diff --git "a/checkpoint-1820/trainer_state.json" "b/checkpoint-1820/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1820/trainer_state.json"
@@ -0,0 +1,12774 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.002200220022002,
+  "eval_steps": 500,
+  "global_step": 1820,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011001100110011,
+      "grad_norm": 6.46875,
+      "learning_rate": 0.0,
+      "loss": 2.144,
+      "step": 1
+    },
+    {
+      "epoch": 0.0022002200220022,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 3.057,
+      "step": 2
+    },
+    {
+      "epoch": 0.0033003300330033004,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.7201,
+      "step": 3
+    },
+    {
+      "epoch": 0.0044004400440044,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.8e-05,
+      "loss": 1.6661,
+      "step": 4
+    },
+    {
+      "epoch": 0.005500550055005501,
+      "grad_norm": 5.21875,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 2.2751,
+      "step": 5
+    },
+    {
+      "epoch": 0.006600660066006601,
+      "grad_norm": 4.5,
+      "learning_rate": 8e-05,
+      "loss": 1.8133,
+      "step": 6
+    },
+    {
+      "epoch": 0.007700770077007701,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.99706098457017e-05,
+      "loss": 2.1569,
+      "step": 7
+    },
+    {
+      "epoch": 0.0088008800880088,
+      "grad_norm": 4.65625,
+      "learning_rate": 7.994121969140339e-05,
+      "loss": 1.7213,
+      "step": 8
+    },
+    {
+      "epoch": 0.009900990099009901,
+      "grad_norm": 4.46875,
+      "learning_rate": 7.991182953710508e-05,
+      "loss": 1.9531,
+      "step": 9
+    },
+    {
+      "epoch": 0.011001100110011002,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.988243938280677e-05,
+      "loss": 1.9228,
+      "step": 10
+    },
+    {
+      "epoch": 0.0121012101210121,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.985304922850846e-05,
+      "loss": 2.028,
+      "step": 11
+    },
+    {
+      "epoch": 0.013201320132013201,
+      "grad_norm": 3.90625,
+      "learning_rate": 7.982365907421014e-05,
+      "loss": 2.0012,
+      "step": 12
+    },
+    {
+      "epoch": 0.014301430143014302,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.979426891991184e-05,
+      "loss": 2.0968,
+      "step": 13
+    },
+    {
+      "epoch": 0.015401540154015401,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.976487876561353e-05,
+      "loss": 1.5875,
+      "step": 14
+    },
+    {
+      "epoch": 0.0165016501650165,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.973548861131522e-05,
+      "loss": 1.9844,
+      "step": 15
+    },
+    {
+      "epoch": 0.0176017601760176,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.97060984570169e-05,
+      "loss": 1.5483,
+      "step": 16
+    },
+    {
+      "epoch": 0.0187018701870187,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.96767083027186e-05,
+      "loss": 1.801,
+      "step": 17
+    },
+    {
+      "epoch": 0.019801980198019802,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.964731814842029e-05,
+      "loss": 1.9239,
+      "step": 18
+    },
+    {
+      "epoch": 0.020902090209020903,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.961792799412197e-05,
+      "loss": 1.8882,
+      "step": 19
+    },
+    {
+      "epoch": 0.022002200220022004,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.958853783982366e-05,
+      "loss": 1.4882,
+      "step": 20
+    },
+    {
+      "epoch": 0.0231023102310231,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.955914768552536e-05,
+      "loss": 1.7706,
+      "step": 21
+    },
+    {
+      "epoch": 0.0242024202420242,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.952975753122704e-05,
+      "loss": 1.4345,
+      "step": 22
+    },
+    {
+      "epoch": 0.025302530253025302,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.950036737692873e-05,
+      "loss": 1.9439,
+      "step": 23
+    },
+    {
+      "epoch": 0.026402640264026403,
+      "grad_norm": 3.25,
+      "learning_rate": 7.947097722263042e-05,
+      "loss": 1.5659,
+      "step": 24
+    },
+    {
+      "epoch": 0.027502750275027504,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.944158706833211e-05,
+      "loss": 1.4751,
+      "step": 25
+    },
+    {
+      "epoch": 0.028602860286028604,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.94121969140338e-05,
+      "loss": 1.7868,
+      "step": 26
+    },
+    {
+      "epoch": 0.0297029702970297,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.93828067597355e-05,
+      "loss": 1.5937,
+      "step": 27
+    },
+    {
+      "epoch": 0.030803080308030802,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.935341660543718e-05,
+      "loss": 1.375,
+      "step": 28
+    },
+    {
+      "epoch": 0.0319031903190319,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.932402645113887e-05,
+      "loss": 1.5409,
+      "step": 29
+    },
+    {
+      "epoch": 0.033003300330033,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.929463629684057e-05,
+      "loss": 2.0894,
+      "step": 30
+    },
+    {
+      "epoch": 0.034103410341034104,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.926524614254226e-05,
+      "loss": 1.5313,
+      "step": 31
+    },
+    {
+      "epoch": 0.0352035203520352,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.923585598824395e-05,
+      "loss": 1.532,
+      "step": 32
+    },
+    {
+      "epoch": 0.036303630363036306,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.920646583394564e-05,
+      "loss": 1.7818,
+      "step": 33
+    },
+    {
+      "epoch": 0.0374037403740374,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.917707567964733e-05,
+      "loss": 1.6455,
+      "step": 34
+    },
+    {
+      "epoch": 0.03850385038503851,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.914768552534902e-05,
+      "loss": 1.8688,
+      "step": 35
+    },
+    {
+      "epoch": 0.039603960396039604,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.911829537105071e-05,
+      "loss": 1.5657,
+      "step": 36
+    },
+    {
+      "epoch": 0.0407040704070407,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.90889052167524e-05,
+      "loss": 1.9176,
+      "step": 37
+    },
+    {
+      "epoch": 0.041804180418041806,
+      "grad_norm": 3.625,
+      "learning_rate": 7.905951506245409e-05,
+      "loss": 2.0095,
+      "step": 38
+    },
+    {
+      "epoch": 0.0429042904290429,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.903012490815578e-05,
+      "loss": 1.6829,
+      "step": 39
+    },
+    {
+      "epoch": 0.04400440044004401,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.900073475385747e-05,
+      "loss": 1.4805,
+      "step": 40
+    },
+    {
+      "epoch": 0.045104510451045104,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.897134459955915e-05,
+      "loss": 1.7129,
+      "step": 41
+    },
+    {
+      "epoch": 0.0462046204620462,
+      "grad_norm": 2.75,
+      "learning_rate": 7.894195444526084e-05,
+      "loss": 1.846,
+      "step": 42
+    },
+    {
+      "epoch": 0.047304730473047306,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.891256429096254e-05,
+      "loss": 1.5123,
+      "step": 43
+    },
+    {
+      "epoch": 0.0484048404840484,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.888317413666422e-05,
+      "loss": 1.5001,
+      "step": 44
+    },
+    {
+      "epoch": 0.04950495049504951,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.885378398236591e-05,
+      "loss": 1.7146,
+      "step": 45
+    },
+    {
+      "epoch": 0.050605060506050605,
+      "grad_norm": 2.75,
+      "learning_rate": 7.88243938280676e-05,
+      "loss": 1.5773,
+      "step": 46
+    },
+    {
+      "epoch": 0.0517051705170517,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.87950036737693e-05,
+      "loss": 1.5799,
+      "step": 47
+    },
+    {
+      "epoch": 0.052805280528052806,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.876561351947098e-05,
+      "loss": 1.8769,
+      "step": 48
+    },
+    {
+      "epoch": 0.0539053905390539,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.873622336517267e-05,
+      "loss": 1.8068,
+      "step": 49
+    },
+    {
+      "epoch": 0.05500550055005501,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.870683321087436e-05,
+      "loss": 1.8403,
+      "step": 50
+    },
+    {
+      "epoch": 0.056105610561056105,
+      "grad_norm": 3.125,
+      "learning_rate": 7.867744305657605e-05,
+      "loss": 1.3685,
+      "step": 51
+    },
+    {
+      "epoch": 0.05720572057205721,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.864805290227774e-05,
+      "loss": 1.983,
+      "step": 52
+    },
+    {
+      "epoch": 0.058305830583058306,
+      "grad_norm": 2.75,
+      "learning_rate": 7.861866274797943e-05,
+      "loss": 1.7412,
+      "step": 53
+    },
+    {
+      "epoch": 0.0594059405940594,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.858927259368112e-05,
+      "loss": 1.8109,
+      "step": 54
+    },
+    {
+      "epoch": 0.06050605060506051,
+      "grad_norm": 2.875,
+      "learning_rate": 7.855988243938281e-05,
+      "loss": 1.7846,
+      "step": 55
+    },
+    {
+      "epoch": 0.061606160616061605,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.85304922850845e-05,
+      "loss": 1.4929,
+      "step": 56
+    },
+    {
+      "epoch": 0.0627062706270627,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.85011021307862e-05,
+      "loss": 1.5486,
+      "step": 57
+    },
+    {
+      "epoch": 0.0638063806380638,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.847171197648789e-05,
+      "loss": 1.7472,
+      "step": 58
+    },
+    {
+      "epoch": 0.06490649064906491,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.844232182218956e-05,
+      "loss": 1.9393,
+      "step": 59
+    },
+    {
+      "epoch": 0.066006600660066,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.841293166789127e-05,
+      "loss": 1.6674,
+      "step": 60
+    },
+    {
+      "epoch": 0.0671067106710671,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.838354151359296e-05,
+      "loss": 1.5629,
+      "step": 61
+    },
+    {
+      "epoch": 0.06820682068206821,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.835415135929465e-05,
+      "loss": 1.4875,
+      "step": 62
+    },
+    {
+      "epoch": 0.06930693069306931,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.832476120499632e-05,
+      "loss": 1.7542,
+      "step": 63
+    },
+    {
+      "epoch": 0.0704070407040704,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.829537105069803e-05,
+      "loss": 1.3666,
+      "step": 64
+    },
+    {
+      "epoch": 0.07150715071507151,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.826598089639972e-05,
+      "loss": 1.8975,
+      "step": 65
+    },
+    {
+      "epoch": 0.07260726072607261,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.82365907421014e-05,
+      "loss": 1.6946,
+      "step": 66
+    },
+    {
+      "epoch": 0.0737073707370737,
+      "grad_norm": 2.625,
+      "learning_rate": 7.820720058780309e-05,
+      "loss": 1.9786,
+      "step": 67
+    },
+    {
+      "epoch": 0.0748074807480748,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.817781043350479e-05,
+      "loss": 1.5855,
+      "step": 68
+    },
+    {
+      "epoch": 0.07590759075907591,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.814842027920647e-05,
+      "loss": 1.4395,
+      "step": 69
+    },
+    {
+      "epoch": 0.07700770077007701,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.811903012490816e-05,
+      "loss": 1.7063,
+      "step": 70
+    },
+    {
+      "epoch": 0.0781078107810781,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.808963997060985e-05,
+      "loss": 1.7596,
+      "step": 71
+    },
+    {
+      "epoch": 0.07920792079207921,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.806024981631155e-05,
+      "loss": 1.6308,
+      "step": 72
+    },
+    {
+      "epoch": 0.08030803080308031,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.803085966201323e-05,
+      "loss": 1.513,
+      "step": 73
+    },
+    {
+      "epoch": 0.0814081408140814,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.800146950771492e-05,
+      "loss": 1.8351,
+      "step": 74
+    },
+    {
+      "epoch": 0.08250825082508251,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.797207935341661e-05,
+      "loss": 1.9086,
+      "step": 75
+    },
+    {
+      "epoch": 0.08360836083608361,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.79426891991183e-05,
+      "loss": 1.7254,
+      "step": 76
+    },
+    {
+      "epoch": 0.0847084708470847,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.791329904481999e-05,
+      "loss": 1.5895,
+      "step": 77
+    },
+    {
+      "epoch": 0.0858085808580858,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.788390889052168e-05,
+      "loss": 1.409,
+      "step": 78
+    },
+    {
+      "epoch": 0.08690869086908691,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.785451873622337e-05,
+      "loss": 1.912,
+      "step": 79
+    },
+    {
+      "epoch": 0.08800880088008801,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.782512858192506e-05,
+      "loss": 1.8979,
+      "step": 80
+    },
+    {
+      "epoch": 0.0891089108910891,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.779573842762675e-05,
+      "loss": 1.772,
+      "step": 81
+    },
+    {
+      "epoch": 0.09020902090209021,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.776634827332844e-05,
+      "loss": 1.7706,
+      "step": 82
+    },
+    {
+      "epoch": 0.09130913091309131,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.773695811903013e-05,
+      "loss": 1.7933,
+      "step": 83
+    },
+    {
+      "epoch": 0.0924092409240924,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.770756796473181e-05,
+      "loss": 1.4408,
+      "step": 84
+    },
+    {
+      "epoch": 0.09350935093509351,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.767817781043352e-05,
+      "loss": 1.6825,
+      "step": 85
+    },
+    {
+      "epoch": 0.09460946094609461,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.76487876561352e-05,
+      "loss": 1.9553,
+      "step": 86
+    },
+    {
+      "epoch": 0.09570957095709572,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.76193975018369e-05,
+      "loss": 1.3714,
+      "step": 87
+    },
+    {
+      "epoch": 0.0968096809680968,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.759000734753857e-05,
+      "loss": 1.8458,
+      "step": 88
+    },
+    {
+      "epoch": 0.09790979097909791,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.756061719324028e-05,
+      "loss": 1.7907,
+      "step": 89
+    },
+    {
+      "epoch": 0.09900990099009901,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.753122703894197e-05,
+      "loss": 1.7021,
+      "step": 90
+    },
+    {
+      "epoch": 0.1001100110011001,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.750183688464365e-05,
+      "loss": 1.7753,
+      "step": 91
+    },
+    {
+      "epoch": 0.10121012101210121,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.747244673034534e-05,
+      "loss": 1.7924,
+      "step": 92
+    },
+    {
+      "epoch": 0.10231023102310231,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.744305657604703e-05,
+      "loss": 1.7719,
+      "step": 93
+    },
+    {
+      "epoch": 0.1034103410341034,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.741366642174872e-05,
+      "loss": 1.476,
+      "step": 94
+    },
+    {
+      "epoch": 0.10451045104510451,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.738427626745041e-05,
+      "loss": 1.588,
+      "step": 95
+    },
+    {
+      "epoch": 0.10561056105610561,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.73548861131521e-05,
+      "loss": 1.614,
+      "step": 96
+    },
+    {
+      "epoch": 0.10671067106710672,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.732549595885379e-05,
+      "loss": 1.3961,
+      "step": 97
+    },
+    {
+      "epoch": 0.1078107810781078,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.729610580455548e-05,
+      "loss": 1.6342,
+      "step": 98
+    },
+    {
+      "epoch": 0.10891089108910891,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.726671565025717e-05,
+      "loss": 1.672,
+      "step": 99
+    },
+    {
+      "epoch": 0.11001100110011001,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.723732549595886e-05,
+      "loss": 1.5051,
+      "step": 100
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.720793534166055e-05,
+      "loss": 1.9235,
+      "step": 101
+    },
+    {
+      "epoch": 0.11221122112211221,
+      "grad_norm": 2.5,
+      "learning_rate": 7.717854518736224e-05,
+      "loss": 1.4361,
+      "step": 102
+    },
+    {
+      "epoch": 0.11331133113311331,
+      "grad_norm": 3.0,
+      "learning_rate": 7.714915503306393e-05,
+      "loss": 1.6646,
+      "step": 103
+    },
+    {
+      "epoch": 0.11441144114411442,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.711976487876562e-05,
+      "loss": 1.4909,
+      "step": 104
+    },
+    {
+      "epoch": 0.11551155115511551,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.709037472446731e-05,
+      "loss": 1.6798,
+      "step": 105
+    },
+    {
+      "epoch": 0.11661166116611661,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.7060984570169e-05,
+      "loss": 1.9486,
+      "step": 106
+    },
+    {
+      "epoch": 0.11771177117711772,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.70315944158707e-05,
+      "loss": 1.5624,
+      "step": 107
+    },
+    {
+      "epoch": 0.1188118811881188,
+      "grad_norm": 2.375,
+      "learning_rate": 7.700220426157238e-05,
+      "loss": 1.4138,
+      "step": 108
+    },
+    {
+      "epoch": 0.11991199119911991,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.697281410727407e-05,
+      "loss": 1.4676,
+      "step": 109
+    },
+    {
+      "epoch": 0.12101210121012101,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.694342395297575e-05,
+      "loss": 1.6737,
+      "step": 110
+    },
+    {
+      "epoch": 0.12211221122112212,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.691403379867746e-05,
+      "loss": 1.3377,
+      "step": 111
+    },
+    {
+      "epoch": 0.12321232123212321,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.688464364437915e-05,
+      "loss": 1.6819,
+      "step": 112
+    },
+    {
+      "epoch": 0.12431243124312431,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.685525349008082e-05,
+      "loss": 1.4974,
+      "step": 113
+    },
+    {
+      "epoch": 0.1254125412541254,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.682586333578251e-05,
+      "loss": 1.61,
+      "step": 114
+    },
+    {
+      "epoch": 0.1265126512651265,
+      "grad_norm": 2.625,
+      "learning_rate": 7.679647318148422e-05,
+      "loss": 1.5181,
+      "step": 115
+    },
+    {
+      "epoch": 0.1276127612761276,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.67670830271859e-05,
+      "loss": 1.5267,
+      "step": 116
+    },
+    {
+      "epoch": 0.12871287128712872,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.673769287288758e-05,
+      "loss": 1.5479,
+      "step": 117
+    },
+    {
+      "epoch": 0.12981298129812982,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.670830271858928e-05,
+      "loss": 1.8229,
+      "step": 118
+    },
+    {
+      "epoch": 0.13091309130913092,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.667891256429098e-05,
+      "loss": 1.6233,
+      "step": 119
+    },
+    {
+      "epoch": 0.132013201320132,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.664952240999266e-05,
+      "loss": 2.0874,
+      "step": 120
+    },
+    {
+      "epoch": 0.1331133113311331,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.662013225569435e-05,
+      "loss": 1.7253,
+      "step": 121
+    },
+    {
+      "epoch": 0.1342134213421342,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.659074210139604e-05,
+      "loss": 1.4686,
+      "step": 122
+    },
+    {
+      "epoch": 0.1353135313531353,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.656135194709773e-05,
+      "loss": 1.7246,
+      "step": 123
+    },
+    {
+      "epoch": 0.13641364136413642,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.653196179279942e-05,
+      "loss": 1.5427,
+      "step": 124
+    },
+    {
+      "epoch": 0.13751375137513752,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.650257163850111e-05,
+      "loss": 1.7827,
+      "step": 125
+    },
+    {
+      "epoch": 0.13861386138613863,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.64731814842028e-05,
+      "loss": 1.8431,
+      "step": 126
+    },
+    {
+      "epoch": 0.1397139713971397,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.644379132990449e-05,
+      "loss": 1.5337,
+      "step": 127
+    },
+    {
+      "epoch": 0.1408140814081408,
+      "grad_norm": 2.625,
+      "learning_rate": 7.641440117560618e-05,
+      "loss": 1.7567,
+      "step": 128
+    },
+    {
+      "epoch": 0.1419141914191419,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.638501102130787e-05,
+      "loss": 1.8852,
+      "step": 129
+    },
+    {
+      "epoch": 0.14301430143014301,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.635562086700956e-05,
+      "loss": 1.6625,
+      "step": 130
+    },
+    {
+      "epoch": 0.14411441144114412,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.632623071271124e-05,
+      "loss": 1.9057,
+      "step": 131
+    },
+    {
+      "epoch": 0.14521452145214522,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.629684055841294e-05,
+      "loss": 1.228,
+      "step": 132
+    },
+    {
+      "epoch": 0.14631463146314633,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.626745040411463e-05,
+      "loss": 1.8245,
+      "step": 133
+    },
+    {
+      "epoch": 0.1474147414741474,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.623806024981632e-05,
+      "loss": 1.5367,
+      "step": 134
+    },
+    {
+      "epoch": 0.1485148514851485,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.6208670095518e-05,
+      "loss": 1.8345,
+      "step": 135
+    },
+    {
+      "epoch": 0.1496149614961496,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.61792799412197e-05,
+      "loss": 1.8428,
+      "step": 136
+    },
+    {
+      "epoch": 0.15071507150715072,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.61498897869214e-05,
+      "loss": 2.0396,
+      "step": 137
+    },
+    {
+      "epoch": 0.15181518151815182,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.612049963262307e-05,
+      "loss": 1.7412,
+      "step": 138
+    },
+    {
+      "epoch": 0.15291529152915292,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.609110947832476e-05,
+      "loss": 1.6583,
+      "step": 139
+    },
+    {
+      "epoch": 0.15401540154015403,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.606171932402647e-05,
+      "loss": 2.0174,
+      "step": 140
+    },
+    {
+      "epoch": 0.1551155115511551,
+      "grad_norm": 2.375,
+      "learning_rate": 7.603232916972814e-05,
+      "loss": 1.5039,
+      "step": 141
+    },
+    {
+      "epoch": 0.1562156215621562,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.600293901542983e-05,
+      "loss": 1.7029,
+      "step": 142
+    },
+    {
+      "epoch": 0.1573157315731573,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.597354886113152e-05,
+      "loss": 2.1031,
+      "step": 143
+    },
+    {
+      "epoch": 0.15841584158415842,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.594415870683321e-05,
+      "loss": 1.7631,
+      "step": 144
+    },
+    {
+      "epoch": 0.15951595159515952,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.59147685525349e-05,
+      "loss": 1.5917,
+      "step": 145
+    },
+    {
+      "epoch": 0.16061606160616063,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.58853783982366e-05,
+      "loss": 1.5669,
+      "step": 146
+    },
+    {
+      "epoch": 0.1617161716171617,
+      "grad_norm": 2.625,
+      "learning_rate": 7.585598824393829e-05,
+      "loss": 1.5334,
+      "step": 147
+    },
+    {
+      "epoch": 0.1628162816281628,
+      "grad_norm": 2.625,
+      "learning_rate": 7.582659808963998e-05,
+      "loss": 1.6525,
+      "step": 148
+    },
+    {
+      "epoch": 0.1639163916391639,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.579720793534167e-05,
+      "loss": 1.6358,
+      "step": 149
+    },
+    {
+      "epoch": 0.16501650165016502,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.576781778104336e-05,
+      "loss": 1.6799,
+      "step": 150
+    },
+    {
+      "epoch": 0.16611661166116612,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.573842762674505e-05,
+      "loss": 1.5884,
+      "step": 151
+    },
+    {
+      "epoch": 0.16721672167216722,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.570903747244674e-05,
+      "loss": 1.6021,
+      "step": 152
+    },
+    {
+      "epoch": 0.16831683168316833,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.567964731814843e-05,
+      "loss": 1.9629,
+      "step": 153
+    },
+    {
+      "epoch": 0.1694169416941694,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.565025716385012e-05,
+      "loss": 1.6163,
+      "step": 154
+    },
+    {
+      "epoch": 0.1705170517051705,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.562086700955181e-05,
+      "loss": 1.7119,
+      "step": 155
+    },
+    {
+      "epoch": 0.1716171617161716,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.559147685525349e-05,
+      "loss": 1.5513,
+      "step": 156
+    },
+    {
+      "epoch": 0.17271727172717272,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.556208670095519e-05,
+      "loss": 1.5609,
+      "step": 157
+    },
+    {
+      "epoch": 0.17381738173817382,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.553269654665688e-05,
+      "loss": 1.7749,
+      "step": 158
+    },
+    {
+      "epoch": 0.17491749174917492,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.550330639235857e-05,
+      "loss": 1.6209,
+      "step": 159
+    },
+    {
+      "epoch": 0.17601760176017603,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.547391623806025e-05,
+      "loss": 1.7637,
+      "step": 160
+    },
+    {
+      "epoch": 0.1771177117711771,
+      "grad_norm": 2.25,
+      "learning_rate": 7.544452608376194e-05,
+      "loss": 1.1374,
+      "step": 161
+    },
+    {
+      "epoch": 0.1782178217821782,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.541513592946364e-05,
+      "loss": 1.3335,
+      "step": 162
+    },
+    {
+      "epoch": 0.1793179317931793,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.538574577516532e-05,
+      "loss": 1.727,
+      "step": 163
+    },
+    {
+      "epoch": 0.18041804180418042,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.535635562086701e-05,
+      "loss": 1.6084,
+      "step": 164
+    },
+    {
+      "epoch": 0.18151815181518152,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.53269654665687e-05,
+      "loss": 2.068,
+      "step": 165
+    },
+    {
+      "epoch": 0.18261826182618263,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.529757531227039e-05,
+      "loss": 1.4939,
+      "step": 166
+    },
+    {
+      "epoch": 0.18371837183718373,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.526818515797208e-05,
+      "loss": 1.4844,
+      "step": 167
+    },
+    {
+      "epoch": 0.1848184818481848,
+      "grad_norm": 2.75,
+      "learning_rate": 7.523879500367377e-05,
+      "loss": 1.4863,
+      "step": 168
+    },
+    {
+      "epoch": 0.1859185918591859,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.520940484937546e-05,
+      "loss": 1.5916,
+      "step": 169
+    },
+    {
+      "epoch": 0.18701870187018702,
+      "grad_norm": 2.875,
+      "learning_rate": 7.518001469507715e-05,
+      "loss": 1.6779,
+      "step": 170
+    },
+    {
+      "epoch": 0.18811881188118812,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.515062454077884e-05,
+      "loss": 1.4509,
+      "step": 171
+    },
+    {
+      "epoch": 0.18921892189218922,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.512123438648054e-05,
+      "loss": 1.5561,
+      "step": 172
+    },
+    {
+      "epoch": 0.19031903190319033,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.509184423218223e-05,
+      "loss": 1.5905,
+      "step": 173
+    },
+    {
+      "epoch": 0.19141914191419143,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.506245407788392e-05,
+      "loss": 1.604,
+      "step": 174
+    },
+    {
+      "epoch": 0.1925192519251925,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.503306392358561e-05,
+      "loss": 1.5499,
+      "step": 175
+    },
+    {
+      "epoch": 0.1936193619361936,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.50036737692873e-05,
+      "loss": 2.0668,
+      "step": 176
+    },
+    {
+      "epoch": 0.19471947194719472,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.497428361498899e-05,
+      "loss": 1.9587,
+      "step": 177
+    },
+    {
+      "epoch": 0.19581958195819582,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.494489346069066e-05,
+      "loss": 1.7427,
+      "step": 178
+    },
+    {
+      "epoch": 0.19691969196919692,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.491550330639237e-05,
+      "loss": 1.9622,
+      "step": 179
+    },
+    {
+      "epoch": 0.19801980198019803,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.488611315209406e-05,
+      "loss": 1.4033,
+      "step": 180
+    },
+    {
+      "epoch": 0.19911991199119913,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.485672299779575e-05,
+      "loss": 1.5019,
+      "step": 181
+    },
+    {
+      "epoch": 0.2002200220022002,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.482733284349743e-05,
+      "loss": 1.9215,
+      "step": 182
+    },
+    {
+      "epoch": 0.20132013201320131,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.479794268919913e-05,
+      "loss": 1.7523,
+      "step": 183
+    },
+    {
+      "epoch": 0.20242024202420242,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.476855253490082e-05,
+      "loss": 1.7564,
+      "step": 184
+    },
+    {
+      "epoch": 0.20352035203520352,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.47391623806025e-05,
+      "loss": 1.6221,
+      "step": 185
+    },
+    {
+      "epoch": 0.20462046204620463,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.470977222630419e-05,
+      "loss": 1.5086,
+      "step": 186
+    },
+    {
+      "epoch": 0.20572057205720573,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.468038207200589e-05,
+      "loss": 1.881,
+      "step": 187
+    },
+    {
+      "epoch": 0.2068206820682068,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.465099191770757e-05,
+      "loss": 1.4287,
+      "step": 188
+    },
+    {
+      "epoch": 0.2079207920792079,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.462160176340926e-05,
+      "loss": 1.6612,
+      "step": 189
+    },
+    {
+      "epoch": 0.20902090209020902,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.459221160911095e-05,
+      "loss": 1.5406,
+      "step": 190
+    },
+    {
+      "epoch": 0.21012101210121012,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.456282145481264e-05,
+      "loss": 1.7948,
+      "step": 191
+    },
+    {
+      "epoch": 0.21122112211221122,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.453343130051433e-05,
+      "loss": 1.7622,
+      "step": 192
+    },
+    {
+      "epoch": 0.21232123212321233,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.450404114621602e-05,
+      "loss": 1.5527,
+      "step": 193
+    },
+    {
+      "epoch": 0.21342134213421343,
+      "grad_norm": 2.75,
+      "learning_rate": 7.447465099191771e-05,
+      "loss": 1.9561,
+      "step": 194
+    },
+    {
+      "epoch": 0.2145214521452145,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.44452608376194e-05,
+      "loss": 1.7265,
+      "step": 195
+    },
+    {
+      "epoch": 0.2156215621562156,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.44158706833211e-05,
+      "loss": 1.6651,
+      "step": 196
+    },
+    {
+      "epoch": 0.21672167216721672,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.438648052902278e-05,
+      "loss": 1.4983,
+      "step": 197
+    },
+    {
+      "epoch": 0.21782178217821782,
+      "grad_norm": 2.375,
+      "learning_rate": 7.435709037472447e-05,
+      "loss": 1.718,
+      "step": 198
+    },
+    {
+      "epoch": 0.21892189218921893,
+      "grad_norm": 2.5,
+      "learning_rate": 7.432770022042617e-05,
+      "loss": 1.7679,
+      "step": 199
+    },
+    {
+      "epoch": 0.22002200220022003,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.429831006612786e-05,
+      "loss": 1.6781,
+      "step": 200
+    },
+    {
+      "epoch": 0.22112211221122113,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.426891991182955e-05,
+      "loss": 1.637,
+      "step": 201
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.423952975753124e-05,
+      "loss": 1.834,
+      "step": 202
+    },
+    {
+      "epoch": 0.22332233223322331,
+      "grad_norm": 2.25,
+      "learning_rate": 7.421013960323291e-05,
+      "loss": 1.5641,
+      "step": 203
+    },
+    {
+      "epoch": 0.22442244224422442,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.418074944893462e-05,
+      "loss": 1.3664,
+      "step": 204
+    },
+    {
+      "epoch": 0.22552255225522552,
+      "grad_norm": 2.5,
+      "learning_rate": 7.415135929463631e-05,
+      "loss": 1.5489,
+      "step": 205
+    },
+    {
+      "epoch": 0.22662266226622663,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.4121969140338e-05,
+      "loss": 1.6292,
+      "step": 206
+    },
+    {
+      "epoch": 0.22772277227722773,
+      "grad_norm": 2.75,
+      "learning_rate": 7.409257898603968e-05,
+      "loss": 1.5634,
+      "step": 207
+    },
+    {
+      "epoch": 0.22882288228822883,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.406318883174137e-05,
+      "loss": 1.5732,
+      "step": 208
+    },
+    {
+      "epoch": 0.2299229922992299,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.403379867744307e-05,
+      "loss": 1.5683,
+      "step": 209
+    },
+    {
+      "epoch": 0.23102310231023102,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.400440852314475e-05,
+      "loss": 1.8769,
+      "step": 210
+    },
+    {
+      "epoch": 0.23212321232123212,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.397501836884644e-05,
+      "loss": 1.6611,
+      "step": 211
+    },
+    {
+      "epoch": 0.23322332233223322,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.394562821454813e-05,
+      "loss": 1.518,
+      "step": 212
+    },
+    {
+      "epoch": 0.23432343234323433,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.391623806024982e-05,
+      "loss": 1.8108,
+      "step": 213
+    },
+    {
+      "epoch": 0.23542354235423543,
+      "grad_norm": 2.625,
+      "learning_rate": 7.388684790595151e-05,
+      "loss": 1.5755,
+      "step": 214
+    },
+    {
+      "epoch": 0.23652365236523654,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.38574577516532e-05,
+      "loss": 1.7679,
+      "step": 215
+    },
+    {
+      "epoch": 0.2376237623762376,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.382806759735489e-05,
+      "loss": 1.8811,
+      "step": 216
+    },
+    {
+      "epoch": 0.23872387238723872,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.379867744305658e-05,
+      "loss": 1.7181,
+      "step": 217
+    },
+    {
+      "epoch": 0.23982398239823982,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.376928728875827e-05,
+      "loss": 1.5724,
+      "step": 218
+    },
+    {
+      "epoch": 0.24092409240924093,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.373989713445996e-05,
+      "loss": 1.6301,
+      "step": 219
+    },
+    {
+      "epoch": 0.24202420242024203,
+      "grad_norm": 2.625,
+      "learning_rate": 7.371050698016165e-05,
+      "loss": 2.0522,
+      "step": 220
+    },
+    {
+      "epoch": 0.24312431243124313,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.368111682586334e-05,
+      "loss": 1.9336,
+      "step": 221
+    },
+    {
+      "epoch": 0.24422442244224424,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.365172667156503e-05,
+      "loss": 1.873,
+      "step": 222
+    },
+    {
+      "epoch": 0.24532453245324531,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.362233651726672e-05,
+      "loss": 1.2278,
+      "step": 223
+    },
+    {
+      "epoch": 0.24642464246424642,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.359294636296841e-05,
+      "loss": 1.7754,
+      "step": 224
+    },
+    {
+      "epoch": 0.24752475247524752,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.356355620867009e-05,
+      "loss": 1.6586,
+      "step": 225
+    },
+    {
+      "epoch": 0.24862486248624863,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.35341660543718e-05,
+      "loss": 1.7999,
+      "step": 226
+    },
+    {
+      "epoch": 0.24972497249724973,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.350477590007349e-05,
+      "loss": 1.7571,
+      "step": 227
+    },
+    {
+      "epoch": 0.2508250825082508,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.347538574577516e-05,
+      "loss": 1.7123,
+      "step": 228
+    },
+    {
+      "epoch": 0.25192519251925194,
+      "grad_norm": 2.25,
+      "learning_rate": 7.344599559147685e-05,
+      "loss": 1.4597,
+      "step": 229
+    },
+    {
+      "epoch": 0.253025302530253,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.341660543717856e-05,
+      "loss": 1.6806,
+      "step": 230
+    },
+    {
+      "epoch": 0.25412541254125415,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.338721528288025e-05,
+      "loss": 1.5726,
+      "step": 231
+    },
+    {
+      "epoch": 0.2552255225522552,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.335782512858192e-05,
+      "loss": 1.7357,
+      "step": 232
+    },
+    {
+      "epoch": 0.2563256325632563,
+      "grad_norm": 3.25,
+      "learning_rate": 7.332843497428362e-05,
+      "loss": 1.5015,
+      "step": 233
+    },
+    {
+      "epoch": 0.25742574257425743,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.329904481998532e-05,
+      "loss": 1.5673,
+      "step": 234
+    },
+    {
+      "epoch": 0.2585258525852585,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.3269654665687e-05,
+      "loss": 1.3667,
+      "step": 235
+    },
+    {
+      "epoch": 0.25962596259625964,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.324026451138869e-05,
+      "loss": 1.8071,
+      "step": 236
+    },
+    {
+      "epoch": 0.2607260726072607,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.321087435709038e-05,
+      "loss": 1.7789,
+      "step": 237
+    },
+    {
+      "epoch": 0.26182618261826185,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.318148420279207e-05,
+      "loss": 1.2831,
+      "step": 238
+    },
+    {
+      "epoch": 0.2629262926292629,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.315209404849376e-05,
+      "loss": 1.5218,
+      "step": 239
+    },
+    {
+      "epoch": 0.264026402640264,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.312270389419545e-05,
+      "loss": 1.6039,
+      "step": 240
+    },
+    {
+      "epoch": 0.26512651265126513,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.309331373989714e-05,
+      "loss": 1.4841,
+      "step": 241
+    },
+    {
+      "epoch": 0.2662266226622662,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.306392358559883e-05,
+      "loss": 1.696,
+      "step": 242
+    },
+    {
+      "epoch": 0.26732673267326734,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.303453343130052e-05,
+      "loss": 1.5164,
+      "step": 243
+    },
+    {
+      "epoch": 0.2684268426842684,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.300514327700221e-05,
+      "loss": 1.4525,
+      "step": 244
+    },
+    {
+      "epoch": 0.26952695269526955,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.29757531227039e-05,
+      "loss": 1.573,
+      "step": 245
+    },
+    {
+      "epoch": 0.2706270627062706,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.294636296840559e-05,
+      "loss": 1.8129,
+      "step": 246
+    },
+    {
+      "epoch": 0.2717271727172717,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.291697281410728e-05,
+      "loss": 1.7507,
+      "step": 247
+    },
+    {
+      "epoch": 0.27282728272827284,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.288758265980897e-05,
+      "loss": 1.5875,
+      "step": 248
+    },
+    {
+      "epoch": 0.2739273927392739,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.285819250551066e-05,
+      "loss": 1.6278,
+      "step": 249
+    },
+    {
+      "epoch": 0.27502750275027504,
+      "grad_norm": 2.375,
+      "learning_rate": 7.282880235121234e-05,
+      "loss": 1.6012,
+      "step": 250
+    },
+    {
+      "epoch": 0.2761276127612761,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.279941219691404e-05,
+      "loss": 1.9209,
+      "step": 251
+    },
+    {
+      "epoch": 0.27722772277227725,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.277002204261574e-05,
+      "loss": 1.7311,
+      "step": 252
+    },
+    {
+      "epoch": 0.27832783278327833,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.274063188831743e-05,
+      "loss": 1.6439,
+      "step": 253
+    },
+    {
+      "epoch": 0.2794279427942794,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.27112417340191e-05,
+      "loss": 1.7193,
+      "step": 254
+    },
+    {
+      "epoch": 0.28052805280528054,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.26818515797208e-05,
+      "loss": 1.5784,
+      "step": 255
+    },
+    {
+      "epoch": 0.2816281628162816,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.26524614254225e-05,
+      "loss": 1.5997,
+      "step": 256
+    },
+    {
+      "epoch": 0.28272827282728275,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.262307127112417e-05,
+      "loss": 1.4495,
+      "step": 257
+    },
+    {
+      "epoch": 0.2838283828382838,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.259368111682586e-05,
+      "loss": 1.8039,
+      "step": 258
+    },
+    {
+      "epoch": 0.28492849284928495,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.256429096252755e-05,
+      "loss": 1.7299,
+      "step": 259
+    },
+    {
+      "epoch": 0.28602860286028603,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.253490080822925e-05,
+      "loss": 1.6577,
+      "step": 260
+    },
+    {
+      "epoch": 0.2871287128712871,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.250551065393094e-05,
+      "loss": 1.5686,
+      "step": 261
+    },
+    {
+      "epoch": 0.28822882288228824,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.247612049963263e-05,
+      "loss": 1.6936,
+      "step": 262
+    },
+    {
+      "epoch": 0.2893289328932893,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.244673034533432e-05,
+      "loss": 1.6112,
+      "step": 263
+    },
+    {
+      "epoch": 0.29042904290429045,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.241734019103601e-05,
+      "loss": 1.6046,
+      "step": 264
+    },
+    {
+      "epoch": 0.2915291529152915,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.23879500367377e-05,
+      "loss": 1.5808,
+      "step": 265
+    },
+    {
+      "epoch": 0.29262926292629265,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.235855988243939e-05,
+      "loss": 1.6693,
+      "step": 266
+    },
+    {
+      "epoch": 0.29372937293729373,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.232916972814108e-05,
+      "loss": 1.5833,
+      "step": 267
+    },
+    {
+      "epoch": 0.2948294829482948,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.229977957384277e-05,
+      "loss": 1.4142,
+      "step": 268
+    },
+    {
+      "epoch": 0.29592959295929594,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.227038941954446e-05,
+      "loss": 1.7591,
+      "step": 269
+    },
+    {
+      "epoch": 0.297029702970297,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.224099926524615e-05,
+      "loss": 1.5419,
+      "step": 270
+    },
+    {
+      "epoch": 0.29812981298129815,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.221160911094784e-05,
+      "loss": 1.7169,
+      "step": 271
+    },
+    {
+      "epoch": 0.2992299229922992,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.218221895664953e-05,
+      "loss": 1.5188,
+      "step": 272
+    },
+    {
+      "epoch": 0.30033003300330036,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.215282880235122e-05,
+      "loss": 1.8599,
+      "step": 273
+    },
+    {
+      "epoch": 0.30143014301430143,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.212343864805291e-05,
+      "loss": 1.7104,
+      "step": 274
+    },
+    {
+      "epoch": 0.3025302530253025,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.209404849375459e-05,
+      "loss": 1.5099,
+      "step": 275
+    },
+    {
+      "epoch": 0.30363036303630364,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.206465833945628e-05,
+      "loss": 1.5086,
+      "step": 276
+    },
+    {
+      "epoch": 0.3047304730473047,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.203526818515798e-05,
+      "loss": 1.734,
+      "step": 277
+    },
+    {
+      "epoch": 0.30583058305830585,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.200587803085967e-05,
+      "loss": 1.7063,
+      "step": 278
+    },
+    {
+      "epoch": 0.3069306930693069,
+      "grad_norm": 2.25,
+      "learning_rate": 7.197648787656135e-05,
+      "loss": 1.5324,
+      "step": 279
+    },
+    {
+      "epoch": 0.30803080308030806,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.194709772226304e-05,
+      "loss": 1.5914,
+      "step": 280
+    },
+    {
+      "epoch": 0.30913091309130913,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.191770756796475e-05,
+      "loss": 1.6797,
+      "step": 281
+    },
+    {
+      "epoch": 0.3102310231023102,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.188831741366642e-05,
+      "loss": 1.4601,
+      "step": 282
+    },
+    {
+      "epoch": 0.31133113311331134,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.185892725936811e-05,
+      "loss": 1.5889,
+      "step": 283
+    },
+    {
+      "epoch": 0.3124312431243124,
+      "grad_norm": 2.375,
+      "learning_rate": 7.18295371050698e-05,
+      "loss": 1.8779,
+      "step": 284
+    },
+    {
+      "epoch": 0.31353135313531355,
+      "grad_norm": 2.375,
+      "learning_rate": 7.18001469507715e-05,
+      "loss": 1.5158,
+      "step": 285
+    },
+    {
+      "epoch": 0.3146314631463146,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.177075679647319e-05,
+      "loss": 1.4873,
+      "step": 286
+    },
+    {
+      "epoch": 0.31573157315731576,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.174136664217488e-05,
+      "loss": 1.7513,
+      "step": 287
+    },
+    {
+      "epoch": 0.31683168316831684,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.171197648787657e-05,
+      "loss": 1.5886,
+      "step": 288
+    },
+    {
+      "epoch": 0.3179317931793179,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.168258633357826e-05,
+      "loss": 1.7757,
+      "step": 289
+    },
+    {
+      "epoch": 0.31903190319031904,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.165319617927995e-05,
+      "loss": 1.6159,
+      "step": 290
+    },
+    {
+      "epoch": 0.3201320132013201,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.162380602498164e-05,
+      "loss": 1.7434,
+      "step": 291
+    },
+    {
+      "epoch": 0.32123212321232125,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.159441587068333e-05,
+      "loss": 1.7193,
+      "step": 292
+    },
+    {
+      "epoch": 0.32233223322332233,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.156502571638502e-05,
+      "loss": 1.6798,
+      "step": 293
+    },
+    {
+      "epoch": 0.3234323432343234,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.153563556208671e-05,
+      "loss": 1.6219,
+      "step": 294
+    },
+    {
+      "epoch": 0.32453245324532454,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.15062454077884e-05,
+      "loss": 1.4986,
+      "step": 295
+    },
+    {
+      "epoch": 0.3256325632563256,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.147685525349009e-05,
+      "loss": 1.6969,
+      "step": 296
+    },
+    {
+      "epoch": 0.32673267326732675,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.144746509919177e-05,
+      "loss": 1.7282,
+      "step": 297
+    },
+    {
+      "epoch": 0.3278327832783278,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.141807494489347e-05,
+      "loss": 1.6309,
+      "step": 298
+    },
+    {
+      "epoch": 0.32893289328932895,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.138868479059516e-05,
+      "loss": 1.6723,
+      "step": 299
+    },
+    {
+      "epoch": 0.33003300330033003,
+      "grad_norm": 2.625,
+      "learning_rate": 7.135929463629684e-05,
+      "loss": 1.6562,
+      "step": 300
+    },
+    {
+      "epoch": 0.3311331133113311,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.132990448199853e-05,
+      "loss": 1.8531,
+      "step": 301
+    },
+    {
+      "epoch": 0.33223322332233224,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.130051432770023e-05,
+      "loss": 1.5951,
+      "step": 302
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.127112417340192e-05,
+      "loss": 1.8763,
+      "step": 303
+    },
+    {
+      "epoch": 0.33443344334433445,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.12417340191036e-05,
+      "loss": 1.5892,
+      "step": 304
+    },
+    {
+      "epoch": 0.3355335533553355,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.121234386480529e-05,
+      "loss": 1.6187,
+      "step": 305
+    },
+    {
+      "epoch": 0.33663366336633666,
+      "grad_norm": 2.5,
+      "learning_rate": 7.1182953710507e-05,
+      "loss": 1.6265,
+      "step": 306
+    },
+    {
+      "epoch": 0.33773377337733773,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.115356355620867e-05,
+      "loss": 1.7981,
+      "step": 307
+    },
+    {
+      "epoch": 0.3388338833883388,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.112417340191036e-05,
+      "loss": 1.6605,
+      "step": 308
+    },
+    {
+      "epoch": 0.33993399339933994,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.109478324761205e-05,
+      "loss": 1.3162,
+      "step": 309
+    },
+    {
+      "epoch": 0.341034103410341,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.106539309331374e-05,
+      "loss": 1.6,
+      "step": 310
+    },
+    {
+      "epoch": 0.34213421342134215,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.103600293901543e-05,
+      "loss": 1.7578,
+      "step": 311
+    },
+    {
+      "epoch": 0.3432343234323432,
+      "grad_norm": 2.375,
+      "learning_rate": 7.100661278471712e-05,
+      "loss": 1.8671,
+      "step": 312
+    },
+    {
+      "epoch": 0.34433443344334436,
+      "grad_norm": 2.625,
+      "learning_rate": 7.097722263041882e-05,
+      "loss": 1.5863,
+      "step": 313
+    },
+    {
+      "epoch": 0.34543454345434543,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.09478324761205e-05,
+      "loss": 1.4452,
+      "step": 314
+    },
+    {
+      "epoch": 0.3465346534653465,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.09184423218222e-05,
+      "loss": 1.7327,
+      "step": 315
+    },
+    {
+      "epoch": 0.34763476347634764,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.088905216752389e-05,
+      "loss": 1.9633,
+      "step": 316
+    },
+    {
+      "epoch": 0.3487348734873487,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.085966201322558e-05,
+      "loss": 1.6835,
+      "step": 317
+    },
+    {
+      "epoch": 0.34983498349834985,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.083027185892727e-05,
+      "loss": 1.699,
+      "step": 318
+    },
+    {
+      "epoch": 0.3509350935093509,
+      "grad_norm": 2.375,
+      "learning_rate": 7.080088170462896e-05,
+      "loss": 1.4307,
+      "step": 319
+    },
+    {
+      "epoch": 0.35203520352035206,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.077149155033065e-05,
+      "loss": 1.4165,
+      "step": 320
+    },
+    {
+      "epoch": 0.35313531353135313,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.074210139603234e-05,
+      "loss": 1.5924,
+      "step": 321
+    },
+    {
+      "epoch": 0.3542354235423542,
+      "grad_norm": 2.5,
+      "learning_rate": 7.071271124173402e-05,
+      "loss": 1.7753,
+      "step": 322
+    },
+    {
+      "epoch": 0.35533553355335534,
+      "grad_norm": 2.25,
+      "learning_rate": 7.068332108743572e-05,
+      "loss": 1.9986,
+      "step": 323
+    },
+    {
+      "epoch": 0.3564356435643564,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.065393093313741e-05,
+      "loss": 1.6479,
+      "step": 324
+    },
+    {
+      "epoch": 0.35753575357535755,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.06245407788391e-05,
+      "loss": 1.7177,
+      "step": 325
+    },
+    {
+      "epoch": 0.3586358635863586,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.059515062454078e-05,
+      "loss": 1.8747,
+      "step": 326
+    },
+    {
+      "epoch": 0.35973597359735976,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.056576047024247e-05,
+      "loss": 1.5444,
+      "step": 327
+    },
+    {
+      "epoch": 0.36083608360836084,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.053637031594417e-05,
+      "loss": 1.6521,
+      "step": 328
+    },
+    {
+      "epoch": 0.3619361936193619,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.050698016164585e-05,
+      "loss": 1.6102,
+      "step": 329
+    },
+    {
+      "epoch": 0.36303630363036304,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.047759000734754e-05,
+      "loss": 1.514,
+      "step": 330
+    },
+    {
+      "epoch": 0.3641364136413641,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.044819985304923e-05,
+      "loss": 1.4824,
+      "step": 331
+    },
+    {
+      "epoch": 0.36523652365236525,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.041880969875092e-05,
+      "loss": 1.3038,
+      "step": 332
+    },
+    {
+      "epoch": 0.36633663366336633,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.038941954445261e-05,
+      "loss": 2.0307,
+      "step": 333
+    },
+    {
+      "epoch": 0.36743674367436746,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.03600293901543e-05,
+      "loss": 1.7726,
+      "step": 334
+    },
+    {
+      "epoch": 0.36853685368536854,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.033063923585599e-05,
+      "loss": 1.6265,
+      "step": 335
+    },
+    {
+      "epoch": 0.3696369636963696,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.030124908155768e-05,
+      "loss": 1.67,
+      "step": 336
+    },
+    {
+      "epoch": 0.37073707370737075,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.027185892725937e-05,
+      "loss": 1.7445,
+      "step": 337
+    },
+    {
+      "epoch": 0.3718371837183718,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.024246877296106e-05,
+      "loss": 1.5664,
+      "step": 338
+    },
+    {
+      "epoch": 0.37293729372937295,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.021307861866275e-05,
+      "loss": 1.395,
+      "step": 339
+    },
+    {
+      "epoch": 0.37403740374037403,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.018368846436445e-05,
+      "loss": 1.5827,
+      "step": 340
+    },
+    {
+      "epoch": 0.37513751375137516,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.015429831006614e-05,
+      "loss": 1.3384,
+      "step": 341
+    },
+    {
+      "epoch": 0.37623762376237624,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.012490815576783e-05,
+      "loss": 1.5161,
+      "step": 342
+    },
+    {
+      "epoch": 0.3773377337733773,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.009551800146952e-05,
+      "loss": 1.9063,
+      "step": 343
+    },
+    {
+      "epoch": 0.37843784378437845,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.00661278471712e-05,
+      "loss": 1.6597,
+      "step": 344
+    },
+    {
+      "epoch": 0.3795379537953795,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.00367376928729e-05,
+      "loss": 1.5507,
+      "step": 345
+    },
+    {
+      "epoch": 0.38063806380638066,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.000734753857459e-05,
+      "loss": 1.6965,
+      "step": 346
+    },
+    {
+      "epoch": 0.38173817381738173,
+      "grad_norm": 2.625,
+      "learning_rate": 6.997795738427627e-05,
+      "loss": 1.4365,
+      "step": 347
+    },
+    {
+      "epoch": 0.38283828382838286,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.994856722997796e-05,
+      "loss": 1.6257,
+      "step": 348
+    },
+    {
+      "epoch": 0.38393839383938394,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.991917707567966e-05,
+      "loss": 1.4717,
+      "step": 349
+    },
+    {
+      "epoch": 0.385038503850385,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.988978692138135e-05,
+      "loss": 1.7631,
+      "step": 350
+    },
+    {
+      "epoch": 0.38613861386138615,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.986039676708303e-05,
+      "loss": 1.4739,
+      "step": 351
+    },
+    {
+      "epoch": 0.3872387238723872,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.983100661278472e-05,
+      "loss": 1.5912,
+      "step": 352
+    },
+    {
+      "epoch": 0.38833883388338836,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.980161645848642e-05,
+      "loss": 1.8626,
+      "step": 353
+    },
+    {
+      "epoch": 0.38943894389438943,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.97722263041881e-05,
+      "loss": 1.6295,
+      "step": 354
+    },
+    {
+      "epoch": 0.39053905390539057,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.974283614988979e-05,
+      "loss": 1.4841,
+      "step": 355
+    },
+    {
+      "epoch": 0.39163916391639164,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.971344599559148e-05,
+      "loss": 1.5922,
+      "step": 356
+    },
+    {
+      "epoch": 0.3927392739273927,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.968405584129317e-05,
+      "loss": 1.556,
+      "step": 357
+    },
+    {
+      "epoch": 0.39383938393839385,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.965466568699486e-05,
+      "loss": 1.7782,
+      "step": 358
+    },
+    {
+      "epoch": 0.3949394939493949,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.962527553269655e-05,
+      "loss": 1.4288,
+      "step": 359
+    },
+    {
+      "epoch": 0.39603960396039606,
+      "grad_norm": 2.625,
+      "learning_rate": 6.959588537839824e-05,
+      "loss": 1.6686,
+      "step": 360
+    },
+    {
+      "epoch": 0.39713971397139713,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.956649522409993e-05,
+      "loss": 1.6766,
+      "step": 361
+    },
+    {
+      "epoch": 0.39823982398239827,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.953710506980162e-05,
+      "loss": 1.4964,
+      "step": 362
+    },
+    {
+      "epoch": 0.39933993399339934,
+      "grad_norm": 2.5,
+      "learning_rate": 6.950771491550331e-05,
+      "loss": 1.8088,
+      "step": 363
+    },
+    {
+      "epoch": 0.4004400440044004,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.9478324761205e-05,
+      "loss": 1.746,
+      "step": 364
+    },
+    {
+      "epoch": 0.40154015401540155,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.94489346069067e-05,
+      "loss": 1.8511,
+      "step": 365
+    },
+    {
+      "epoch": 0.40264026402640263,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.941954445260838e-05,
+      "loss": 1.5927,
+      "step": 366
+    },
+    {
+      "epoch": 0.40374037403740376,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.939015429831008e-05,
+      "loss": 1.6274,
+      "step": 367
+    },
+    {
+      "epoch": 0.40484048404840484,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.936076414401177e-05,
+      "loss": 1.7494,
+      "step": 368
+    },
+    {
+      "epoch": 0.40594059405940597,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.933137398971344e-05,
+      "loss": 1.3937,
+      "step": 369
+    },
+    {
+      "epoch": 0.40704070407040704,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.930198383541515e-05,
+      "loss": 1.6917,
+      "step": 370
+    },
+    {
+      "epoch": 0.4081408140814081,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.927259368111684e-05,
+      "loss": 1.7453,
+      "step": 371
+    },
+    {
+      "epoch": 0.40924092409240925,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.924320352681853e-05,
+      "loss": 1.6261,
+      "step": 372
+    },
+    {
+      "epoch": 0.41034103410341033,
+      "grad_norm": 2.375,
+      "learning_rate": 6.92138133725202e-05,
+      "loss": 1.7451,
+      "step": 373
+    },
+    {
+      "epoch": 0.41144114411441146,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.918442321822191e-05,
+      "loss": 1.5594,
+      "step": 374
+    },
+    {
+      "epoch": 0.41254125412541254,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.91550330639236e-05,
+      "loss": 1.6174,
+      "step": 375
+    },
+    {
+      "epoch": 0.4136413641364136,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.912564290962528e-05,
+      "loss": 1.6909,
+      "step": 376
+    },
+    {
+      "epoch": 0.41474147414741475,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.909625275532697e-05,
+      "loss": 1.4864,
+      "step": 377
+    },
+    {
+      "epoch": 0.4158415841584158,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.906686260102866e-05,
+      "loss": 1.6935,
+      "step": 378
+    },
+    {
+      "epoch": 0.41694169416941695,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.903747244673035e-05,
+      "loss": 1.3741,
+      "step": 379
+    },
+    {
+      "epoch": 0.41804180418041803,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.900808229243204e-05,
+      "loss": 1.6303,
+      "step": 380
+    },
+    {
+      "epoch": 0.41914191419141916,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.897869213813373e-05,
+      "loss": 1.7352,
+      "step": 381
+    },
+    {
+      "epoch": 0.42024202420242024,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.894930198383542e-05,
+      "loss": 1.7933,
+      "step": 382
+    },
+    {
+      "epoch": 0.4213421342134213,
+      "grad_norm": 2.5,
+      "learning_rate": 6.891991182953711e-05,
+      "loss": 1.7936,
+      "step": 383
+    },
+    {
+      "epoch": 0.42244224422442245,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.88905216752388e-05,
+      "loss": 1.8693,
+      "step": 384
+    },
+    {
+      "epoch": 0.4235423542354235,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.886113152094049e-05,
+      "loss": 1.5018,
+      "step": 385
+    },
+    {
+      "epoch": 0.42464246424642466,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.883174136664218e-05,
+      "loss": 1.8069,
+      "step": 386
+    },
+    {
+      "epoch": 0.42574257425742573,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.880235121234387e-05,
+      "loss": 1.8319,
+      "step": 387
+    },
+    {
+      "epoch": 0.42684268426842686,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.877296105804556e-05,
+      "loss": 1.5653,
+      "step": 388
+    },
+    {
+      "epoch": 0.42794279427942794,
+      "grad_norm": 2.25,
+      "learning_rate": 6.874357090374725e-05,
+      "loss": 1.6618,
+      "step": 389
+    },
+    {
+      "epoch": 0.429042904290429,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.871418074944894e-05,
+      "loss": 1.8716,
+      "step": 390
+    },
+    {
+      "epoch": 0.43014301430143015,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.868479059515063e-05,
+      "loss": 1.5454,
+      "step": 391
+    },
+    {
+      "epoch": 0.4312431243124312,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.865540044085232e-05,
+      "loss": 1.8307,
+      "step": 392
+    },
+    {
+      "epoch": 0.43234323432343236,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.862601028655401e-05,
+      "loss": 1.6726,
+      "step": 393
+    },
+    {
+      "epoch": 0.43344334433443343,
+      "grad_norm": 2.625,
+      "learning_rate": 6.859662013225569e-05,
+      "loss": 1.7123,
+      "step": 394
+    },
+    {
+      "epoch": 0.43454345434543457,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.856722997795738e-05,
+      "loss": 1.5337,
+      "step": 395
+    },
+    {
+      "epoch": 0.43564356435643564,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.853783982365909e-05,
+      "loss": 1.7549,
+      "step": 396
+    },
+    {
+      "epoch": 0.4367436743674367,
+      "grad_norm": 2.75,
+      "learning_rate": 6.850844966936078e-05,
+      "loss": 1.5537,
+      "step": 397
+    },
+    {
+      "epoch": 0.43784378437843785,
+      "grad_norm": 2.25,
+      "learning_rate": 6.847905951506245e-05,
+      "loss": 1.6732,
+      "step": 398
+    },
+    {
+      "epoch": 0.4389438943894389,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.844966936076414e-05,
+      "loss": 1.8663,
+      "step": 399
+    },
+    {
+      "epoch": 0.44004400440044006,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.842027920646585e-05,
+      "loss": 1.7585,
+      "step": 400
+    },
+    {
+      "epoch": 0.44114411441144114,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.839088905216753e-05,
+      "loss": 1.6756,
+      "step": 401
+    },
+    {
+      "epoch": 0.44224422442244227,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.836149889786922e-05,
+      "loss": 1.7382,
+      "step": 402
+    },
+    {
+      "epoch": 0.44334433443344334,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.83321087435709e-05,
+      "loss": 1.6183,
+      "step": 403
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.83027185892726e-05,
+      "loss": 1.6384,
+      "step": 404
+    },
+    {
+      "epoch": 0.44554455445544555,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.827332843497429e-05,
+      "loss": 1.6465,
+      "step": 405
+    },
+    {
+      "epoch": 0.44664466446644663,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.824393828067598e-05,
+      "loss": 1.6681,
+      "step": 406
+    },
+    {
+      "epoch": 0.44774477447744776,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.821454812637767e-05,
+      "loss": 1.5549,
+      "step": 407
+    },
+    {
+      "epoch": 0.44884488448844884,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.818515797207936e-05,
+      "loss": 1.7307,
+      "step": 408
+    },
+    {
+      "epoch": 0.44994499449944997,
+      "grad_norm": 2.75,
+      "learning_rate": 6.815576781778105e-05,
+      "loss": 1.4536,
+      "step": 409
+    },
+    {
+      "epoch": 0.45104510451045104,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.812637766348274e-05,
+      "loss": 1.5995,
+      "step": 410
+    },
+    {
+      "epoch": 0.4521452145214521,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.809698750918443e-05,
+      "loss": 1.3394,
+      "step": 411
+    },
+    {
+      "epoch": 0.45324532453245325,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.806759735488612e-05,
+      "loss": 1.6423,
+      "step": 412
+    },
+    {
+      "epoch": 0.45434543454345433,
+      "grad_norm": 2.625,
+      "learning_rate": 6.803820720058781e-05,
+      "loss": 1.5503,
+      "step": 413
+    },
+    {
+      "epoch": 0.45544554455445546,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.80088170462895e-05,
+      "loss": 1.6648,
+      "step": 414
+    },
+    {
+      "epoch": 0.45654565456545654,
+      "grad_norm": 2.25,
+      "learning_rate": 6.797942689199119e-05,
+      "loss": 1.6802,
+      "step": 415
+    },
+    {
+      "epoch": 0.45764576457645767,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.795003673769287e-05,
+      "loss": 1.5534,
+      "step": 416
+    },
+    {
+      "epoch": 0.45874587458745875,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.792064658339457e-05,
+      "loss": 1.8407,
+      "step": 417
+    },
+    {
+      "epoch": 0.4598459845984598,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.789125642909626e-05,
+      "loss": 1.7772,
+      "step": 418
+    },
+    {
+      "epoch": 0.46094609460946095,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.786186627479794e-05,
+      "loss": 1.4179,
+      "step": 419
+    },
+    {
+      "epoch": 0.46204620462046203,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.783247612049963e-05,
+      "loss": 1.526,
+      "step": 420
+    },
+    {
+      "epoch": 0.46314631463146316,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.780308596620134e-05,
+      "loss": 1.3918,
+      "step": 421
+    },
+    {
+      "epoch": 0.46424642464246424,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.777369581190303e-05,
+      "loss": 1.5415,
+      "step": 422
+    },
+    {
+      "epoch": 0.46534653465346537,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.77443056576047e-05,
+      "loss": 1.434,
+      "step": 423
+    },
+    {
+      "epoch": 0.46644664466446645,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.77149155033064e-05,
+      "loss": 1.7102,
+      "step": 424
+    },
+    {
+      "epoch": 0.4675467546754675,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.76855253490081e-05,
+      "loss": 1.787,
+      "step": 425
+    },
+    {
+      "epoch": 0.46864686468646866,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.765613519470977e-05,
+      "loss": 1.6816,
+      "step": 426
+    },
+    {
+      "epoch": 0.46974697469746973,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.762674504041146e-05,
+      "loss": 1.4694,
+      "step": 427
+    },
+    {
+      "epoch": 0.47084708470847086,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.759735488611316e-05,
+      "loss": 1.5861,
+      "step": 428
+    },
+    {
+      "epoch": 0.47194719471947194,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.756796473181485e-05,
+      "loss": 1.8261,
+      "step": 429
+    },
+    {
+      "epoch": 0.4730473047304731,
+      "grad_norm": 2.625,
+      "learning_rate": 6.753857457751654e-05,
+      "loss": 1.6531,
+      "step": 430
+    },
+    {
+      "epoch": 0.47414741474147415,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.750918442321823e-05,
+      "loss": 1.5926,
+      "step": 431
+    },
+    {
+      "epoch": 0.4752475247524752,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.747979426891992e-05,
+      "loss": 1.8642,
+      "step": 432
+    },
+    {
+      "epoch": 0.47634763476347636,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.745040411462161e-05,
+      "loss": 1.4616,
+      "step": 433
+    },
+    {
+      "epoch": 0.47744774477447743,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.74210139603233e-05,
+      "loss": 1.6287,
+      "step": 434
+    },
+    {
+      "epoch": 0.47854785478547857,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.739162380602499e-05,
+      "loss": 1.6636,
+      "step": 435
+    },
+    {
+      "epoch": 0.47964796479647964,
+      "grad_norm": 2.25,
+      "learning_rate": 6.736223365172668e-05,
+      "loss": 1.4892,
+      "step": 436
+    },
+    {
+      "epoch": 0.4807480748074808,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.733284349742837e-05,
+      "loss": 1.5949,
+      "step": 437
+    },
+    {
+      "epoch": 0.48184818481848185,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.730345334313006e-05,
+      "loss": 1.697,
+      "step": 438
+    },
+    {
+      "epoch": 0.4829482948294829,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.727406318883175e-05,
+      "loss": 1.4523,
+      "step": 439
+    },
+    {
+      "epoch": 0.48404840484048406,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.724467303453344e-05,
+      "loss": 1.7106,
+      "step": 440
+    },
+    {
+      "epoch": 0.48514851485148514,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.721528288023512e-05,
+      "loss": 1.4608,
+      "step": 441
+    },
+    {
+      "epoch": 0.48624862486248627,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.718589272593682e-05,
+      "loss": 1.6342,
+      "step": 442
+    },
+    {
+      "epoch": 0.48734873487348734,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.715650257163851e-05,
+      "loss": 1.7359,
+      "step": 443
+    },
+    {
+      "epoch": 0.4884488448844885,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.71271124173402e-05,
+      "loss": 1.7394,
+      "step": 444
+    },
+    {
+      "epoch": 0.48954895489548955,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.709772226304188e-05,
+      "loss": 1.4434,
+      "step": 445
+    },
+    {
+      "epoch": 0.49064906490649063,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.706833210874357e-05,
+      "loss": 1.6439,
+      "step": 446
+    },
+    {
+      "epoch": 0.49174917491749176,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.703894195444527e-05,
+      "loss": 1.5515,
+      "step": 447
+    },
+    {
+      "epoch": 0.49284928492849284,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.700955180014695e-05,
+      "loss": 1.9121,
+      "step": 448
+    },
+    {
+      "epoch": 0.49394939493949397,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.698016164584864e-05,
+      "loss": 1.442,
+      "step": 449
+    },
+    {
+      "epoch": 0.49504950495049505,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.695077149155033e-05,
+      "loss": 2.0113,
+      "step": 450
+    },
+    {
+      "epoch": 0.4961496149614962,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.692138133725202e-05,
+      "loss": 1.4442,
+      "step": 451
+    },
+    {
+      "epoch": 0.49724972497249725,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.689199118295371e-05,
+      "loss": 1.552,
+      "step": 452
+    },
+    {
+      "epoch": 0.49834983498349833,
+      "grad_norm": 2.25,
+      "learning_rate": 6.68626010286554e-05,
+      "loss": 1.3807,
+      "step": 453
+    },
+    {
+      "epoch": 0.49944994499449946,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.68332108743571e-05,
+      "loss": 1.5911,
+      "step": 454
+    },
+    {
+      "epoch": 0.5005500550055005,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.680382072005879e-05,
+      "loss": 1.6458,
+      "step": 455
+    },
+    {
+      "epoch": 0.5016501650165016,
+      "grad_norm": 1.9296875,
+      "learning_rate": 6.677443056576048e-05,
+      "loss": 1.4319,
+      "step": 456
+    },
+    {
+      "epoch": 0.5027502750275028,
+      "grad_norm": 2.125,
+      "learning_rate": 6.674504041146217e-05,
+      "loss": 1.5848,
+      "step": 457
+    },
+    {
+      "epoch": 0.5038503850385039,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.671565025716386e-05,
+      "loss": 1.6517,
+      "step": 458
+    },
+    {
+      "epoch": 0.504950495049505,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.668626010286555e-05,
+      "loss": 1.6322,
+      "step": 459
+    },
+    {
+      "epoch": 0.506050605060506,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.665686994856724e-05,
+      "loss": 1.3613,
+      "step": 460
+    },
+    {
+      "epoch": 0.5071507150715071,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.662747979426893e-05,
+      "loss": 1.8262,
+      "step": 461
+    },
+    {
+      "epoch": 0.5082508250825083,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.659808963997062e-05,
+      "loss": 1.4084,
+      "step": 462
+    },
+    {
+      "epoch": 0.5093509350935094,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.65686994856723e-05,
+      "loss": 1.4215,
+      "step": 463
+    },
+    {
+      "epoch": 0.5104510451045104,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.6539309331374e-05,
+      "loss": 1.4242,
+      "step": 464
+    },
+    {
+      "epoch": 0.5115511551155115,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.650991917707569e-05,
+      "loss": 1.8834,
+      "step": 465
+    },
+    {
+      "epoch": 0.5126512651265126,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.648052902277737e-05,
+      "loss": 1.5671,
+      "step": 466
+    },
+    {
+      "epoch": 0.5137513751375138,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.645113886847906e-05,
+      "loss": 1.4935,
+      "step": 467
+    },
+    {
+      "epoch": 0.5148514851485149,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.642174871418076e-05,
+      "loss": 1.6638,
+      "step": 468
+    },
+    {
+      "epoch": 0.5159515951595159,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.639235855988245e-05,
+      "loss": 1.8999,
+      "step": 469
+    },
+    {
+      "epoch": 0.517051705170517,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.636296840558413e-05,
+      "loss": 1.4005,
+      "step": 470
+    },
+    {
+      "epoch": 0.5181518151815182,
+      "grad_norm": 2.5,
+      "learning_rate": 6.633357825128582e-05,
+      "loss": 1.7815,
+      "step": 471
+    },
+    {
+      "epoch": 0.5192519251925193,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.630418809698752e-05,
+      "loss": 1.5491,
+      "step": 472
+    },
+    {
+      "epoch": 0.5203520352035204,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.62747979426892e-05,
+      "loss": 1.61,
+      "step": 473
+    },
+    {
+      "epoch": 0.5214521452145214,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.624540778839089e-05,
+      "loss": 1.6878,
+      "step": 474
+    },
+    {
+      "epoch": 0.5225522552255225,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.621601763409258e-05,
+      "loss": 1.5463,
+      "step": 475
+    },
+    {
+      "epoch": 0.5236523652365237,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.618662747979427e-05,
+      "loss": 1.457,
+      "step": 476
+    },
+    {
+      "epoch": 0.5247524752475248,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.615723732549596e-05,
+      "loss": 1.5677,
+      "step": 477
+    },
+    {
+      "epoch": 0.5258525852585259,
+      "grad_norm": 2.25,
+      "learning_rate": 6.612784717119765e-05,
+      "loss": 1.4645,
+      "step": 478
+    },
+    {
+      "epoch": 0.5269526952695269,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.609845701689934e-05,
+      "loss": 1.4323,
+      "step": 479
+    },
+    {
+      "epoch": 0.528052805280528,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.606906686260103e-05,
+      "loss": 1.5941,
+      "step": 480
+    },
+    {
+      "epoch": 0.5291529152915292,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.603967670830272e-05,
+      "loss": 1.8619,
+      "step": 481
+    },
+    {
+      "epoch": 0.5302530253025303,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.601028655400442e-05,
+      "loss": 1.5176,
+      "step": 482
+    },
+    {
+      "epoch": 0.5313531353135313,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.59808963997061e-05,
+      "loss": 1.714,
+      "step": 483
+    },
+    {
+      "epoch": 0.5324532453245324,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.59515062454078e-05,
+      "loss": 1.3819,
+      "step": 484
+    },
+    {
+      "epoch": 0.5335533553355336,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.592211609110949e-05,
+      "loss": 1.651,
+      "step": 485
+    },
+    {
+      "epoch": 0.5346534653465347,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.589272593681118e-05,
+      "loss": 1.5377,
+      "step": 486
+    },
+    {
+      "epoch": 0.5357535753575358,
+      "grad_norm": 1.890625,
+      "learning_rate": 6.586333578251287e-05,
+      "loss": 1.6605,
+      "step": 487
+    },
+    {
+      "epoch": 0.5368536853685368,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.583394562821454e-05,
+      "loss": 1.5816,
+      "step": 488
+    },
+    {
+      "epoch": 0.5379537953795379,
+      "grad_norm": 2.125,
+      "learning_rate": 6.580455547391625e-05,
+      "loss": 1.6929,
+      "step": 489
+    },
+    {
+      "epoch": 0.5390539053905391,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.577516531961794e-05,
+      "loss": 1.7521,
+      "step": 490
+    },
+    {
+      "epoch": 0.5401540154015402,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.574577516531962e-05,
+      "loss": 1.487,
+      "step": 491
+    },
+    {
+      "epoch": 0.5412541254125413,
+      "grad_norm": 2.75,
+      "learning_rate": 6.571638501102131e-05,
+      "loss": 1.7397,
+      "step": 492
+    },
+    {
+      "epoch": 0.5423542354235423,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.568699485672301e-05,
+      "loss": 1.1897,
+      "step": 493
+    },
+    {
+      "epoch": 0.5434543454345434,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.56576047024247e-05,
+      "loss": 1.6536,
+      "step": 494
+    },
+    {
+      "epoch": 0.5445544554455446,
+      "grad_norm": 2.25,
+      "learning_rate": 6.562821454812638e-05,
+      "loss": 1.4894,
+      "step": 495
+    },
+    {
+      "epoch": 0.5456545654565457,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.559882439382807e-05,
+      "loss": 1.6685,
+      "step": 496
+    },
+    {
+      "epoch": 0.5467546754675467,
+      "grad_norm": 2.375,
+      "learning_rate": 6.556943423952976e-05,
+      "loss": 1.5364,
+      "step": 497
+    },
+    {
+      "epoch": 0.5478547854785478,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.554004408523145e-05,
+      "loss": 1.2747,
+      "step": 498
+    },
+    {
+      "epoch": 0.5489548954895489,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.551065393093314e-05,
+      "loss": 1.9298,
+      "step": 499
+    },
+    {
+      "epoch": 0.5500550055005501,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.548126377663483e-05,
+      "loss": 1.523,
+      "step": 500
+    },
+    {
+      "epoch": 0.5511551155115512,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.545187362233652e-05,
+      "loss": 1.6715,
+      "step": 501
+    },
+    {
+      "epoch": 0.5522552255225522,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.542248346803821e-05,
+      "loss": 1.422,
+      "step": 502
+    },
+    {
+      "epoch": 0.5533553355335533,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.53930933137399e-05,
+      "loss": 1.6306,
+      "step": 503
+    },
+    {
+      "epoch": 0.5544554455445545,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.536370315944159e-05,
+      "loss": 1.6758,
+      "step": 504
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.533431300514328e-05,
+      "loss": 1.5846,
+      "step": 505
+    },
+    {
+      "epoch": 0.5566556655665567,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.530492285084497e-05,
+      "loss": 1.7384,
+      "step": 506
+    },
+    {
+      "epoch": 0.5577557755775577,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.527553269654666e-05,
+      "loss": 1.6483,
+      "step": 507
+    },
+    {
+      "epoch": 0.5588558855885588,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.524614254224835e-05,
+      "loss": 1.3783,
+      "step": 508
+    },
+    {
+      "epoch": 0.55995599559956,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.521675238795005e-05,
+      "loss": 1.4983,
+      "step": 509
+    },
+    {
+      "epoch": 0.5610561056105611,
+      "grad_norm": 2.25,
+      "learning_rate": 6.518736223365174e-05,
+      "loss": 1.5189,
+      "step": 510
+    },
+    {
+      "epoch": 0.5621562156215621,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.515797207935343e-05,
+      "loss": 1.449,
+      "step": 511
+    },
+    {
+      "epoch": 0.5632563256325632,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.512858192505512e-05,
+      "loss": 1.6381,
+      "step": 512
+    },
+    {
+      "epoch": 0.5643564356435643,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.50991917707568e-05,
+      "loss": 1.35,
+      "step": 513
+    },
+    {
+      "epoch": 0.5654565456545655,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.506980161645848e-05,
+      "loss": 1.5687,
+      "step": 514
+    },
+    {
+      "epoch": 0.5665566556655666,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.504041146216019e-05,
+      "loss": 1.551,
+      "step": 515
+    },
+    {
+      "epoch": 0.5676567656765676,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.501102130786188e-05,
+      "loss": 1.4412,
+      "step": 516
+    },
+    {
+      "epoch": 0.5687568756875687,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.498163115356356e-05,
+      "loss": 1.5595,
+      "step": 517
+    },
+    {
+      "epoch": 0.5698569856985699,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.495224099926525e-05,
+      "loss": 1.8032,
+      "step": 518
+    },
+    {
+      "epoch": 0.570957095709571,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.492285084496695e-05,
+      "loss": 1.5501,
+      "step": 519
+    },
+    {
+      "epoch": 0.5720572057205721,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.489346069066863e-05,
+      "loss": 1.4089,
+      "step": 520
+    },
+    {
+      "epoch": 0.5731573157315731,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.486407053637032e-05,
+      "loss": 1.7413,
+      "step": 521
+    },
+    {
+      "epoch": 0.5742574257425742,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.483468038207201e-05,
+      "loss": 1.5463,
+      "step": 522
+    },
+    {
+      "epoch": 0.5753575357535754,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.48052902277737e-05,
+      "loss": 1.9864,
+      "step": 523
+    },
+    {
+      "epoch": 0.5764576457645765,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.477590007347539e-05,
+      "loss": 1.566,
+      "step": 524
+    },
+    {
+      "epoch": 0.5775577557755776,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.474650991917708e-05,
+      "loss": 1.8475,
+      "step": 525
+    },
+    {
+      "epoch": 0.5786578657865786,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.471711976487877e-05,
+      "loss": 1.7088,
+      "step": 526
+    },
+    {
+      "epoch": 0.5797579757975797,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.468772961058046e-05,
+      "loss": 1.518,
+      "step": 527
+    },
+    {
+      "epoch": 0.5808580858085809,
+      "grad_norm": 2.125,
+      "learning_rate": 6.465833945628215e-05,
+      "loss": 1.5894,
+      "step": 528
+    },
+    {
+      "epoch": 0.581958195819582,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.462894930198384e-05,
+      "loss": 1.6634,
+      "step": 529
+    },
+    {
+      "epoch": 0.583058305830583,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.459955914768553e-05,
+      "loss": 1.5625,
+      "step": 530
+    },
+    {
+      "epoch": 0.5841584158415841,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.457016899338722e-05,
+      "loss": 1.6221,
+      "step": 531
+    },
+    {
+      "epoch": 0.5852585258525853,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.454077883908891e-05,
+      "loss": 1.5927,
+      "step": 532
+    },
+    {
+      "epoch": 0.5863586358635864,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.45113886847906e-05,
+      "loss": 1.5614,
+      "step": 533
+    },
+    {
+      "epoch": 0.5874587458745875,
+      "grad_norm": 2.25,
+      "learning_rate": 6.44819985304923e-05,
+      "loss": 1.4454,
+      "step": 534
+    },
+    {
+      "epoch": 0.5885588558855885,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.445260837619397e-05,
+      "loss": 1.7025,
+      "step": 535
+    },
+    {
+      "epoch": 0.5896589658965896,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.442321822189568e-05,
+      "loss": 1.649,
+      "step": 536
+    },
+    {
+      "epoch": 0.5907590759075908,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.439382806759737e-05,
+      "loss": 1.6797,
+      "step": 537
+    },
+    {
+      "epoch": 0.5918591859185919,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.436443791329904e-05,
+      "loss": 1.7956,
+      "step": 538
+    },
+    {
+      "epoch": 0.592959295929593,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.433504775900073e-05,
+      "loss": 1.9168,
+      "step": 539
+    },
+    {
+      "epoch": 0.594059405940594,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.430565760470244e-05,
+      "loss": 1.5967,
+      "step": 540
+    },
+    {
+      "epoch": 0.5951595159515951,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.427626745040413e-05,
+      "loss": 1.8711,
+      "step": 541
+    },
+    {
+      "epoch": 0.5962596259625963,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.42468772961058e-05,
+      "loss": 1.5065,
+      "step": 542
+    },
+    {
+      "epoch": 0.5973597359735974,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.42174871418075e-05,
+      "loss": 1.4933,
+      "step": 543
+    },
+    {
+      "epoch": 0.5984598459845984,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.418809698750919e-05,
+      "loss": 1.753,
+      "step": 544
+    },
+    {
+      "epoch": 0.5995599559955995,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.415870683321088e-05,
+      "loss": 1.4709,
+      "step": 545
+    },
+    {
+      "epoch": 0.6006600660066007,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.412931667891257e-05,
+      "loss": 1.5334,
+      "step": 546
+    },
+    {
+      "epoch": 0.6017601760176018,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.409992652461426e-05,
+      "loss": 1.7762,
+      "step": 547
+    },
+    {
+      "epoch": 0.6028602860286029,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.407053637031595e-05,
+      "loss": 1.4819,
+      "step": 548
+    },
+    {
+      "epoch": 0.6039603960396039,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.404114621601764e-05,
+      "loss": 1.3427,
+      "step": 549
+    },
+    {
+      "epoch": 0.605060506050605,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.401175606171933e-05,
+      "loss": 1.7221,
+      "step": 550
+    },
+    {
+      "epoch": 0.6061606160616062,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.398236590742102e-05,
+      "loss": 1.7561,
+      "step": 551
+    },
+    {
+      "epoch": 0.6072607260726073,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.395297575312271e-05,
+      "loss": 1.5356,
+      "step": 552
+    },
+    {
+      "epoch": 0.6083608360836084,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.39235855988244e-05,
+      "loss": 1.7293,
+      "step": 553
+    },
+    {
+      "epoch": 0.6094609460946094,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.389419544452609e-05,
+      "loss": 1.5879,
+      "step": 554
+    },
+    {
+      "epoch": 0.6105610561056105,
+      "grad_norm": 2.125,
+      "learning_rate": 6.386480529022778e-05,
+      "loss": 1.4298,
+      "step": 555
+    },
+    {
+      "epoch": 0.6116611661166117,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.383541513592947e-05,
+      "loss": 1.7145,
+      "step": 556
+    },
+    {
+      "epoch": 0.6127612761276128,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.380602498163116e-05,
+      "loss": 1.5959,
+      "step": 557
+    },
+    {
+      "epoch": 0.6138613861386139,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.377663482733285e-05,
+      "loss": 1.6782,
+      "step": 558
+    },
+    {
+      "epoch": 0.6149614961496149,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.374724467303454e-05,
+      "loss": 1.315,
+      "step": 559
+    },
+    {
+      "epoch": 0.6160616061606161,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.371785451873622e-05,
+      "loss": 1.6165,
+      "step": 560
+    },
+    {
+      "epoch": 0.6171617161716172,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.368846436443791e-05,
+      "loss": 1.5327,
+      "step": 561
+    },
+    {
+      "epoch": 0.6182618261826183,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.365907421013962e-05,
+      "loss": 1.6079,
+      "step": 562
+    },
+    {
+      "epoch": 0.6193619361936193,
+      "grad_norm": 2.375,
+      "learning_rate": 6.362968405584129e-05,
+      "loss": 1.5346,
+      "step": 563
+    },
+    {
+      "epoch": 0.6204620462046204,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.360029390154298e-05,
+      "loss": 1.5389,
+      "step": 564
+    },
+    {
+      "epoch": 0.6215621562156216,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.357090374724467e-05,
+      "loss": 1.5751,
+      "step": 565
+    },
+    {
+      "epoch": 0.6226622662266227,
+      "grad_norm": 2.375,
+      "learning_rate": 6.354151359294638e-05,
+      "loss": 1.5874,
+      "step": 566
+    },
+    {
+      "epoch": 0.6237623762376238,
+      "grad_norm": 2.375,
+      "learning_rate": 6.351212343864805e-05,
+      "loss": 1.5609,
+      "step": 567
+    },
+    {
+      "epoch": 0.6248624862486248,
+      "grad_norm": 2.5,
+      "learning_rate": 6.348273328434974e-05,
+      "loss": 1.5016,
+      "step": 568
+    },
+    {
+      "epoch": 0.6259625962596259,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.345334313005144e-05,
+      "loss": 1.6998,
+      "step": 569
+    },
+    {
+      "epoch": 0.6270627062706271,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.342395297575313e-05,
+      "loss": 1.6603,
+      "step": 570
+    },
+    {
+      "epoch": 0.6281628162816282,
+      "grad_norm": 2.125,
+      "learning_rate": 6.339456282145482e-05,
+      "loss": 1.6487,
+      "step": 571
+    },
+    {
+      "epoch": 0.6292629262926293,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.33651726671565e-05,
+      "loss": 1.3722,
+      "step": 572
+    },
+    {
+      "epoch": 0.6303630363036303,
+      "grad_norm": 2.25,
+      "learning_rate": 6.33357825128582e-05,
+      "loss": 1.5305,
+      "step": 573
+    },
+    {
+      "epoch": 0.6314631463146315,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.330639235855989e-05,
+      "loss": 1.576,
+      "step": 574
+    },
+    {
+      "epoch": 0.6325632563256326,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.327700220426158e-05,
+      "loss": 1.5673,
+      "step": 575
+    },
+    {
+      "epoch": 0.6336633663366337,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.324761204996327e-05,
+      "loss": 1.5882,
+      "step": 576
+    },
+    {
+      "epoch": 0.6347634763476347,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.321822189566496e-05,
+      "loss": 1.7532,
+      "step": 577
+    },
+    {
+      "epoch": 0.6358635863586358,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.318883174136665e-05,
+      "loss": 1.7783,
+      "step": 578
+    },
+    {
+      "epoch": 0.636963696369637,
+      "grad_norm": 2.25,
+      "learning_rate": 6.315944158706834e-05,
+      "loss": 1.6383,
+      "step": 579
+    },
+    {
+      "epoch": 0.6380638063806381,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.313005143277003e-05,
+      "loss": 1.4386,
+      "step": 580
+    },
+    {
+      "epoch": 0.6391639163916392,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.310066127847172e-05,
+      "loss": 1.5814,
+      "step": 581
+    },
+    {
+      "epoch": 0.6402640264026402,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.30712711241734e-05,
+      "loss": 1.6569,
+      "step": 582
+    },
+    {
+      "epoch": 0.6413641364136413,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.30418809698751e-05,
+      "loss": 1.7379,
+      "step": 583
+    },
+    {
+      "epoch": 0.6424642464246425,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.301249081557679e-05,
+      "loss": 1.5544,
+      "step": 584
+    },
+    {
+      "epoch": 0.6435643564356436,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.298310066127847e-05,
+      "loss": 1.5163,
+      "step": 585
+    },
+    {
+      "epoch": 0.6446644664466447,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.295371050698016e-05,
+      "loss": 1.5387,
+      "step": 586
+    },
+    {
+      "epoch": 0.6457645764576457,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.292432035268186e-05,
+      "loss": 1.5613,
+      "step": 587
+    },
+    {
+      "epoch": 0.6468646864686468,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.289493019838355e-05,
+      "loss": 1.5041,
+      "step": 588
+    },
+    {
+      "epoch": 0.647964796479648,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.286554004408523e-05,
+      "loss": 1.627,
+      "step": 589
+    },
+    {
+      "epoch": 0.6490649064906491,
+      "grad_norm": 2.25,
+      "learning_rate": 6.283614988978692e-05,
+      "loss": 1.4932,
+      "step": 590
+    },
+    {
+      "epoch": 0.6501650165016502,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.280675973548863e-05,
+      "loss": 1.5794,
+      "step": 591
+    },
+    {
+      "epoch": 0.6512651265126512,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.27773695811903e-05,
+      "loss": 1.9137,
+      "step": 592
+    },
+    {
+      "epoch": 0.6523652365236524,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.2747979426892e-05,
+      "loss": 1.5973,
+      "step": 593
+    },
+    {
+      "epoch": 0.6534653465346535,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.271858927259368e-05,
+      "loss": 1.6684,
+      "step": 594
+    },
+    {
+      "epoch": 0.6545654565456546,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.268919911829537e-05,
+      "loss": 1.6799,
+      "step": 595
+    },
+    {
+      "epoch": 0.6556655665566556,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.265980896399707e-05,
+      "loss": 1.5359,
+      "step": 596
+    },
+    {
+      "epoch": 0.6567656765676567,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.263041880969876e-05,
+      "loss": 1.558,
+      "step": 597
+    },
+    {
+      "epoch": 0.6578657865786579,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.260102865540045e-05,
+      "loss": 1.8665,
+      "step": 598
+    },
+    {
+      "epoch": 0.658965896589659,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.257163850110214e-05,
+      "loss": 1.359,
+      "step": 599
+    },
+    {
+      "epoch": 0.6600660066006601,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.254224834680383e-05,
+      "loss": 1.4225,
+      "step": 600
+    },
+    {
+      "epoch": 0.6611661166116611,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.251285819250552e-05,
+      "loss": 1.6186,
+      "step": 601
+    },
+    {
+      "epoch": 0.6622662266226622,
+      "grad_norm": 2.125,
+      "learning_rate": 6.248346803820721e-05,
+      "loss": 1.6425,
+      "step": 602
+    },
+    {
+      "epoch": 0.6633663366336634,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.24540778839089e-05,
+      "loss": 1.6694,
+      "step": 603
+    },
+    {
+      "epoch": 0.6644664466446645,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.242468772961059e-05,
+      "loss": 1.3447,
+      "step": 604
+    },
+    {
+      "epoch": 0.6655665566556656,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.239529757531228e-05,
+      "loss": 1.6121,
+      "step": 605
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.236590742101397e-05,
+      "loss": 1.6802,
+      "step": 606
+    },
+    {
+      "epoch": 0.6677667766776678,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.233651726671565e-05,
+      "loss": 1.6062,
+      "step": 607
+    },
+    {
+      "epoch": 0.6688668866886689,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.230712711241735e-05,
+      "loss": 1.6884,
+      "step": 608
+    },
+    {
+      "epoch": 0.66996699669967,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.227773695811904e-05,
+      "loss": 1.5522,
+      "step": 609
+    },
+    {
+      "epoch": 0.671067106710671,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.224834680382072e-05,
+      "loss": 1.7008,
+      "step": 610
+    },
+    {
+      "epoch": 0.6721672167216721,
+      "grad_norm": 2.125,
+      "learning_rate": 6.221895664952241e-05,
+      "loss": 1.4858,
+      "step": 611
+    },
+    {
+      "epoch": 0.6732673267326733,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.21895664952241e-05,
+      "loss": 1.7735,
+      "step": 612
+    },
+    {
+      "epoch": 0.6743674367436744,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.21601763409258e-05,
+      "loss": 1.6625,
+      "step": 613
+    },
+    {
+      "epoch": 0.6754675467546755,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.213078618662748e-05,
+      "loss": 1.6534,
+      "step": 614
+    },
+    {
+      "epoch": 0.6765676567656765,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.210139603232917e-05,
+      "loss": 1.5599,
+      "step": 615
+    },
+    {
+      "epoch": 0.6776677667766776,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.207200587803086e-05,
+      "loss": 1.4886,
+      "step": 616
+    },
+    {
+      "epoch": 0.6787678767876788,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.204261572373255e-05,
+      "loss": 1.6381,
+      "step": 617
+    },
+    {
+      "epoch": 0.6798679867986799,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.201322556943424e-05,
+      "loss": 1.4443,
+      "step": 618
+    },
+    {
+      "epoch": 0.680968096809681,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.198383541513593e-05,
+      "loss": 1.6981,
+      "step": 619
+    },
+    {
+      "epoch": 0.682068206820682,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.195444526083762e-05,
+      "loss": 1.6879,
+      "step": 620
+    },
+    {
+      "epoch": 0.6831683168316832,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.192505510653931e-05,
+      "loss": 1.7068,
+      "step": 621
+    },
+    {
+      "epoch": 0.6842684268426843,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.1895664952241e-05,
+      "loss": 1.6888,
+      "step": 622
+    },
+    {
+      "epoch": 0.6853685368536854,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.18662747979427e-05,
+      "loss": 1.6878,
+      "step": 623
+    },
+    {
+      "epoch": 0.6864686468646864,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.183688464364439e-05,
+      "loss": 1.6992,
+      "step": 624
+    },
+    {
+      "epoch": 0.6875687568756875,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.180749448934608e-05,
+      "loss": 1.3723,
+      "step": 625
+    },
+    {
+      "epoch": 0.6886688668866887,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.177810433504777e-05,
+      "loss": 1.4349,
+      "step": 626
+    },
+    {
+      "epoch": 0.6897689768976898,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.174871418074946e-05,
+      "loss": 1.4981,
+      "step": 627
+    },
+    {
+      "epoch": 0.6908690869086909,
+      "grad_norm": 2.25,
+      "learning_rate": 6.171932402645115e-05,
+      "loss": 1.408,
+      "step": 628
+    },
+    {
+      "epoch": 0.6919691969196919,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.168993387215282e-05,
+      "loss": 1.7976,
+      "step": 629
+    },
+    {
+      "epoch": 0.693069306930693,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.166054371785453e-05,
+      "loss": 1.846,
+      "step": 630
+    },
+    {
+      "epoch": 0.6941694169416942,
+      "grad_norm": 2.25,
+      "learning_rate": 6.163115356355622e-05,
+      "loss": 1.2452,
+      "step": 631
+    },
+    {
+      "epoch": 0.6952695269526953,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.16017634092579e-05,
+      "loss": 1.5504,
+      "step": 632
+    },
+    {
+      "epoch": 0.6963696369636964,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.157237325495959e-05,
+      "loss": 1.572,
+      "step": 633
+    },
+    {
+      "epoch": 0.6974697469746974,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.154298310066129e-05,
+      "loss": 1.2571,
+      "step": 634
+    },
+    {
+      "epoch": 0.6985698569856986,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.151359294636297e-05,
+      "loss": 1.4102,
+      "step": 635
+    },
+    {
+      "epoch": 0.6996699669966997,
+      "grad_norm": 2.375,
+      "learning_rate": 6.148420279206466e-05,
+      "loss": 1.5268,
+      "step": 636
+    },
+    {
+      "epoch": 0.7007700770077008,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.145481263776635e-05,
+      "loss": 1.5699,
+      "step": 637
+    },
+    {
+      "epoch": 0.7018701870187019,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.142542248346805e-05,
+      "loss": 1.6892,
+      "step": 638
+    },
+    {
+      "epoch": 0.7029702970297029,
+      "grad_norm": 2.375,
+      "learning_rate": 6.139603232916973e-05,
+      "loss": 1.4365,
+      "step": 639
+    },
+    {
+      "epoch": 0.7040704070407041,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.136664217487142e-05,
+      "loss": 1.4789,
+      "step": 640
+    },
+    {
+      "epoch": 0.7051705170517052,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.133725202057311e-05,
+      "loss": 1.5648,
+      "step": 641
+    },
+    {
+      "epoch": 0.7062706270627063,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.13078618662748e-05,
+      "loss": 1.6063,
+      "step": 642
+    },
+    {
+      "epoch": 0.7073707370737073,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.127847171197649e-05,
+      "loss": 1.5852,
+      "step": 643
+    },
+    {
+      "epoch": 0.7084708470847084,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.124908155767818e-05,
+      "loss": 1.7057,
+      "step": 644
+    },
+    {
+      "epoch": 0.7095709570957096,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.121969140337987e-05,
+      "loss": 1.6184,
+      "step": 645
+    },
+    {
+      "epoch": 0.7106710671067107,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.119030124908156e-05,
+      "loss": 1.5353,
+      "step": 646
+    },
+    {
+      "epoch": 0.7117711771177118,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.116091109478325e-05,
+      "loss": 1.3631,
+      "step": 647
+    },
+    {
+      "epoch": 0.7128712871287128,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.113152094048494e-05,
+      "loss": 1.4375,
+      "step": 648
+    },
+    {
+      "epoch": 0.713971397139714,
+      "grad_norm": 2.125,
+      "learning_rate": 6.110213078618663e-05,
+      "loss": 1.6513,
+      "step": 649
+    },
+    {
+      "epoch": 0.7150715071507151,
+      "grad_norm": 2.25,
+      "learning_rate": 6.107274063188833e-05,
+      "loss": 1.3335,
+      "step": 650
+    },
+    {
+      "epoch": 0.7161716171617162,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.104335047759002e-05,
+      "loss": 1.4353,
+      "step": 651
+    },
+    {
+      "epoch": 0.7172717271727173,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.1013960323291706e-05,
+      "loss": 1.5369,
+      "step": 652
+    },
+    {
+      "epoch": 0.7183718371837183,
+      "grad_norm": 2.375,
+      "learning_rate": 6.098457016899339e-05,
+      "loss": 1.565,
+      "step": 653
+    },
+    {
+      "epoch": 0.7194719471947195,
+      "grad_norm": 1.9375,
+      "learning_rate": 6.095518001469508e-05,
+      "loss": 1.6488,
+      "step": 654
+    },
+    {
+      "epoch": 0.7205720572057206,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.092578986039678e-05,
+      "loss": 1.8179,
+      "step": 655
+    },
+    {
+      "epoch": 0.7216721672167217,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.089639970609846e-05,
+      "loss": 1.3994,
+      "step": 656
+    },
+    {
+      "epoch": 0.7227722772277227,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.086700955180015e-05,
+      "loss": 1.6511,
+      "step": 657
+    },
+    {
+      "epoch": 0.7238723872387238,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.083761939750184e-05,
+      "loss": 1.479,
+      "step": 658
+    },
+    {
+      "epoch": 0.724972497249725,
+      "grad_norm": 1.9765625,
+      "learning_rate": 6.080822924320353e-05,
+      "loss": 1.3285,
+      "step": 659
+    },
+    {
+      "epoch": 0.7260726072607261,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.0778839088905223e-05,
+      "loss": 1.5376,
+      "step": 660
+    },
+    {
+      "epoch": 0.7271727172717272,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.0749448934606914e-05,
+      "loss": 1.4111,
+      "step": 661
+    },
+    {
+      "epoch": 0.7282728272827282,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.07200587803086e-05,
+      "loss": 1.7049,
+      "step": 662
+    },
+    {
+      "epoch": 0.7293729372937293,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.069066862601029e-05,
+      "loss": 1.7659,
+      "step": 663
+    },
+    {
+      "epoch": 0.7304730473047305,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.0661278471711986e-05,
+      "loss": 1.6799,
+      "step": 664
+    },
+    {
+      "epoch": 0.7315731573157316,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.063188831741367e-05,
+      "loss": 1.6455,
+      "step": 665
+    },
+    {
+      "epoch": 0.7326732673267327,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.060249816311536e-05,
+      "loss": 1.6416,
+      "step": 666
+    },
+    {
+      "epoch": 0.7337733773377337,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.057310800881705e-05,
+      "loss": 1.3102,
+      "step": 667
+    },
+    {
+      "epoch": 0.7348734873487349,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.054371785451875e-05,
+      "loss": 1.6933,
+      "step": 668
+    },
+    {
+      "epoch": 0.735973597359736,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.051432770022043e-05,
+      "loss": 1.4144,
+      "step": 669
+    },
+    {
+      "epoch": 0.7370737073707371,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.048493754592212e-05,
+      "loss": 1.6709,
+      "step": 670
+    },
+    {
+      "epoch": 0.7381738173817382,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.0455547391623805e-05,
+      "loss": 1.5223,
+      "step": 671
+    },
+    {
+      "epoch": 0.7392739273927392,
+      "grad_norm": 1.9140625,
+      "learning_rate": 6.04261572373255e-05,
+      "loss": 1.6362,
+      "step": 672
+    },
+    {
+      "epoch": 0.7403740374037404,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.039676708302719e-05,
+      "loss": 1.5531,
+      "step": 673
+    },
+    {
+      "epoch": 0.7414741474147415,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.036737692872888e-05,
+      "loss": 1.4168,
+      "step": 674
+    },
+    {
+      "epoch": 0.7425742574257426,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.033798677443057e-05,
+      "loss": 1.5238,
+      "step": 675
+    },
+    {
+      "epoch": 0.7436743674367436,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.0308596620132265e-05,
+      "loss": 1.8466,
+      "step": 676
+    },
+    {
+      "epoch": 0.7447744774477447,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.0279206465833955e-05,
+      "loss": 1.6457,
+      "step": 677
+    },
+    {
+      "epoch": 0.7458745874587459,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.024981631153564e-05,
+      "loss": 1.7251,
+      "step": 678
+    },
+    {
+      "epoch": 0.746974697469747,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.022042615723733e-05,
+      "loss": 1.7786,
+      "step": 679
+    },
+    {
+      "epoch": 0.7480748074807481,
+      "grad_norm": 2.25,
+      "learning_rate": 6.019103600293901e-05,
+      "loss": 1.4986,
+      "step": 680
+    },
+    {
+      "epoch": 0.7491749174917491,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.016164584864071e-05,
+      "loss": 1.6529,
+      "step": 681
+    },
+    {
+      "epoch": 0.7502750275027503,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.01322556943424e-05,
+      "loss": 1.6746,
+      "step": 682
+    },
+    {
+      "epoch": 0.7513751375137514,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.010286554004409e-05,
+      "loss": 1.4736,
+      "step": 683
+    },
+    {
+      "epoch": 0.7524752475247525,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.0073475385745775e-05,
+      "loss": 1.6419,
+      "step": 684
+    },
+    {
+      "epoch": 0.7535753575357536,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.004408523144747e-05,
+      "loss": 1.6457,
+      "step": 685
+    },
+    {
+      "epoch": 0.7546754675467546,
+      "grad_norm": 2.25,
+      "learning_rate": 6.001469507714916e-05,
+      "loss": 1.5347,
+      "step": 686
+    },
+    {
+      "epoch": 0.7557755775577558,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.998530492285085e-05,
+      "loss": 1.6255,
+      "step": 687
+    },
+    {
+      "epoch": 0.7568756875687569,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.995591476855254e-05,
+      "loss": 1.6088,
+      "step": 688
+    },
+    {
+      "epoch": 0.757975797579758,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.9926524614254235e-05,
+      "loss": 1.5574,
+      "step": 689
+    },
+    {
+      "epoch": 0.759075907590759,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.989713445995592e-05,
+      "loss": 1.6363,
+      "step": 690
+    },
+    {
+      "epoch": 0.7601760176017601,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.986774430565761e-05,
+      "loss": 1.3214,
+      "step": 691
+    },
+    {
+      "epoch": 0.7612761276127613,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.98383541513593e-05,
+      "loss": 1.5883,
+      "step": 692
+    },
+    {
+      "epoch": 0.7623762376237624,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.9808963997060997e-05,
+      "loss": 1.6644,
+      "step": 693
+    },
+    {
+      "epoch": 0.7634763476347635,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.977957384276268e-05,
+      "loss": 1.7116,
+      "step": 694
+    },
+    {
+      "epoch": 0.7645764576457645,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.975018368846437e-05,
+      "loss": 1.7597,
+      "step": 695
+    },
+    {
+      "epoch": 0.7656765676567657,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.9720793534166055e-05,
+      "loss": 1.7127,
+      "step": 696
+    },
+    {
+      "epoch": 0.7667766776677668,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.9691403379867745e-05,
+      "loss": 1.5249,
+      "step": 697
+    },
+    {
+      "epoch": 0.7678767876787679,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.966201322556944e-05,
+      "loss": 1.6108,
+      "step": 698
+    },
+    {
+      "epoch": 0.768976897689769,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.963262307127113e-05,
+      "loss": 1.6445,
+      "step": 699
+    },
+    {
+      "epoch": 0.77007700770077,
+      "grad_norm": 2.5,
+      "learning_rate": 5.9603232916972817e-05,
+      "loss": 1.4276,
+      "step": 700
+    },
+    {
+      "epoch": 0.7711771177117712,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.957384276267451e-05,
+      "loss": 1.2891,
+      "step": 701
+    },
+    {
+      "epoch": 0.7722772277227723,
+      "grad_norm": 5.34375,
+      "learning_rate": 5.9544452608376204e-05,
+      "loss": 1.423,
+      "step": 702
+    },
+    {
+      "epoch": 0.7733773377337734,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.951506245407789e-05,
+      "loss": 1.3971,
+      "step": 703
+    },
+    {
+      "epoch": 0.7744774477447744,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.948567229977958e-05,
+      "loss": 1.4835,
+      "step": 704
+    },
+    {
+      "epoch": 0.7755775577557755,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.945628214548126e-05,
+      "loss": 1.6282,
+      "step": 705
+    },
+    {
+      "epoch": 0.7766776677667767,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.942689199118296e-05,
+      "loss": 1.6325,
+      "step": 706
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 2.25,
+      "learning_rate": 5.939750183688465e-05,
+      "loss": 1.8523,
+      "step": 707
+    },
+    {
+      "epoch": 0.7788778877887789,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.936811168258634e-05,
+      "loss": 1.2488,
+      "step": 708
+    },
+    {
+      "epoch": 0.7799779977997799,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.9338721528288024e-05,
+      "loss": 1.3927,
+      "step": 709
+    },
+    {
+      "epoch": 0.7810781078107811,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.930933137398972e-05,
+      "loss": 1.7919,
+      "step": 710
+    },
+    {
+      "epoch": 0.7821782178217822,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.927994121969141e-05,
+      "loss": 1.4676,
+      "step": 711
+    },
+    {
+      "epoch": 0.7832783278327833,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.9250551065393096e-05,
+      "loss": 1.4834,
+      "step": 712
+    },
+    {
+      "epoch": 0.7843784378437844,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.9221160911094786e-05,
+      "loss": 1.4333,
+      "step": 713
+    },
+    {
+      "epoch": 0.7854785478547854,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.919177075679648e-05,
+      "loss": 1.5365,
+      "step": 714
+    },
+    {
+      "epoch": 0.7865786578657866,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.916238060249817e-05,
+      "loss": 1.6095,
+      "step": 715
+    },
+    {
+      "epoch": 0.7876787678767877,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.913299044819986e-05,
+      "loss": 1.5687,
+      "step": 716
+    },
+    {
+      "epoch": 0.7887788778877888,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.910360029390155e-05,
+      "loss": 1.6749,
+      "step": 717
+    },
+    {
+      "epoch": 0.7898789878987899,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.907421013960323e-05,
+      "loss": 1.6895,
+      "step": 718
+    },
+    {
+      "epoch": 0.7909790979097909,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.904481998530493e-05,
+      "loss": 1.7855,
+      "step": 719
+    },
+    {
+      "epoch": 0.7920792079207921,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.901542983100662e-05,
+      "loss": 1.3495,
+      "step": 720
+    },
+    {
+      "epoch": 0.7931793179317932,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.8986039676708304e-05,
+      "loss": 1.557,
+      "step": 721
+    },
+    {
+      "epoch": 0.7942794279427943,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.8956649522409994e-05,
+      "loss": 1.5275,
+      "step": 722
+    },
+    {
+      "epoch": 0.7953795379537953,
+      "grad_norm": 1.9609375,
+      "learning_rate": 5.892725936811169e-05,
+      "loss": 1.4305,
+      "step": 723
+    },
+    {
+      "epoch": 0.7964796479647965,
+      "grad_norm": 1.9765625,
+      "learning_rate": 5.889786921381338e-05,
+      "loss": 1.7755,
+      "step": 724
+    },
+    {
+      "epoch": 0.7975797579757976,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.8868479059515066e-05,
+      "loss": 1.5996,
+      "step": 725
+    },
+    {
+      "epoch": 0.7986798679867987,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.8839088905216756e-05,
+      "loss": 1.4994,
+      "step": 726
+    },
+    {
+      "epoch": 0.7997799779977998,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.880969875091845e-05,
+      "loss": 1.638,
+      "step": 727
+    },
+    {
+      "epoch": 0.8008800880088008,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.878030859662014e-05,
+      "loss": 1.4262,
+      "step": 728
+    },
+    {
+      "epoch": 0.801980198019802,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.875091844232183e-05,
+      "loss": 1.4648,
+      "step": 729
+    },
+    {
+      "epoch": 0.8030803080308031,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.872152828802352e-05,
+      "loss": 1.683,
+      "step": 730
+    },
+    {
+      "epoch": 0.8041804180418042,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.86921381337252e-05,
+      "loss": 1.4112,
+      "step": 731
+    },
+    {
+      "epoch": 0.8052805280528053,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.86627479794269e-05,
+      "loss": 1.5181,
+      "step": 732
+    },
+    {
+      "epoch": 0.8063806380638063,
+      "grad_norm": 2.25,
+      "learning_rate": 5.863335782512859e-05,
+      "loss": 1.5265,
+      "step": 733
+    },
+    {
+      "epoch": 0.8074807480748075,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.860396767083027e-05,
+      "loss": 1.777,
+      "step": 734
+    },
+    {
+      "epoch": 0.8085808580858086,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.8574577516531964e-05,
+      "loss": 1.4209,
+      "step": 735
+    },
+    {
+      "epoch": 0.8096809680968097,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.854518736223366e-05,
+      "loss": 1.5664,
+      "step": 736
+    },
+    {
+      "epoch": 0.8107810781078107,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.8515797207935345e-05,
+      "loss": 1.8181,
+      "step": 737
+    },
+    {
+      "epoch": 0.8118811881188119,
+      "grad_norm": 1.953125,
+      "learning_rate": 5.8486407053637035e-05,
+      "loss": 1.3897,
+      "step": 738
+    },
+    {
+      "epoch": 0.812981298129813,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.8457016899338726e-05,
+      "loss": 1.6942,
+      "step": 739
+    },
+    {
+      "epoch": 0.8140814081408141,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.842762674504042e-05,
+      "loss": 1.4986,
+      "step": 740
+    },
+    {
+      "epoch": 0.8151815181518152,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.839823659074211e-05,
+      "loss": 1.4796,
+      "step": 741
+    },
+    {
+      "epoch": 0.8162816281628162,
+      "grad_norm": 1.875,
+      "learning_rate": 5.83688464364438e-05,
+      "loss": 1.4076,
+      "step": 742
+    },
+    {
+      "epoch": 0.8173817381738174,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.833945628214548e-05,
+      "loss": 1.6396,
+      "step": 743
+    },
+    {
+      "epoch": 0.8184818481848185,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.831006612784718e-05,
+      "loss": 1.5624,
+      "step": 744
+    },
+    {
+      "epoch": 0.8195819581958196,
+      "grad_norm": 2.375,
+      "learning_rate": 5.828067597354887e-05,
+      "loss": 1.8003,
+      "step": 745
+    },
+    {
+      "epoch": 0.8206820682068207,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.825128581925055e-05,
+      "loss": 1.4028,
+      "step": 746
+    },
+    {
+      "epoch": 0.8217821782178217,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.822189566495224e-05,
+      "loss": 1.4568,
+      "step": 747
+    },
+    {
+      "epoch": 0.8228822882288229,
+      "grad_norm": 2.5,
+      "learning_rate": 5.8192505510653934e-05,
+      "loss": 1.8554,
+      "step": 748
+    },
+    {
+      "epoch": 0.823982398239824,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.816311535635563e-05,
+      "loss": 1.617,
+      "step": 749
+    },
+    {
+      "epoch": 0.8250825082508251,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.8133725202057315e-05,
+      "loss": 1.2893,
+      "step": 750
+    },
+    {
+      "epoch": 0.8261826182618262,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.8104335047759005e-05,
+      "loss": 1.8837,
+      "step": 751
+    },
+    {
+      "epoch": 0.8272827282728272,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.807494489346069e-05,
+      "loss": 1.4506,
+      "step": 752
+    },
+    {
+      "epoch": 0.8283828382838284,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.8045554739162386e-05,
+      "loss": 1.4465,
+      "step": 753
+    },
+    {
+      "epoch": 0.8294829482948295,
+      "grad_norm": 2.125,
+      "learning_rate": 5.801616458486408e-05,
+      "loss": 1.5741,
+      "step": 754
+    },
+    {
+      "epoch": 0.8305830583058306,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.798677443056577e-05,
+      "loss": 1.6751,
+      "step": 755
+    },
+    {
+      "epoch": 0.8316831683168316,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.795738427626745e-05,
+      "loss": 1.6346,
+      "step": 756
+    },
+    {
+      "epoch": 0.8327832783278328,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.792799412196915e-05,
+      "loss": 1.6761,
+      "step": 757
+    },
+    {
+      "epoch": 0.8338833883388339,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.789860396767084e-05,
+      "loss": 1.6316,
+      "step": 758
+    },
+    {
+      "epoch": 0.834983498349835,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.786921381337252e-05,
+      "loss": 1.3671,
+      "step": 759
+    },
+    {
+      "epoch": 0.8360836083608361,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.783982365907421e-05,
+      "loss": 1.5949,
+      "step": 760
+    },
+    {
+      "epoch": 0.8371837183718371,
+      "grad_norm": 2.375,
+      "learning_rate": 5.781043350477591e-05,
+      "loss": 1.474,
+      "step": 761
+    },
+    {
+      "epoch": 0.8382838283828383,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.7781043350477594e-05,
+      "loss": 1.5606,
+      "step": 762
+    },
+    {
+      "epoch": 0.8393839383938394,
+      "grad_norm": 2.125,
+      "learning_rate": 5.7751653196179284e-05,
+      "loss": 1.697,
+      "step": 763
+    },
+    {
+      "epoch": 0.8404840484048405,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.7722263041880975e-05,
+      "loss": 1.4388,
+      "step": 764
+    },
+    {
+      "epoch": 0.8415841584158416,
+      "grad_norm": 2.125,
+      "learning_rate": 5.769287288758266e-05,
+      "loss": 1.7886,
+      "step": 765
+    },
+    {
+      "epoch": 0.8426842684268426,
+      "grad_norm": 2.375,
+      "learning_rate": 5.7663482733284356e-05,
+      "loss": 1.7736,
+      "step": 766
+    },
+    {
+      "epoch": 0.8437843784378438,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.7634092578986046e-05,
+      "loss": 1.4872,
+      "step": 767
+    },
+    {
+      "epoch": 0.8448844884488449,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.760470242468773e-05,
+      "loss": 1.343,
+      "step": 768
+    },
+    {
+      "epoch": 0.845984598459846,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.757531227038942e-05,
+      "loss": 1.4557,
+      "step": 769
+    },
+    {
+      "epoch": 0.847084708470847,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.754592211609112e-05,
+      "loss": 1.7495,
+      "step": 770
+    },
+    {
+      "epoch": 0.8481848184818482,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.751653196179281e-05,
+      "loss": 1.7764,
+      "step": 771
+    },
+    {
+      "epoch": 0.8492849284928493,
+      "grad_norm": 2.375,
+      "learning_rate": 5.748714180749449e-05,
+      "loss": 1.5884,
+      "step": 772
+    },
+    {
+      "epoch": 0.8503850385038504,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.745775165319618e-05,
+      "loss": 1.4816,
+      "step": 773
+    },
+    {
+      "epoch": 0.8514851485148515,
+      "grad_norm": 2.125,
+      "learning_rate": 5.742836149889788e-05,
+      "loss": 1.4588,
+      "step": 774
+    },
+    {
+      "epoch": 0.8525852585258525,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.7398971344599564e-05,
+      "loss": 1.618,
+      "step": 775
+    },
+    {
+      "epoch": 0.8536853685368537,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.7369581190301254e-05,
+      "loss": 1.7487,
+      "step": 776
+    },
+    {
+      "epoch": 0.8547854785478548,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.734019103600294e-05,
+      "loss": 1.6769,
+      "step": 777
+    },
+    {
+      "epoch": 0.8558855885588559,
+      "grad_norm": 2.375,
+      "learning_rate": 5.7310800881704635e-05,
+      "loss": 1.487,
+      "step": 778
+    },
+    {
+      "epoch": 0.856985698569857,
+      "grad_norm": 2.375,
+      "learning_rate": 5.7281410727406326e-05,
+      "loss": 1.5438,
+      "step": 779
+    },
+    {
+      "epoch": 0.858085808580858,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.7252020573108016e-05,
+      "loss": 1.1473,
+      "step": 780
+    },
+    {
+      "epoch": 0.8591859185918592,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.72226304188097e-05,
+      "loss": 1.6146,
+      "step": 781
+    },
+    {
+      "epoch": 0.8602860286028603,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.719324026451139e-05,
+      "loss": 1.6389,
+      "step": 782
+    },
+    {
+      "epoch": 0.8613861386138614,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.716385011021309e-05,
+      "loss": 1.6052,
+      "step": 783
+    },
+    {
+      "epoch": 0.8624862486248625,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.713445995591477e-05,
+      "loss": 1.6683,
+      "step": 784
+    },
+    {
+      "epoch": 0.8635863586358636,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.710506980161646e-05,
+      "loss": 1.4732,
+      "step": 785
+    },
+    {
+      "epoch": 0.8646864686468647,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.707567964731815e-05,
+      "loss": 1.743,
+      "step": 786
+    },
+    {
+      "epoch": 0.8657865786578658,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.704628949301984e-05,
+      "loss": 1.3659,
+      "step": 787
+    },
+    {
+      "epoch": 0.8668866886688669,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.7016899338721533e-05,
+      "loss": 1.6343,
+      "step": 788
+    },
+    {
+      "epoch": 0.8679867986798679,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.6987509184423224e-05,
+      "loss": 1.71,
+      "step": 789
+    },
+    {
+      "epoch": 0.8690869086908691,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.695811903012491e-05,
+      "loss": 1.7739,
+      "step": 790
+    },
+    {
+      "epoch": 0.8701870187018702,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.6928728875826605e-05,
+      "loss": 1.6741,
+      "step": 791
+    },
+    {
+      "epoch": 0.8712871287128713,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.6899338721528295e-05,
+      "loss": 1.5435,
+      "step": 792
+    },
+    {
+      "epoch": 0.8723872387238724,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.686994856722998e-05,
+      "loss": 1.6073,
+      "step": 793
+    },
+    {
+      "epoch": 0.8734873487348734,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.684055841293167e-05,
+      "loss": 1.5007,
+      "step": 794
+    },
+    {
+      "epoch": 0.8745874587458746,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.681116825863337e-05,
+      "loss": 1.4593,
+      "step": 795
+    },
+    {
+      "epoch": 0.8756875687568757,
+      "grad_norm": 3.671875,
+      "learning_rate": 5.678177810433506e-05,
+      "loss": 1.6745,
+      "step": 796
+    },
+    {
+      "epoch": 0.8767876787678768,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.675238795003674e-05,
+      "loss": 1.5353,
+      "step": 797
+    },
+    {
+      "epoch": 0.8778877887788779,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.672299779573843e-05,
+      "loss": 1.4888,
+      "step": 798
+    },
+    {
+      "epoch": 0.878987898789879,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.6693607641440115e-05,
+      "loss": 1.2896,
+      "step": 799
+    },
+    {
+      "epoch": 0.8800880088008801,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.666421748714181e-05,
+      "loss": 1.5834,
+      "step": 800
+    },
+    {
+      "epoch": 0.8811881188118812,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.66348273328435e-05,
+      "loss": 1.4749,
+      "step": 801
+    },
+    {
+      "epoch": 0.8822882288228823,
+      "grad_norm": 2.25,
+      "learning_rate": 5.6605437178545194e-05,
+      "loss": 1.632,
+      "step": 802
+    },
+    {
+      "epoch": 0.8833883388338833,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.657604702424688e-05,
+      "loss": 1.5413,
+      "step": 803
+    },
+    {
+      "epoch": 0.8844884488448845,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.6546656869948575e-05,
+      "loss": 1.7308,
+      "step": 804
+    },
+    {
+      "epoch": 0.8855885588558856,
+      "grad_norm": 2.25,
+      "learning_rate": 5.6517266715650265e-05,
+      "loss": 1.6322,
+      "step": 805
+    },
+    {
+      "epoch": 0.8866886688668867,
+      "grad_norm": 2.375,
+      "learning_rate": 5.648787656135195e-05,
+      "loss": 1.541,
+      "step": 806
+    },
+    {
+      "epoch": 0.8877887788778878,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.645848640705364e-05,
+      "loss": 1.5616,
+      "step": 807
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.642909625275534e-05,
+      "loss": 1.7956,
+      "step": 808
+    },
+    {
+      "epoch": 0.88998899889989,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.639970609845702e-05,
+      "loss": 1.7087,
+      "step": 809
+    },
+    {
+      "epoch": 0.8910891089108911,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.637031594415871e-05,
+      "loss": 1.3398,
+      "step": 810
+    },
+    {
+      "epoch": 0.8921892189218922,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.63409257898604e-05,
+      "loss": 1.6365,
+      "step": 811
+    },
+    {
+      "epoch": 0.8932893289328933,
+      "grad_norm": 2.125,
+      "learning_rate": 5.63115356355621e-05,
+      "loss": 1.2201,
+      "step": 812
+    },
+    {
+      "epoch": 0.8943894389438944,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.628214548126378e-05,
+      "loss": 1.7129,
+      "step": 813
+    },
+    {
+      "epoch": 0.8954895489548955,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.625275532696547e-05,
+      "loss": 1.4734,
+      "step": 814
+    },
+    {
+      "epoch": 0.8965896589658966,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.622336517266716e-05,
+      "loss": 1.7215,
+      "step": 815
+    },
+    {
+      "epoch": 0.8976897689768977,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.619397501836885e-05,
+      "loss": 1.6865,
+      "step": 816
+    },
+    {
+      "epoch": 0.8987898789878987,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.6164584864070545e-05,
+      "loss": 1.5027,
+      "step": 817
+    },
+    {
+      "epoch": 0.8998899889988999,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.613519470977223e-05,
+      "loss": 1.4826,
+      "step": 818
+    },
+    {
+      "epoch": 0.900990099009901,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.610580455547392e-05,
+      "loss": 1.4865,
+      "step": 819
+    },
+    {
+      "epoch": 0.9020902090209021,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.607641440117561e-05,
+      "loss": 1.5752,
+      "step": 820
+    },
+    {
+      "epoch": 0.9031903190319032,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.6047024246877307e-05,
+      "loss": 1.4319,
+      "step": 821
+    },
+    {
+      "epoch": 0.9042904290429042,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.601763409257899e-05,
+      "loss": 1.5401,
+      "step": 822
+    },
+    {
+      "epoch": 0.9053905390539054,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.598824393828068e-05,
+      "loss": 1.4957,
+      "step": 823
+    },
+    {
+      "epoch": 0.9064906490649065,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.5958853783982365e-05,
+      "loss": 1.1597,
+      "step": 824
+    },
+    {
+      "epoch": 0.9075907590759076,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.592946362968406e-05,
+      "loss": 1.8065,
+      "step": 825
+    },
+    {
+      "epoch": 0.9086908690869087,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.590007347538575e-05,
+      "loss": 1.5861,
+      "step": 826
+    },
+    {
+      "epoch": 0.9097909790979097,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.587068332108744e-05,
+      "loss": 1.652,
+      "step": 827
+    },
+    {
+      "epoch": 0.9108910891089109,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.5841293166789127e-05,
+      "loss": 1.5176,
+      "step": 828
+    },
+    {
+      "epoch": 0.911991199119912,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.5811903012490824e-05,
+      "loss": 1.4969,
+      "step": 829
+    },
+    {
+      "epoch": 0.9130913091309131,
+      "grad_norm": 2.25,
+      "learning_rate": 5.5782512858192514e-05,
+      "loss": 1.6083,
+      "step": 830
+    },
+    {
+      "epoch": 0.9141914191419142,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.57531227038942e-05,
+      "loss": 1.6042,
+      "step": 831
+    },
+    {
+      "epoch": 0.9152915291529153,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.572373254959589e-05,
+      "loss": 1.8022,
+      "step": 832
+    },
+    {
+      "epoch": 0.9163916391639164,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.569434239529758e-05,
+      "loss": 1.4958,
+      "step": 833
+    },
+    {
+      "epoch": 0.9174917491749175,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.566495224099927e-05,
+      "loss": 1.4334,
+      "step": 834
+    },
+    {
+      "epoch": 0.9185918591859186,
+      "grad_norm": 1.8828125,
+      "learning_rate": 5.563556208670096e-05,
+      "loss": 1.5565,
+      "step": 835
+    },
+    {
+      "epoch": 0.9196919691969196,
+      "grad_norm": 2.125,
+      "learning_rate": 5.560617193240265e-05,
+      "loss": 1.7164,
+      "step": 836
+    },
+    {
+      "epoch": 0.9207920792079208,
+      "grad_norm": 1.8359375,
+      "learning_rate": 5.5576781778104334e-05,
+      "loss": 1.258,
+      "step": 837
+    },
+    {
+      "epoch": 0.9218921892189219,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.554739162380603e-05,
+      "loss": 1.4457,
+      "step": 838
+    },
+    {
+      "epoch": 0.922992299229923,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.551800146950772e-05,
+      "loss": 1.6521,
+      "step": 839
+    },
+    {
+      "epoch": 0.9240924092409241,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.5488611315209406e-05,
+      "loss": 1.4665,
+      "step": 840
+    },
+    {
+      "epoch": 0.9251925192519251,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.5459221160911096e-05,
+      "loss": 1.6125,
+      "step": 841
+    },
+    {
+      "epoch": 0.9262926292629263,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.5429831006612794e-05,
+      "loss": 1.5871,
+      "step": 842
+    },
+    {
+      "epoch": 0.9273927392739274,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.5400440852314484e-05,
+      "loss": 1.6145,
+      "step": 843
+    },
+    {
+      "epoch": 0.9284928492849285,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.537105069801617e-05,
+      "loss": 1.8115,
+      "step": 844
+    },
+    {
+      "epoch": 0.9295929592959296,
+      "grad_norm": 2.375,
+      "learning_rate": 5.534166054371786e-05,
+      "loss": 1.5685,
+      "step": 845
+    },
+    {
+      "epoch": 0.9306930693069307,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.5312270389419556e-05,
+      "loss": 1.372,
+      "step": 846
+    },
+    {
+      "epoch": 0.9317931793179318,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.528288023512124e-05,
+      "loss": 1.6581,
+      "step": 847
+    },
+    {
+      "epoch": 0.9328932893289329,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.525349008082293e-05,
+      "loss": 1.741,
+      "step": 848
+    },
+    {
+      "epoch": 0.933993399339934,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.5224099926524614e-05,
+      "loss": 1.5709,
+      "step": 849
+    },
+    {
+      "epoch": 0.935093509350935,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.5194709772226304e-05,
+      "loss": 1.5417,
+      "step": 850
+    },
+    {
+      "epoch": 0.9361936193619362,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.5165319617928e-05,
+      "loss": 1.5576,
+      "step": 851
+    },
+    {
+      "epoch": 0.9372937293729373,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.513592946362969e-05,
+      "loss": 1.472,
+      "step": 852
+    },
+    {
+      "epoch": 0.9383938393839384,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.5106539309331376e-05,
+      "loss": 1.6572,
+      "step": 853
+    },
+    {
+      "epoch": 0.9394939493949395,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.5077149155033066e-05,
+      "loss": 1.5672,
+      "step": 854
+    },
+    {
+      "epoch": 0.9405940594059405,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.504775900073476e-05,
+      "loss": 1.5213,
+      "step": 855
+    },
+    {
+      "epoch": 0.9416941694169417,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.501836884643645e-05,
+      "loss": 1.5937,
+      "step": 856
+    },
+    {
+      "epoch": 0.9427942794279428,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.498897869213814e-05,
+      "loss": 1.4235,
+      "step": 857
+    },
+    {
+      "epoch": 0.9438943894389439,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.495958853783983e-05,
+      "loss": 1.7409,
+      "step": 858
+    },
+    {
+      "epoch": 0.944994499449945,
+      "grad_norm": 1.9609375,
+      "learning_rate": 5.4930198383541525e-05,
+      "loss": 1.6493,
+      "step": 859
+    },
+    {
+      "epoch": 0.9460946094609461,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.490080822924321e-05,
+      "loss": 1.3586,
+      "step": 860
+    },
+    {
+      "epoch": 0.9471947194719472,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.48714180749449e-05,
+      "loss": 1.582,
+      "step": 861
+    },
+    {
+      "epoch": 0.9482948294829483,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.484202792064658e-05,
+      "loss": 1.6701,
+      "step": 862
+    },
+    {
+      "epoch": 0.9493949394939494,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.4812637766348274e-05,
+      "loss": 1.743,
+      "step": 863
+    },
+    {
+      "epoch": 0.9504950495049505,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.478324761204997e-05,
+      "loss": 1.5275,
+      "step": 864
+    },
+    {
+      "epoch": 0.9515951595159516,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.4753857457751655e-05,
+      "loss": 1.6513,
+      "step": 865
+    },
+    {
+      "epoch": 0.9526952695269527,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.4724467303453345e-05,
+      "loss": 1.5936,
+      "step": 866
+    },
+    {
+      "epoch": 0.9537953795379538,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.4695077149155036e-05,
+      "loss": 1.6711,
+      "step": 867
+    },
+    {
+      "epoch": 0.9548954895489549,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.466568699485673e-05,
+      "loss": 1.7695,
+      "step": 868
+    },
+    {
+      "epoch": 0.9559955995599559,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.463629684055842e-05,
+      "loss": 1.6498,
+      "step": 869
+    },
+    {
+      "epoch": 0.9570957095709571,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.460690668626011e-05,
+      "loss": 1.5715,
+      "step": 870
+    },
+    {
+      "epoch": 0.9581958195819582,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.457751653196179e-05,
+      "loss": 1.6479,
+      "step": 871
+    },
+    {
+      "epoch": 0.9592959295929593,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.454812637766349e-05,
+      "loss": 1.6561,
+      "step": 872
+    },
+    {
+      "epoch": 0.9603960396039604,
+      "grad_norm": 1.984375,
+      "learning_rate": 5.451873622336518e-05,
+      "loss": 1.3865,
+      "step": 873
+    },
+    {
+      "epoch": 0.9614961496149615,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.448934606906687e-05,
+      "loss": 1.7027,
+      "step": 874
+    },
+    {
+      "epoch": 0.9625962596259626,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.445995591476855e-05,
+      "loss": 1.8046,
+      "step": 875
+    },
+    {
+      "epoch": 0.9636963696369637,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.443056576047025e-05,
+      "loss": 1.4968,
+      "step": 876
+    },
+    {
+      "epoch": 0.9647964796479648,
+      "grad_norm": 2.375,
+      "learning_rate": 5.440117560617194e-05,
+      "loss": 1.5226,
+      "step": 877
+    },
+    {
+      "epoch": 0.9658965896589659,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.4371785451873625e-05,
+      "loss": 1.4835,
+      "step": 878
+    },
+    {
+      "epoch": 0.966996699669967,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.4342395297575315e-05,
+      "loss": 1.5699,
+      "step": 879
+    },
+    {
+      "epoch": 0.9680968096809681,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.4313005143277e-05,
+      "loss": 1.6416,
+      "step": 880
+    },
+    {
+      "epoch": 0.9691969196919692,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.4283614988978696e-05,
+      "loss": 1.7116,
+      "step": 881
+    },
+    {
+      "epoch": 0.9702970297029703,
+      "grad_norm": 2.375,
+      "learning_rate": 5.425422483468039e-05,
+      "loss": 1.5741,
+      "step": 882
+    },
+    {
+      "epoch": 0.9713971397139713,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.422483468038208e-05,
+      "loss": 1.8149,
+      "step": 883
+    },
+    {
+      "epoch": 0.9724972497249725,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.419544452608376e-05,
+      "loss": 1.3872,
+      "step": 884
+    },
+    {
+      "epoch": 0.9735973597359736,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.416605437178546e-05,
+      "loss": 1.3464,
+      "step": 885
+    },
+    {
+      "epoch": 0.9746974697469747,
+      "grad_norm": 2.0,
+      "learning_rate": 5.413666421748715e-05,
+      "loss": 1.7328,
+      "step": 886
+    },
+    {
+      "epoch": 0.9757975797579758,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.410727406318883e-05,
+      "loss": 1.5121,
+      "step": 887
+    },
+    {
+      "epoch": 0.976897689768977,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.407788390889052e-05,
+      "loss": 1.5179,
+      "step": 888
+    },
+    {
+      "epoch": 0.977997799779978,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.404849375459222e-05,
+      "loss": 1.6404,
+      "step": 889
+    },
+    {
+      "epoch": 0.9790979097909791,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.401910360029391e-05,
+      "loss": 1.6263,
+      "step": 890
+    },
+    {
+      "epoch": 0.9801980198019802,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.3989713445995594e-05,
+      "loss": 1.6783,
+      "step": 891
+    },
+    {
+      "epoch": 0.9812981298129813,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.3960323291697285e-05,
+      "loss": 1.3568,
+      "step": 892
+    },
+    {
+      "epoch": 0.9823982398239824,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.393093313739898e-05,
+      "loss": 1.491,
+      "step": 893
+    },
+    {
+      "epoch": 0.9834983498349835,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.3901542983100666e-05,
+      "loss": 1.7797,
+      "step": 894
+    },
+    {
+      "epoch": 0.9845984598459846,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.3872152828802356e-05,
+      "loss": 1.4643,
+      "step": 895
+    },
+    {
+      "epoch": 0.9856985698569857,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.384276267450404e-05,
+      "loss": 1.7508,
+      "step": 896
+    },
+    {
+      "epoch": 0.9867986798679867,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.381337252020573e-05,
+      "loss": 1.5216,
+      "step": 897
+    },
+    {
+      "epoch": 0.9878987898789879,
+      "grad_norm": 2.75,
+      "learning_rate": 5.378398236590743e-05,
+      "loss": 1.3264,
+      "step": 898
+    },
+    {
+      "epoch": 0.988998899889989,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.375459221160912e-05,
+      "loss": 1.9772,
+      "step": 899
+    },
+    {
+      "epoch": 0.9900990099009901,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.37252020573108e-05,
+      "loss": 1.3596,
+      "step": 900
+    },
+    {
+      "epoch": 0.9911991199119912,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.369581190301249e-05,
+      "loss": 1.8375,
+      "step": 901
+    },
+    {
+      "epoch": 0.9922992299229924,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.366642174871419e-05,
+      "loss": 1.5125,
+      "step": 902
+    },
+    {
+      "epoch": 0.9933993399339934,
+      "grad_norm": 2.375,
+      "learning_rate": 5.3637031594415874e-05,
+      "loss": 1.2587,
+      "step": 903
+    },
+    {
+      "epoch": 0.9944994499449945,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.3607641440117564e-05,
+      "loss": 1.582,
+      "step": 904
+    },
+    {
+      "epoch": 0.9955995599559956,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.3578251285819255e-05,
+      "loss": 1.2471,
+      "step": 905
+    },
+    {
+      "epoch": 0.9966996699669967,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.3548861131520945e-05,
+      "loss": 1.6905,
+      "step": 906
+    },
+    {
+      "epoch": 0.9977997799779978,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.3519470977222636e-05,
+      "loss": 1.4588,
+      "step": 907
+    },
+    {
+      "epoch": 0.9988998899889989,
+      "grad_norm": 1.9921875,
+      "learning_rate": 5.3490080822924326e-05,
+      "loss": 1.6387,
+      "step": 908
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.346069066862601e-05,
+      "loss": 1.4002,
+      "step": 909
+    },
+    {
+      "epoch": 1.001100110011001,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.343130051432771e-05,
+      "loss": 1.4269,
+      "step": 910
+    },
+    {
+      "epoch": 1.0022002200220022,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.34019103600294e-05,
+      "loss": 1.1121,
+      "step": 911
+    },
+    {
+      "epoch": 1.0033003300330032,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.337252020573108e-05,
+      "loss": 1.2428,
+      "step": 912
+    },
+    {
+      "epoch": 1.0044004400440043,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.334313005143277e-05,
+      "loss": 1.1478,
+      "step": 913
+    },
+    {
+      "epoch": 1.0055005500550056,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.331373989713446e-05,
+      "loss": 1.2539,
+      "step": 914
+    },
+    {
+      "epoch": 1.0066006600660067,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.328434974283616e-05,
+      "loss": 1.3182,
+      "step": 915
+    },
+    {
+      "epoch": 1.0077007700770078,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.3254959588537843e-05,
+      "loss": 1.1508,
+      "step": 916
+    },
+    {
+      "epoch": 1.0088008800880088,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.3225569434239534e-05,
+      "loss": 1.2783,
+      "step": 917
+    },
+    {
+      "epoch": 1.00990099009901,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.319617927994122e-05,
+      "loss": 1.0719,
+      "step": 918
+    },
+    {
+      "epoch": 1.011001100110011,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.3166789125642915e-05,
+      "loss": 1.3897,
+      "step": 919
+    },
+    {
+      "epoch": 1.012101210121012,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.3137398971344605e-05,
+      "loss": 1.1849,
+      "step": 920
+    },
+    {
+      "epoch": 1.0132013201320131,
+      "grad_norm": 2.875,
+      "learning_rate": 5.3108008817046296e-05,
+      "loss": 0.9525,
+      "step": 921
+    },
+    {
+      "epoch": 1.0143014301430142,
+      "grad_norm": 2.875,
+      "learning_rate": 5.307861866274798e-05,
+      "loss": 1.2745,
+      "step": 922
+    },
+    {
+      "epoch": 1.0154015401540153,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.304922850844968e-05,
+      "loss": 1.3548,
+      "step": 923
+    },
+    {
+      "epoch": 1.0165016501650166,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.301983835415137e-05,
+      "loss": 1.2586,
+      "step": 924
+    },
+    {
+      "epoch": 1.0176017601760177,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.299044819985305e-05,
+      "loss": 1.0806,
+      "step": 925
+    },
+    {
+      "epoch": 1.0187018701870187,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.296105804555474e-05,
+      "loss": 1.3606,
+      "step": 926
+    },
+    {
+      "epoch": 1.0198019801980198,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.293166789125644e-05,
+      "loss": 1.0979,
+      "step": 927
+    },
+    {
+      "epoch": 1.020902090209021,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.290227773695812e-05,
+      "loss": 1.1317,
+      "step": 928
+    },
+    {
+      "epoch": 1.022002200220022,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.287288758265981e-05,
+      "loss": 1.2155,
+      "step": 929
+    },
+    {
+      "epoch": 1.023102310231023,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.2843497428361504e-05,
+      "loss": 1.1456,
+      "step": 930
+    },
+    {
+      "epoch": 1.0242024202420241,
+      "grad_norm": 2.0,
+      "learning_rate": 5.281410727406319e-05,
+      "loss": 1.0199,
+      "step": 931
+    },
+    {
+      "epoch": 1.0253025302530252,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.2784717119764885e-05,
+      "loss": 1.4554,
+      "step": 932
+    },
+    {
+      "epoch": 1.0264026402640265,
+      "grad_norm": 2.375,
+      "learning_rate": 5.2755326965466575e-05,
+      "loss": 1.2992,
+      "step": 933
+    },
+    {
+      "epoch": 1.0275027502750276,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.272593681116826e-05,
+      "loss": 1.0207,
+      "step": 934
+    },
+    {
+      "epoch": 1.0286028602860287,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.269654665686995e-05,
+      "loss": 1.1568,
+      "step": 935
+    },
+    {
+      "epoch": 1.0297029702970297,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.266715650257165e-05,
+      "loss": 0.9997,
+      "step": 936
+    },
+    {
+      "epoch": 1.0308030803080308,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.263776634827333e-05,
+      "loss": 1.208,
+      "step": 937
+    },
+    {
+      "epoch": 1.0319031903190319,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.260837619397502e-05,
+      "loss": 1.0617,
+      "step": 938
+    },
+    {
+      "epoch": 1.033003300330033,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.257898603967671e-05,
+      "loss": 1.1478,
+      "step": 939
+    },
+    {
+      "epoch": 1.034103410341034,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.254959588537841e-05,
+      "loss": 1.0133,
+      "step": 940
+    },
+    {
+      "epoch": 1.0352035203520351,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.252020573108009e-05,
+      "loss": 0.894,
+      "step": 941
+    },
+    {
+      "epoch": 1.0363036303630364,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.249081557678178e-05,
+      "loss": 1.3769,
+      "step": 942
+    },
+    {
+      "epoch": 1.0374037403740375,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.246142542248347e-05,
+      "loss": 1.061,
+      "step": 943
+    },
+    {
+      "epoch": 1.0385038503850386,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.2432035268185164e-05,
+      "loss": 0.999,
+      "step": 944
+    },
+    {
+      "epoch": 1.0396039603960396,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.2402645113886855e-05,
+      "loss": 1.3555,
+      "step": 945
+    },
+    {
+      "epoch": 1.0407040704070407,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.2373254959588545e-05,
+      "loss": 1.3337,
+      "step": 946
+    },
+    {
+      "epoch": 1.0418041804180418,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.234386480529023e-05,
+      "loss": 1.4254,
+      "step": 947
+    },
+    {
+      "epoch": 1.0429042904290429,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.231447465099192e-05,
+      "loss": 1.283,
+      "step": 948
+    },
+    {
+      "epoch": 1.044004400440044,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.2285084496693617e-05,
+      "loss": 1.1668,
+      "step": 949
+    },
+    {
+      "epoch": 1.045104510451045,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.22556943423953e-05,
+      "loss": 1.0541,
+      "step": 950
+    },
+    {
+      "epoch": 1.046204620462046,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.222630418809699e-05,
+      "loss": 1.2134,
+      "step": 951
+    },
+    {
+      "epoch": 1.0473047304730474,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.219691403379868e-05,
+      "loss": 1.2902,
+      "step": 952
+    },
+    {
+      "epoch": 1.0484048404840485,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.216752387950037e-05,
+      "loss": 1.1353,
+      "step": 953
+    },
+    {
+      "epoch": 1.0495049504950495,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.213813372520206e-05,
+      "loss": 1.1466,
+      "step": 954
+    },
+    {
+      "epoch": 1.0506050605060506,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.210874357090375e-05,
+      "loss": 1.2522,
+      "step": 955
+    },
+    {
+      "epoch": 1.0517051705170517,
+      "grad_norm": 2.5,
+      "learning_rate": 5.2079353416605437e-05,
+      "loss": 1.2851,
+      "step": 956
+    },
+    {
+      "epoch": 1.0528052805280528,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.2049963262307134e-05,
+      "loss": 1.555,
+      "step": 957
+    },
+    {
+      "epoch": 1.0539053905390539,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.2020573108008824e-05,
+      "loss": 1.2545,
+      "step": 958
+    },
+    {
+      "epoch": 1.055005500550055,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.199118295371051e-05,
+      "loss": 1.118,
+      "step": 959
+    },
+    {
+      "epoch": 1.056105610561056,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.19617927994122e-05,
+      "loss": 1.3316,
+      "step": 960
+    },
+    {
+      "epoch": 1.0572057205720573,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.1932402645113896e-05,
+      "loss": 1.565,
+      "step": 961
+    },
+    {
+      "epoch": 1.0583058305830584,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.1903012490815586e-05,
+      "loss": 1.1789,
+      "step": 962
+    },
+    {
+      "epoch": 1.0594059405940595,
+      "grad_norm": 2.375,
+      "learning_rate": 5.187362233651727e-05,
+      "loss": 1.4611,
+      "step": 963
+    },
+    {
+      "epoch": 1.0605060506050605,
+      "grad_norm": 2.375,
+      "learning_rate": 5.184423218221896e-05,
+      "loss": 1.0217,
+      "step": 964
+    },
+    {
+      "epoch": 1.0616061606160616,
+      "grad_norm": 2.0,
+      "learning_rate": 5.1814842027920644e-05,
+      "loss": 1.3831,
+      "step": 965
+    },
+    {
+      "epoch": 1.0627062706270627,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.178545187362234e-05,
+      "loss": 1.3578,
+      "step": 966
+    },
+    {
+      "epoch": 1.0638063806380638,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.175606171932403e-05,
+      "loss": 1.315,
+      "step": 967
+    },
+    {
+      "epoch": 1.0649064906490648,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.1726671565025716e-05,
+      "loss": 1.2475,
+      "step": 968
+    },
+    {
+      "epoch": 1.066006600660066,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.1697281410727406e-05,
+      "loss": 1.3253,
+      "step": 969
+    },
+    {
+      "epoch": 1.0671067106710672,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.1667891256429104e-05,
+      "loss": 1.1204,
+      "step": 970
+    },
+    {
+      "epoch": 1.0682068206820683,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.1638501102130794e-05,
+      "loss": 1.3817,
+      "step": 971
+    },
+    {
+      "epoch": 1.0693069306930694,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.160911094783248e-05,
+      "loss": 1.1561,
+      "step": 972
+    },
+    {
+      "epoch": 1.0704070407040704,
+      "grad_norm": 2.25,
+      "learning_rate": 5.157972079353417e-05,
+      "loss": 1.0891,
+      "step": 973
+    },
+    {
+      "epoch": 1.0715071507150715,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.1550330639235866e-05,
+      "loss": 1.2216,
+      "step": 974
+    },
+    {
+      "epoch": 1.0726072607260726,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.152094048493755e-05,
+      "loss": 1.0594,
+      "step": 975
+    },
+    {
+      "epoch": 1.0737073707370737,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.149155033063924e-05,
+      "loss": 1.2216,
+      "step": 976
+    },
+    {
+      "epoch": 1.0748074807480748,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.146216017634093e-05,
+      "loss": 1.3661,
+      "step": 977
+    },
+    {
+      "epoch": 1.0759075907590758,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.143277002204262e-05,
+      "loss": 0.9648,
+      "step": 978
+    },
+    {
+      "epoch": 1.0770077007700771,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.140337986774431e-05,
+      "loss": 1.0614,
+      "step": 979
+    },
+    {
+      "epoch": 1.0781078107810782,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.1373989713446e-05,
+      "loss": 1.2266,
+      "step": 980
+    },
+    {
+      "epoch": 1.0792079207920793,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.1344599559147686e-05,
+      "loss": 1.3718,
+      "step": 981
+    },
+    {
+      "epoch": 1.0803080308030804,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.1315209404849376e-05,
+      "loss": 1.1942,
+      "step": 982
+    },
+    {
+      "epoch": 1.0814081408140814,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.128581925055107e-05,
+      "loss": 1.105,
+      "step": 983
+    },
+    {
+      "epoch": 1.0825082508250825,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.125642909625276e-05,
+      "loss": 1.1801,
+      "step": 984
+    },
+    {
+      "epoch": 1.0836083608360836,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.122703894195445e-05,
+      "loss": 1.0894,
+      "step": 985
+    },
+    {
+      "epoch": 1.0847084708470847,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.119764878765614e-05,
+      "loss": 1.0531,
+      "step": 986
+    },
+    {
+      "epoch": 1.0858085808580857,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.1168258633357835e-05,
+      "loss": 0.9903,
+      "step": 987
+    },
+    {
+      "epoch": 1.0869086908690868,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.113886847905952e-05,
+      "loss": 0.9502,
+      "step": 988
+    },
+    {
+      "epoch": 1.0880088008800881,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.110947832476121e-05,
+      "loss": 1.2628,
+      "step": 989
+    },
+    {
+      "epoch": 1.0891089108910892,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.108008817046289e-05,
+      "loss": 1.2676,
+      "step": 990
+    },
+    {
+      "epoch": 1.0902090209020903,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.105069801616459e-05,
+      "loss": 1.1489,
+      "step": 991
+    },
+    {
+      "epoch": 1.0913091309130913,
+      "grad_norm": 2.0,
+      "learning_rate": 5.102130786186628e-05,
+      "loss": 1.1527,
+      "step": 992
+    },
+    {
+      "epoch": 1.0924092409240924,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.099191770756797e-05,
+      "loss": 1.5156,
+      "step": 993
+    },
+    {
+      "epoch": 1.0935093509350935,
+      "grad_norm": 2.375,
+      "learning_rate": 5.0962527553269655e-05,
+      "loss": 1.0121,
+      "step": 994
+    },
+    {
+      "epoch": 1.0946094609460946,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.093313739897135e-05,
+      "loss": 1.4556,
+      "step": 995
+    },
+    {
+      "epoch": 1.0957095709570956,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.090374724467304e-05,
+      "loss": 1.1344,
+      "step": 996
+    },
+    {
+      "epoch": 1.0968096809680967,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.087435709037473e-05,
+      "loss": 1.0032,
+      "step": 997
+    },
+    {
+      "epoch": 1.0979097909790978,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.084496693607642e-05,
+      "loss": 1.312,
+      "step": 998
+    },
+    {
+      "epoch": 1.099009900990099,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.08155767817781e-05,
+      "loss": 1.2078,
+      "step": 999
+    },
+    {
+      "epoch": 1.1001100110011002,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.07861866274798e-05,
+      "loss": 1.3494,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1012101210121013,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.075679647318149e-05,
+      "loss": 1.3328,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1023102310231023,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.072740631888318e-05,
+      "loss": 1.2453,
+      "step": 1002
+    },
+    {
+      "epoch": 1.1034103410341034,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.069801616458486e-05,
+      "loss": 1.2597,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1045104510451045,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.066862601028656e-05,
+      "loss": 1.109,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1056105610561056,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.063923585598825e-05,
+      "loss": 1.2462,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1067106710671066,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.0609845701689935e-05,
+      "loss": 1.381,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1078107810781077,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.0580455547391625e-05,
+      "loss": 0.9699,
+      "step": 1007
+    },
+    {
+      "epoch": 1.108910891089109,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.055106539309332e-05,
+      "loss": 1.5846,
+      "step": 1008
+    },
+    {
+      "epoch": 1.11001100110011,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.0521675238795006e-05,
+      "loss": 1.4307,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.04922850844967e-05,
+      "loss": 1.2259,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1122112211221122,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.046289493019839e-05,
+      "loss": 1.1772,
+      "step": 1011
+    },
+    {
+      "epoch": 1.1133113311331133,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.0433504775900084e-05,
+      "loss": 1.1651,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1144114411441144,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.040411462160177e-05,
+      "loss": 1.3773,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1155115511551155,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.037472446730346e-05,
+      "loss": 1.0995,
+      "step": 1014
+    },
+    {
+      "epoch": 1.1166116611661165,
+      "grad_norm": 2.125,
+      "learning_rate": 5.034533431300514e-05,
+      "loss": 1.2973,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1177117711771176,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.031594415870683e-05,
+      "loss": 1.3272,
+      "step": 1016
+    },
+    {
+      "epoch": 1.118811881188119,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.028655400440853e-05,
+      "loss": 1.2021,
+      "step": 1017
+    },
+    {
+      "epoch": 1.11991199119912,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.025716385011022e-05,
+      "loss": 1.2301,
+      "step": 1018
+    },
+    {
+      "epoch": 1.121012101210121,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.0227773695811904e-05,
+      "loss": 1.0296,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1221122112211221,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.0198383541513595e-05,
+      "loss": 1.1173,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1232123212321232,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.016899338721529e-05,
+      "loss": 1.4074,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1243124312431243,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.0139603232916976e-05,
+      "loss": 1.2705,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1254125412541254,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.0110213078618666e-05,
+      "loss": 1.2953,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1265126512651265,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.008082292432036e-05,
+      "loss": 1.2869,
+      "step": 1024
+    },
+    {
+      "epoch": 1.1276127612761275,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.005143277002205e-05,
+      "loss": 1.1114,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1287128712871288,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.002204261572374e-05,
+      "loss": 1.1888,
+      "step": 1026
+    },
+    {
+      "epoch": 1.12981298129813,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.999265246142543e-05,
+      "loss": 1.1731,
+      "step": 1027
+    },
+    {
+      "epoch": 1.130913091309131,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.996326230712711e-05,
+      "loss": 1.2443,
+      "step": 1028
+    },
+    {
+      "epoch": 1.132013201320132,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.993387215282881e-05,
+      "loss": 1.3778,
+      "step": 1029
+    },
+    {
+      "epoch": 1.1331133113311331,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.99044819985305e-05,
+      "loss": 1.2769,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1342134213421342,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.9875091844232184e-05,
+      "loss": 1.1871,
+      "step": 1031
+    },
+    {
+      "epoch": 1.1353135313531353,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.9845701689933874e-05,
+      "loss": 1.2167,
+      "step": 1032
+    },
+    {
+      "epoch": 1.1364136413641364,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.9816311535635565e-05,
+      "loss": 1.2439,
+      "step": 1033
+    },
+    {
+      "epoch": 1.1375137513751374,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.978692138133726e-05,
+      "loss": 1.4802,
+      "step": 1034
+    },
+    {
+      "epoch": 1.1386138613861387,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.9757531227038946e-05,
+      "loss": 1.2965,
+      "step": 1035
+    },
+    {
+      "epoch": 1.1397139713971396,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.9728141072740636e-05,
+      "loss": 1.3417,
+      "step": 1036
+    },
+    {
+      "epoch": 1.140814081408141,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.969875091844232e-05,
+      "loss": 1.144,
+      "step": 1037
+    },
+    {
+      "epoch": 1.141914191419142,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.966936076414402e-05,
+      "loss": 1.4449,
+      "step": 1038
+    },
+    {
+      "epoch": 1.143014301430143,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.963997060984571e-05,
+      "loss": 1.2326,
+      "step": 1039
+    },
+    {
+      "epoch": 1.1441144114411441,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.961058045554739e-05,
+      "loss": 1.412,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1452145214521452,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.958119030124908e-05,
+      "loss": 1.2254,
+      "step": 1041
+    },
+    {
+      "epoch": 1.1463146314631463,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.955180014695078e-05,
+      "loss": 0.9923,
+      "step": 1042
+    },
+    {
+      "epoch": 1.1474147414741473,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.952240999265247e-05,
+      "loss": 1.3785,
+      "step": 1043
+    },
+    {
+      "epoch": 1.1485148514851484,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.9493019838354153e-05,
+      "loss": 1.2895,
+      "step": 1044
+    },
+    {
+      "epoch": 1.1496149614961495,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.9463629684055844e-05,
+      "loss": 1.0586,
+      "step": 1045
+    },
+    {
+      "epoch": 1.1507150715071508,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.943423952975754e-05,
+      "loss": 1.1182,
+      "step": 1046
+    },
+    {
+      "epoch": 1.1518151815181519,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.9404849375459225e-05,
+      "loss": 1.5135,
+      "step": 1047
+    },
+    {
+      "epoch": 1.152915291529153,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.9375459221160915e-05,
+      "loss": 1.0663,
+      "step": 1048
+    },
+    {
+      "epoch": 1.154015401540154,
+      "grad_norm": 2.375,
+      "learning_rate": 4.9346069066862606e-05,
+      "loss": 0.9709,
+      "step": 1049
+    },
+    {
+      "epoch": 1.155115511551155,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.931667891256429e-05,
+      "loss": 1.1705,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1562156215621562,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.928728875826599e-05,
+      "loss": 0.9862,
+      "step": 1051
+    },
+    {
+      "epoch": 1.1573157315731573,
+      "grad_norm": 2.5,
+      "learning_rate": 4.925789860396768e-05,
+      "loss": 1.3157,
+      "step": 1052
+    },
+    {
+      "epoch": 1.1584158415841583,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.922850844966936e-05,
+      "loss": 1.0843,
+      "step": 1053
+    },
+    {
+      "epoch": 1.1595159515951594,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.919911829537105e-05,
+      "loss": 1.4297,
+      "step": 1054
+    },
+    {
+      "epoch": 1.1606160616061607,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.916972814107275e-05,
+      "loss": 1.044,
+      "step": 1055
+    },
+    {
+      "epoch": 1.1617161716171618,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.914033798677443e-05,
+      "loss": 1.3456,
+      "step": 1056
+    },
+    {
+      "epoch": 1.1628162816281629,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.911094783247612e-05,
+      "loss": 0.9411,
+      "step": 1057
+    },
+    {
+      "epoch": 1.163916391639164,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.9081557678177814e-05,
+      "loss": 1.332,
+      "step": 1058
+    },
+    {
+      "epoch": 1.165016501650165,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.905216752387951e-05,
+      "loss": 1.1876,
+      "step": 1059
+    },
+    {
+      "epoch": 1.166116611661166,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.9022777369581195e-05,
+      "loss": 0.949,
+      "step": 1060
+    },
+    {
+      "epoch": 1.1672167216721672,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.8993387215282885e-05,
+      "loss": 1.289,
+      "step": 1061
+    },
+    {
+      "epoch": 1.1683168316831682,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.896399706098457e-05,
+      "loss": 1.1725,
+      "step": 1062
+    },
+    {
+      "epoch": 1.1694169416941693,
+      "grad_norm": 2.5,
+      "learning_rate": 4.8934606906686266e-05,
+      "loss": 1.085,
+      "step": 1063
+    },
+    {
+      "epoch": 1.1705170517051706,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.890521675238796e-05,
+      "loss": 0.9612,
+      "step": 1064
+    },
+    {
+      "epoch": 1.1716171617161717,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.887582659808965e-05,
+      "loss": 1.2199,
+      "step": 1065
+    },
+    {
+      "epoch": 1.1727172717271728,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.884643644379133e-05,
+      "loss": 1.2503,
+      "step": 1066
+    },
+    {
+      "epoch": 1.1738173817381738,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.881704628949302e-05,
+      "loss": 1.2211,
+      "step": 1067
+    },
+    {
+      "epoch": 1.174917491749175,
+      "grad_norm": 2.25,
+      "learning_rate": 4.878765613519472e-05,
+      "loss": 1.1428,
+      "step": 1068
+    },
+    {
+      "epoch": 1.176017601760176,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.87582659808964e-05,
+      "loss": 1.175,
+      "step": 1069
+    },
+    {
+      "epoch": 1.177117711771177,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.872887582659809e-05,
+      "loss": 1.3779,
+      "step": 1070
+    },
+    {
+      "epoch": 1.1782178217821782,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.869948567229978e-05,
+      "loss": 1.233,
+      "step": 1071
+    },
+    {
+      "epoch": 1.1793179317931792,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.8670095518001474e-05,
+      "loss": 1.2291,
+      "step": 1072
+    },
+    {
+      "epoch": 1.1804180418041805,
+      "grad_norm": 2.375,
+      "learning_rate": 4.8640705363703164e-05,
+      "loss": 1.0313,
+      "step": 1073
+    },
+    {
+      "epoch": 1.1815181518151816,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.8611315209404855e-05,
+      "loss": 1.1511,
+      "step": 1074
+    },
+    {
+      "epoch": 1.1826182618261827,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.858192505510654e-05,
+      "loss": 1.3597,
+      "step": 1075
+    },
+    {
+      "epoch": 1.1837183718371838,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.8552534900808236e-05,
+      "loss": 1.0628,
+      "step": 1076
+    },
+    {
+      "epoch": 1.1848184818481848,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.8523144746509927e-05,
+      "loss": 1.2451,
+      "step": 1077
+    },
+    {
+      "epoch": 1.185918591859186,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.849375459221161e-05,
+      "loss": 1.1172,
+      "step": 1078
+    },
+    {
+      "epoch": 1.187018701870187,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.84643644379133e-05,
+      "loss": 1.1285,
+      "step": 1079
+    },
+    {
+      "epoch": 1.188118811881188,
+      "grad_norm": 2.25,
+      "learning_rate": 4.8434974283615e-05,
+      "loss": 1.058,
+      "step": 1080
+    },
+    {
+      "epoch": 1.1892189218921891,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.840558412931668e-05,
+      "loss": 1.4405,
+      "step": 1081
+    },
+    {
+      "epoch": 1.1903190319031904,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.837619397501837e-05,
+      "loss": 1.1415,
+      "step": 1082
+    },
+    {
+      "epoch": 1.1914191419141915,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.834680382072006e-05,
+      "loss": 1.1366,
+      "step": 1083
+    },
+    {
+      "epoch": 1.1925192519251926,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.8317413666421746e-05,
+      "loss": 1.4537,
+      "step": 1084
+    },
+    {
+      "epoch": 1.1936193619361937,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.8288023512123444e-05,
+      "loss": 1.212,
+      "step": 1085
+    },
+    {
+      "epoch": 1.1947194719471947,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.8258633357825134e-05,
+      "loss": 1.3101,
+      "step": 1086
+    },
+    {
+      "epoch": 1.1958195819581958,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.822924320352682e-05,
+      "loss": 1.1174,
+      "step": 1087
+    },
+    {
+      "epoch": 1.196919691969197,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.819985304922851e-05,
+      "loss": 1.1863,
+      "step": 1088
+    },
+    {
+      "epoch": 1.198019801980198,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.8170462894930206e-05,
+      "loss": 1.063,
+      "step": 1089
+    },
+    {
+      "epoch": 1.199119911991199,
+      "grad_norm": 2.375,
+      "learning_rate": 4.8141072740631896e-05,
+      "loss": 1.1504,
+      "step": 1090
+    },
+    {
+      "epoch": 1.2002200220022001,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.811168258633358e-05,
+      "loss": 1.2864,
+      "step": 1091
+    },
+    {
+      "epoch": 1.2013201320132012,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.808229243203527e-05,
+      "loss": 1.1613,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2024202420242025,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.805290227773697e-05,
+      "loss": 1.2041,
+      "step": 1093
+    },
+    {
+      "epoch": 1.2035203520352036,
+      "grad_norm": 2.625,
+      "learning_rate": 4.802351212343865e-05,
+      "loss": 1.3243,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2046204620462047,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.799412196914034e-05,
+      "loss": 1.011,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2057205720572057,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.796473181484203e-05,
+      "loss": 1.2029,
+      "step": 1096
+    },
+    {
+      "epoch": 1.2068206820682068,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.793534166054372e-05,
+      "loss": 1.2578,
+      "step": 1097
+    },
+    {
+      "epoch": 1.2079207920792079,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.7905951506245414e-05,
+      "loss": 1.3469,
+      "step": 1098
+    },
+    {
+      "epoch": 1.209020902090209,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.7876561351947104e-05,
+      "loss": 0.9905,
+      "step": 1099
+    },
+    {
+      "epoch": 1.21012101210121,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.784717119764879e-05,
+      "loss": 1.4448,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2112211221122111,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.781778104335048e-05,
+      "loss": 1.266,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2123212321232124,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.7788390889052176e-05,
+      "loss": 1.078,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2134213421342135,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.775900073475386e-05,
+      "loss": 1.1223,
+      "step": 1103
+    },
+    {
+      "epoch": 1.2145214521452146,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.772961058045555e-05,
+      "loss": 1.1073,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2156215621562156,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.770022042615724e-05,
+      "loss": 1.3109,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2167216721672167,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.767083027185894e-05,
+      "loss": 1.0902,
+      "step": 1106
+    },
+    {
+      "epoch": 1.2178217821782178,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.764144011756062e-05,
+      "loss": 1.2626,
+      "step": 1107
+    },
+    {
+      "epoch": 1.2189218921892189,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.761204996326231e-05,
+      "loss": 1.2154,
+      "step": 1108
+    },
+    {
+      "epoch": 1.22002200220022,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.7582659808963996e-05,
+      "loss": 1.5045,
+      "step": 1109
+    },
+    {
+      "epoch": 1.221122112211221,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.755326965466569e-05,
+      "loss": 1.1823,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.752387950036738e-05,
+      "loss": 1.4075,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2233223322332234,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.749448934606907e-05,
+      "loss": 1.2221,
+      "step": 1112
+    },
+    {
+      "epoch": 1.2244224422442245,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.746509919177076e-05,
+      "loss": 1.044,
+      "step": 1113
+    },
+    {
+      "epoch": 1.2255225522552256,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.7435709037472455e-05,
+      "loss": 1.3007,
+      "step": 1114
+    },
+    {
+      "epoch": 1.2266226622662266,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.7406318883174145e-05,
+      "loss": 1.1053,
+      "step": 1115
+    },
+    {
+      "epoch": 1.2277227722772277,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.737692872887583e-05,
+      "loss": 1.2035,
+      "step": 1116
+    },
+    {
+      "epoch": 1.2288228822882288,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.734753857457752e-05,
+      "loss": 1.1674,
+      "step": 1117
+    },
+    {
+      "epoch": 1.2299229922992299,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.73181484202792e-05,
+      "loss": 1.1442,
+      "step": 1118
+    },
+    {
+      "epoch": 1.231023102310231,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.72887582659809e-05,
+      "loss": 1.2645,
+      "step": 1119
+    },
+    {
+      "epoch": 1.2321232123212322,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.725936811168259e-05,
+      "loss": 1.3937,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2332233223322333,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.722997795738428e-05,
+      "loss": 1.4413,
+      "step": 1121
+    },
+    {
+      "epoch": 1.2343234323432344,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.7200587803085965e-05,
+      "loss": 1.2714,
+      "step": 1122
+    },
+    {
+      "epoch": 1.2354235423542355,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.717119764878766e-05,
+      "loss": 1.135,
+      "step": 1123
+    },
+    {
+      "epoch": 1.2365236523652365,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.714180749448935e-05,
+      "loss": 1.2184,
+      "step": 1124
+    },
+    {
+      "epoch": 1.2376237623762376,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.711241734019104e-05,
+      "loss": 1.2429,
+      "step": 1125
+    },
+    {
+      "epoch": 1.2387238723872387,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.708302718589273e-05,
+      "loss": 1.2204,
+      "step": 1126
+    },
+    {
+      "epoch": 1.2398239823982398,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.7053637031594425e-05,
+      "loss": 1.1624,
+      "step": 1127
+    },
+    {
+      "epoch": 1.2409240924092408,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.702424687729611e-05,
+      "loss": 1.2407,
+      "step": 1128
+    },
+    {
+      "epoch": 1.2420242024202421,
+      "grad_norm": 2.875,
+      "learning_rate": 4.69948567229978e-05,
+      "loss": 1.3829,
+      "step": 1129
+    },
+    {
+      "epoch": 1.2431243124312432,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.696546656869949e-05,
+      "loss": 1.269,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2442244224422443,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.693607641440119e-05,
+      "loss": 1.2564,
+      "step": 1131
+    },
+    {
+      "epoch": 1.2453245324532454,
+      "grad_norm": 2.25,
+      "learning_rate": 4.690668626010287e-05,
+      "loss": 1.2967,
+      "step": 1132
+    },
+    {
+      "epoch": 1.2464246424642464,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.687729610580456e-05,
+      "loss": 1.045,
+      "step": 1133
+    },
+    {
+      "epoch": 1.2475247524752475,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.6847905951506245e-05,
+      "loss": 1.1003,
+      "step": 1134
+    },
+    {
+      "epoch": 1.2486248624862486,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.6818515797207935e-05,
+      "loss": 1.2076,
+      "step": 1135
+    },
+    {
+      "epoch": 1.2497249724972497,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.678912564290963e-05,
+      "loss": 1.4209,
+      "step": 1136
+    },
+    {
+      "epoch": 1.2508250825082508,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.675973548861132e-05,
+      "loss": 1.1369,
+      "step": 1137
+    },
+    {
+      "epoch": 1.251925192519252,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.6730345334313007e-05,
+      "loss": 0.9744,
+      "step": 1138
+    },
+    {
+      "epoch": 1.253025302530253,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.67009551800147e-05,
+      "loss": 1.4391,
+      "step": 1139
+    },
+    {
+      "epoch": 1.2541254125412542,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.6671565025716394e-05,
+      "loss": 1.2242,
+      "step": 1140
+    },
+    {
+      "epoch": 1.2552255225522553,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.664217487141808e-05,
+      "loss": 1.3415,
+      "step": 1141
+    },
+    {
+      "epoch": 1.2563256325632564,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.661278471711977e-05,
+      "loss": 1.2063,
+      "step": 1142
+    },
+    {
+      "epoch": 1.2574257425742574,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.658339456282145e-05,
+      "loss": 1.1285,
+      "step": 1143
+    },
+    {
+      "epoch": 1.2585258525852585,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.655400440852315e-05,
+      "loss": 1.0923,
+      "step": 1144
+    },
+    {
+      "epoch": 1.2596259625962596,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.652461425422484e-05,
+      "loss": 1.1163,
+      "step": 1145
+    },
+    {
+      "epoch": 1.2607260726072607,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.649522409992653e-05,
+      "loss": 1.3582,
+      "step": 1146
+    },
+    {
+      "epoch": 1.261826182618262,
+      "grad_norm": 2.5,
+      "learning_rate": 4.6465833945628214e-05,
+      "loss": 1.1407,
+      "step": 1147
+    },
+    {
+      "epoch": 1.2629262926292628,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.643644379132991e-05,
+      "loss": 1.3831,
+      "step": 1148
+    },
+    {
+      "epoch": 1.2640264026402641,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.64070536370316e-05,
+      "loss": 1.3663,
+      "step": 1149
+    },
+    {
+      "epoch": 1.2651265126512652,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.6377663482733286e-05,
+      "loss": 1.2085,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2662266226622663,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6348273328434976e-05,
+      "loss": 0.9514,
+      "step": 1151
+    },
+    {
+      "epoch": 1.2673267326732673,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.631888317413667e-05,
+      "loss": 1.2943,
+      "step": 1152
+    },
+    {
+      "epoch": 1.2684268426842684,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.628949301983836e-05,
+      "loss": 1.1978,
+      "step": 1153
+    },
+    {
+      "epoch": 1.2695269526952695,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.626010286554005e-05,
+      "loss": 1.5028,
+      "step": 1154
+    },
+    {
+      "epoch": 1.2706270627062706,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.623071271124174e-05,
+      "loss": 1.2143,
+      "step": 1155
+    },
+    {
+      "epoch": 1.2717271727172716,
+      "grad_norm": 1.8671875,
+      "learning_rate": 4.620132255694342e-05,
+      "loss": 1.1177,
+      "step": 1156
+    },
+    {
+      "epoch": 1.2728272827282727,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.617193240264512e-05,
+      "loss": 1.3139,
+      "step": 1157
+    },
+    {
+      "epoch": 1.273927392739274,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.614254224834681e-05,
+      "loss": 1.0174,
+      "step": 1158
+    },
+    {
+      "epoch": 1.275027502750275,
+      "grad_norm": 5.125,
+      "learning_rate": 4.6113152094048494e-05,
+      "loss": 1.4015,
+      "step": 1159
+    },
+    {
+      "epoch": 1.2761276127612762,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.6083761939750184e-05,
+      "loss": 1.2651,
+      "step": 1160
+    },
+    {
+      "epoch": 1.2772277227722773,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.605437178545188e-05,
+      "loss": 1.1728,
+      "step": 1161
+    },
+    {
+      "epoch": 1.2783278327832783,
+      "grad_norm": 2.25,
+      "learning_rate": 4.602498163115357e-05,
+      "loss": 1.1789,
+      "step": 1162
+    },
+    {
+      "epoch": 1.2794279427942794,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.5995591476855256e-05,
+      "loss": 1.3452,
+      "step": 1163
+    },
+    {
+      "epoch": 1.2805280528052805,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.5966201322556946e-05,
+      "loss": 1.2877,
+      "step": 1164
+    },
+    {
+      "epoch": 1.2816281628162816,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.5936811168258643e-05,
+      "loss": 0.9833,
+      "step": 1165
+    },
+    {
+      "epoch": 1.2827282728272826,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.590742101396033e-05,
+      "loss": 1.2037,
+      "step": 1166
+    },
+    {
+      "epoch": 1.283828382838284,
+      "grad_norm": 2.5,
+      "learning_rate": 4.587803085966202e-05,
+      "loss": 1.3993,
+      "step": 1167
+    },
+    {
+      "epoch": 1.284928492849285,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.584864070536371e-05,
+      "loss": 1.3893,
+      "step": 1168
+    },
+    {
+      "epoch": 1.286028602860286,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.581925055106539e-05,
+      "loss": 1.4178,
+      "step": 1169
+    },
+    {
+      "epoch": 1.2871287128712872,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.578986039676709e-05,
+      "loss": 1.3377,
+      "step": 1170
+    },
+    {
+      "epoch": 1.2882288228822882,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.576047024246878e-05,
+      "loss": 1.1213,
+      "step": 1171
+    },
+    {
+      "epoch": 1.2893289328932893,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.5731080088170463e-05,
+      "loss": 1.3061,
+      "step": 1172
+    },
+    {
+      "epoch": 1.2904290429042904,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.5701689933872154e-05,
+      "loss": 1.1868,
+      "step": 1173
+    },
+    {
+      "epoch": 1.2915291529152915,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.567229977957385e-05,
+      "loss": 1.3013,
+      "step": 1174
+    },
+    {
+      "epoch": 1.2926292629262925,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.5642909625275535e-05,
+      "loss": 1.077,
+      "step": 1175
+    },
+    {
+      "epoch": 1.2937293729372938,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.5613519470977225e-05,
+      "loss": 1.3166,
+      "step": 1176
+    },
+    {
+      "epoch": 1.2948294829482947,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.5584129316678916e-05,
+      "loss": 1.3223,
+      "step": 1177
+    },
+    {
+      "epoch": 1.295929592959296,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.555473916238061e-05,
+      "loss": 1.0972,
+      "step": 1178
+    },
+    {
+      "epoch": 1.297029702970297,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.55253490080823e-05,
+      "loss": 1.2414,
+      "step": 1179
+    },
+    {
+      "epoch": 1.2981298129812981,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.549595885378399e-05,
+      "loss": 1.3277,
+      "step": 1180
+    },
+    {
+      "epoch": 1.2992299229922992,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.546656869948567e-05,
+      "loss": 1.3855,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3003300330033003,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.543717854518737e-05,
+      "loss": 1.2207,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3014301430143014,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.540778839088906e-05,
+      "loss": 1.1645,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3025302530253025,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.537839823659074e-05,
+      "loss": 1.2004,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3036303630363038,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.534900808229243e-05,
+      "loss": 1.4946,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3047304730473046,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.5319617927994124e-05,
+      "loss": 1.1027,
+      "step": 1186
+    },
+    {
+      "epoch": 1.305830583058306,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.529022777369582e-05,
+      "loss": 1.1728,
+      "step": 1187
+    },
+    {
+      "epoch": 1.306930693069307,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.5260837619397505e-05,
+      "loss": 1.3997,
+      "step": 1188
+    },
+    {
+      "epoch": 1.308030803080308,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.5231447465099195e-05,
+      "loss": 1.0449,
+      "step": 1189
+    },
+    {
+      "epoch": 1.3091309130913091,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.520205731080088e-05,
+      "loss": 1.0546,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3102310231023102,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.5172667156502576e-05,
+      "loss": 1.0996,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3113311331133113,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.514327700220427e-05,
+      "loss": 1.4026,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3124312431243124,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.511388684790596e-05,
+      "loss": 1.4221,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3135313531353137,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.508449669360764e-05,
+      "loss": 1.4213,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3146314631463145,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.505510653930934e-05,
+      "loss": 1.0952,
+      "step": 1195
+    },
+    {
+      "epoch": 1.3157315731573158,
+      "grad_norm": 2.5,
+      "learning_rate": 4.502571638501103e-05,
+      "loss": 1.3559,
+      "step": 1196
+    },
+    {
+      "epoch": 1.316831683168317,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.499632623071271e-05,
+      "loss": 1.1425,
+      "step": 1197
+    },
+    {
+      "epoch": 1.317931793179318,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.49669360764144e-05,
+      "loss": 1.1916,
+      "step": 1198
+    },
+    {
+      "epoch": 1.319031903190319,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.4937545922116093e-05,
+      "loss": 1.4053,
+      "step": 1199
+    },
+    {
+      "epoch": 1.3201320132013201,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.4908155767817784e-05,
+      "loss": 1.4082,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3212321232123212,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.4878765613519474e-05,
+      "loss": 1.2766,
+      "step": 1201
+    },
+    {
+      "epoch": 1.3223322332233223,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.4849375459221165e-05,
+      "loss": 1.376,
+      "step": 1202
+    },
+    {
+      "epoch": 1.3234323432343233,
+      "grad_norm": 2.25,
+      "learning_rate": 4.481998530492285e-05,
+      "loss": 0.988,
+      "step": 1203
+    },
+    {
+      "epoch": 1.3245324532453244,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.4790595150624546e-05,
+      "loss": 1.1424,
+      "step": 1204
+    },
+    {
+      "epoch": 1.3256325632563257,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.4761204996326236e-05,
+      "loss": 1.4422,
+      "step": 1205
+    },
+    {
+      "epoch": 1.3267326732673268,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.473181484202792e-05,
+      "loss": 1.2835,
+      "step": 1206
+    },
+    {
+      "epoch": 1.3278327832783279,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.470242468772961e-05,
+      "loss": 1.218,
+      "step": 1207
+    },
+    {
+      "epoch": 1.328932893289329,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.467303453343131e-05,
+      "loss": 1.2439,
+      "step": 1208
+    },
+    {
+      "epoch": 1.33003300330033,
+      "grad_norm": 2.625,
+      "learning_rate": 4.4643644379133e-05,
+      "loss": 1.0372,
+      "step": 1209
+    },
+    {
+      "epoch": 1.331133113311331,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.461425422483468e-05,
+      "loss": 1.3484,
+      "step": 1210
+    },
+    {
+      "epoch": 1.3322332233223322,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.458486407053637e-05,
+      "loss": 1.1253,
+      "step": 1211
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.455547391623807e-05,
+      "loss": 1.0815,
+      "step": 1212
+    },
+    {
+      "epoch": 1.3344334433443343,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.4526083761939754e-05,
+      "loss": 1.4208,
+      "step": 1213
+    },
+    {
+      "epoch": 1.3355335533553356,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.4496693607641444e-05,
+      "loss": 1.1481,
+      "step": 1214
+    },
+    {
+      "epoch": 1.3366336633663367,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.4467303453343135e-05,
+      "loss": 1.0272,
+      "step": 1215
+    },
+    {
+      "epoch": 1.3377337733773378,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.443791329904482e-05,
+      "loss": 1.0857,
+      "step": 1216
+    },
+    {
+      "epoch": 1.3388338833883389,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.4408523144746516e-05,
+      "loss": 1.4816,
+      "step": 1217
+    },
+    {
+      "epoch": 1.33993399339934,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.4379132990448206e-05,
+      "loss": 1.2985,
+      "step": 1218
+    },
+    {
+      "epoch": 1.341034103410341,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.434974283614989e-05,
+      "loss": 1.3713,
+      "step": 1219
+    },
+    {
+      "epoch": 1.342134213421342,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.432035268185158e-05,
+      "loss": 1.2553,
+      "step": 1220
+    },
+    {
+      "epoch": 1.3432343234323432,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.429096252755328e-05,
+      "loss": 1.1298,
+      "step": 1221
+    },
+    {
+      "epoch": 1.3443344334433442,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.426157237325496e-05,
+      "loss": 1.1663,
+      "step": 1222
+    },
+    {
+      "epoch": 1.3454345434543455,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.423218221895665e-05,
+      "loss": 1.0309,
+      "step": 1223
+    },
+    {
+      "epoch": 1.3465346534653464,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.420279206465834e-05,
+      "loss": 1.2194,
+      "step": 1224
+    },
+    {
+      "epoch": 1.3476347634763477,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.417340191036004e-05,
+      "loss": 1.4533,
+      "step": 1225
+    },
+    {
+      "epoch": 1.3487348734873488,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.4144011756061724e-05,
+      "loss": 1.3851,
+      "step": 1226
+    },
+    {
+      "epoch": 1.3498349834983498,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.4114621601763414e-05,
+      "loss": 1.1178,
+      "step": 1227
+    },
+    {
+      "epoch": 1.350935093509351,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.40852314474651e-05,
+      "loss": 1.0999,
+      "step": 1228
+    },
+    {
+      "epoch": 1.352035203520352,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.4055841293166795e-05,
+      "loss": 1.3034,
+      "step": 1229
+    },
+    {
+      "epoch": 1.353135313531353,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.4026451138868486e-05,
+      "loss": 1.33,
+      "step": 1230
+    },
+    {
+      "epoch": 1.3542354235423542,
+      "grad_norm": 2.25,
+      "learning_rate": 4.399706098457017e-05,
+      "loss": 1.3627,
+      "step": 1231
+    },
+    {
+      "epoch": 1.3553355335533555,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.396767083027186e-05,
+      "loss": 1.2069,
+      "step": 1232
+    },
+    {
+      "epoch": 1.3564356435643563,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.393828067597355e-05,
+      "loss": 1.0895,
+      "step": 1233
+    },
+    {
+      "epoch": 1.3575357535753576,
+      "grad_norm": 2.375,
+      "learning_rate": 4.390889052167525e-05,
+      "loss": 1.076,
+      "step": 1234
+    },
+    {
+      "epoch": 1.3586358635863587,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.387950036737693e-05,
+      "loss": 1.1,
+      "step": 1235
+    },
+    {
+      "epoch": 1.3597359735973598,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.385011021307862e-05,
+      "loss": 1.2034,
+      "step": 1236
+    },
+    {
+      "epoch": 1.3608360836083608,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.3820720058780306e-05,
+      "loss": 1.3919,
+      "step": 1237
+    },
+    {
+      "epoch": 1.361936193619362,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.3791329904482e-05,
+      "loss": 1.2147,
+      "step": 1238
+    },
+    {
+      "epoch": 1.363036303630363,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.376193975018369e-05,
+      "loss": 1.1465,
+      "step": 1239
+    },
+    {
+      "epoch": 1.364136413641364,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.3732549595885384e-05,
+      "loss": 1.0461,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3652365236523654,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.370315944158707e-05,
+      "loss": 1.211,
+      "step": 1241
+    },
+    {
+      "epoch": 1.3663366336633662,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.3673769287288765e-05,
+      "loss": 1.0133,
+      "step": 1242
+    },
+    {
+      "epoch": 1.3674367436743675,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.3644379132990455e-05,
+      "loss": 1.2368,
+      "step": 1243
+    },
+    {
+      "epoch": 1.3685368536853686,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.361498897869214e-05,
+      "loss": 1.0974,
+      "step": 1244
+    },
+    {
+      "epoch": 1.3696369636963697,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.358559882439383e-05,
+      "loss": 1.0904,
+      "step": 1245
+    },
+    {
+      "epoch": 1.3707370737073707,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.355620867009553e-05,
+      "loss": 1.1639,
+      "step": 1246
+    },
+    {
+      "epoch": 1.3718371837183718,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.352681851579721e-05,
+      "loss": 1.0106,
+      "step": 1247
+    },
+    {
+      "epoch": 1.372937293729373,
+      "grad_norm": 2.375,
+      "learning_rate": 4.34974283614989e-05,
+      "loss": 1.116,
+      "step": 1248
+    },
+    {
+      "epoch": 1.374037403740374,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.346803820720059e-05,
+      "loss": 1.2619,
+      "step": 1249
+    },
+    {
+      "epoch": 1.3751375137513753,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.3438648052902275e-05,
+      "loss": 1.0481,
+      "step": 1250
+    },
+    {
+      "epoch": 1.3762376237623761,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.340925789860397e-05,
+      "loss": 1.0278,
+      "step": 1251
+    },
+    {
+      "epoch": 1.3773377337733774,
+      "grad_norm": 2.25,
+      "learning_rate": 4.337986774430566e-05,
+      "loss": 1.3989,
+      "step": 1252
+    },
+    {
+      "epoch": 1.3784378437843785,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.335047759000735e-05,
+      "loss": 1.304,
+      "step": 1253
+    },
+    {
+      "epoch": 1.3795379537953796,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.332108743570904e-05,
+      "loss": 1.0466,
+      "step": 1254
+    },
+    {
+      "epoch": 1.3806380638063807,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.3291697281410735e-05,
+      "loss": 1.3401,
+      "step": 1255
+    },
+    {
+      "epoch": 1.3817381738173817,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.3262307127112425e-05,
+      "loss": 1.318,
+      "step": 1256
+    },
+    {
+      "epoch": 1.3828382838283828,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.323291697281411e-05,
+      "loss": 1.2518,
+      "step": 1257
+    },
+    {
+      "epoch": 1.3839383938393839,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.32035268185158e-05,
+      "loss": 1.396,
+      "step": 1258
+    },
+    {
+      "epoch": 1.385038503850385,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.3174136664217497e-05,
+      "loss": 1.2654,
+      "step": 1259
+    },
+    {
+      "epoch": 1.386138613861386,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.314474650991918e-05,
+      "loss": 1.3256,
+      "step": 1260
+    },
+    {
+      "epoch": 1.3872387238723873,
+      "grad_norm": 2.25,
+      "learning_rate": 4.311535635562087e-05,
+      "loss": 1.0458,
+      "step": 1261
+    },
+    {
+      "epoch": 1.3883388338833884,
+      "grad_norm": 2.25,
+      "learning_rate": 4.3085966201322555e-05,
+      "loss": 1.4471,
+      "step": 1262
+    },
+    {
+      "epoch": 1.3894389438943895,
+      "grad_norm": 2.375,
+      "learning_rate": 4.305657604702425e-05,
+      "loss": 1.0932,
+      "step": 1263
+    },
+    {
+      "epoch": 1.3905390539053906,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.302718589272594e-05,
+      "loss": 1.1142,
+      "step": 1264
+    },
+    {
+      "epoch": 1.3916391639163916,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.299779573842763e-05,
+      "loss": 1.3771,
+      "step": 1265
+    },
+    {
+      "epoch": 1.3927392739273927,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.2968405584129317e-05,
+      "loss": 1.6955,
+      "step": 1266
+    },
+    {
+      "epoch": 1.3938393839383938,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.293901542983101e-05,
+      "loss": 1.1692,
+      "step": 1267
+    },
+    {
+      "epoch": 1.3949394939493949,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.2909625275532704e-05,
+      "loss": 1.2859,
+      "step": 1268
+    },
+    {
+      "epoch": 1.396039603960396,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.288023512123439e-05,
+      "loss": 1.1804,
+      "step": 1269
+    },
+    {
+      "epoch": 1.3971397139713972,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.285084496693608e-05,
+      "loss": 1.1999,
+      "step": 1270
+    },
+    {
+      "epoch": 1.3982398239823983,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.282145481263777e-05,
+      "loss": 1.1631,
+      "step": 1271
+    },
+    {
+      "epoch": 1.3993399339933994,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.279206465833946e-05,
+      "loss": 1.2156,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4004400440044005,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.276267450404115e-05,
+      "loss": 1.2094,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4015401540154016,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.273328434974284e-05,
+      "loss": 1.3376,
+      "step": 1274
+    },
+    {
+      "epoch": 1.4026402640264026,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.2703894195444524e-05,
+      "loss": 1.271,
+      "step": 1275
+    },
+    {
+      "epoch": 1.4037403740374037,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.267450404114622e-05,
+      "loss": 1.2078,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4048404840484048,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.264511388684791e-05,
+      "loss": 1.0953,
+      "step": 1277
+    },
+    {
+      "epoch": 1.4059405940594059,
+      "grad_norm": 2.5,
+      "learning_rate": 4.2615723732549596e-05,
+      "loss": 1.292,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4070407040704072,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.2586333578251286e-05,
+      "loss": 1.376,
+      "step": 1279
+    },
+    {
+      "epoch": 1.408140814081408,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.2556943423952984e-05,
+      "loss": 1.3155,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4092409240924093,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.2527553269654674e-05,
+      "loss": 1.2564,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4103410341034104,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.249816311535636e-05,
+      "loss": 1.4339,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4114411441144115,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.246877296105805e-05,
+      "loss": 0.9847,
+      "step": 1283
+    },
+    {
+      "epoch": 1.4125412541254125,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.243938280675973e-05,
+      "loss": 1.2788,
+      "step": 1284
+    },
+    {
+      "epoch": 1.4136413641364136,
+      "grad_norm": 2.5,
+      "learning_rate": 4.240999265246143e-05,
+      "loss": 1.3318,
+      "step": 1285
+    },
+    {
+      "epoch": 1.4147414741474147,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.238060249816312e-05,
+      "loss": 1.3238,
+      "step": 1286
+    },
+    {
+      "epoch": 1.4158415841584158,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.235121234386481e-05,
+      "loss": 1.2619,
+      "step": 1287
+    },
+    {
+      "epoch": 1.416941694169417,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.2321822189566494e-05,
+      "loss": 1.0036,
+      "step": 1288
+    },
+    {
+      "epoch": 1.418041804180418,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.229243203526819e-05,
+      "loss": 1.2271,
+      "step": 1289
+    },
+    {
+      "epoch": 1.4191419141914192,
+      "grad_norm": 2.125,
+      "learning_rate": 4.226304188096988e-05,
+      "loss": 1.4129,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4202420242024203,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.2233651726671566e-05,
+      "loss": 1.3156,
+      "step": 1291
+    },
+    {
+      "epoch": 1.4213421342134214,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.2204261572373256e-05,
+      "loss": 1.2971,
+      "step": 1292
+    },
+    {
+      "epoch": 1.4224422442244224,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.2174871418074953e-05,
+      "loss": 1.3583,
+      "step": 1293
+    },
+    {
+      "epoch": 1.4235423542354235,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.214548126377664e-05,
+      "loss": 1.3134,
+      "step": 1294
+    },
+    {
+      "epoch": 1.4246424642464246,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.211609110947833e-05,
+      "loss": 1.2964,
+      "step": 1295
+    },
+    {
+      "epoch": 1.4257425742574257,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.208670095518002e-05,
+      "loss": 1.3721,
+      "step": 1296
+    },
+    {
+      "epoch": 1.426842684268427,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.2057310800881715e-05,
+      "loss": 1.3412,
+      "step": 1297
+    },
+    {
+      "epoch": 1.4279427942794278,
+      "grad_norm": 2.375,
+      "learning_rate": 4.20279206465834e-05,
+      "loss": 1.2519,
+      "step": 1298
+    },
+    {
+      "epoch": 1.4290429042904291,
+      "grad_norm": 2.25,
+      "learning_rate": 4.199853049228509e-05,
+      "loss": 1.3367,
+      "step": 1299
+    },
+    {
+      "epoch": 1.4301430143014302,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.196914033798677e-05,
+      "loss": 0.9375,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4312431243124313,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.1939750183688464e-05,
+      "loss": 1.1473,
+      "step": 1301
+    },
+    {
+      "epoch": 1.4323432343234324,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.191036002939016e-05,
+      "loss": 1.123,
+      "step": 1302
+    },
+    {
+      "epoch": 1.4334433443344334,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.1880969875091845e-05,
+      "loss": 1.3949,
+      "step": 1303
+    },
+    {
+      "epoch": 1.4345434543454345,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.1851579720793535e-05,
+      "loss": 1.0643,
+      "step": 1304
+    },
+    {
+      "epoch": 1.4356435643564356,
+      "grad_norm": 2.375,
+      "learning_rate": 4.1822189566495226e-05,
+      "loss": 1.259,
+      "step": 1305
+    },
+    {
+      "epoch": 1.4367436743674367,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.179279941219692e-05,
+      "loss": 1.157,
+      "step": 1306
+    },
+    {
+      "epoch": 1.4378437843784377,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.176340925789861e-05,
+      "loss": 1.2918,
+      "step": 1307
+    },
+    {
+      "epoch": 1.438943894389439,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.17340191036003e-05,
+      "loss": 1.3229,
+      "step": 1308
+    },
+    {
+      "epoch": 1.4400440044004401,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.170462894930198e-05,
+      "loss": 1.3424,
+      "step": 1309
+    },
+    {
+      "epoch": 1.4411441144114412,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.167523879500368e-05,
+      "loss": 1.4766,
+      "step": 1310
+    },
+    {
+      "epoch": 1.4422442244224423,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.164584864070537e-05,
+      "loss": 1.1963,
+      "step": 1311
+    },
+    {
+      "epoch": 1.4433443344334433,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.161645848640706e-05,
+      "loss": 1.3786,
+      "step": 1312
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.158706833210874e-05,
+      "loss": 1.4402,
+      "step": 1313
+    },
+    {
+      "epoch": 1.4455445544554455,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.155767817781044e-05,
+      "loss": 1.4877,
+      "step": 1314
+    },
+    {
+      "epoch": 1.4466446644664466,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.152828802351213e-05,
+      "loss": 1.2154,
+      "step": 1315
+    },
+    {
+      "epoch": 1.4477447744774476,
+      "grad_norm": 2.375,
+      "learning_rate": 4.1498897869213815e-05,
+      "loss": 1.095,
+      "step": 1316
+    },
+    {
+      "epoch": 1.448844884488449,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.1469507714915505e-05,
+      "loss": 1.0042,
+      "step": 1317
+    },
+    {
+      "epoch": 1.44994499449945,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.1440117560617196e-05,
+      "loss": 1.2506,
+      "step": 1318
+    },
+    {
+      "epoch": 1.451045104510451,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.1410727406318886e-05,
+      "loss": 1.1411,
+      "step": 1319
+    },
+    {
+      "epoch": 1.4521452145214522,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.138133725202058e-05,
+      "loss": 1.3284,
+      "step": 1320
+    },
+    {
+      "epoch": 1.4532453245324533,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.135194709772227e-05,
+      "loss": 0.9431,
+      "step": 1321
+    },
+    {
+      "epoch": 1.4543454345434543,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.132255694342395e-05,
+      "loss": 1.1058,
+      "step": 1322
+    },
+    {
+      "epoch": 1.4554455445544554,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.129316678912565e-05,
+      "loss": 1.0375,
+      "step": 1323
+    },
+    {
+      "epoch": 1.4565456545654565,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.126377663482734e-05,
+      "loss": 1.0439,
+      "step": 1324
+    },
+    {
+      "epoch": 1.4576457645764576,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.123438648052902e-05,
+      "loss": 1.3758,
+      "step": 1325
+    },
+    {
+      "epoch": 1.4587458745874589,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.120499632623071e-05,
+      "loss": 1.2355,
+      "step": 1326
+    },
+    {
+      "epoch": 1.4598459845984597,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.117560617193241e-05,
+      "loss": 1.2276,
+      "step": 1327
+    },
+    {
+      "epoch": 1.460946094609461,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.11462160176341e-05,
+      "loss": 1.2068,
+      "step": 1328
+    },
+    {
+      "epoch": 1.462046204620462,
+      "grad_norm": 2.875,
+      "learning_rate": 4.1116825863335784e-05,
+      "loss": 0.963,
+      "step": 1329
+    },
+    {
+      "epoch": 1.4631463146314632,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.1087435709037475e-05,
+      "loss": 1.2328,
+      "step": 1330
+    },
+    {
+      "epoch": 1.4642464246424642,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.105804555473917e-05,
+      "loss": 1.3298,
+      "step": 1331
+    },
+    {
+      "epoch": 1.4653465346534653,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.1028655400440856e-05,
+      "loss": 1.1159,
+      "step": 1332
+    },
+    {
+      "epoch": 1.4664466446644664,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.0999265246142546e-05,
+      "loss": 1.1887,
+      "step": 1333
+    },
+    {
+      "epoch": 1.4675467546754675,
+      "grad_norm": 2.25,
+      "learning_rate": 4.096987509184423e-05,
+      "loss": 1.3676,
+      "step": 1334
+    },
+    {
+      "epoch": 1.4686468646864688,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.094048493754592e-05,
+      "loss": 1.2954,
+      "step": 1335
+    },
+    {
+      "epoch": 1.4697469746974696,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.091109478324762e-05,
+      "loss": 1.2628,
+      "step": 1336
+    },
+    {
+      "epoch": 1.470847084708471,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.088170462894931e-05,
+      "loss": 1.1175,
+      "step": 1337
+    },
+    {
+      "epoch": 1.471947194719472,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.085231447465099e-05,
+      "loss": 1.234,
+      "step": 1338
+    },
+    {
+      "epoch": 1.473047304730473,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.082292432035268e-05,
+      "loss": 1.4255,
+      "step": 1339
+    },
+    {
+      "epoch": 1.4741474147414741,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.079353416605438e-05,
+      "loss": 1.33,
+      "step": 1340
+    },
+    {
+      "epoch": 1.4752475247524752,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.0764144011756064e-05,
+      "loss": 1.5489,
+      "step": 1341
+    },
+    {
+      "epoch": 1.4763476347634763,
+      "grad_norm": 2.125,
+      "learning_rate": 4.0734753857457754e-05,
+      "loss": 1.3025,
+      "step": 1342
+    },
+    {
+      "epoch": 1.4774477447744774,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.0705363703159445e-05,
+      "loss": 1.1529,
+      "step": 1343
+    },
+    {
+      "epoch": 1.4785478547854787,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.0675973548861135e-05,
+      "loss": 1.287,
+      "step": 1344
+    },
+    {
+      "epoch": 1.4796479647964795,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.0646583394562826e-05,
+      "loss": 1.3194,
+      "step": 1345
+    },
+    {
+      "epoch": 1.4807480748074808,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.0617193240264516e-05,
+      "loss": 1.2789,
+      "step": 1346
+    },
+    {
+      "epoch": 1.481848184818482,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.05878030859662e-05,
+      "loss": 1.4058,
+      "step": 1347
+    },
+    {
+      "epoch": 1.482948294829483,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.05584129316679e-05,
+      "loss": 1.1109,
+      "step": 1348
+    },
+    {
+      "epoch": 1.484048404840484,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.052902277736959e-05,
+      "loss": 1.2515,
+      "step": 1349
+    },
+    {
+      "epoch": 1.4851485148514851,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.049963262307127e-05,
+      "loss": 1.0229,
+      "step": 1350
+    },
+    {
+      "epoch": 1.4862486248624862,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.047024246877296e-05,
+      "loss": 1.0549,
+      "step": 1351
+    },
+    {
+      "epoch": 1.4873487348734873,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.044085231447465e-05,
+      "loss": 0.969,
+      "step": 1352
+    },
+    {
+      "epoch": 1.4884488448844886,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.041146216017635e-05,
+      "loss": 1.2102,
+      "step": 1353
+    },
+    {
+      "epoch": 1.4895489548954894,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.0382072005878033e-05,
+      "loss": 1.294,
+      "step": 1354
+    },
+    {
+      "epoch": 1.4906490649064907,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.0352681851579724e-05,
+      "loss": 1.0932,
+      "step": 1355
+    },
+    {
+      "epoch": 1.4917491749174918,
+      "grad_norm": 1.9921875,
+      "learning_rate": 4.032329169728141e-05,
+      "loss": 1.0233,
+      "step": 1356
+    },
+    {
+      "epoch": 1.492849284928493,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.0293901542983105e-05,
+      "loss": 1.2011,
+      "step": 1357
+    },
+    {
+      "epoch": 1.493949394939494,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.0264511388684796e-05,
+      "loss": 1.1045,
+      "step": 1358
+    },
+    {
+      "epoch": 1.495049504950495,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.0235121234386486e-05,
+      "loss": 1.3033,
+      "step": 1359
+    },
+    {
+      "epoch": 1.4961496149614961,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.020573108008817e-05,
+      "loss": 1.3977,
+      "step": 1360
+    },
+    {
+      "epoch": 1.4972497249724972,
+      "grad_norm": 2.375,
+      "learning_rate": 4.017634092578987e-05,
+      "loss": 1.0913,
+      "step": 1361
+    },
+    {
+      "epoch": 1.4983498349834983,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.014695077149156e-05,
+      "loss": 1.244,
+      "step": 1362
+    },
+    {
+      "epoch": 1.4994499449944994,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.011756061719324e-05,
+      "loss": 1.35,
+      "step": 1363
+    },
+    {
+      "epoch": 1.5005500550055006,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.008817046289493e-05,
+      "loss": 1.3057,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5016501650165015,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.005878030859663e-05,
+      "loss": 1.1611,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5027502750275028,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.002939015429831e-05,
+      "loss": 1.0236,
+      "step": 1366
+    },
+    {
+      "epoch": 1.5038503850385039,
+      "grad_norm": 2.40625,
+      "learning_rate": 4e-05,
+      "loss": 1.0669,
+      "step": 1367
+    },
+    {
+      "epoch": 1.504950495049505,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.9970609845701694e-05,
+      "loss": 1.3915,
+      "step": 1368
+    },
+    {
+      "epoch": 1.506050605060506,
+      "grad_norm": 2.375,
+      "learning_rate": 3.9941219691403384e-05,
+      "loss": 1.1884,
+      "step": 1369
+    },
+    {
+      "epoch": 1.507150715071507,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.991182953710507e-05,
+      "loss": 1.2127,
+      "step": 1370
+    },
+    {
+      "epoch": 1.5082508250825084,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.9882439382806765e-05,
+      "loss": 1.1282,
+      "step": 1371
+    },
+    {
+      "epoch": 1.5093509350935093,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.985304922850845e-05,
+      "loss": 1.3427,
+      "step": 1372
+    },
+    {
+      "epoch": 1.5104510451045106,
+      "grad_norm": 2.25,
+      "learning_rate": 3.9823659074210146e-05,
+      "loss": 1.3927,
+      "step": 1373
+    },
+    {
+      "epoch": 1.5115511551155114,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.979426891991183e-05,
+      "loss": 1.0987,
+      "step": 1374
+    },
+    {
+      "epoch": 1.5126512651265127,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.976487876561352e-05,
+      "loss": 1.1355,
+      "step": 1375
+    },
+    {
+      "epoch": 1.5137513751375138,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.973548861131521e-05,
+      "loss": 1.2147,
+      "step": 1376
+    },
+    {
+      "epoch": 1.5148514851485149,
+      "grad_norm": 2.375,
+      "learning_rate": 3.97060984570169e-05,
+      "loss": 1.2304,
+      "step": 1377
+    },
+    {
+      "epoch": 1.515951595159516,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.967670830271859e-05,
+      "loss": 1.2153,
+      "step": 1378
+    },
+    {
+      "epoch": 1.517051705170517,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.964731814842028e-05,
+      "loss": 1.2249,
+      "step": 1379
+    },
+    {
+      "epoch": 1.5181518151815183,
+      "grad_norm": 2.25,
+      "learning_rate": 3.961792799412197e-05,
+      "loss": 1.419,
+      "step": 1380
+    },
+    {
+      "epoch": 1.5192519251925192,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.9588537839823664e-05,
+      "loss": 1.1026,
+      "step": 1381
+    },
+    {
+      "epoch": 1.5203520352035205,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.9559147685525354e-05,
+      "loss": 1.1099,
+      "step": 1382
+    },
+    {
+      "epoch": 1.5214521452145213,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.9529757531227045e-05,
+      "loss": 1.2736,
+      "step": 1383
+    },
+    {
+      "epoch": 1.5225522552255226,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.9500367376928735e-05,
+      "loss": 1.0199,
+      "step": 1384
+    },
+    {
+      "epoch": 1.5236523652365237,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.947097722263042e-05,
+      "loss": 1.3659,
+      "step": 1385
+    },
+    {
+      "epoch": 1.5247524752475248,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.944158706833211e-05,
+      "loss": 1.3952,
+      "step": 1386
+    },
+    {
+      "epoch": 1.5258525852585259,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.94121969140338e-05,
+      "loss": 1.2145,
+      "step": 1387
+    },
+    {
+      "epoch": 1.526952695269527,
+      "grad_norm": 2.375,
+      "learning_rate": 3.938280675973549e-05,
+      "loss": 1.0888,
+      "step": 1388
+    },
+    {
+      "epoch": 1.528052805280528,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.935341660543718e-05,
+      "loss": 1.0338,
+      "step": 1389
+    },
+    {
+      "epoch": 1.529152915291529,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.932402645113887e-05,
+      "loss": 1.517,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5302530253025304,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.929463629684056e-05,
+      "loss": 1.0935,
+      "step": 1391
+    },
+    {
+      "epoch": 1.5313531353135312,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.926524614254225e-05,
+      "loss": 1.0982,
+      "step": 1392
+    },
+    {
+      "epoch": 1.5324532453245325,
+      "grad_norm": 2.375,
+      "learning_rate": 3.923585598824394e-05,
+      "loss": 1.2742,
+      "step": 1393
+    },
+    {
+      "epoch": 1.5335533553355336,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.920646583394563e-05,
+      "loss": 1.0714,
+      "step": 1394
+    },
+    {
+      "epoch": 1.5346534653465347,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.9177075679647324e-05,
+      "loss": 1.3674,
+      "step": 1395
+    },
+    {
+      "epoch": 1.5357535753575358,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.9147685525349014e-05,
+      "loss": 1.3619,
+      "step": 1396
+    },
+    {
+      "epoch": 1.5368536853685368,
+      "grad_norm": 2.625,
+      "learning_rate": 3.91182953710507e-05,
+      "loss": 1.3988,
+      "step": 1397
+    },
+    {
+      "epoch": 1.537953795379538,
+      "grad_norm": 4.0,
+      "learning_rate": 3.9088905216752395e-05,
+      "loss": 1.3786,
+      "step": 1398
+    },
+    {
+      "epoch": 1.539053905390539,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.905951506245408e-05,
+      "loss": 1.0037,
+      "step": 1399
+    },
+    {
+      "epoch": 1.5401540154015403,
+      "grad_norm": 2.375,
+      "learning_rate": 3.9030124908155776e-05,
+      "loss": 1.2625,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5412541254125411,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.900073475385746e-05,
+      "loss": 1.173,
+      "step": 1401
+    },
+    {
+      "epoch": 1.5423542354235424,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.897134459955915e-05,
+      "loss": 1.2469,
+      "step": 1402
+    },
+    {
+      "epoch": 1.5434543454345433,
+      "grad_norm": 2.375,
+      "learning_rate": 3.894195444526084e-05,
+      "loss": 1.3606,
+      "step": 1403
+    },
+    {
+      "epoch": 1.5445544554455446,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.891256429096253e-05,
+      "loss": 1.1695,
+      "step": 1404
+    },
+    {
+      "epoch": 1.5456545654565457,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.888317413666422e-05,
+      "loss": 1.1132,
+      "step": 1405
+    },
+    {
+      "epoch": 1.5467546754675467,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.8853783982365906e-05,
+      "loss": 1.1727,
+      "step": 1406
+    },
+    {
+      "epoch": 1.5478547854785478,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.88243938280676e-05,
+      "loss": 1.2232,
+      "step": 1407
+    },
+    {
+      "epoch": 1.548954895489549,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.879500367376929e-05,
+      "loss": 1.1348,
+      "step": 1408
+    },
+    {
+      "epoch": 1.5500550055005502,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.8765613519470984e-05,
+      "loss": 1.4215,
+      "step": 1409
+    },
+    {
+      "epoch": 1.551155115511551,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.873622336517267e-05,
+      "loss": 1.2157,
+      "step": 1410
+    },
+    {
+      "epoch": 1.5522552255225524,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.870683321087436e-05,
+      "loss": 1.1548,
+      "step": 1411
+    },
+    {
+      "epoch": 1.5533553355335532,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.867744305657605e-05,
+      "loss": 1.0821,
+      "step": 1412
+    },
+    {
+      "epoch": 1.5544554455445545,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.864805290227774e-05,
+      "loss": 1.2611,
+      "step": 1413
+    },
+    {
+      "epoch": 1.5555555555555556,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.861866274797943e-05,
+      "loss": 1.1675,
+      "step": 1414
+    },
+    {
+      "epoch": 1.5566556655665567,
+      "grad_norm": 2.5,
+      "learning_rate": 3.858927259368112e-05,
+      "loss": 1.0726,
+      "step": 1415
+    },
+    {
+      "epoch": 1.5577557755775577,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.855988243938281e-05,
+      "loss": 1.1488,
+      "step": 1416
+    },
+    {
+      "epoch": 1.5588558855885588,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.85304922850845e-05,
+      "loss": 1.1306,
+      "step": 1417
+    },
+    {
+      "epoch": 1.55995599559956,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.850110213078619e-05,
+      "loss": 1.1276,
+      "step": 1418
+    },
+    {
+      "epoch": 1.561056105610561,
+      "grad_norm": 2.5,
+      "learning_rate": 3.8471711976487876e-05,
+      "loss": 1.2062,
+      "step": 1419
+    },
+    {
+      "epoch": 1.5621562156215623,
+      "grad_norm": 2.25,
+      "learning_rate": 3.844232182218957e-05,
+      "loss": 1.2177,
+      "step": 1420
+    },
+    {
+      "epoch": 1.5632563256325631,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.8412931667891257e-05,
+      "loss": 1.1128,
+      "step": 1421
+    },
+    {
+      "epoch": 1.5643564356435644,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.838354151359295e-05,
+      "loss": 1.1548,
+      "step": 1422
+    },
+    {
+      "epoch": 1.5654565456545655,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.835415135929464e-05,
+      "loss": 1.5635,
+      "step": 1423
+    },
+    {
+      "epoch": 1.5665566556655666,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.832476120499633e-05,
+      "loss": 1.0381,
+      "step": 1424
+    },
+    {
+      "epoch": 1.5676567656765676,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.829537105069802e-05,
+      "loss": 1.1078,
+      "step": 1425
+    },
+    {
+      "epoch": 1.5687568756875687,
+      "grad_norm": 2.5,
+      "learning_rate": 3.826598089639971e-05,
+      "loss": 1.4226,
+      "step": 1426
+    },
+    {
+      "epoch": 1.56985698569857,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.82365907421014e-05,
+      "loss": 1.219,
+      "step": 1427
+    },
+    {
+      "epoch": 1.5709570957095709,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.820720058780309e-05,
+      "loss": 1.0448,
+      "step": 1428
+    },
+    {
+      "epoch": 1.5720572057205722,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.817781043350478e-05,
+      "loss": 1.223,
+      "step": 1429
+    },
+    {
+      "epoch": 1.573157315731573,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.814842027920647e-05,
+      "loss": 1.0755,
+      "step": 1430
+    },
+    {
+      "epoch": 1.5742574257425743,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.811903012490816e-05,
+      "loss": 1.332,
+      "step": 1431
+    },
+    {
+      "epoch": 1.5753575357535754,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.808963997060985e-05,
+      "loss": 1.1,
+      "step": 1432
+    },
+    {
+      "epoch": 1.5764576457645765,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.8060249816311536e-05,
+      "loss": 1.2519,
+      "step": 1433
+    },
+    {
+      "epoch": 1.5775577557755776,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.803085966201323e-05,
+      "loss": 1.3513,
+      "step": 1434
+    },
+    {
+      "epoch": 1.5786578657865786,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.800146950771492e-05,
+      "loss": 1.3471,
+      "step": 1435
+    },
+    {
+      "epoch": 1.5797579757975797,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.797207935341661e-05,
+      "loss": 1.1264,
+      "step": 1436
+    },
+    {
+      "epoch": 1.5808580858085808,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.79426891991183e-05,
+      "loss": 1.2462,
+      "step": 1437
+    },
+    {
+      "epoch": 1.581958195819582,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.791329904481999e-05,
+      "loss": 1.1093,
+      "step": 1438
+    },
+    {
+      "epoch": 1.583058305830583,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.788390889052168e-05,
+      "loss": 1.0633,
+      "step": 1439
+    },
+    {
+      "epoch": 1.5841584158415842,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.785451873622337e-05,
+      "loss": 1.1621,
+      "step": 1440
+    },
+    {
+      "epoch": 1.5852585258525853,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.782512858192506e-05,
+      "loss": 1.188,
+      "step": 1441
+    },
+    {
+      "epoch": 1.5863586358635864,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.7795738427626744e-05,
+      "loss": 1.2578,
+      "step": 1442
+    },
+    {
+      "epoch": 1.5874587458745875,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.776634827332844e-05,
+      "loss": 1.4288,
+      "step": 1443
+    },
+    {
+      "epoch": 1.5885588558855885,
+      "grad_norm": 2.375,
+      "learning_rate": 3.7736958119030125e-05,
+      "loss": 1.0755,
+      "step": 1444
+    },
+    {
+      "epoch": 1.5896589658965896,
+      "grad_norm": 2.5,
+      "learning_rate": 3.770756796473182e-05,
+      "loss": 1.1749,
+      "step": 1445
+    },
+    {
+      "epoch": 1.5907590759075907,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.7678177810433506e-05,
+      "loss": 1.1452,
+      "step": 1446
+    },
+    {
+      "epoch": 1.591859185918592,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.7648787656135196e-05,
+      "loss": 1.1905,
+      "step": 1447
+    },
+    {
+      "epoch": 1.5929592959295928,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.761939750183689e-05,
+      "loss": 1.159,
+      "step": 1448
+    },
+    {
+      "epoch": 1.5940594059405941,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.759000734753858e-05,
+      "loss": 1.1143,
+      "step": 1449
+    },
+    {
+      "epoch": 1.595159515951595,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.756061719324027e-05,
+      "loss": 0.9495,
+      "step": 1450
+    },
+    {
+      "epoch": 1.5962596259625963,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.753122703894196e-05,
+      "loss": 1.0586,
+      "step": 1451
+    },
+    {
+      "epoch": 1.5973597359735974,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.750183688464365e-05,
+      "loss": 1.0861,
+      "step": 1452
+    },
+    {
+      "epoch": 1.5984598459845984,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.747244673034533e-05,
+      "loss": 1.1171,
+      "step": 1453
+    },
+    {
+      "epoch": 1.5995599559955995,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.744305657604703e-05,
+      "loss": 1.1062,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6006600660066006,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.7413666421748713e-05,
+      "loss": 1.2746,
+      "step": 1455
+    },
+    {
+      "epoch": 1.601760176017602,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.738427626745041e-05,
+      "loss": 1.2563,
+      "step": 1456
+    },
+    {
+      "epoch": 1.6028602860286028,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.7354886113152094e-05,
+      "loss": 1.2414,
+      "step": 1457
+    },
+    {
+      "epoch": 1.603960396039604,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.7325495958853785e-05,
+      "loss": 1.0779,
+      "step": 1458
+    },
+    {
+      "epoch": 1.605060506050605,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.7296105804555475e-05,
+      "loss": 1.2416,
+      "step": 1459
+    },
+    {
+      "epoch": 1.6061606160616062,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.7266715650257166e-05,
+      "loss": 1.4363,
+      "step": 1460
+    },
+    {
+      "epoch": 1.6072607260726073,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.7237325495958856e-05,
+      "loss": 1.3637,
+      "step": 1461
+    },
+    {
+      "epoch": 1.6083608360836084,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.720793534166055e-05,
+      "loss": 1.1108,
+      "step": 1462
+    },
+    {
+      "epoch": 1.6094609460946094,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.717854518736224e-05,
+      "loss": 1.3172,
+      "step": 1463
+    },
+    {
+      "epoch": 1.6105610561056105,
+      "grad_norm": 2.625,
+      "learning_rate": 3.714915503306393e-05,
+      "loss": 1.1269,
+      "step": 1464
+    },
+    {
+      "epoch": 1.6116611661166118,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.711976487876562e-05,
+      "loss": 1.319,
+      "step": 1465
+    },
+    {
+      "epoch": 1.6127612761276127,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.709037472446731e-05,
+      "loss": 1.0529,
+      "step": 1466
+    },
+    {
+      "epoch": 1.613861386138614,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.7060984570169e-05,
+      "loss": 1.3648,
+      "step": 1467
+    },
+    {
+      "epoch": 1.6149614961496148,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.703159441587068e-05,
+      "loss": 1.1279,
+      "step": 1468
+    },
+    {
+      "epoch": 1.6160616061606161,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.7002204261572374e-05,
+      "loss": 1.2482,
+      "step": 1469
+    },
+    {
+      "epoch": 1.6171617161716172,
+      "grad_norm": 2.5,
+      "learning_rate": 3.6972814107274064e-05,
+      "loss": 1.1917,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6182618261826183,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.6943423952975755e-05,
+      "loss": 1.2901,
+      "step": 1471
+    },
+    {
+      "epoch": 1.6193619361936193,
+      "grad_norm": 2.125,
+      "learning_rate": 3.6914033798677445e-05,
+      "loss": 1.3877,
+      "step": 1472
+    },
+    {
+      "epoch": 1.6204620462046204,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.6884643644379136e-05,
+      "loss": 1.0429,
+      "step": 1473
+    },
+    {
+      "epoch": 1.6215621562156217,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.6855253490080826e-05,
+      "loss": 1.2245,
+      "step": 1474
+    },
+    {
+      "epoch": 1.6226622662266226,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.682586333578252e-05,
+      "loss": 1.3594,
+      "step": 1475
+    },
+    {
+      "epoch": 1.6237623762376239,
+      "grad_norm": 2.9375,
+      "learning_rate": 3.679647318148421e-05,
+      "loss": 1.2449,
+      "step": 1476
+    },
+    {
+      "epoch": 1.6248624862486247,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.67670830271859e-05,
+      "loss": 1.2531,
+      "step": 1477
+    },
+    {
+      "epoch": 1.625962596259626,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.673769287288758e-05,
+      "loss": 1.3515,
+      "step": 1478
+    },
+    {
+      "epoch": 1.627062706270627,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.670830271858928e-05,
+      "loss": 1.471,
+      "step": 1479
+    },
+    {
+      "epoch": 1.6281628162816282,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.667891256429096e-05,
+      "loss": 1.1144,
+      "step": 1480
+    },
+    {
+      "epoch": 1.6292629262926293,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.664952240999266e-05,
+      "loss": 1.1425,
+      "step": 1481
+    },
+    {
+      "epoch": 1.6303630363036303,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.6620132255694343e-05,
+      "loss": 1.2757,
+      "step": 1482
+    },
+    {
+      "epoch": 1.6314631463146316,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.6590742101396034e-05,
+      "loss": 1.0956,
+      "step": 1483
+    },
+    {
+      "epoch": 1.6325632563256325,
+      "grad_norm": 2.125,
+      "learning_rate": 3.6561351947097724e-05,
+      "loss": 1.3943,
+      "step": 1484
+    },
+    {
+      "epoch": 1.6336633663366338,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.6531961792799415e-05,
+      "loss": 1.3176,
+      "step": 1485
+    },
+    {
+      "epoch": 1.6347634763476346,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.6502571638501105e-05,
+      "loss": 1.3096,
+      "step": 1486
+    },
+    {
+      "epoch": 1.635863586358636,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.6473181484202796e-05,
+      "loss": 1.0698,
+      "step": 1487
+    },
+    {
+      "epoch": 1.636963696369637,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.6443791329904486e-05,
+      "loss": 1.4113,
+      "step": 1488
+    },
+    {
+      "epoch": 1.638063806380638,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.641440117560617e-05,
+      "loss": 0.9938,
+      "step": 1489
+    },
+    {
+      "epoch": 1.6391639163916392,
+      "grad_norm": 2.25,
+      "learning_rate": 3.638501102130787e-05,
+      "loss": 0.9903,
+      "step": 1490
+    },
+    {
+      "epoch": 1.6402640264026402,
+      "grad_norm": 2.25,
+      "learning_rate": 3.635562086700955e-05,
+      "loss": 1.3115,
+      "step": 1491
+    },
+    {
+      "epoch": 1.6413641364136413,
+      "grad_norm": 2.625,
+      "learning_rate": 3.632623071271125e-05,
+      "loss": 1.2446,
+      "step": 1492
+    },
+    {
+      "epoch": 1.6424642464246424,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.629684055841293e-05,
+      "loss": 1.3085,
+      "step": 1493
+    },
+    {
+      "epoch": 1.6435643564356437,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.626745040411462e-05,
+      "loss": 1.3824,
+      "step": 1494
+    },
+    {
+      "epoch": 1.6446644664466445,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.623806024981631e-05,
+      "loss": 1.2736,
+      "step": 1495
+    },
+    {
+      "epoch": 1.6457645764576458,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.6208670095518004e-05,
+      "loss": 1.1342,
+      "step": 1496
+    },
+    {
+      "epoch": 1.6468646864686467,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.6179279941219694e-05,
+      "loss": 1.3361,
+      "step": 1497
+    },
+    {
+      "epoch": 1.647964796479648,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.6149889786921385e-05,
+      "loss": 1.0793,
+      "step": 1498
+    },
+    {
+      "epoch": 1.649064906490649,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.6120499632623075e-05,
+      "loss": 1.1777,
+      "step": 1499
+    },
+    {
+      "epoch": 1.6501650165016502,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.6091109478324766e-05,
+      "loss": 1.2501,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6512651265126512,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.6061719324026456e-05,
+      "loss": 1.4469,
+      "step": 1501
+    },
+    {
+      "epoch": 1.6523652365236523,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.603232916972814e-05,
+      "loss": 1.2929,
+      "step": 1502
+    },
+    {
+      "epoch": 1.6534653465346536,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.600293901542984e-05,
+      "loss": 0.9163,
+      "step": 1503
+    },
+    {
+      "epoch": 1.6545654565456545,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.597354886113152e-05,
+      "loss": 1.1553,
+      "step": 1504
+    },
+    {
+      "epoch": 1.6556655665566558,
+      "grad_norm": 2.5,
+      "learning_rate": 3.594415870683321e-05,
+      "loss": 1.1433,
+      "step": 1505
+    },
+    {
+      "epoch": 1.6567656765676566,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.59147685525349e-05,
+      "loss": 1.2179,
+      "step": 1506
+    },
+    {
+      "epoch": 1.657865786578658,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.588537839823659e-05,
+      "loss": 1.1712,
+      "step": 1507
+    },
+    {
+      "epoch": 1.658965896589659,
+      "grad_norm": 2.375,
+      "learning_rate": 3.585598824393828e-05,
+      "loss": 1.246,
+      "step": 1508
+    },
+    {
+      "epoch": 1.66006600660066,
+      "grad_norm": 2.625,
+      "learning_rate": 3.5826598089639974e-05,
+      "loss": 1.2963,
+      "step": 1509
+    },
+    {
+      "epoch": 1.6611661166116611,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.5797207935341664e-05,
+      "loss": 1.2695,
+      "step": 1510
+    },
+    {
+      "epoch": 1.6622662266226622,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.5767817781043355e-05,
+      "loss": 1.2565,
+      "step": 1511
+    },
+    {
+      "epoch": 1.6633663366336635,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.5738427626745045e-05,
+      "loss": 1.2391,
+      "step": 1512
+    },
+    {
+      "epoch": 1.6644664466446644,
+      "grad_norm": 2.5,
+      "learning_rate": 3.5709037472446736e-05,
+      "loss": 1.1843,
+      "step": 1513
+    },
+    {
+      "epoch": 1.6655665566556657,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.567964731814842e-05,
+      "loss": 1.3728,
+      "step": 1514
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.5650257163850117e-05,
+      "loss": 1.0671,
+      "step": 1515
+    },
+    {
+      "epoch": 1.6677667766776678,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.56208670095518e-05,
+      "loss": 1.369,
+      "step": 1516
+    },
+    {
+      "epoch": 1.668866886688669,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.55914768552535e-05,
+      "loss": 1.1731,
+      "step": 1517
+    },
+    {
+      "epoch": 1.66996699669967,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.556208670095518e-05,
+      "loss": 1.1016,
+      "step": 1518
+    },
+    {
+      "epoch": 1.671067106710671,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.553269654665687e-05,
+      "loss": 1.1873,
+      "step": 1519
+    },
+    {
+      "epoch": 1.6721672167216721,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.550330639235856e-05,
+      "loss": 1.1535,
+      "step": 1520
+    },
+    {
+      "epoch": 1.6732673267326734,
+      "grad_norm": 2.375,
+      "learning_rate": 3.547391623806025e-05,
+      "loss": 0.9728,
+      "step": 1521
+    },
+    {
+      "epoch": 1.6743674367436743,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.544452608376194e-05,
+      "loss": 1.2827,
+      "step": 1522
+    },
+    {
+      "epoch": 1.6754675467546756,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.5415135929463634e-05,
+      "loss": 1.1011,
+      "step": 1523
+    },
+    {
+      "epoch": 1.6765676567656764,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.5385745775165324e-05,
+      "loss": 1.249,
+      "step": 1524
+    },
+    {
+      "epoch": 1.6776677667766777,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.535635562086701e-05,
+      "loss": 1.3856,
+      "step": 1525
+    },
+    {
+      "epoch": 1.6787678767876788,
+      "grad_norm": 2.625,
+      "learning_rate": 3.5326965466568705e-05,
+      "loss": 1.0956,
+      "step": 1526
+    },
+    {
+      "epoch": 1.6798679867986799,
+      "grad_norm": 2.5,
+      "learning_rate": 3.529757531227039e-05,
+      "loss": 1.1473,
+      "step": 1527
+    },
+    {
+      "epoch": 1.680968096809681,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.5268185157972086e-05,
+      "loss": 1.0785,
+      "step": 1528
+    },
+    {
+      "epoch": 1.682068206820682,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.523879500367377e-05,
+      "loss": 1.3808,
+      "step": 1529
+    },
+    {
+      "epoch": 1.6831683168316833,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.520940484937546e-05,
+      "loss": 1.4087,
+      "step": 1530
+    },
+    {
+      "epoch": 1.6842684268426842,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.518001469507715e-05,
+      "loss": 1.284,
+      "step": 1531
+    },
+    {
+      "epoch": 1.6853685368536855,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.515062454077884e-05,
+      "loss": 1.2225,
+      "step": 1532
+    },
+    {
+      "epoch": 1.6864686468646863,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.512123438648053e-05,
+      "loss": 1.1956,
+      "step": 1533
+    },
+    {
+      "epoch": 1.6875687568756876,
+      "grad_norm": 2.625,
+      "learning_rate": 3.509184423218222e-05,
+      "loss": 0.9187,
+      "step": 1534
+    },
+    {
+      "epoch": 1.6886688668866887,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.506245407788391e-05,
+      "loss": 1.1082,
+      "step": 1535
+    },
+    {
+      "epoch": 1.6897689768976898,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.50330639235856e-05,
+      "loss": 1.1982,
+      "step": 1536
+    },
+    {
+      "epoch": 1.6908690869086909,
+      "grad_norm": 2.75,
+      "learning_rate": 3.5003673769287294e-05,
+      "loss": 1.1606,
+      "step": 1537
+    },
+    {
+      "epoch": 1.691969196919692,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.497428361498898e-05,
+      "loss": 1.1692,
+      "step": 1538
+    },
+    {
+      "epoch": 1.693069306930693,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.4944893460690675e-05,
+      "loss": 1.2153,
+      "step": 1539
+    },
+    {
+      "epoch": 1.694169416941694,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.491550330639236e-05,
+      "loss": 1.2613,
+      "step": 1540
+    },
+    {
+      "epoch": 1.6952695269526954,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.488611315209405e-05,
+      "loss": 1.4035,
+      "step": 1541
+    },
+    {
+      "epoch": 1.6963696369636962,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.485672299779574e-05,
+      "loss": 1.181,
+      "step": 1542
+    },
+    {
+      "epoch": 1.6974697469746975,
+      "grad_norm": 2.5,
+      "learning_rate": 3.482733284349743e-05,
+      "loss": 1.0401,
+      "step": 1543
+    },
+    {
+      "epoch": 1.6985698569856986,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.479794268919912e-05,
+      "loss": 1.1333,
+      "step": 1544
+    },
+    {
+      "epoch": 1.6996699669966997,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.476855253490081e-05,
+      "loss": 1.1148,
+      "step": 1545
+    },
+    {
+      "epoch": 1.7007700770077008,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.47391623806025e-05,
+      "loss": 1.0417,
+      "step": 1546
+    },
+    {
+      "epoch": 1.7018701870187019,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.470977222630419e-05,
+      "loss": 1.2791,
+      "step": 1547
+    },
+    {
+      "epoch": 1.702970297029703,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.468038207200588e-05,
+      "loss": 1.2158,
+      "step": 1548
+    },
+    {
+      "epoch": 1.704070407040704,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.465099191770757e-05,
+      "loss": 1.1986,
+      "step": 1549
+    },
+    {
+      "epoch": 1.7051705170517053,
+      "grad_norm": 2.625,
+      "learning_rate": 3.4621601763409264e-05,
+      "loss": 1.4321,
+      "step": 1550
+    },
+    {
+      "epoch": 1.7062706270627062,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.4592211609110954e-05,
+      "loss": 1.1341,
+      "step": 1551
+    },
+    {
+      "epoch": 1.7073707370737075,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.456282145481264e-05,
+      "loss": 1.0625,
+      "step": 1552
+    },
+    {
+      "epoch": 1.7084708470847083,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.453343130051433e-05,
+      "loss": 1.256,
+      "step": 1553
+    },
+    {
+      "epoch": 1.7095709570957096,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.450404114621602e-05,
+      "loss": 1.2314,
+      "step": 1554
+    },
+    {
+      "epoch": 1.7106710671067107,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.447465099191771e-05,
+      "loss": 1.1262,
+      "step": 1555
+    },
+    {
+      "epoch": 1.7117711771177118,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.44452608376194e-05,
+      "loss": 1.1461,
+      "step": 1556
+    },
+    {
+      "epoch": 1.7128712871287128,
+      "grad_norm": 2.5,
+      "learning_rate": 3.441587068332109e-05,
+      "loss": 1.5179,
+      "step": 1557
+    },
+    {
+      "epoch": 1.713971397139714,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.438648052902278e-05,
+      "loss": 1.062,
+      "step": 1558
+    },
+    {
+      "epoch": 1.7150715071507152,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.435709037472447e-05,
+      "loss": 0.8953,
+      "step": 1559
+    },
+    {
+      "epoch": 1.716171617161716,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.432770022042616e-05,
+      "loss": 1.0676,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7172717271727174,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.4298310066127846e-05,
+      "loss": 1.2773,
+      "step": 1561
+    },
+    {
+      "epoch": 1.7183718371837182,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.426891991182954e-05,
+      "loss": 1.2314,
+      "step": 1562
+    },
+    {
+      "epoch": 1.7194719471947195,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.423952975753123e-05,
+      "loss": 1.2141,
+      "step": 1563
+    },
+    {
+      "epoch": 1.7205720572057206,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.4210139603232924e-05,
+      "loss": 1.205,
+      "step": 1564
+    },
+    {
+      "epoch": 1.7216721672167217,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.418074944893461e-05,
+      "loss": 1.413,
+      "step": 1565
+    },
+    {
+      "epoch": 1.7227722772277227,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.41513592946363e-05,
+      "loss": 1.5004,
+      "step": 1566
+    },
+    {
+      "epoch": 1.7238723872387238,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.412196914033799e-05,
+      "loss": 1.168,
+      "step": 1567
+    },
+    {
+      "epoch": 1.7249724972497251,
+      "grad_norm": 2.375,
+      "learning_rate": 3.409257898603968e-05,
+      "loss": 1.2436,
+      "step": 1568
+    },
+    {
+      "epoch": 1.726072607260726,
+      "grad_norm": 2.5,
+      "learning_rate": 3.406318883174137e-05,
+      "loss": 1.1563,
+      "step": 1569
+    },
+    {
+      "epoch": 1.7271727172717273,
+      "grad_norm": 2.375,
+      "learning_rate": 3.403379867744306e-05,
+      "loss": 1.4049,
+      "step": 1570
+    },
+    {
+      "epoch": 1.7282728272827281,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.400440852314475e-05,
+      "loss": 1.3548,
+      "step": 1571
+    },
+    {
+      "epoch": 1.7293729372937294,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.3975018368846435e-05,
+      "loss": 1.3346,
+      "step": 1572
+    },
+    {
+      "epoch": 1.7304730473047305,
+      "grad_norm": 2.375,
+      "learning_rate": 3.394562821454813e-05,
+      "loss": 1.2033,
+      "step": 1573
+    },
+    {
+      "epoch": 1.7315731573157316,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.3916238060249816e-05,
+      "loss": 1.1843,
+      "step": 1574
+    },
+    {
+      "epoch": 1.7326732673267327,
+      "grad_norm": 2.125,
+      "learning_rate": 3.388684790595151e-05,
+      "loss": 1.2961,
+      "step": 1575
+    },
+    {
+      "epoch": 1.7337733773377337,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.38574577516532e-05,
+      "loss": 1.3601,
+      "step": 1576
+    },
+    {
+      "epoch": 1.734873487348735,
+      "grad_norm": 2.375,
+      "learning_rate": 3.382806759735489e-05,
+      "loss": 1.1102,
+      "step": 1577
+    },
+    {
+      "epoch": 1.7359735973597359,
+      "grad_norm": 2.625,
+      "learning_rate": 3.379867744305658e-05,
+      "loss": 1.3475,
+      "step": 1578
+    },
+    {
+      "epoch": 1.7370737073707372,
+      "grad_norm": 2.375,
+      "learning_rate": 3.376928728875827e-05,
+      "loss": 1.1934,
+      "step": 1579
+    },
+    {
+      "epoch": 1.738173817381738,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.373989713445996e-05,
+      "loss": 1.0664,
+      "step": 1580
+    },
+    {
+      "epoch": 1.7392739273927393,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.371050698016165e-05,
+      "loss": 1.3347,
+      "step": 1581
+    },
+    {
+      "epoch": 1.7403740374037404,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.368111682586334e-05,
+      "loss": 1.3882,
+      "step": 1582
+    },
+    {
+      "epoch": 1.7414741474147415,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.365172667156503e-05,
+      "loss": 1.0899,
+      "step": 1583
+    },
+    {
+      "epoch": 1.7425742574257426,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.362233651726672e-05,
+      "loss": 1.1579,
+      "step": 1584
+    },
+    {
+      "epoch": 1.7436743674367436,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.359294636296841e-05,
+      "loss": 1.3066,
+      "step": 1585
+    },
+    {
+      "epoch": 1.7447744774477447,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.35635562086701e-05,
+      "loss": 1.2844,
+      "step": 1586
+    },
+    {
+      "epoch": 1.7458745874587458,
+      "grad_norm": 2.625,
+      "learning_rate": 3.3534166054371785e-05,
+      "loss": 0.9718,
+      "step": 1587
+    },
+    {
+      "epoch": 1.746974697469747,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.3504775900073476e-05,
+      "loss": 1.0999,
+      "step": 1588
+    },
+    {
+      "epoch": 1.748074807480748,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.3475385745775166e-05,
+      "loss": 1.0059,
+      "step": 1589
+    },
+    {
+      "epoch": 1.7491749174917492,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.344599559147686e-05,
+      "loss": 1.2678,
+      "step": 1590
+    },
+    {
+      "epoch": 1.7502750275027503,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.341660543717855e-05,
+      "loss": 1.1554,
+      "step": 1591
+    },
+    {
+      "epoch": 1.7513751375137514,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.338721528288024e-05,
+      "loss": 1.2437,
+      "step": 1592
+    },
+    {
+      "epoch": 1.7524752475247525,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.335782512858193e-05,
+      "loss": 1.0923,
+      "step": 1593
+    },
+    {
+      "epoch": 1.7535753575357536,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.332843497428362e-05,
+      "loss": 1.1158,
+      "step": 1594
+    },
+    {
+      "epoch": 1.7546754675467546,
+      "grad_norm": 2.125,
+      "learning_rate": 3.329904481998531e-05,
+      "loss": 1.2843,
+      "step": 1595
+    },
+    {
+      "epoch": 1.7557755775577557,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.3269654665687e-05,
+      "loss": 1.2662,
+      "step": 1596
+    },
+    {
+      "epoch": 1.756875687568757,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.3240264511388684e-05,
+      "loss": 0.9132,
+      "step": 1597
+    },
+    {
+      "epoch": 1.7579757975797579,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.321087435709038e-05,
+      "loss": 1.2893,
+      "step": 1598
+    },
+    {
+      "epoch": 1.7590759075907592,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.3181484202792065e-05,
+      "loss": 1.4667,
+      "step": 1599
+    },
+    {
+      "epoch": 1.76017601760176,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.315209404849376e-05,
+      "loss": 1.1352,
+      "step": 1600
+    },
+    {
+      "epoch": 1.7612761276127613,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.3122703894195446e-05,
+      "loss": 1.1806,
+      "step": 1601
+    },
+    {
+      "epoch": 1.7623762376237624,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.3093313739897136e-05,
+      "loss": 1.017,
+      "step": 1602
+    },
+    {
+      "epoch": 1.7634763476347635,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.306392358559883e-05,
+      "loss": 1.135,
+      "step": 1603
+    },
+    {
+      "epoch": 1.7645764576457645,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.303453343130052e-05,
+      "loss": 1.365,
+      "step": 1604
+    },
+    {
+      "epoch": 1.7656765676567656,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.300514327700221e-05,
+      "loss": 1.3115,
+      "step": 1605
+    },
+    {
+      "epoch": 1.766776677667767,
+      "grad_norm": 2.625,
+      "learning_rate": 3.29757531227039e-05,
+      "loss": 1.22,
+      "step": 1606
+    },
+    {
+      "epoch": 1.7678767876787678,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.294636296840559e-05,
+      "loss": 1.0921,
+      "step": 1607
+    },
+    {
+      "epoch": 1.768976897689769,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.291697281410727e-05,
+      "loss": 1.1911,
+      "step": 1608
+    },
+    {
+      "epoch": 1.77007700770077,
+      "grad_norm": 2.625,
+      "learning_rate": 3.288758265980897e-05,
+      "loss": 1.1434,
+      "step": 1609
+    },
+    {
+      "epoch": 1.7711771177117712,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.2858192505510653e-05,
+      "loss": 1.1679,
+      "step": 1610
+    },
+    {
+      "epoch": 1.7722772277227723,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.282880235121235e-05,
+      "loss": 1.1204,
+      "step": 1611
+    },
+    {
+      "epoch": 1.7733773377337734,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.2799412196914034e-05,
+      "loss": 1.2861,
+      "step": 1612
+    },
+    {
+      "epoch": 1.7744774477447744,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.2770022042615725e-05,
+      "loss": 1.1229,
+      "step": 1613
+    },
+    {
+      "epoch": 1.7755775577557755,
+      "grad_norm": 2.375,
+      "learning_rate": 3.2740631888317415e-05,
+      "loss": 1.3173,
+      "step": 1614
+    },
+    {
+      "epoch": 1.7766776677667768,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.2711241734019106e-05,
+      "loss": 1.1328,
+      "step": 1615
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.2681851579720796e-05,
+      "loss": 1.085,
+      "step": 1616
+    },
+    {
+      "epoch": 1.778877887788779,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.265246142542249e-05,
+      "loss": 1.1556,
+      "step": 1617
+    },
+    {
+      "epoch": 1.7799779977997798,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.262307127112418e-05,
+      "loss": 1.1751,
+      "step": 1618
+    },
+    {
+      "epoch": 1.7810781078107811,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.259368111682587e-05,
+      "loss": 1.2924,
+      "step": 1619
+    },
+    {
+      "epoch": 1.7821782178217822,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.256429096252756e-05,
+      "loss": 1.1055,
+      "step": 1620
+    },
+    {
+      "epoch": 1.7832783278327833,
+      "grad_norm": 2.625,
+      "learning_rate": 3.253490080822924e-05,
+      "loss": 1.3167,
+      "step": 1621
+    },
+    {
+      "epoch": 1.7843784378437844,
+      "grad_norm": 2.5,
+      "learning_rate": 3.250551065393094e-05,
+      "loss": 1.5045,
+      "step": 1622
+    },
+    {
+      "epoch": 1.7854785478547854,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.247612049963262e-05,
+      "loss": 1.0809,
+      "step": 1623
+    },
+    {
+      "epoch": 1.7865786578657867,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.2446730345334314e-05,
+      "loss": 1.0763,
+      "step": 1624
+    },
+    {
+      "epoch": 1.7876787678767876,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.2417340191036004e-05,
+      "loss": 1.273,
+      "step": 1625
+    },
+    {
+      "epoch": 1.7887788778877889,
+      "grad_norm": 2.0,
+      "learning_rate": 3.2387950036737695e-05,
+      "loss": 1.126,
+      "step": 1626
+    },
+    {
+      "epoch": 1.7898789878987897,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.2358559882439385e-05,
+      "loss": 1.2906,
+      "step": 1627
+    },
+    {
+      "epoch": 1.790979097909791,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.2329169728141076e-05,
+      "loss": 1.0994,
+      "step": 1628
+    },
+    {
+      "epoch": 1.7920792079207921,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.2299779573842766e-05,
+      "loss": 1.1772,
+      "step": 1629
+    },
+    {
+      "epoch": 1.7931793179317932,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.227038941954446e-05,
+      "loss": 1.0692,
+      "step": 1630
+    },
+    {
+      "epoch": 1.7942794279427943,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.224099926524615e-05,
+      "loss": 1.2394,
+      "step": 1631
+    },
+    {
+      "epoch": 1.7953795379537953,
+      "grad_norm": 2.5,
+      "learning_rate": 3.221160911094784e-05,
+      "loss": 1.4307,
+      "step": 1632
+    },
+    {
+      "epoch": 1.7964796479647966,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.218221895664952e-05,
+      "loss": 1.4457,
+      "step": 1633
+    },
+    {
+      "epoch": 1.7975797579757975,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.215282880235122e-05,
+      "loss": 1.0984,
+      "step": 1634
+    },
+    {
+      "epoch": 1.7986798679867988,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.21234386480529e-05,
+      "loss": 1.1262,
+      "step": 1635
+    },
+    {
+      "epoch": 1.7997799779977997,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.209404849375459e-05,
+      "loss": 1.309,
+      "step": 1636
+    },
+    {
+      "epoch": 1.800880088008801,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.2064658339456284e-05,
+      "loss": 1.2252,
+      "step": 1637
+    },
+    {
+      "epoch": 1.801980198019802,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2035268185157974e-05,
+      "loss": 1.2133,
+      "step": 1638
+    },
+    {
+      "epoch": 1.803080308030803,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.2005878030859665e-05,
+      "loss": 1.3172,
+      "step": 1639
+    },
+    {
+      "epoch": 1.8041804180418042,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.1976487876561355e-05,
+      "loss": 1.3874,
+      "step": 1640
+    },
+    {
+      "epoch": 1.8052805280528053,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.1947097722263046e-05,
+      "loss": 1.0443,
+      "step": 1641
+    },
+    {
+      "epoch": 1.8063806380638063,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.1917707567964736e-05,
+      "loss": 1.1812,
+      "step": 1642
+    },
+    {
+      "epoch": 1.8074807480748074,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.1888317413666427e-05,
+      "loss": 1.3933,
+      "step": 1643
+    },
+    {
+      "epoch": 1.8085808580858087,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.185892725936811e-05,
+      "loss": 1.3339,
+      "step": 1644
+    },
+    {
+      "epoch": 1.8096809680968096,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.182953710506981e-05,
+      "loss": 1.2864,
+      "step": 1645
+    },
+    {
+      "epoch": 1.8107810781078109,
+      "grad_norm": 2.25,
+      "learning_rate": 3.180014695077149e-05,
+      "loss": 1.1447,
+      "step": 1646
+    },
+    {
+      "epoch": 1.811881188118812,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.177075679647319e-05,
+      "loss": 1.1564,
+      "step": 1647
+    },
+    {
+      "epoch": 1.812981298129813,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.174136664217487e-05,
+      "loss": 1.3902,
+      "step": 1648
+    },
+    {
+      "epoch": 1.814081408140814,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.171197648787656e-05,
+      "loss": 1.2058,
+      "step": 1649
+    },
+    {
+      "epoch": 1.8151815181518152,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.168258633357825e-05,
+      "loss": 1.2919,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8162816281628162,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.1653196179279944e-05,
+      "loss": 1.1101,
+      "step": 1651
+    },
+    {
+      "epoch": 1.8173817381738173,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.1623806024981634e-05,
+      "loss": 1.0016,
+      "step": 1652
+    },
+    {
+      "epoch": 1.8184818481848186,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.1594415870683325e-05,
+      "loss": 1.1192,
+      "step": 1653
+    },
+    {
+      "epoch": 1.8195819581958195,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.1565025716385015e-05,
+      "loss": 1.5158,
+      "step": 1654
+    },
+    {
+      "epoch": 1.8206820682068208,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.15356355620867e-05,
+      "loss": 1.1344,
+      "step": 1655
+    },
+    {
+      "epoch": 1.8217821782178216,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.1506245407788396e-05,
+      "loss": 1.2837,
+      "step": 1656
+    },
+    {
+      "epoch": 1.822882288228823,
+      "grad_norm": 2.25,
+      "learning_rate": 3.147685525349008e-05,
+      "loss": 1.1829,
+      "step": 1657
+    },
+    {
+      "epoch": 1.823982398239824,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.144746509919178e-05,
+      "loss": 1.2356,
+      "step": 1658
+    },
+    {
+      "epoch": 1.825082508250825,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.141807494489346e-05,
+      "loss": 1.0054,
+      "step": 1659
+    },
+    {
+      "epoch": 1.8261826182618262,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.138868479059515e-05,
+      "loss": 1.296,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8272827282728272,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.135929463629684e-05,
+      "loss": 1.0589,
+      "step": 1661
+    },
+    {
+      "epoch": 1.8283828382838285,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.132990448199853e-05,
+      "loss": 1.3331,
+      "step": 1662
+    },
+    {
+      "epoch": 1.8294829482948294,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.130051432770022e-05,
+      "loss": 1.2329,
+      "step": 1663
+    },
+    {
+      "epoch": 1.8305830583058307,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.1271124173401914e-05,
+      "loss": 1.1558,
+      "step": 1664
+    },
+    {
+      "epoch": 1.8316831683168315,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.1241734019103604e-05,
+      "loss": 1.5006,
+      "step": 1665
+    },
+    {
+      "epoch": 1.8327832783278328,
+      "grad_norm": 2.5,
+      "learning_rate": 3.1212343864805295e-05,
+      "loss": 1.1453,
+      "step": 1666
+    },
+    {
+      "epoch": 1.833883388338834,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.1182953710506985e-05,
+      "loss": 1.2051,
+      "step": 1667
+    },
+    {
+      "epoch": 1.834983498349835,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.1153563556208676e-05,
+      "loss": 1.4527,
+      "step": 1668
+    },
+    {
+      "epoch": 1.836083608360836,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.112417340191036e-05,
+      "loss": 1.4406,
+      "step": 1669
+    },
+    {
+      "epoch": 1.8371837183718371,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.109478324761205e-05,
+      "loss": 1.4381,
+      "step": 1670
+    },
+    {
+      "epoch": 1.8382838283828384,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.106539309331374e-05,
+      "loss": 1.1546,
+      "step": 1671
+    },
+    {
+      "epoch": 1.8393839383938393,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.103600293901543e-05,
+      "loss": 1.1383,
+      "step": 1672
+    },
+    {
+      "epoch": 1.8404840484048406,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.100661278471712e-05,
+      "loss": 0.9154,
+      "step": 1673
+    },
+    {
+      "epoch": 1.8415841584158414,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.097722263041881e-05,
+      "loss": 1.2372,
+      "step": 1674
+    },
+    {
+      "epoch": 1.8426842684268427,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.09478324761205e-05,
+      "loss": 1.0533,
+      "step": 1675
+    },
+    {
+      "epoch": 1.8437843784378438,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.091844232182219e-05,
+      "loss": 1.3827,
+      "step": 1676
+    },
+    {
+      "epoch": 1.844884488448845,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.088905216752388e-05,
+      "loss": 0.9401,
+      "step": 1677
+    },
+    {
+      "epoch": 1.845984598459846,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.0859662013225574e-05,
+      "loss": 1.2382,
+      "step": 1678
+    },
+    {
+      "epoch": 1.847084708470847,
+      "grad_norm": 2.375,
+      "learning_rate": 3.0830271858927264e-05,
+      "loss": 1.313,
+      "step": 1679
+    },
+    {
+      "epoch": 1.8481848184818483,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.080088170462895e-05,
+      "loss": 1.43,
+      "step": 1680
+    },
+    {
+      "epoch": 1.8492849284928492,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.0771491550330645e-05,
+      "loss": 1.2136,
+      "step": 1681
+    },
+    {
+      "epoch": 1.8503850385038505,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.074210139603233e-05,
+      "loss": 1.2278,
+      "step": 1682
+    },
+    {
+      "epoch": 1.8514851485148514,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.0712711241734026e-05,
+      "loss": 1.0048,
+      "step": 1683
+    },
+    {
+      "epoch": 1.8525852585258527,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.068332108743571e-05,
+      "loss": 1.2847,
+      "step": 1684
+    },
+    {
+      "epoch": 1.8536853685368537,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.06539309331374e-05,
+      "loss": 1.237,
+      "step": 1685
+    },
+    {
+      "epoch": 1.8547854785478548,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.062454077883909e-05,
+      "loss": 1.3331,
+      "step": 1686
+    },
+    {
+      "epoch": 1.8558855885588559,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.059515062454078e-05,
+      "loss": 1.2909,
+      "step": 1687
+    },
+    {
+      "epoch": 1.856985698569857,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.056576047024247e-05,
+      "loss": 1.0499,
+      "step": 1688
+    },
+    {
+      "epoch": 1.858085808580858,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.053637031594416e-05,
+      "loss": 1.0572,
+      "step": 1689
+    },
+    {
+      "epoch": 1.859185918591859,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.0506980161645853e-05,
+      "loss": 1.5574,
+      "step": 1690
+    },
+    {
+      "epoch": 1.8602860286028604,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.047759000734754e-05,
+      "loss": 1.3047,
+      "step": 1691
+    },
+    {
+      "epoch": 1.8613861386138613,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.044819985304923e-05,
+      "loss": 1.1164,
+      "step": 1692
+    },
+    {
+      "epoch": 1.8624862486248626,
+      "grad_norm": 2.375,
+      "learning_rate": 3.041880969875092e-05,
+      "loss": 1.2488,
+      "step": 1693
+    },
+    {
+      "epoch": 1.8635863586358636,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.0389419544452612e-05,
+      "loss": 1.1098,
+      "step": 1694
+    },
+    {
+      "epoch": 1.8646864686468647,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.03600293901543e-05,
+      "loss": 1.2167,
+      "step": 1695
+    },
+    {
+      "epoch": 1.8657865786578658,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.0330639235855993e-05,
+      "loss": 1.1934,
+      "step": 1696
+    },
+    {
+      "epoch": 1.8668866886688669,
+      "grad_norm": 2.375,
+      "learning_rate": 3.030124908155768e-05,
+      "loss": 1.0567,
+      "step": 1697
+    },
+    {
+      "epoch": 1.867986798679868,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.0271858927259374e-05,
+      "loss": 1.1064,
+      "step": 1698
+    },
+    {
+      "epoch": 1.869086908690869,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.024246877296106e-05,
+      "loss": 1.1842,
+      "step": 1699
+    },
+    {
+      "epoch": 1.8701870187018703,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.021307861866275e-05,
+      "loss": 1.0996,
+      "step": 1700
+    },
+    {
+      "epoch": 1.8712871287128712,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.018368846436444e-05,
+      "loss": 1.0199,
+      "step": 1701
+    },
+    {
+      "epoch": 1.8723872387238725,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.0154298310066132e-05,
+      "loss": 1.1437,
+      "step": 1702
+    },
+    {
+      "epoch": 1.8734873487348733,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.012490815576782e-05,
+      "loss": 1.2167,
+      "step": 1703
+    },
+    {
+      "epoch": 1.8745874587458746,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.0095518001469507e-05,
+      "loss": 1.2737,
+      "step": 1704
+    },
+    {
+      "epoch": 1.8756875687568757,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.00661278471712e-05,
+      "loss": 1.262,
+      "step": 1705
+    },
+    {
+      "epoch": 1.8767876787678768,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.0036737692872888e-05,
+      "loss": 1.1948,
+      "step": 1706
+    },
+    {
+      "epoch": 1.8778877887788779,
+      "grad_norm": 3.21875,
+      "learning_rate": 3.000734753857458e-05,
+      "loss": 1.0379,
+      "step": 1707
+    },
+    {
+      "epoch": 1.878987898789879,
+      "grad_norm": 2.375,
+      "learning_rate": 2.997795738427627e-05,
+      "loss": 1.383,
+      "step": 1708
+    },
+    {
+      "epoch": 1.8800880088008802,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.994856722997796e-05,
+      "loss": 1.2402,
+      "step": 1709
+    },
+    {
+      "epoch": 1.881188118811881,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.991917707567965e-05,
+      "loss": 1.2044,
+      "step": 1710
+    },
+    {
+      "epoch": 1.8822882288228824,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.988978692138134e-05,
+      "loss": 1.1497,
+      "step": 1711
+    },
+    {
+      "epoch": 1.8833883388338832,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.9860396767083027e-05,
+      "loss": 1.3766,
+      "step": 1712
+    },
+    {
+      "epoch": 1.8844884488448845,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.983100661278472e-05,
+      "loss": 1.3858,
+      "step": 1713
+    },
+    {
+      "epoch": 1.8855885588558856,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.9801616458486408e-05,
+      "loss": 1.307,
+      "step": 1714
+    },
+    {
+      "epoch": 1.8866886688668867,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.9772226304188102e-05,
+      "loss": 1.3307,
+      "step": 1715
+    },
+    {
+      "epoch": 1.8877887788778878,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.974283614988979e-05,
+      "loss": 1.4586,
+      "step": 1716
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.971344599559148e-05,
+      "loss": 1.2674,
+      "step": 1717
+    },
+    {
+      "epoch": 1.8899889988998901,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.968405584129317e-05,
+      "loss": 1.1686,
+      "step": 1718
+    },
+    {
+      "epoch": 1.891089108910891,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.965466568699486e-05,
+      "loss": 1.2669,
+      "step": 1719
+    },
+    {
+      "epoch": 1.8921892189218923,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.9625275532696548e-05,
+      "loss": 1.2401,
+      "step": 1720
+    },
+    {
+      "epoch": 1.8932893289328931,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.959588537839824e-05,
+      "loss": 1.1122,
+      "step": 1721
+    },
+    {
+      "epoch": 1.8943894389438944,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.956649522409993e-05,
+      "loss": 1.0152,
+      "step": 1722
+    },
+    {
+      "epoch": 1.8954895489548955,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9537105069801616e-05,
+      "loss": 1.1875,
+      "step": 1723
+    },
+    {
+      "epoch": 1.8965896589658966,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.950771491550331e-05,
+      "loss": 1.223,
+      "step": 1724
+    },
+    {
+      "epoch": 1.8976897689768977,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.9478324761204997e-05,
+      "loss": 1.0765,
+      "step": 1725
+    },
+    {
+      "epoch": 1.8987898789878987,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.944893460690669e-05,
+      "loss": 1.0747,
+      "step": 1726
+    },
+    {
+      "epoch": 1.8998899889989,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.9419544452608378e-05,
+      "loss": 1.1732,
+      "step": 1727
+    },
+    {
+      "epoch": 1.900990099009901,
+      "grad_norm": 2.375,
+      "learning_rate": 2.939015429831007e-05,
+      "loss": 1.4294,
+      "step": 1728
+    },
+    {
+      "epoch": 1.9020902090209022,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.936076414401176e-05,
+      "loss": 1.0625,
+      "step": 1729
+    },
+    {
+      "epoch": 1.903190319031903,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.933137398971345e-05,
+      "loss": 1.2338,
+      "step": 1730
+    },
+    {
+      "epoch": 1.9042904290429044,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.9301983835415137e-05,
+      "loss": 1.0537,
+      "step": 1731
+    },
+    {
+      "epoch": 1.9053905390539054,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.927259368111683e-05,
+      "loss": 1.1773,
+      "step": 1732
+    },
+    {
+      "epoch": 1.9064906490649065,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.9243203526818518e-05,
+      "loss": 1.1045,
+      "step": 1733
+    },
+    {
+      "epoch": 1.9075907590759076,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.921381337252021e-05,
+      "loss": 1.0802,
+      "step": 1734
+    },
+    {
+      "epoch": 1.9086908690869087,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.91844232182219e-05,
+      "loss": 1.4034,
+      "step": 1735
+    },
+    {
+      "epoch": 1.9097909790979097,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.915503306392359e-05,
+      "loss": 1.1439,
+      "step": 1736
+    },
+    {
+      "epoch": 1.9108910891089108,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.9125642909625276e-05,
+      "loss": 1.188,
+      "step": 1737
+    },
+    {
+      "epoch": 1.911991199119912,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9096252755326967e-05,
+      "loss": 1.2863,
+      "step": 1738
+    },
+    {
+      "epoch": 1.913091309130913,
+      "grad_norm": 2.5,
+      "learning_rate": 2.9066862601028657e-05,
+      "loss": 0.9819,
+      "step": 1739
+    },
+    {
+      "epoch": 1.9141914191419143,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.9037472446730344e-05,
+      "loss": 1.3277,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9152915291529153,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.900808229243204e-05,
+      "loss": 1.0285,
+      "step": 1741
+    },
+    {
+      "epoch": 1.9163916391639164,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.8978692138133725e-05,
+      "loss": 1.4419,
+      "step": 1742
+    },
+    {
+      "epoch": 1.9174917491749175,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.894930198383542e-05,
+      "loss": 0.8625,
+      "step": 1743
+    },
+    {
+      "epoch": 1.9185918591859186,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8919911829537106e-05,
+      "loss": 1.2662,
+      "step": 1744
+    },
+    {
+      "epoch": 1.9196919691969196,
+      "grad_norm": 2.375,
+      "learning_rate": 2.8890521675238797e-05,
+      "loss": 1.2673,
+      "step": 1745
+    },
+    {
+      "epoch": 1.9207920792079207,
+      "grad_norm": 2.375,
+      "learning_rate": 2.8861131520940487e-05,
+      "loss": 1.0652,
+      "step": 1746
+    },
+    {
+      "epoch": 1.921892189218922,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.8831741366642178e-05,
+      "loss": 1.2595,
+      "step": 1747
+    },
+    {
+      "epoch": 1.9229922992299229,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8802351212343865e-05,
+      "loss": 0.9257,
+      "step": 1748
+    },
+    {
+      "epoch": 1.9240924092409242,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.877296105804556e-05,
+      "loss": 1.3216,
+      "step": 1749
+    },
+    {
+      "epoch": 1.925192519251925,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.8743570903747246e-05,
+      "loss": 1.1868,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9262926292629263,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.871418074944894e-05,
+      "loss": 1.0613,
+      "step": 1751
+    },
+    {
+      "epoch": 1.9273927392739274,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.8684790595150627e-05,
+      "loss": 1.24,
+      "step": 1752
+    },
+    {
+      "epoch": 1.9284928492849285,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.8655400440852318e-05,
+      "loss": 1.1782,
+      "step": 1753
+    },
+    {
+      "epoch": 1.9295929592959296,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.8626010286554008e-05,
+      "loss": 1.2147,
+      "step": 1754
+    },
+    {
+      "epoch": 1.9306930693069306,
+      "grad_norm": 2.0,
+      "learning_rate": 2.8596620132255695e-05,
+      "loss": 1.1156,
+      "step": 1755
+    },
+    {
+      "epoch": 1.931793179317932,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8567229977957386e-05,
+      "loss": 1.1029,
+      "step": 1756
+    },
+    {
+      "epoch": 1.9328932893289328,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.8537839823659076e-05,
+      "loss": 1.2712,
+      "step": 1757
+    },
+    {
+      "epoch": 1.933993399339934,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.8508449669360767e-05,
+      "loss": 1.1975,
+      "step": 1758
+    },
+    {
+      "epoch": 1.935093509350935,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.8479059515062454e-05,
+      "loss": 1.4663,
+      "step": 1759
+    },
+    {
+      "epoch": 1.9361936193619362,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.8449669360764148e-05,
+      "loss": 1.3989,
+      "step": 1760
+    },
+    {
+      "epoch": 1.9372937293729373,
+      "grad_norm": 2.375,
+      "learning_rate": 2.8420279206465835e-05,
+      "loss": 1.0726,
+      "step": 1761
+    },
+    {
+      "epoch": 1.9383938393839384,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.839088905216753e-05,
+      "loss": 1.2323,
+      "step": 1762
+    },
+    {
+      "epoch": 1.9394939493949395,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.8361498897869216e-05,
+      "loss": 1.1034,
+      "step": 1763
+    },
+    {
+      "epoch": 1.9405940594059405,
+      "grad_norm": 2.5,
+      "learning_rate": 2.8332108743570906e-05,
+      "loss": 1.1885,
+      "step": 1764
+    },
+    {
+      "epoch": 1.9416941694169418,
+      "grad_norm": 2.375,
+      "learning_rate": 2.8302718589272597e-05,
+      "loss": 1.111,
+      "step": 1765
+    },
+    {
+      "epoch": 1.9427942794279427,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.8273328434974287e-05,
+      "loss": 1.2391,
+      "step": 1766
+    },
+    {
+      "epoch": 1.943894389438944,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.8243938280675974e-05,
+      "loss": 1.2338,
+      "step": 1767
+    },
+    {
+      "epoch": 1.9449944994499448,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.821454812637767e-05,
+      "loss": 1.2085,
+      "step": 1768
+    },
+    {
+      "epoch": 1.9460946094609461,
+      "grad_norm": 2.25,
+      "learning_rate": 2.8185157972079356e-05,
+      "loss": 1.2699,
+      "step": 1769
+    },
+    {
+      "epoch": 1.9471947194719472,
+      "grad_norm": 2.375,
+      "learning_rate": 2.815576781778105e-05,
+      "loss": 1.4165,
+      "step": 1770
+    },
+    {
+      "epoch": 1.9482948294829483,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.8126377663482737e-05,
+      "loss": 1.144,
+      "step": 1771
+    },
+    {
+      "epoch": 1.9493949394939494,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.8096987509184424e-05,
+      "loss": 1.3715,
+      "step": 1772
+    },
+    {
+      "epoch": 1.9504950495049505,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.8067597354886114e-05,
+      "loss": 1.3049,
+      "step": 1773
+    },
+    {
+      "epoch": 1.9515951595159517,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.8038207200587805e-05,
+      "loss": 1.2757,
+      "step": 1774
+    },
+    {
+      "epoch": 1.9526952695269526,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.8008817046289495e-05,
+      "loss": 1.2818,
+      "step": 1775
+    },
+    {
+      "epoch": 1.953795379537954,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.7979426891991182e-05,
+      "loss": 1.316,
+      "step": 1776
+    },
+    {
+      "epoch": 1.9548954895489548,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.7950036737692876e-05,
+      "loss": 1.1306,
+      "step": 1777
+    },
+    {
+      "epoch": 1.955995599559956,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.7920646583394563e-05,
+      "loss": 1.0581,
+      "step": 1778
+    },
+    {
+      "epoch": 1.9570957095709571,
+      "grad_norm": 2.75,
+      "learning_rate": 2.7891256429096257e-05,
+      "loss": 1.1463,
+      "step": 1779
+    },
+    {
+      "epoch": 1.9581958195819582,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.7861866274797944e-05,
+      "loss": 1.0238,
+      "step": 1780
+    },
+    {
+      "epoch": 1.9592959295929593,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.7832476120499635e-05,
+      "loss": 1.2044,
+      "step": 1781
+    },
+    {
+      "epoch": 1.9603960396039604,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.7803085966201325e-05,
+      "loss": 1.1006,
+      "step": 1782
+    },
+    {
+      "epoch": 1.9614961496149617,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.7773695811903016e-05,
+      "loss": 1.2047,
+      "step": 1783
+    },
+    {
+      "epoch": 1.9625962596259625,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.7744305657604703e-05,
+      "loss": 1.2663,
+      "step": 1784
+    },
+    {
+      "epoch": 1.9636963696369638,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.7714915503306397e-05,
+      "loss": 1.3644,
+      "step": 1785
+    },
+    {
+      "epoch": 1.9647964796479647,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.7685525349008084e-05,
+      "loss": 1.2156,
+      "step": 1786
+    },
+    {
+      "epoch": 1.965896589658966,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.7656135194709778e-05,
+      "loss": 1.3024,
+      "step": 1787
+    },
+    {
+      "epoch": 1.966996699669967,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.7626745040411465e-05,
+      "loss": 1.0227,
+      "step": 1788
+    },
+    {
+      "epoch": 1.9680968096809681,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.7597354886113152e-05,
+      "loss": 1.0698,
+      "step": 1789
+    },
+    {
+      "epoch": 1.9691969196919692,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.7567964731814846e-05,
+      "loss": 1.1479,
+      "step": 1790
+    },
+    {
+      "epoch": 1.9702970297029703,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.7538574577516533e-05,
+      "loss": 1.2113,
+      "step": 1791
+    },
+    {
+      "epoch": 1.9713971397139713,
+      "grad_norm": 30.0,
+      "learning_rate": 2.7509184423218224e-05,
+      "loss": 2.2909,
+      "step": 1792
+    },
+    {
+      "epoch": 1.9724972497249724,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.7479794268919914e-05,
+      "loss": 1.3478,
+      "step": 1793
+    },
+    {
+      "epoch": 1.9735973597359737,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.7450404114621605e-05,
+      "loss": 1.1221,
+      "step": 1794
+    },
+    {
+      "epoch": 1.9746974697469746,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.742101396032329e-05,
+      "loss": 1.2099,
+      "step": 1795
+    },
+    {
+      "epoch": 1.9757975797579759,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.7391623806024986e-05,
+      "loss": 1.1937,
+      "step": 1796
+    },
+    {
+      "epoch": 1.976897689768977,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.7362233651726673e-05,
+      "loss": 1.136,
+      "step": 1797
+    },
+    {
+      "epoch": 1.977997799779978,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.7332843497428367e-05,
+      "loss": 1.1615,
+      "step": 1798
+    },
+    {
+      "epoch": 1.979097909790979,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.7303453343130054e-05,
+      "loss": 1.269,
+      "step": 1799
+    },
+    {
+      "epoch": 1.9801980198019802,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.7274063188831744e-05,
+      "loss": 1.2401,
+      "step": 1800
+    },
+    {
+      "epoch": 1.9812981298129813,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.7244673034533435e-05,
+      "loss": 1.1832,
+      "step": 1801
+    },
+    {
+      "epoch": 1.9823982398239823,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.7215282880235125e-05,
+      "loss": 1.233,
+      "step": 1802
+    },
+    {
+      "epoch": 1.9834983498349836,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.7185892725936812e-05,
+      "loss": 1.4824,
+      "step": 1803
+    },
+    {
+      "epoch": 1.9845984598459845,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.71565025716385e-05,
+      "loss": 1.0727,
+      "step": 1804
+    },
+    {
+      "epoch": 1.9856985698569858,
+      "grad_norm": 2.5,
+      "learning_rate": 2.7127112417340193e-05,
+      "loss": 1.0226,
+      "step": 1805
+    },
+    {
+      "epoch": 1.9867986798679866,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.709772226304188e-05,
+      "loss": 1.3801,
+      "step": 1806
+    },
+    {
+      "epoch": 1.987898789878988,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.7068332108743574e-05,
+      "loss": 1.2423,
+      "step": 1807
+    },
+    {
+      "epoch": 1.988998899889989,
+      "grad_norm": 2.75,
+      "learning_rate": 2.703894195444526e-05,
+      "loss": 1.2382,
+      "step": 1808
+    },
+    {
+      "epoch": 1.99009900990099,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.7009551800146955e-05,
+      "loss": 1.2121,
+      "step": 1809
+    },
+    {
+      "epoch": 1.9911991199119912,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.6980161645848642e-05,
+      "loss": 1.1106,
+      "step": 1810
+    },
+    {
+      "epoch": 1.9922992299229922,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.6950771491550333e-05,
+      "loss": 1.3594,
+      "step": 1811
+    },
+    {
+      "epoch": 1.9933993399339935,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.692138133725202e-05,
+      "loss": 1.3398,
+      "step": 1812
+    },
+    {
+      "epoch": 1.9944994499449944,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.6891991182953714e-05,
+      "loss": 1.2078,
+      "step": 1813
+    },
+    {
+      "epoch": 1.9955995599559957,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.68626010286554e-05,
+      "loss": 1.2074,
+      "step": 1814
+    },
+    {
+      "epoch": 1.9966996699669965,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.6833210874357095e-05,
+      "loss": 1.0701,
+      "step": 1815
+    },
+    {
+      "epoch": 1.9977997799779978,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.6803820720058782e-05,
+      "loss": 1.0547,
+      "step": 1816
+    },
+    {
+      "epoch": 1.998899889988999,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.6774430565760473e-05,
+      "loss": 1.1877,
+      "step": 1817
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.6745040411462163e-05,
+      "loss": 1.3667,
+      "step": 1818
+    },
+    {
+      "epoch": 2.0011001100110013,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.6715650257163854e-05,
+      "loss": 0.931,
+      "step": 1819
+    },
+    {
+      "epoch": 2.002200220022002,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.668626010286554e-05,
+      "loss": 0.9089,
+      "step": 1820
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2727,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 455,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0951904215058842e+17,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}