| { | |
| "best_metric": 206.38494873046875, | |
| "best_model_checkpoint": "./ckpts/tinyllama_v1.1/int2-g128/checkpoint-400", | |
| "epoch": 4.0, | |
| "eval_steps": 40, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0, | |
| "loss": 5217.7002, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 2e-05, | |
| "loss": 5011.7588, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 2e-05, | |
| "loss": 4812.8281, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2e-05, | |
| "loss": 1860.1274, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2e-05, | |
| "loss": 1713.5734, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2e-05, | |
| "loss": 1103.2999, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 2e-05, | |
| "loss": 675.6068, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 2e-05, | |
| "loss": 683.2965, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 2e-05, | |
| "loss": 734.9794, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 2e-05, | |
| "loss": 585.422, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 2e-05, | |
| "loss": 570.1306, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2e-05, | |
| "loss": 599.2966, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 2e-05, | |
| "loss": 574.2372, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 2e-05, | |
| "loss": 481.7502, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 2e-05, | |
| "loss": 531.738, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 2e-05, | |
| "loss": 448.2928, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 2e-05, | |
| "loss": 490.8231, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 2e-05, | |
| "loss": 448.1083, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 2e-05, | |
| "loss": 506.6299, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 2e-05, | |
| "loss": 432.8912, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 2e-05, | |
| "loss": 449.625, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 2e-05, | |
| "loss": 588.6897, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 2e-05, | |
| "loss": 405.68, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 2e-05, | |
| "loss": 516.0303, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 2e-05, | |
| "loss": 401.7353, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 2e-05, | |
| "loss": 399.9238, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 2e-05, | |
| "loss": 351.5388, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 2e-05, | |
| "loss": 401.7846, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 2e-05, | |
| "loss": 385.5488, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 2e-05, | |
| "loss": 378.391, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 2e-05, | |
| "loss": 446.2048, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 2e-05, | |
| "loss": 354.5841, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 2e-05, | |
| "loss": 370.7014, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 2e-05, | |
| "loss": 397.8215, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 2e-05, | |
| "loss": 400.33, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 2e-05, | |
| "loss": 379.8638, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 2e-05, | |
| "loss": 282.1729, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 2e-05, | |
| "loss": 288.713, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 2e-05, | |
| "loss": 362.3285, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 2e-05, | |
| "loss": 345.252, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 349.94061279296875, | |
| "eval_runtime": 91.4064, | |
| "eval_samples_per_second": 17.504, | |
| "eval_steps_per_second": 1.094, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 2e-05, | |
| "loss": 391.5078, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 2e-05, | |
| "loss": 332.1484, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 2e-05, | |
| "loss": 352.9743, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 2e-05, | |
| "loss": 314.8037, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 2e-05, | |
| "loss": 386.3977, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 2e-05, | |
| "loss": 359.9244, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 2e-05, | |
| "loss": 376.9478, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 2e-05, | |
| "loss": 307.694, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2e-05, | |
| "loss": 359.4525, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2e-05, | |
| "loss": 319.51, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2e-05, | |
| "loss": 349.2659, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2e-05, | |
| "loss": 332.9238, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2e-05, | |
| "loss": 324.871, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2e-05, | |
| "loss": 305.993, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2e-05, | |
| "loss": 334.1832, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2e-05, | |
| "loss": 393.5037, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2e-05, | |
| "loss": 453.1027, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2e-05, | |
| "loss": 306.5744, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2e-05, | |
| "loss": 343.3282, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2e-05, | |
| "loss": 367.3992, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2e-05, | |
| "loss": 252.5841, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 2e-05, | |
| "loss": 332.6815, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 2e-05, | |
| "loss": 260.7815, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 2e-05, | |
| "loss": 322.11, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 2e-05, | |
| "loss": 316.3943, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 2e-05, | |
| "loss": 282.9459, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 2e-05, | |
| "loss": 389.4407, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 2e-05, | |
| "loss": 266.971, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 2e-05, | |
| "loss": 309.2886, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 2e-05, | |
| "loss": 248.2584, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 2e-05, | |
| "loss": 311.3903, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 2e-05, | |
| "loss": 281.4355, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 2e-05, | |
| "loss": 348.8126, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 2e-05, | |
| "loss": 335.4451, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 2e-05, | |
| "loss": 252.2587, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 2e-05, | |
| "loss": 280.6372, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 2e-05, | |
| "loss": 276.1567, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 2e-05, | |
| "loss": 295.8489, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 2e-05, | |
| "loss": 258.2193, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 2e-05, | |
| "loss": 303.292, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 290.92047119140625, | |
| "eval_runtime": 91.4002, | |
| "eval_samples_per_second": 17.505, | |
| "eval_steps_per_second": 1.094, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 2e-05, | |
| "loss": 292.2969, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 2e-05, | |
| "loss": 292.5847, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 2e-05, | |
| "loss": 287.2512, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 2e-05, | |
| "loss": 274.7219, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 2e-05, | |
| "loss": 259.4895, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2e-05, | |
| "loss": 308.7735, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2e-05, | |
| "loss": 358.5125, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2e-05, | |
| "loss": 228.2059, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 2e-05, | |
| "loss": 221.5, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 2e-05, | |
| "loss": 256.0698, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 2e-05, | |
| "loss": 280.6437, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 2e-05, | |
| "loss": 265.0312, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 2e-05, | |
| "loss": 276.1062, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 2e-05, | |
| "loss": 332.605, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 2e-05, | |
| "loss": 241.6841, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2e-05, | |
| "loss": 290.7553, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 2e-05, | |
| "loss": 242.8853, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 2e-05, | |
| "loss": 267.5484, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2e-05, | |
| "loss": 263.2162, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 2e-05, | |
| "loss": 226.8806, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 2e-05, | |
| "loss": 175.9632, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 2e-05, | |
| "loss": 293.8885, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 2e-05, | |
| "loss": 264.4063, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 2e-05, | |
| "loss": 268.8441, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 2e-05, | |
| "loss": 251.1409, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 2e-05, | |
| "loss": 222.4796, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 2e-05, | |
| "loss": 259.7393, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 2e-05, | |
| "loss": 247.0995, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 2e-05, | |
| "loss": 228.6188, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 2e-05, | |
| "loss": 238.4029, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 2e-05, | |
| "loss": 249.7835, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 2e-05, | |
| "loss": 255.0745, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 2e-05, | |
| "loss": 281.3386, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 2e-05, | |
| "loss": 258.1128, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 2e-05, | |
| "loss": 258.487, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 2e-05, | |
| "loss": 252.1913, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 2e-05, | |
| "loss": 222.6366, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 2e-05, | |
| "loss": 247.7612, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 2e-05, | |
| "loss": 212.2664, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 2e-05, | |
| "loss": 260.1885, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 268.6912841796875, | |
| "eval_runtime": 91.4506, | |
| "eval_samples_per_second": 17.496, | |
| "eval_steps_per_second": 1.093, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 2e-05, | |
| "loss": 225.2303, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 2e-05, | |
| "loss": 208.434, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 2e-05, | |
| "loss": 260.5778, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 2e-05, | |
| "loss": 256.1859, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 2e-05, | |
| "loss": 225.3799, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 2e-05, | |
| "loss": 242.6659, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 2e-05, | |
| "loss": 218.2521, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 2e-05, | |
| "loss": 237.711, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 2e-05, | |
| "loss": 228.6392, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 2e-05, | |
| "loss": 257.0567, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 2e-05, | |
| "loss": 225.4318, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 2e-05, | |
| "loss": 255.6358, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 2e-05, | |
| "loss": 243.6262, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 2e-05, | |
| "loss": 235.9305, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 2e-05, | |
| "loss": 238.0324, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 2e-05, | |
| "loss": 239.2688, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 2e-05, | |
| "loss": 234.8799, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 2e-05, | |
| "loss": 249.6847, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 2e-05, | |
| "loss": 259.0303, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 2e-05, | |
| "loss": 230.0663, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 2e-05, | |
| "loss": 312.8887, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 2e-05, | |
| "loss": 214.6919, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 2e-05, | |
| "loss": 204.0403, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 2e-05, | |
| "loss": 219.8406, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 2e-05, | |
| "loss": 229.476, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 2e-05, | |
| "loss": 222.8145, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 2e-05, | |
| "loss": 257.3806, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 2e-05, | |
| "loss": 206.661, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 2e-05, | |
| "loss": 244.2539, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 2e-05, | |
| "loss": 219.9999, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 2e-05, | |
| "loss": 186.9665, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 2e-05, | |
| "loss": 246.9571, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 2e-05, | |
| "loss": 296.5907, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 2e-05, | |
| "loss": 235.987, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 2e-05, | |
| "loss": 232.2841, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "learning_rate": 2e-05, | |
| "loss": 257.2687, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 2e-05, | |
| "loss": 229.2959, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 2e-05, | |
| "loss": 204.7547, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 2e-05, | |
| "loss": 229.0461, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 2e-05, | |
| "loss": 208.6121, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_loss": 251.8457794189453, | |
| "eval_runtime": 91.4542, | |
| "eval_samples_per_second": 17.495, | |
| "eval_steps_per_second": 1.093, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 2e-05, | |
| "loss": 213.7581, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 2e-05, | |
| "loss": 250.1387, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 2e-05, | |
| "loss": 216.0783, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 2e-05, | |
| "loss": 223.1108, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 2e-05, | |
| "loss": 223.4337, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 2e-05, | |
| "loss": 216.0298, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 2e-05, | |
| "loss": 210.1397, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 2e-05, | |
| "loss": 255.6102, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 2e-05, | |
| "loss": 206.8196, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 2e-05, | |
| "loss": 225.3016, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 2e-05, | |
| "loss": 204.423, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "learning_rate": 2e-05, | |
| "loss": 200.3793, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 2e-05, | |
| "loss": 254.3165, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 2e-05, | |
| "loss": 228.116, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 2e-05, | |
| "loss": 215.9781, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 2e-05, | |
| "loss": 240.427, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 2e-05, | |
| "loss": 285.4974, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 2e-05, | |
| "loss": 241.3725, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 2e-05, | |
| "loss": 208.2607, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 2e-05, | |
| "loss": 189.7236, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 2e-05, | |
| "loss": 251.2979, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 2e-05, | |
| "loss": 221.9034, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 2e-05, | |
| "loss": 212.9315, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "learning_rate": 2e-05, | |
| "loss": 269.8028, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 2e-05, | |
| "loss": 234.7929, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 2e-05, | |
| "loss": 232.632, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 2e-05, | |
| "loss": 234.5318, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 2e-05, | |
| "loss": 217.2522, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 2e-05, | |
| "loss": 212.185, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 2e-05, | |
| "loss": 193.634, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 2e-05, | |
| "loss": 194.4389, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 2e-05, | |
| "loss": 239.7684, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 2e-05, | |
| "loss": 214.6728, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 2e-05, | |
| "loss": 238.8546, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 2e-05, | |
| "loss": 252.4766, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 2e-05, | |
| "loss": 213.9979, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 2e-05, | |
| "loss": 226.8183, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "learning_rate": 2e-05, | |
| "loss": 178.3002, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 2e-05, | |
| "loss": 226.1548, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 2e-05, | |
| "loss": 197.3228, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 234.64894104003906, | |
| "eval_runtime": 91.475, | |
| "eval_samples_per_second": 17.491, | |
| "eval_steps_per_second": 1.093, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 2e-05, | |
| "loss": 177.7395, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 2e-05, | |
| "loss": 207.6397, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 2e-05, | |
| "loss": 234.7, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 2e-05, | |
| "loss": 161.9427, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 2e-05, | |
| "loss": 152.5065, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 2e-05, | |
| "loss": 209.7177, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 2e-05, | |
| "loss": 180.4935, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 2e-05, | |
| "loss": 189.8789, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 2e-05, | |
| "loss": 179.2588, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 2e-05, | |
| "loss": 199.4266, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 2e-05, | |
| "loss": 217.2014, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 2e-05, | |
| "loss": 194.8955, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 2e-05, | |
| "loss": 199.1684, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 2e-05, | |
| "loss": 184.8403, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "learning_rate": 2e-05, | |
| "loss": 181.7052, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 2e-05, | |
| "loss": 178.4005, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 2e-05, | |
| "loss": 166.0668, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 2e-05, | |
| "loss": 200.5868, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 2e-05, | |
| "loss": 164.0996, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 2e-05, | |
| "loss": 215.2086, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 2e-05, | |
| "loss": 182.4766, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 2e-05, | |
| "loss": 265.9225, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 2e-05, | |
| "loss": 170.2314, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 2e-05, | |
| "loss": 168.8915, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 2e-05, | |
| "loss": 202.6422, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 2e-05, | |
| "loss": 193.6124, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 2e-05, | |
| "loss": 191.7759, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 2e-05, | |
| "loss": 172.5128, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 2e-05, | |
| "loss": 174.589, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 2e-05, | |
| "loss": 190.2146, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 2e-05, | |
| "loss": 206.5455, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 2e-05, | |
| "loss": 212.3613, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 2e-05, | |
| "loss": 196.8155, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "learning_rate": 2e-05, | |
| "loss": 175.7169, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 2e-05, | |
| "loss": 246.1433, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 2e-05, | |
| "loss": 273.7065, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 2e-05, | |
| "loss": 158.33, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 2e-05, | |
| "loss": 159.6902, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "learning_rate": 2e-05, | |
| "loss": 260.6693, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 2e-05, | |
| "loss": 191.4345, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_loss": 228.75088500976562, | |
| "eval_runtime": 91.7222, | |
| "eval_samples_per_second": 17.444, | |
| "eval_steps_per_second": 1.09, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "learning_rate": 2e-05, | |
| "loss": 194.6978, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 2e-05, | |
| "loss": 170.113, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "learning_rate": 2e-05, | |
| "loss": 206.9311, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 2e-05, | |
| "loss": 195.2319, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 2e-05, | |
| "loss": 163.4541, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "learning_rate": 2e-05, | |
| "loss": 194.2114, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "learning_rate": 2e-05, | |
| "loss": 204.1492, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 2e-05, | |
| "loss": 202.1168, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 2e-05, | |
| "loss": 188.9232, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 2e-05, | |
| "loss": 183.1904, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "learning_rate": 2e-05, | |
| "loss": 171.6944, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 2e-05, | |
| "loss": 218.1628, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 2e-05, | |
| "loss": 178.1614, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "learning_rate": 2e-05, | |
| "loss": 175.8137, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 2e-05, | |
| "loss": 176.3016, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 2e-05, | |
| "loss": 195.611, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 2e-05, | |
| "loss": 154.6473, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "learning_rate": 2e-05, | |
| "loss": 175.7625, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 2e-05, | |
| "loss": 180.9702, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 2e-05, | |
| "loss": 172.006, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 2e-05, | |
| "loss": 166.616, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "learning_rate": 2e-05, | |
| "loss": 205.9087, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "learning_rate": 2e-05, | |
| "loss": 195.5401, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 2e-05, | |
| "loss": 182.7327, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 2e-05, | |
| "loss": 187.6268, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 2e-05, | |
| "loss": 150.9506, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 2e-05, | |
| "loss": 187.1612, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "learning_rate": 2e-05, | |
| "loss": 199.4861, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "learning_rate": 2e-05, | |
| "loss": 197.6736, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 2e-05, | |
| "loss": 204.7334, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "learning_rate": 2e-05, | |
| "loss": 186.2923, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 2e-05, | |
| "loss": 191.4558, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 2e-05, | |
| "loss": 195.1405, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 2e-05, | |
| "loss": 193.3551, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "learning_rate": 2e-05, | |
| "loss": 191.0934, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "learning_rate": 2e-05, | |
| "loss": 181.389, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 2e-05, | |
| "loss": 175.3716, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 2e-05, | |
| "loss": 172.3194, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 2e-05, | |
| "loss": 210.0355, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 2e-05, | |
| "loss": 151.5427, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "eval_loss": 220.8135528564453, | |
| "eval_runtime": 91.7827, | |
| "eval_samples_per_second": 17.432, | |
| "eval_steps_per_second": 1.09, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 2e-05, | |
| "loss": 216.8114, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "learning_rate": 2e-05, | |
| "loss": 204.5617, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 2e-05, | |
| "loss": 170.5889, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "learning_rate": 2e-05, | |
| "loss": 207.3868, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 2e-05, | |
| "loss": 181.0243, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 2e-05, | |
| "loss": 181.3605, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 2e-05, | |
| "loss": 151.9068, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 2e-05, | |
| "loss": 181.1088, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 2e-05, | |
| "loss": 168.5044, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 2e-05, | |
| "loss": 169.9193, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "learning_rate": 2e-05, | |
| "loss": 163.9978, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 2e-05, | |
| "loss": 160.5624, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 2e-05, | |
| "loss": 168.6609, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "learning_rate": 2e-05, | |
| "loss": 176.1334, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "learning_rate": 2e-05, | |
| "loss": 183.0596, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 2e-05, | |
| "loss": 137.6645, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "learning_rate": 2e-05, | |
| "loss": 207.355, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "learning_rate": 2e-05, | |
| "loss": 142.0311, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "learning_rate": 2e-05, | |
| "loss": 220.0689, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 2e-05, | |
| "loss": 188.8789, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "learning_rate": 2e-05, | |
| "loss": 125.5, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 2e-05, | |
| "loss": 135.3729, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 2e-05, | |
| "loss": 138.8647, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "learning_rate": 2e-05, | |
| "loss": 179.7612, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "learning_rate": 2e-05, | |
| "loss": 155.7191, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 2e-05, | |
| "loss": 142.4713, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "learning_rate": 2e-05, | |
| "loss": 153.5572, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "learning_rate": 2e-05, | |
| "loss": 158.0588, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "learning_rate": 2e-05, | |
| "loss": 188.1761, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "learning_rate": 2e-05, | |
| "loss": 149.4881, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "learning_rate": 2e-05, | |
| "loss": 143.047, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 2e-05, | |
| "loss": 149.3641, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "learning_rate": 2e-05, | |
| "loss": 157.5219, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 2e-05, | |
| "loss": 160.7546, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "learning_rate": 2e-05, | |
| "loss": 204.2928, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "learning_rate": 2e-05, | |
| "loss": 162.4894, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "learning_rate": 2e-05, | |
| "loss": 142.8077, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "learning_rate": 2e-05, | |
| "loss": 189.9061, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "learning_rate": 2e-05, | |
| "loss": 160.648, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 2e-05, | |
| "loss": 195.4514, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "eval_loss": 214.12860107421875, | |
| "eval_runtime": 91.6917, | |
| "eval_samples_per_second": 17.45, | |
| "eval_steps_per_second": 1.091, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 2e-05, | |
| "loss": 165.5777, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 2e-05, | |
| "loss": 157.4871, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "learning_rate": 2e-05, | |
| "loss": 156.9734, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "learning_rate": 2e-05, | |
| "loss": 171.0179, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "learning_rate": 2e-05, | |
| "loss": 158.758, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "learning_rate": 2e-05, | |
| "loss": 180.9356, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "learning_rate": 2e-05, | |
| "loss": 132.6712, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "learning_rate": 2e-05, | |
| "loss": 142.0085, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "learning_rate": 2e-05, | |
| "loss": 122.8798, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "learning_rate": 2e-05, | |
| "loss": 166.3639, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "learning_rate": 2e-05, | |
| "loss": 126.7612, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "learning_rate": 2e-05, | |
| "loss": 188.9171, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 2e-05, | |
| "loss": 151.0345, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "learning_rate": 2e-05, | |
| "loss": 159.7177, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "learning_rate": 2e-05, | |
| "loss": 150.0818, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "learning_rate": 2e-05, | |
| "loss": 162.671, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "learning_rate": 2e-05, | |
| "loss": 129.0101, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "learning_rate": 2e-05, | |
| "loss": 187.8155, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "learning_rate": 2e-05, | |
| "loss": 136.3083, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 2e-05, | |
| "loss": 139.6308, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 2e-05, | |
| "loss": 158.1506, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "learning_rate": 2e-05, | |
| "loss": 152.7077, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "learning_rate": 2e-05, | |
| "loss": 173.8158, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "learning_rate": 2e-05, | |
| "loss": 128.805, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "learning_rate": 2e-05, | |
| "loss": 152.3222, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "learning_rate": 2e-05, | |
| "loss": 163.3487, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "learning_rate": 2e-05, | |
| "loss": 169.0825, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 2e-05, | |
| "loss": 156.5232, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "learning_rate": 2e-05, | |
| "loss": 188.6721, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "learning_rate": 2e-05, | |
| "loss": 201.6223, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "learning_rate": 2e-05, | |
| "loss": 294.7936, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "learning_rate": 2e-05, | |
| "loss": 155.2639, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "learning_rate": 2e-05, | |
| "loss": 148.6182, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "learning_rate": 2e-05, | |
| "loss": 207.8028, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "learning_rate": 2e-05, | |
| "loss": 163.1711, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "learning_rate": 2e-05, | |
| "loss": 162.5552, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 2e-05, | |
| "loss": 167.8712, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "learning_rate": 2e-05, | |
| "loss": 155.6208, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "learning_rate": 2e-05, | |
| "loss": 178.2028, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 2e-05, | |
| "loss": 174.9905, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "eval_loss": 211.39349365234375, | |
| "eval_runtime": 91.4666, | |
| "eval_samples_per_second": 17.493, | |
| "eval_steps_per_second": 1.093, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "learning_rate": 2e-05, | |
| "loss": 149.9229, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "learning_rate": 2e-05, | |
| "loss": 166.3071, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "learning_rate": 2e-05, | |
| "loss": 124.036, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "learning_rate": 2e-05, | |
| "loss": 172.7505, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "learning_rate": 2e-05, | |
| "loss": 148.0122, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "learning_rate": 2e-05, | |
| "loss": 185.0846, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "learning_rate": 2e-05, | |
| "loss": 153.4166, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "learning_rate": 2e-05, | |
| "loss": 144.2338, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "learning_rate": 2e-05, | |
| "loss": 158.5771, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 2e-05, | |
| "loss": 163.8886, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "learning_rate": 2e-05, | |
| "loss": 151.2742, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "learning_rate": 2e-05, | |
| "loss": 169.2691, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "learning_rate": 2e-05, | |
| "loss": 125.0493, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "learning_rate": 2e-05, | |
| "loss": 144.3527, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 2e-05, | |
| "loss": 210.2006, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "learning_rate": 2e-05, | |
| "loss": 162.8882, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "learning_rate": 2e-05, | |
| "loss": 163.0425, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "learning_rate": 2e-05, | |
| "loss": 144.6404, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 2e-05, | |
| "loss": 169.7259, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 2e-05, | |
| "loss": 117.309, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "learning_rate": 2e-05, | |
| "loss": 179.2435, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "learning_rate": 2e-05, | |
| "loss": 156.4202, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "learning_rate": 2e-05, | |
| "loss": 212.8055, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 2e-05, | |
| "loss": 130.1424, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "learning_rate": 2e-05, | |
| "loss": 143.6542, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "learning_rate": 2e-05, | |
| "loss": 193.6444, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "learning_rate": 2e-05, | |
| "loss": 176.6723, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "learning_rate": 2e-05, | |
| "loss": 150.6032, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "learning_rate": 2e-05, | |
| "loss": 146.1843, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "learning_rate": 2e-05, | |
| "loss": 152.6586, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "learning_rate": 2e-05, | |
| "loss": 162.0343, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "learning_rate": 2e-05, | |
| "loss": 157.9043, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "learning_rate": 2e-05, | |
| "loss": 140.6674, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "learning_rate": 2e-05, | |
| "loss": 186.9754, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "learning_rate": 2e-05, | |
| "loss": 157.4324, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "learning_rate": 2e-05, | |
| "loss": 151.3968, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "learning_rate": 2e-05, | |
| "loss": 182.1434, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "learning_rate": 2e-05, | |
| "loss": 164.4491, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "learning_rate": 2e-05, | |
| "loss": 151.253, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 2e-05, | |
| "loss": 136.0101, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 206.38494873046875, | |
| "eval_runtime": 91.5142, | |
| "eval_samples_per_second": 17.484, | |
| "eval_steps_per_second": 1.093, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 40, | |
| "total_flos": 1.627150507573248e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |