288, "learning_rate": 2.5104598584349543e-07, "loss": 0.4688, "step": 12552 }, { "epoch": 5.907294117647059, "grad_norm": 1.0016090473179295, "learning_rate": 2.4851727006627297e-07, "loss": 0.6367, "step": 12553 }, { "epoch": 5.907764705882353, "grad_norm": 0.9866726076050838, "learning_rate": 2.460013438002173e-07, "loss": 0.3027, "step": 12554 }, { "epoch": 5.908235294117647, "grad_norm": 1.0693823783756027, "learning_rate": 2.434982072601732e-07, "loss": 0.8164, "step": 12555 }, { "epoch": 5.908705882352941, "grad_norm": 0.9248020902090147, "learning_rate": 2.4100786065991975e-07, "loss": 0.707, "step": 12556 }, { "epoch": 5.9091764705882355, "grad_norm": 0.7323617690841003, "learning_rate": 2.3853030421218667e-07, "loss": 0.2295, "step": 12557 }, { "epoch": 5.90964705882353, "grad_norm": 0.6507671025011238, "learning_rate": 2.3606553812852146e-07, "loss": 0.3613, "step": 12558 }, { "epoch": 5.910117647058824, "grad_norm": 0.5699921812523259, "learning_rate": 2.3361356261945575e-07, "loss": 0.0942, "step": 12559 }, { "epoch": 5.910588235294117, "grad_norm": 1.4055210400169593, "learning_rate": 2.3117437789438865e-07, "loss": 0.9531, "step": 12560 }, { "epoch": 5.9110588235294115, "grad_norm": 1.1435682943407013, "learning_rate": 2.2874798416163686e-07, "loss": 0.7656, "step": 12561 }, { "epoch": 5.911529411764706, "grad_norm": 1.306904836046272, "learning_rate": 2.2633438162843465e-07, "loss": 0.5625, "step": 12562 }, { "epoch": 5.912, "grad_norm": 1.1121258735701058, "learning_rate": 2.239335705009171e-07, "loss": 1.3203, "step": 12563 }, { "epoch": 5.912470588235294, "grad_norm": 1.019985148298382, "learning_rate": 2.215455509841202e-07, "loss": 0.3887, "step": 12564 }, { "epoch": 5.912941176470588, "grad_norm": 1.1568148029052063, "learning_rate": 2.1917032328199747e-07, "loss": 0.9062, "step": 12565 }, { "epoch": 5.913411764705883, "grad_norm": 1.0121750711565587, "learning_rate": 2.1680788759738664e-07, "loss": 0.8203, "step": 12566 }, { "epoch": 5.913882352941177, "grad_norm": 1.3122861783867397, "learning_rate": 2.144582441320597e-07, "loss": 1.125, "step": 12567 }, { "epoch": 5.914352941176471, "grad_norm": 0.7400515674137375, "learning_rate": 2.1212139308670607e-07, "loss": 0.1602, "step": 12568 }, { "epoch": 5.914823529411764, "grad_norm": 0.8227459463831297, "learning_rate": 2.0979733466088277e-07, "loss": 0.2393, "step": 12569 }, { "epoch": 5.915294117647059, "grad_norm": 0.7980056234233993, "learning_rate": 2.0748606905309774e-07, "loss": 0.21, "step": 12570 }, { "epoch": 5.915764705882353, "grad_norm": 1.0857187988837729, "learning_rate": 2.051875964607097e-07, "loss": 0.5625, "step": 12571 }, { "epoch": 5.916235294117647, "grad_norm": 0.651749646998653, "learning_rate": 2.0290191708006165e-07, "loss": 0.1406, "step": 12572 }, { "epoch": 5.916705882352941, "grad_norm": 1.266925253265888, "learning_rate": 2.0062903110633077e-07, "loss": 0.6914, "step": 12573 }, { "epoch": 5.9171764705882355, "grad_norm": 1.1205216040213082, "learning_rate": 1.9836893873364513e-07, "loss": 0.5781, "step": 12574 }, { "epoch": 5.91764705882353, "grad_norm": 1.44309494091061, "learning_rate": 1.9612164015501696e-07, "loss": 1.1953, "step": 12575 }, { "epoch": 5.918117647058823, "grad_norm": 1.5260190380536773, "learning_rate": 1.9388713556239277e-07, "loss": 1.5078, "step": 12576 }, { "epoch": 5.918588235294117, "grad_norm": 0.8275123469876035, "learning_rate": 1.9166542514658656e-07, "loss": 0.2402, "step": 12577 }, { "epoch": 5.9190588235294115, "grad_norm": 1.1696119818873534, "learning_rate": 1.8945650909737985e-07, "loss": 0.7461, "step": 12578 }, { "epoch": 5.919529411764706, "grad_norm": 1.526429771426538, "learning_rate": 1.8726038760338847e-07, "loss": 1.0938, "step": 12579 }, { "epoch": 5.92, "grad_norm": 0.8320307369677721, "learning_rate": 1.8507706085219566e-07, "loss": 0.2676, "step": 12580 }, { "epoch": 5.920470588235294, "grad_norm": 0.8381619385868586, "learning_rate": 1.8290652903025228e-07, "loss": 0.832, "step": 12581 }, { "epoch": 5.920941176470588, "grad_norm": 0.9258576794645085, "learning_rate": 1.8074879232296002e-07, "loss": 0.6836, "step": 12582 }, { "epoch": 5.921411764705883, "grad_norm": 0.6272011692897577, "learning_rate": 1.7860385091455488e-07, "loss": 0.1631, "step": 12583 }, { "epoch": 5.921882352941177, "grad_norm": 1.264543305110536, "learning_rate": 1.764717049882569e-07, "loss": 1.1172, "step": 12584 }, { "epoch": 5.92235294117647, "grad_norm": 0.6483608854426903, "learning_rate": 1.743523547261705e-07, "loss": 0.2158, "step": 12585 }, { "epoch": 5.922823529411764, "grad_norm": 1.0146538130666485, "learning_rate": 1.722458003092675e-07, "loss": 0.3242, "step": 12586 }, { "epoch": 5.923294117647059, "grad_norm": 1.1205025134965083, "learning_rate": 1.701520419174873e-07, "loss": 0.6523, "step": 12587 }, { "epoch": 5.923764705882353, "grad_norm": 0.7123927175477619, "learning_rate": 1.680710797296203e-07, "loss": 0.1738, "step": 12588 }, { "epoch": 5.924235294117647, "grad_norm": 0.9853978523091296, "learning_rate": 1.660029139234076e-07, "loss": 0.8086, "step": 12589 }, { "epoch": 5.924705882352941, "grad_norm": 1.487324090352723, "learning_rate": 1.639475446754912e-07, "loss": 1.5078, "step": 12590 }, { "epoch": 5.9251764705882355, "grad_norm": 1.4235238685986444, "learning_rate": 1.6190497216138076e-07, "loss": 1.2109, "step": 12591 }, { "epoch": 5.92564705882353, "grad_norm": 1.253669557066292, "learning_rate": 1.598751965555367e-07, "loss": 0.7969, "step": 12592 }, { "epoch": 5.926117647058824, "grad_norm": 1.1127646087546277, "learning_rate": 1.5785821803132037e-07, "loss": 0.6875, "step": 12593 }, { "epoch": 5.926588235294117, "grad_norm": 0.5280208503478915, "learning_rate": 1.558540367609773e-07, "loss": 0.0947, "step": 12594 }, { "epoch": 5.9270588235294115, "grad_norm": 0.6773537090350347, "learning_rate": 1.5386265291568723e-07, "loss": 0.2021, "step": 12595 }, { "epoch": 5.927529411764706, "grad_norm": 0.5279834296953876, "learning_rate": 1.5188406666551412e-07, "loss": 0.3242, "step": 12596 }, { "epoch": 5.928, "grad_norm": 1.0376995875892898, "learning_rate": 1.4991827817943947e-07, "loss": 1.4219, "step": 12597 }, { "epoch": 5.928470588235294, "grad_norm": 1.1708962462037937, "learning_rate": 1.479652876253623e-07, "loss": 0.4668, "step": 12598 }, { "epoch": 5.928941176470588, "grad_norm": 0.9958866141372222, "learning_rate": 1.4602509517006588e-07, "loss": 0.4688, "step": 12599 }, { "epoch": 5.929411764705883, "grad_norm": 1.2293324639414853, "learning_rate": 1.4409770097926765e-07, "loss": 1.3203, "step": 12600 }, { "epoch": 5.929882352941177, "grad_norm": 1.3684352261834, "learning_rate": 1.421831052175526e-07, "loss": 1.2188, "step": 12601 }, { "epoch": 5.930352941176471, "grad_norm": 1.18614942952494, "learning_rate": 1.4028130804843995e-07, "loss": 0.6445, "step": 12602 }, { "epoch": 5.930823529411764, "grad_norm": 0.9105997789038628, "learning_rate": 1.3839230963436642e-07, "loss": 0.3652, "step": 12603 }, { "epoch": 5.931294117647059, "grad_norm": 1.4896817054500913, "learning_rate": 1.3651611013666964e-07, "loss": 0.9688, "step": 12604 }, { "epoch": 5.931764705882353, "grad_norm": 1.3422486203755877, "learning_rate": 1.346527097155714e-07, "loss": 1.0469, "step": 12605 }, { "epoch": 5.932235294117647, "grad_norm": 1.0041690528325744, "learning_rate": 1.3280210853019447e-07, "loss": 0.3203, "step": 12606 }, { "epoch": 5.932705882352941, "grad_norm": 0.9708626366348806, "learning_rate": 1.3096430673862901e-07, "loss": 0.3672, "step": 12607 }, { "epoch": 5.9331764705882355, "grad_norm": 1.5013402340420698, "learning_rate": 1.2913930449781619e-07, "loss": 1.1641, "step": 12608 }, { "epoch": 5.93364705882353, "grad_norm": 0.6739527724266109, "learning_rate": 1.2732710196361463e-07, "loss": 0.1465, "step": 12609 }, { "epoch": 5.934117647058823, "grad_norm": 1.1809012208137624, "learning_rate": 1.2552769929081717e-07, "loss": 0.5547, "step": 12610 }, { "epoch": 5.934588235294117, "grad_norm": 0.6506075816209989, "learning_rate": 1.2374109663306763e-07, "loss": 0.4922, "step": 12611 }, { "epoch": 5.9350588235294115, "grad_norm": 1.013383744744174, "learning_rate": 1.2196729414296058e-07, "loss": 0.4961, "step": 12612 }, { "epoch": 5.935529411764706, "grad_norm": 0.5615494858611219, "learning_rate": 1.2020629197200814e-07, "loss": 0.0933, "step": 12613 }, { "epoch": 5.936, "grad_norm": 1.102109940618975, "learning_rate": 1.1845809027059006e-07, "loss": 0.6016, "step": 12614 }, { "epoch": 5.936470588235294, "grad_norm": 0.5825404091263454, "learning_rate": 1.1672268918800353e-07, "loss": 0.1406, "step": 12615 }, { "epoch": 5.936941176470588, "grad_norm": 1.2303564534170208, "learning_rate": 1.1500008887249668e-07, "loss": 0.4824, "step": 12616 }, { "epoch": 5.937411764705883, "grad_norm": 1.1081970927446585, "learning_rate": 1.1329028947115181e-07, "loss": 0.3828, "step": 12617 }, { "epoch": 5.937882352941177, "grad_norm": 1.2341672579772318, "learning_rate": 1.1159329113000215e-07, "loss": 0.9023, "step": 12618 }, { "epoch": 5.93835294117647, "grad_norm": 0.8112379660726124, "learning_rate": 1.0990909399398173e-07, "loss": 0.2207, "step": 12619 }, { "epoch": 5.938823529411764, "grad_norm": 1.3207288061767948, "learning_rate": 1.0823769820694218e-07, "loss": 0.8086, "step": 12620 }, { "epoch": 5.939294117647059, "grad_norm": 0.6527306826440229, "learning_rate": 1.0657910391161928e-07, "loss": 0.1021, "step": 12621 }, { "epoch": 5.939764705882353, "grad_norm": 0.8368514434306601, "learning_rate": 1.0493331124964976e-07, "loss": 0.3242, "step": 12622 }, { "epoch": 5.940235294117647, "grad_norm": 1.0221286536446892, "learning_rate": 1.0330032036162117e-07, "loss": 0.875, "step": 12623 }, { "epoch": 5.940705882352941, "grad_norm": 1.6099084236584735, "learning_rate": 1.0168013138697196e-07, "loss": 1.2188, "step": 12624 }, { "epoch": 5.9411764705882355, "grad_norm": 0.9237683329970581, "learning_rate": 1.0007274446409141e-07, "loss": 0.668, "step": 12625 }, { "epoch": 5.94164705882353, "grad_norm": 0.7191548916936596, "learning_rate": 9.847815973025308e-08, "loss": 0.4336, "step": 12626 }, { "epoch": 5.942117647058824, "grad_norm": 1.315487966697844, "learning_rate": 9.689637732164802e-08, "loss": 0.9648, "step": 12627 }, { "epoch": 5.942588235294117, "grad_norm": 1.0650617300648202, "learning_rate": 9.532739737336814e-08, "loss": 0.6641, "step": 12628 }, { "epoch": 5.9430588235294115, "grad_norm": 1.690619685247048, "learning_rate": 9.377122001938963e-08, "loss": 0.8359, "step": 12629 }, { "epoch": 5.943529411764706, "grad_norm": 1.234365453069399, "learning_rate": 9.222784539263949e-08, "loss": 0.3691, "step": 12630 }, { "epoch": 5.944, "grad_norm": 1.0545725765579839, "learning_rate": 9.069727362492896e-08, "loss": 0.3984, "step": 12631 }, { "epoch": 5.944470588235294, "grad_norm": 0.9482787802543933, "learning_rate": 8.917950484697012e-08, "loss": 0.8633, "step": 12632 }, { "epoch": 5.944941176470588, "grad_norm": 0.7698045408299354, "learning_rate": 8.767453918839263e-08, "loss": 0.1836, "step": 12633 }, { "epoch": 5.945411764705883, "grad_norm": 1.3050673479589523, "learning_rate": 8.618237677772699e-08, "loss": 1.0469, "step": 12634 }, { "epoch": 5.945882352941177, "grad_norm": 1.1005656649231923, "learning_rate": 8.470301774240463e-08, "loss": 0.4277, "step": 12635 }, { "epoch": 5.946352941176471, "grad_norm": 1.3516623001235648, "learning_rate": 8.323646220877444e-08, "loss": 0.8867, "step": 12636 }, { "epoch": 5.946823529411764, "grad_norm": 1.335027606030182, "learning_rate": 8.178271030210292e-08, "loss": 0.9688, "step": 12637 }, { "epoch": 5.947294117647059, "grad_norm": 0.6206752084850191, "learning_rate": 8.034176214650745e-08, "loss": 0.3574, "step": 12638 }, { "epoch": 5.947764705882353, "grad_norm": 1.4271949912639332, "learning_rate": 7.891361786508953e-08, "loss": 0.6953, "step": 12639 }, { "epoch": 5.948235294117647, "grad_norm": 1.1800542835168322, "learning_rate": 7.749827757981831e-08, "loss": 0.6406, "step": 12640 }, { "epoch": 5.948705882352941, "grad_norm": 0.8158553681778703, "learning_rate": 7.609574141154706e-08, "loss": 0.2676, "step": 12641 }, { "epoch": 5.9491764705882355, "grad_norm": 1.1856444005858102, "learning_rate": 7.470600948006333e-08, "loss": 0.5742, "step": 12642 }, { "epoch": 5.94964705882353, "grad_norm": 1.2099122228803987, "learning_rate": 7.332908190408882e-08, "loss": 0.4395, "step": 12643 }, { "epoch": 5.950117647058823, "grad_norm": 1.0392507112675016, "learning_rate": 7.196495880117947e-08, "loss": 0.5039, "step": 12644 }, { "epoch": 5.950588235294117, "grad_norm": 1.121553277676518, "learning_rate": 7.061364028785876e-08, "loss": 0.6797, "step": 12645 }, { "epoch": 5.9510588235294115, "grad_norm": 1.2629837116381273, "learning_rate": 6.927512647953437e-08, "loss": 0.7852, "step": 12646 }, { "epoch": 5.951529411764706, "grad_norm": 1.498761551226136, "learning_rate": 6.794941749053151e-08, "loss": 1.5547, "step": 12647 }, { "epoch": 5.952, "grad_norm": 1.107059804563499, "learning_rate": 6.663651343404297e-08, "loss": 0.9883, "step": 12648 }, { "epoch": 5.952470588235294, "grad_norm": 1.5068293826326575, "learning_rate": 6.533641442222904e-08, "loss": 0.6758, "step": 12649 }, { "epoch": 5.952941176470588, "grad_norm": 0.7709334034036143, "learning_rate": 6.40491205661009e-08, "loss": 0.459, "step": 12650 }, { "epoch": 5.953411764705883, "grad_norm": 0.7451907662461345, "learning_rate": 6.277463197560395e-08, "loss": 0.167, "step": 12651 }, { "epoch": 5.953882352941177, "grad_norm": 0.6005502906533136, "learning_rate": 6.151294875960111e-08, "loss": 0.0889, "step": 12652 }, { "epoch": 5.95435294117647, "grad_norm": 1.4746848124813434, "learning_rate": 6.026407102583952e-08, "loss": 0.9414, "step": 12653 }, { "epoch": 5.954823529411764, "grad_norm": 1.4805519841207238, "learning_rate": 5.902799888096721e-08, "loss": 0.9062, "step": 12654 }, { "epoch": 5.955294117647059, "grad_norm": 0.9205055173415281, "learning_rate": 5.780473243056638e-08, "loss": 1.0625, "step": 12655 }, { "epoch": 5.955764705882353, "grad_norm": 1.0754591639170605, "learning_rate": 5.659427177910347e-08, "loss": 0.7734, "step": 12656 }, { "epoch": 5.956235294117647, "grad_norm": 1.5252972267095195, "learning_rate": 5.539661702996245e-08, "loss": 1.6953, "step": 12657 }, { "epoch": 5.956705882352941, "grad_norm": 1.363568926097053, "learning_rate": 5.421176828541152e-08, "loss": 1.0312, "step": 12658 }, { "epoch": 5.9571764705882355, "grad_norm": 1.418795259603906, "learning_rate": 5.30397256466697e-08, "loss": 0.8242, "step": 12659 }, { "epoch": 5.95764705882353, "grad_norm": 1.0433576212740425, "learning_rate": 5.1880489213806944e-08, "loss": 0.6484, "step": 12660 }, { "epoch": 5.958117647058824, "grad_norm": 1.2133069781002448, "learning_rate": 5.0734059085860704e-08, "loss": 0.4453, "step": 12661 }, { "epoch": 5.958588235294117, "grad_norm": 1.0356967254193299, "learning_rate": 4.9600435360702685e-08, "loss": 0.4102, "step": 12662 }, { "epoch": 5.9590588235294115, "grad_norm": 1.1395740021098482, "learning_rate": 4.847961813517209e-08, "loss": 0.2871, "step": 12663 }, { "epoch": 5.959529411764706, "grad_norm": 0.6874680850681866, "learning_rate": 4.737160750500901e-08, "loss": 0.1484, "step": 12664 }, { "epoch": 5.96, "grad_norm": 1.1818909763608736, "learning_rate": 4.6276403564804444e-08, "loss": 0.6445, "step": 12665 }, { "epoch": 5.960470588235294, "grad_norm": 1.0742324818575324, "learning_rate": 4.519400640811688e-08, "loss": 0.4316, "step": 12666 }, { "epoch": 5.960941176470588, "grad_norm": 0.9355774275341409, "learning_rate": 4.412441612738904e-08, "loss": 0.2891, "step": 12667 }, { "epoch": 5.961411764705883, "grad_norm": 1.0058534633206575, "learning_rate": 4.3067632813964524e-08, "loss": 0.6211, "step": 12668 }, { "epoch": 5.961882352941177, "grad_norm": 1.4845946356172754, "learning_rate": 4.2023656558104466e-08, "loss": 1.125, "step": 12669 }, { "epoch": 5.962352941176471, "grad_norm": 1.125610500829404, "learning_rate": 4.099248744897088e-08, "loss": 1.0078, "step": 12670 }, { "epoch": 5.962823529411764, "grad_norm": 0.9798410624073431, "learning_rate": 3.9974125574609994e-08, "loss": 0.3027, "step": 12671 }, { "epoch": 5.963294117647059, "grad_norm": 0.8456772629538976, "learning_rate": 3.896857102200224e-08, "loss": 0.207, "step": 12672 }, { "epoch": 5.963764705882353, "grad_norm": 1.0038656762758889, "learning_rate": 3.797582387704556e-08, "loss": 0.3047, "step": 12673 }, { "epoch": 5.964235294117647, "grad_norm": 0.8153700051551258, "learning_rate": 3.69958842245055e-08, "loss": 0.2168, "step": 12674 }, { "epoch": 5.964705882352941, "grad_norm": 1.3167578281333512, "learning_rate": 3.6028752148081766e-08, "loss": 0.6758, "step": 12675 }, { "epoch": 5.9651764705882355, "grad_norm": 1.1121456728242758, "learning_rate": 3.5074427730374945e-08, "loss": 0.7812, "step": 12676 }, { "epoch": 5.96564705882353, "grad_norm": 1.153277676788943, "learning_rate": 3.4132911052886514e-08, "loss": 0.5078, "step": 12677 }, { "epoch": 5.966117647058823, "grad_norm": 1.0693828369848724, "learning_rate": 3.320420219603548e-08, "loss": 0.7422, "step": 12678 }, { "epoch": 5.966588235294117, "grad_norm": 1.2924178633184664, "learning_rate": 3.228830123910842e-08, "loss": 0.7109, "step": 12679 }, { "epoch": 5.9670588235294115, "grad_norm": 1.3821407050031722, "learning_rate": 3.13852082603594e-08, "loss": 1.1797, "step": 12680 }, { "epoch": 5.967529411764706, "grad_norm": 1.25039041859568, "learning_rate": 3.049492333691006e-08, "loss": 0.8633, "step": 12681 }, { "epoch": 5.968, "grad_norm": 0.7016154095917496, "learning_rate": 2.961744654478293e-08, "loss": 0.5078, "step": 12682 }, { "epoch": 5.968470588235294, "grad_norm": 1.9941568452818186, "learning_rate": 2.875277795891806e-08, "loss": 1.3281, "step": 12683 }, { "epoch": 5.968941176470588, "grad_norm": 0.8692055756590115, "learning_rate": 2.7900917653189692e-08, "loss": 0.9531, "step": 12684 }, { "epoch": 5.969411764705883, "grad_norm": 1.2561866243187678, "learning_rate": 2.706186570032298e-08, "loss": 0.3652, "step": 12685 }, { "epoch": 5.969882352941177, "grad_norm": 1.0267614288854774, "learning_rate": 2.6235622171977278e-08, "loss": 0.3145, "step": 12686 }, { "epoch": 5.97035294117647, "grad_norm": 1.0906346988069568, "learning_rate": 2.542218713874611e-08, "loss": 0.957, "step": 12687 }, { "epoch": 5.970823529411764, "grad_norm": 0.3401751801746105, "learning_rate": 2.462156067005727e-08, "loss": 0.0684, "step": 12688 }, { "epoch": 5.971294117647059, "grad_norm": 1.0006481650653396, "learning_rate": 2.3833742834339365e-08, "loss": 0.3027, "step": 12689 }, { "epoch": 5.971764705882353, "grad_norm": 1.2980352349732112, "learning_rate": 2.305873369882194e-08, "loss": 1.1719, "step": 12690 }, { "epoch": 5.972235294117647, "grad_norm": 1.4888248099317685, "learning_rate": 2.2296533329735356e-08, "loss": 1.5078, "step": 12691 }, { "epoch": 5.972705882352941, "grad_norm": 0.6971340934620651, "learning_rate": 2.1547141792160882e-08, "loss": 0.0894, "step": 12692 }, { "epoch": 5.9731764705882355, "grad_norm": 1.0244516313020182, "learning_rate": 2.0810559150097327e-08, "loss": 0.5547, "step": 12693 }, { "epoch": 5.97364705882353, "grad_norm": 0.9932412815904089, "learning_rate": 2.008678546646103e-08, "loss": 0.3184, "step": 12694 }, { "epoch": 5.974117647058824, "grad_norm": 1.395236150754639, "learning_rate": 1.9375820803052555e-08, "loss": 0.5352, "step": 12695 }, { "epoch": 5.974588235294117, "grad_norm": 1.0365249815886, "learning_rate": 1.8677665220606653e-08, "loss": 0.543, "step": 12696 }, { "epoch": 5.9750588235294115, "grad_norm": 0.9691196388801285, "learning_rate": 1.7992318778742298e-08, "loss": 0.4434, "step": 12697 }, { "epoch": 5.975529411764706, "grad_norm": 0.896797858950832, "learning_rate": 1.7319781535996002e-08, "loss": 0.5312, "step": 12698 }, { "epoch": 5.976, "grad_norm": 1.265078107147593, "learning_rate": 1.6660053549788498e-08, "loss": 0.4277, "step": 12699 }, { "epoch": 5.976470588235294, "grad_norm": 1.134242592347702, "learning_rate": 1.6013134876491362e-08, "loss": 0.9062, "step": 12700 }, { "epoch": 5.976941176470588, "grad_norm": 1.1521090826677098, "learning_rate": 1.5379025571327088e-08, "loss": 0.7227, "step": 12701 }, { "epoch": 5.977411764705883, "grad_norm": 0.3047606077407533, "learning_rate": 1.4757725688469002e-08, "loss": 0.0298, "step": 12702 }, { "epoch": 5.977882352941176, "grad_norm": 0.9163899675663503, "learning_rate": 1.414923528097467e-08, "loss": 0.5586, "step": 12703 }, { "epoch": 5.97835294117647, "grad_norm": 0.3998664605841859, "learning_rate": 1.3553554400802524e-08, "loss": 0.0479, "step": 12704 }, { "epoch": 5.978823529411764, "grad_norm": 0.3702690977232195, "learning_rate": 1.2970683098845192e-08, "loss": 0.0398, "step": 12705 }, { "epoch": 5.979294117647059, "grad_norm": 1.293991483910677, "learning_rate": 1.240062142487952e-08, "loss": 0.7109, "step": 12706 }, { "epoch": 5.979764705882353, "grad_norm": 0.7949133987812284, "learning_rate": 1.1843369427583238e-08, "loss": 0.2246, "step": 12707 }, { "epoch": 5.980235294117647, "grad_norm": 0.5373359744269951, "learning_rate": 1.1298927154551608e-08, "loss": 0.0708, "step": 12708 }, { "epoch": 5.980705882352941, "grad_norm": 1.4448178133466667, "learning_rate": 1.0767294652280767e-08, "loss": 1.4219, "step": 12709 }, { "epoch": 5.9811764705882355, "grad_norm": 1.240724664766229, "learning_rate": 1.0248471966167737e-08, "loss": 0.5195, "step": 12710 }, { "epoch": 5.98164705882353, "grad_norm": 0.9969090967533077, "learning_rate": 9.74245914052707e-09, "loss": 0.3184, "step": 12711 }, { "epoch": 5.982117647058823, "grad_norm": 0.776643355275697, "learning_rate": 9.24925621859085e-09, "loss": 0.1445, "step": 12712 }, { "epoch": 5.982588235294117, "grad_norm": 1.1813084425364284, "learning_rate": 8.768863242475389e-09, "loss": 0.6641, "step": 12713 }, { "epoch": 5.9830588235294115, "grad_norm": 1.173326262898294, "learning_rate": 8.30128025318122e-09, "loss": 0.9922, "step": 12714 }, { "epoch": 5.983529411764706, "grad_norm": 1.0894067805299683, "learning_rate": 7.846507290676373e-09, "loss": 0.8008, "step": 12715 }, { "epoch": 5.984, "grad_norm": 1.3229584818125335, "learning_rate": 7.404544393779799e-09, "loss": 0.9648, "step": 12716 }, { "epoch": 5.984470588235294, "grad_norm": 1.0447481982680233, "learning_rate": 6.975391600261282e-09, "loss": 0.6914, "step": 12717 }, { "epoch": 5.984941176470588, "grad_norm": 0.7674443376874556, "learning_rate": 6.55904894674153e-09, "loss": 0.1729, "step": 12718 }, { "epoch": 5.985411764705883, "grad_norm": 1.0118618203542364, "learning_rate": 6.155516468792088e-09, "loss": 0.7891, "step": 12719 }, { "epoch": 5.985882352941177, "grad_norm": 1.2556515264211592, "learning_rate": 5.764794200885381e-09, "loss": 0.5391, "step": 12720 }, { "epoch": 5.98635294117647, "grad_norm": 1.3332806578075596, "learning_rate": 5.38688217637806e-09, "loss": 0.875, "step": 12721 }, { "epoch": 5.986823529411764, "grad_norm": 1.0129671210476698, "learning_rate": 5.0217804275443084e-09, "loss": 0.6016, "step": 12722 }, { "epoch": 5.987294117647059, "grad_norm": 0.7548275043724263, "learning_rate": 4.669488985575842e-09, "loss": 0.1226, "step": 12723 }, { "epoch": 5.987764705882353, "grad_norm": 0.8011073229241636, "learning_rate": 4.3300078805652564e-09, "loss": 0.1865, "step": 12724 }, { "epoch": 5.988235294117647, "grad_norm": 0.6513194296893943, "learning_rate": 4.003337141489371e-09, "loss": 0.2734, "step": 12725 }, { "epoch": 5.988705882352941, "grad_norm": 0.5680443853130991, "learning_rate": 3.689476796242541e-09, "loss": 0.1406, "step": 12726 }, { "epoch": 5.9891764705882355, "grad_norm": 1.1417275632833854, "learning_rate": 3.388426871653305e-09, "loss": 0.5859, "step": 12727 }, { "epoch": 5.98964705882353, "grad_norm": 1.0750950688825196, "learning_rate": 3.100187393434428e-09, "loss": 0.75, "step": 12728 }, { "epoch": 5.990117647058824, "grad_norm": 0.7550297438945731, "learning_rate": 2.8247583861662484e-09, "loss": 0.126, "step": 12729 }, { "epoch": 5.990588235294117, "grad_norm": 0.9255681157971888, "learning_rate": 2.562139873413249e-09, "loss": 0.3672, "step": 12730 }, { "epoch": 5.9910588235294115, "grad_norm": 1.0339685029528505, "learning_rate": 2.3123318775741805e-09, "loss": 0.6953, "step": 12731 }, { "epoch": 5.991529411764706, "grad_norm": 1.0392728345401423, "learning_rate": 2.075334419998631e-09, "loss": 0.4453, "step": 12732 }, { "epoch": 5.992, "grad_norm": 1.1607662801618952, "learning_rate": 1.8511475209370685e-09, "loss": 0.8516, "step": 12733 }, { "epoch": 5.992470588235294, "grad_norm": 0.8522513014731825, "learning_rate": 1.6397711995241869e-09, "loss": 0.3223, "step": 12734 }, { "epoch": 5.992941176470588, "grad_norm": 1.323758158274033, "learning_rate": 1.441205473795559e-09, "loss": 1.0859, "step": 12735 }, { "epoch": 5.993411764705883, "grad_norm": 0.8733202090378864, "learning_rate": 1.2554503607375976e-09, "loss": 0.3281, "step": 12736 }, { "epoch": 5.993882352941176, "grad_norm": 1.3472502337155956, "learning_rate": 1.082505876204287e-09, "loss": 0.9023, "step": 12737 }, { "epoch": 5.99435294117647, "grad_norm": 1.8210389635763096, "learning_rate": 9.223720349671449e-10, "loss": 1.5547, "step": 12738 }, { "epoch": 5.994823529411764, "grad_norm": 1.1442185421160178, "learning_rate": 7.750488506985674e-10, "loss": 0.3496, "step": 12739 }, { "epoch": 5.995294117647059, "grad_norm": 1.216645256320794, "learning_rate": 6.405363359718308e-10, "loss": 0.6641, "step": 12740 }, { "epoch": 5.995764705882353, "grad_norm": 1.1401752285627758, "learning_rate": 5.188345022943963e-10, "loss": 0.9922, "step": 12741 }, { "epoch": 5.996235294117647, "grad_norm": 1.1064682145575377, "learning_rate": 4.099433600579516e-10, "loss": 0.4062, "step": 12742 }, { "epoch": 5.996705882352941, "grad_norm": 1.0705574932851354, "learning_rate": 3.138629185550634e-10, "loss": 0.6328, "step": 12743 }, { "epoch": 5.9971764705882356, "grad_norm": 1.0956062089848477, "learning_rate": 2.3059318597917763e-10, "loss": 0.4141, "step": 12744 }, { "epoch": 5.99764705882353, "grad_norm": 1.4120245921625696, "learning_rate": 1.6013416947457946e-10, "loss": 0.9805, "step": 12745 }, { "epoch": 5.998117647058823, "grad_norm": 0.993229448101261, "learning_rate": 1.0248587501981986e-10, "loss": 0.6836, "step": 12746 }, { "epoch": 5.998588235294117, "grad_norm": 0.9167536587487773, "learning_rate": 5.7648307577595756e-11, "loss": 0.6953, "step": 12747 }, { "epoch": 5.9990588235294116, "grad_norm": 1.3161267259242666, "learning_rate": 2.5621470944869883e-11, "loss": 0.6875, "step": 12748 }, { "epoch": 5.999529411764706, "grad_norm": 1.0056115854987304, "learning_rate": 6.4053678694442335e-12, "loss": 0.7305, "step": 12749 }, { "epoch": 6.0, "grad_norm": 0.934282787854844, "learning_rate": 0.0, "loss": 0.6992, "step": 12750 }, { "epoch": 6.0, "eval_loss": 0.7636064291000366, "eval_runtime": 37.0893, "eval_samples_per_second": 20.221, "eval_steps_per_second": 10.111, "step": 12750 }, { "epoch": 6.0, "step": 12750, "total_flos": 10012651683840.0, "train_loss": 0.973898250804228, "train_runtime": 242278.5658, "train_samples_per_second": 0.105, "train_steps_per_second": 0.053 } ], "logging_steps": 1.0, "max_steps": 12750, "num_input_tokens_seen": 0, "num_train_epochs": 6, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 10012651683840.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }