| { | |
| "best_metric": 0.8883299798792756, | |
| "best_model_checkpoint": "PhoWhisper-small-vispeech-classifier-v4/checkpoint-490", | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 980, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01020408163265306, | |
| "grad_norm": 51640.66015625, | |
| "learning_rate": 5.1020408163265303e-08, | |
| "loss": 2.0778, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02040816326530612, | |
| "grad_norm": 46773.35546875, | |
| "learning_rate": 1.0204081632653061e-07, | |
| "loss": 2.0809, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030612244897959183, | |
| "grad_norm": 45343.12109375, | |
| "learning_rate": 1.5306122448979592e-07, | |
| "loss": 2.0771, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04081632653061224, | |
| "grad_norm": 49285.91796875, | |
| "learning_rate": 2.0408163265306121e-07, | |
| "loss": 2.077, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05102040816326531, | |
| "grad_norm": 63271.82421875, | |
| "learning_rate": 2.5510204081632656e-07, | |
| "loss": 2.0755, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.061224489795918366, | |
| "grad_norm": 51649.79296875, | |
| "learning_rate": 3.0612244897959183e-07, | |
| "loss": 2.0764, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 51633.98828125, | |
| "learning_rate": 3.5714285714285716e-07, | |
| "loss": 2.0763, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08163265306122448, | |
| "grad_norm": 61777.23828125, | |
| "learning_rate": 4.0816326530612243e-07, | |
| "loss": 2.0772, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09183673469387756, | |
| "grad_norm": 58467.15625, | |
| "learning_rate": 4.591836734693878e-07, | |
| "loss": 2.0743, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10204081632653061, | |
| "grad_norm": 48974.37890625, | |
| "learning_rate": 5.102040816326531e-07, | |
| "loss": 2.072, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11224489795918367, | |
| "grad_norm": 71007.515625, | |
| "learning_rate": 5.612244897959184e-07, | |
| "loss": 2.0751, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12244897959183673, | |
| "grad_norm": 58237.77734375, | |
| "learning_rate": 6.122448979591837e-07, | |
| "loss": 2.0702, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1326530612244898, | |
| "grad_norm": 58785.83203125, | |
| "learning_rate": 6.632653061224491e-07, | |
| "loss": 2.0713, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 52771.265625, | |
| "learning_rate": 7.142857142857143e-07, | |
| "loss": 2.069, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15306122448979592, | |
| "grad_norm": 69647.078125, | |
| "learning_rate": 7.653061224489796e-07, | |
| "loss": 2.0684, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16326530612244897, | |
| "grad_norm": 59148.47265625, | |
| "learning_rate": 8.163265306122449e-07, | |
| "loss": 2.0634, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17346938775510204, | |
| "grad_norm": 66723.921875, | |
| "learning_rate": 8.673469387755103e-07, | |
| "loss": 2.064, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1836734693877551, | |
| "grad_norm": 72545.203125, | |
| "learning_rate": 9.183673469387756e-07, | |
| "loss": 2.0593, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19387755102040816, | |
| "grad_norm": 59112.55859375, | |
| "learning_rate": 9.69387755102041e-07, | |
| "loss": 2.0553, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.20408163265306123, | |
| "grad_norm": 68369.265625, | |
| "learning_rate": 1.0204081632653063e-06, | |
| "loss": 2.0566, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 51080.62890625, | |
| "learning_rate": 1.0714285714285714e-06, | |
| "loss": 2.0505, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.22448979591836735, | |
| "grad_norm": 78760.0859375, | |
| "learning_rate": 1.122448979591837e-06, | |
| "loss": 2.0496, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.23469387755102042, | |
| "grad_norm": 57198.7890625, | |
| "learning_rate": 1.1734693877551022e-06, | |
| "loss": 2.0458, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.24489795918367346, | |
| "grad_norm": 98369.4609375, | |
| "learning_rate": 1.2244897959183673e-06, | |
| "loss": 2.0336, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.25510204081632654, | |
| "grad_norm": 80801.5703125, | |
| "learning_rate": 1.2755102040816329e-06, | |
| "loss": 2.0354, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2653061224489796, | |
| "grad_norm": 82179.9765625, | |
| "learning_rate": 1.3265306122448982e-06, | |
| "loss": 2.0263, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2755102040816326, | |
| "grad_norm": 83225.078125, | |
| "learning_rate": 1.3775510204081633e-06, | |
| "loss": 2.0168, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 62451.64453125, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 2.023, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.29591836734693877, | |
| "grad_norm": 75567.9453125, | |
| "learning_rate": 1.479591836734694e-06, | |
| "loss": 2.0206, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.30612244897959184, | |
| "grad_norm": 98461.484375, | |
| "learning_rate": 1.5306122448979593e-06, | |
| "loss": 1.9984, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3163265306122449, | |
| "grad_norm": 75078.1484375, | |
| "learning_rate": 1.5816326530612248e-06, | |
| "loss": 1.9899, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.32653061224489793, | |
| "grad_norm": 97895.671875, | |
| "learning_rate": 1.6326530612244897e-06, | |
| "loss": 1.9758, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.336734693877551, | |
| "grad_norm": 65880.2421875, | |
| "learning_rate": 1.6836734693877552e-06, | |
| "loss": 1.9918, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3469387755102041, | |
| "grad_norm": 70675.578125, | |
| "learning_rate": 1.7346938775510206e-06, | |
| "loss": 1.9715, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 64327.19921875, | |
| "learning_rate": 1.7857142857142859e-06, | |
| "loss": 1.9597, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3673469387755102, | |
| "grad_norm": 80610.6171875, | |
| "learning_rate": 1.8367346938775512e-06, | |
| "loss": 1.9632, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.37755102040816324, | |
| "grad_norm": 110899.375, | |
| "learning_rate": 1.8877551020408163e-06, | |
| "loss": 1.9476, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.3877551020408163, | |
| "grad_norm": 111132.5703125, | |
| "learning_rate": 1.938775510204082e-06, | |
| "loss": 1.9369, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3979591836734694, | |
| "grad_norm": 88722.625, | |
| "learning_rate": 1.989795918367347e-06, | |
| "loss": 1.9452, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.40816326530612246, | |
| "grad_norm": 105874.296875, | |
| "learning_rate": 2.0408163265306125e-06, | |
| "loss": 1.9459, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.41836734693877553, | |
| "grad_norm": 97286.0078125, | |
| "learning_rate": 2.0918367346938776e-06, | |
| "loss": 1.918, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 82848.0625, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 1.9135, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4387755102040816, | |
| "grad_norm": 114854.4609375, | |
| "learning_rate": 2.1938775510204083e-06, | |
| "loss": 1.8869, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4489795918367347, | |
| "grad_norm": 86661.25, | |
| "learning_rate": 2.244897959183674e-06, | |
| "loss": 1.8861, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.45918367346938777, | |
| "grad_norm": 127591.3984375, | |
| "learning_rate": 2.295918367346939e-06, | |
| "loss": 1.8396, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.46938775510204084, | |
| "grad_norm": 71038.8671875, | |
| "learning_rate": 2.3469387755102044e-06, | |
| "loss": 1.8995, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.47959183673469385, | |
| "grad_norm": 101402.421875, | |
| "learning_rate": 2.3979591836734696e-06, | |
| "loss": 1.8699, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.4897959183673469, | |
| "grad_norm": 113157.3828125, | |
| "learning_rate": 2.4489795918367347e-06, | |
| "loss": 1.8599, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 115485.0546875, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.8214, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5102040816326531, | |
| "grad_norm": 139369.8125, | |
| "learning_rate": 2.5510204081632657e-06, | |
| "loss": 1.8539, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5204081632653061, | |
| "grad_norm": 141401.46875, | |
| "learning_rate": 2.602040816326531e-06, | |
| "loss": 1.7507, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5306122448979592, | |
| "grad_norm": 125767.3984375, | |
| "learning_rate": 2.6530612244897964e-06, | |
| "loss": 1.7948, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5408163265306123, | |
| "grad_norm": 148848.5625, | |
| "learning_rate": 2.7040816326530615e-06, | |
| "loss": 1.7871, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5510204081632653, | |
| "grad_norm": 152509.4375, | |
| "learning_rate": 2.7551020408163266e-06, | |
| "loss": 1.8054, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5612244897959183, | |
| "grad_norm": 114681.078125, | |
| "learning_rate": 2.8061224489795917e-06, | |
| "loss": 1.7711, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 119104.984375, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 1.7139, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5816326530612245, | |
| "grad_norm": 123160.5, | |
| "learning_rate": 2.908163265306123e-06, | |
| "loss": 1.703, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5918367346938775, | |
| "grad_norm": 157218.203125, | |
| "learning_rate": 2.959183673469388e-06, | |
| "loss": 1.7685, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6020408163265306, | |
| "grad_norm": 134804.828125, | |
| "learning_rate": 3.0102040816326534e-06, | |
| "loss": 1.7053, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6122448979591837, | |
| "grad_norm": 153977.625, | |
| "learning_rate": 3.0612244897959185e-06, | |
| "loss": 1.6711, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6224489795918368, | |
| "grad_norm": 204690.125, | |
| "learning_rate": 3.112244897959184e-06, | |
| "loss": 1.5873, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6326530612244898, | |
| "grad_norm": 158604.1875, | |
| "learning_rate": 3.1632653061224496e-06, | |
| "loss": 1.6642, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 181877.484375, | |
| "learning_rate": 3.2142857142857147e-06, | |
| "loss": 1.6204, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6530612244897959, | |
| "grad_norm": 157735.734375, | |
| "learning_rate": 3.2653061224489794e-06, | |
| "loss": 1.6152, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6632653061224489, | |
| "grad_norm": 121935.3828125, | |
| "learning_rate": 3.316326530612245e-06, | |
| "loss": 1.6357, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.673469387755102, | |
| "grad_norm": 133579.859375, | |
| "learning_rate": 3.3673469387755105e-06, | |
| "loss": 1.6085, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6836734693877551, | |
| "grad_norm": 154886.953125, | |
| "learning_rate": 3.4183673469387756e-06, | |
| "loss": 1.5609, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.6938775510204082, | |
| "grad_norm": 151655.0625, | |
| "learning_rate": 3.469387755102041e-06, | |
| "loss": 1.5905, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7040816326530612, | |
| "grad_norm": 198471.609375, | |
| "learning_rate": 3.5204081632653062e-06, | |
| "loss": 1.5052, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 191516.453125, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 1.5798, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7244897959183674, | |
| "grad_norm": 207511.296875, | |
| "learning_rate": 3.6224489795918373e-06, | |
| "loss": 1.5382, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7346938775510204, | |
| "grad_norm": 170820.609375, | |
| "learning_rate": 3.6734693877551024e-06, | |
| "loss": 1.485, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7448979591836735, | |
| "grad_norm": 183925.796875, | |
| "learning_rate": 3.724489795918368e-06, | |
| "loss": 1.427, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7551020408163265, | |
| "grad_norm": 302829.75, | |
| "learning_rate": 3.7755102040816327e-06, | |
| "loss": 1.4242, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7653061224489796, | |
| "grad_norm": 292990.5625, | |
| "learning_rate": 3.826530612244898e-06, | |
| "loss": 1.484, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7755102040816326, | |
| "grad_norm": 189517.421875, | |
| "learning_rate": 3.877551020408164e-06, | |
| "loss": 1.5039, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 239271.890625, | |
| "learning_rate": 3.928571428571429e-06, | |
| "loss": 1.3332, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.7959183673469388, | |
| "grad_norm": 167386.28125, | |
| "learning_rate": 3.979591836734694e-06, | |
| "loss": 1.429, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8061224489795918, | |
| "grad_norm": 165701.9375, | |
| "learning_rate": 4.03061224489796e-06, | |
| "loss": 1.3167, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 284793.46875, | |
| "learning_rate": 4.081632653061225e-06, | |
| "loss": 1.3867, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.826530612244898, | |
| "grad_norm": 159807.671875, | |
| "learning_rate": 4.13265306122449e-06, | |
| "loss": 1.404, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8367346938775511, | |
| "grad_norm": 225698.859375, | |
| "learning_rate": 4.183673469387755e-06, | |
| "loss": 1.3842, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8469387755102041, | |
| "grad_norm": 148945.90625, | |
| "learning_rate": 4.234693877551021e-06, | |
| "loss": 1.4605, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 301350.15625, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 1.4279, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8673469387755102, | |
| "grad_norm": 216519.640625, | |
| "learning_rate": 4.336734693877551e-06, | |
| "loss": 1.3122, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.8775510204081632, | |
| "grad_norm": 243937.1875, | |
| "learning_rate": 4.3877551020408165e-06, | |
| "loss": 1.2084, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8877551020408163, | |
| "grad_norm": 195260.390625, | |
| "learning_rate": 4.438775510204082e-06, | |
| "loss": 1.3024, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.8979591836734694, | |
| "grad_norm": 218123.65625, | |
| "learning_rate": 4.489795918367348e-06, | |
| "loss": 1.2303, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9081632653061225, | |
| "grad_norm": 343990.03125, | |
| "learning_rate": 4.540816326530613e-06, | |
| "loss": 1.2012, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9183673469387755, | |
| "grad_norm": 157196.390625, | |
| "learning_rate": 4.591836734693878e-06, | |
| "loss": 1.2632, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 238103.46875, | |
| "learning_rate": 4.642857142857144e-06, | |
| "loss": 1.3131, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9387755102040817, | |
| "grad_norm": 158442.625, | |
| "learning_rate": 4.693877551020409e-06, | |
| "loss": 1.2849, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9489795918367347, | |
| "grad_norm": 178316.640625, | |
| "learning_rate": 4.744897959183674e-06, | |
| "loss": 1.0987, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9591836734693877, | |
| "grad_norm": 240848.03125, | |
| "learning_rate": 4.795918367346939e-06, | |
| "loss": 1.154, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9693877551020408, | |
| "grad_norm": 677730.8125, | |
| "learning_rate": 4.846938775510204e-06, | |
| "loss": 1.1682, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.9795918367346939, | |
| "grad_norm": 300622.875, | |
| "learning_rate": 4.897959183673469e-06, | |
| "loss": 1.1212, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.9897959183673469, | |
| "grad_norm": 234745.03125, | |
| "learning_rate": 4.948979591836735e-06, | |
| "loss": 1.1859, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 254966.984375, | |
| "learning_rate": 5e-06, | |
| "loss": 1.1763, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8883299798792756, | |
| "eval_loss": 0.968666136264801, | |
| "eval_runtime": 208.5533, | |
| "eval_samples_per_second": 14.299, | |
| "eval_steps_per_second": 0.599, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.010204081632653, | |
| "grad_norm": 227501.953125, | |
| "learning_rate": 4.994331065759638e-06, | |
| "loss": 1.1466, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.0204081632653061, | |
| "grad_norm": 184310.828125, | |
| "learning_rate": 4.9886621315192745e-06, | |
| "loss": 1.0584, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.030612244897959, | |
| "grad_norm": 568293.25, | |
| "learning_rate": 4.982993197278912e-06, | |
| "loss": 1.0574, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.0408163265306123, | |
| "grad_norm": 208197.4375, | |
| "learning_rate": 4.9773242630385495e-06, | |
| "loss": 1.0964, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0510204081632653, | |
| "grad_norm": 302428.34375, | |
| "learning_rate": 4.971655328798186e-06, | |
| "loss": 1.1017, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.0612244897959184, | |
| "grad_norm": 331227.1875, | |
| "learning_rate": 4.965986394557824e-06, | |
| "loss": 1.0954, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "grad_norm": 525813.5625, | |
| "learning_rate": 4.960317460317461e-06, | |
| "loss": 1.061, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.0816326530612246, | |
| "grad_norm": 578634.125, | |
| "learning_rate": 4.954648526077098e-06, | |
| "loss": 1.135, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.0918367346938775, | |
| "grad_norm": 208531.90625, | |
| "learning_rate": 4.948979591836735e-06, | |
| "loss": 0.9104, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.1020408163265305, | |
| "grad_norm": 621899.75, | |
| "learning_rate": 4.943310657596373e-06, | |
| "loss": 1.0909, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.1122448979591837, | |
| "grad_norm": 362684.3125, | |
| "learning_rate": 4.9376417233560094e-06, | |
| "loss": 0.8759, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.1224489795918366, | |
| "grad_norm": 274587.8125, | |
| "learning_rate": 4.931972789115647e-06, | |
| "loss": 0.9977, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.1326530612244898, | |
| "grad_norm": 379287.3125, | |
| "learning_rate": 4.9263038548752836e-06, | |
| "loss": 1.013, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 277623.0, | |
| "learning_rate": 4.920634920634921e-06, | |
| "loss": 0.9425, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.153061224489796, | |
| "grad_norm": 341604.53125, | |
| "learning_rate": 4.914965986394558e-06, | |
| "loss": 0.9697, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.163265306122449, | |
| "grad_norm": 742070.625, | |
| "learning_rate": 4.909297052154195e-06, | |
| "loss": 1.1181, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.1734693877551021, | |
| "grad_norm": 225296.84375, | |
| "learning_rate": 4.903628117913833e-06, | |
| "loss": 0.9465, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.183673469387755, | |
| "grad_norm": 952885.3125, | |
| "learning_rate": 4.897959183673469e-06, | |
| "loss": 1.1135, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.193877551020408, | |
| "grad_norm": 527028.875, | |
| "learning_rate": 4.892290249433107e-06, | |
| "loss": 1.109, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.2040816326530612, | |
| "grad_norm": 243262.0625, | |
| "learning_rate": 4.886621315192744e-06, | |
| "loss": 0.9313, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.2142857142857142, | |
| "grad_norm": 287492.28125, | |
| "learning_rate": 4.880952380952381e-06, | |
| "loss": 0.9603, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.2244897959183674, | |
| "grad_norm": 400522.90625, | |
| "learning_rate": 4.8752834467120185e-06, | |
| "loss": 0.9105, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2346938775510203, | |
| "grad_norm": 190966.40625, | |
| "learning_rate": 4.869614512471656e-06, | |
| "loss": 0.9654, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.2448979591836735, | |
| "grad_norm": 223976.515625, | |
| "learning_rate": 4.863945578231293e-06, | |
| "loss": 0.9511, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.2551020408163265, | |
| "grad_norm": 345172.09375, | |
| "learning_rate": 4.85827664399093e-06, | |
| "loss": 0.8092, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.2653061224489797, | |
| "grad_norm": 378395.75, | |
| "learning_rate": 4.852607709750568e-06, | |
| "loss": 0.9777, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.2755102040816326, | |
| "grad_norm": 408057.8125, | |
| "learning_rate": 4.846938775510204e-06, | |
| "loss": 1.006, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 254345.203125, | |
| "learning_rate": 4.841269841269842e-06, | |
| "loss": 0.9107, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.2959183673469388, | |
| "grad_norm": 465929.21875, | |
| "learning_rate": 4.835600907029479e-06, | |
| "loss": 1.0237, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.306122448979592, | |
| "grad_norm": 665911.5625, | |
| "learning_rate": 4.829931972789116e-06, | |
| "loss": 0.886, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.316326530612245, | |
| "grad_norm": 541784.25, | |
| "learning_rate": 4.824263038548753e-06, | |
| "loss": 0.9266, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.3265306122448979, | |
| "grad_norm": 181121.25, | |
| "learning_rate": 4.818594104308391e-06, | |
| "loss": 0.9289, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.336734693877551, | |
| "grad_norm": 465088.90625, | |
| "learning_rate": 4.8129251700680275e-06, | |
| "loss": 0.8976, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.346938775510204, | |
| "grad_norm": 309529.28125, | |
| "learning_rate": 4.807256235827665e-06, | |
| "loss": 0.9182, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.3571428571428572, | |
| "grad_norm": 672721.125, | |
| "learning_rate": 4.8015873015873025e-06, | |
| "loss": 0.8919, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.3673469387755102, | |
| "grad_norm": 300983.96875, | |
| "learning_rate": 4.795918367346939e-06, | |
| "loss": 0.7491, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.3775510204081631, | |
| "grad_norm": 168024.203125, | |
| "learning_rate": 4.790249433106577e-06, | |
| "loss": 0.803, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.3877551020408163, | |
| "grad_norm": 252603.5, | |
| "learning_rate": 4.784580498866213e-06, | |
| "loss": 0.8481, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.3979591836734695, | |
| "grad_norm": 495847.71875, | |
| "learning_rate": 4.778911564625851e-06, | |
| "loss": 0.9796, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.4081632653061225, | |
| "grad_norm": 213929.046875, | |
| "learning_rate": 4.773242630385488e-06, | |
| "loss": 0.8509, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.4183673469387754, | |
| "grad_norm": 203737.90625, | |
| "learning_rate": 4.767573696145125e-06, | |
| "loss": 0.8133, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 340089.6875, | |
| "learning_rate": 4.761904761904762e-06, | |
| "loss": 0.8628, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.4387755102040816, | |
| "grad_norm": 292163.90625, | |
| "learning_rate": 4.756235827664399e-06, | |
| "loss": 0.9341, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.4489795918367347, | |
| "grad_norm": 266707.96875, | |
| "learning_rate": 4.7505668934240365e-06, | |
| "loss": 1.0158, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.4591836734693877, | |
| "grad_norm": 389876.40625, | |
| "learning_rate": 4.744897959183674e-06, | |
| "loss": 0.9575, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.469387755102041, | |
| "grad_norm": 699534.0625, | |
| "learning_rate": 4.739229024943311e-06, | |
| "loss": 0.9293, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.4795918367346939, | |
| "grad_norm": 495432.5, | |
| "learning_rate": 4.733560090702948e-06, | |
| "loss": 0.7233, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.489795918367347, | |
| "grad_norm": 297580.71875, | |
| "learning_rate": 4.727891156462586e-06, | |
| "loss": 0.7624, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 214458.125, | |
| "learning_rate": 4.722222222222222e-06, | |
| "loss": 0.8704, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.510204081632653, | |
| "grad_norm": 483185.90625, | |
| "learning_rate": 4.71655328798186e-06, | |
| "loss": 0.7841, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.5204081632653061, | |
| "grad_norm": 279528.75, | |
| "learning_rate": 4.710884353741497e-06, | |
| "loss": 0.8634, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.5306122448979593, | |
| "grad_norm": 398520.15625, | |
| "learning_rate": 4.705215419501134e-06, | |
| "loss": 0.902, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.5408163265306123, | |
| "grad_norm": 343986.875, | |
| "learning_rate": 4.699546485260771e-06, | |
| "loss": 0.8896, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.5510204081632653, | |
| "grad_norm": 941463.0625, | |
| "learning_rate": 4.693877551020409e-06, | |
| "loss": 0.8483, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.5612244897959182, | |
| "grad_norm": 382110.125, | |
| "learning_rate": 4.6882086167800455e-06, | |
| "loss": 0.944, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 321413.15625, | |
| "learning_rate": 4.682539682539683e-06, | |
| "loss": 0.8019, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.5816326530612246, | |
| "grad_norm": 404689.875, | |
| "learning_rate": 4.6768707482993205e-06, | |
| "loss": 0.7069, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.5918367346938775, | |
| "grad_norm": 653981.25, | |
| "learning_rate": 4.671201814058957e-06, | |
| "loss": 0.9424, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.6020408163265305, | |
| "grad_norm": 387232.34375, | |
| "learning_rate": 4.665532879818595e-06, | |
| "loss": 0.9593, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.6122448979591837, | |
| "grad_norm": 527808.625, | |
| "learning_rate": 4.659863945578232e-06, | |
| "loss": 0.8147, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.6224489795918369, | |
| "grad_norm": 350020.8125, | |
| "learning_rate": 4.654195011337869e-06, | |
| "loss": 1.0195, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.6326530612244898, | |
| "grad_norm": 510409.9375, | |
| "learning_rate": 4.648526077097506e-06, | |
| "loss": 0.865, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.6428571428571428, | |
| "grad_norm": 238838.65625, | |
| "learning_rate": 4.642857142857144e-06, | |
| "loss": 0.8115, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.6530612244897958, | |
| "grad_norm": 266616.84375, | |
| "learning_rate": 4.6371882086167804e-06, | |
| "loss": 0.8091, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.663265306122449, | |
| "grad_norm": 398578.84375, | |
| "learning_rate": 4.631519274376418e-06, | |
| "loss": 0.6485, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.6734693877551021, | |
| "grad_norm": 242011.765625, | |
| "learning_rate": 4.6258503401360546e-06, | |
| "loss": 0.8242, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.683673469387755, | |
| "grad_norm": 646366.6875, | |
| "learning_rate": 4.620181405895692e-06, | |
| "loss": 0.7616, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.693877551020408, | |
| "grad_norm": 447861.9375, | |
| "learning_rate": 4.614512471655329e-06, | |
| "loss": 0.8694, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.7040816326530612, | |
| "grad_norm": 422529.3125, | |
| "learning_rate": 4.608843537414966e-06, | |
| "loss": 0.9344, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 448338.375, | |
| "learning_rate": 4.603174603174604e-06, | |
| "loss": 0.8459, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.7244897959183674, | |
| "grad_norm": 528423.75, | |
| "learning_rate": 4.59750566893424e-06, | |
| "loss": 0.8312, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.7346938775510203, | |
| "grad_norm": 422879.84375, | |
| "learning_rate": 4.591836734693878e-06, | |
| "loss": 0.8151, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.7448979591836735, | |
| "grad_norm": 543066.5, | |
| "learning_rate": 4.5861678004535145e-06, | |
| "loss": 0.8013, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.7551020408163265, | |
| "grad_norm": 362556.6875, | |
| "learning_rate": 4.580498866213152e-06, | |
| "loss": 0.9636, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.7653061224489797, | |
| "grad_norm": 305676.09375, | |
| "learning_rate": 4.5748299319727895e-06, | |
| "loss": 0.8208, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.7755102040816326, | |
| "grad_norm": 477851.96875, | |
| "learning_rate": 4.569160997732427e-06, | |
| "loss": 0.8109, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 380819.1875, | |
| "learning_rate": 4.563492063492064e-06, | |
| "loss": 0.8019, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.7959183673469388, | |
| "grad_norm": 276121.21875, | |
| "learning_rate": 4.557823129251701e-06, | |
| "loss": 0.7365, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.806122448979592, | |
| "grad_norm": 807004.5625, | |
| "learning_rate": 4.5521541950113386e-06, | |
| "loss": 0.8284, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.816326530612245, | |
| "grad_norm": 612872.5625, | |
| "learning_rate": 4.546485260770975e-06, | |
| "loss": 0.9153, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.8265306122448979, | |
| "grad_norm": 384264.40625, | |
| "learning_rate": 4.540816326530613e-06, | |
| "loss": 0.8697, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.836734693877551, | |
| "grad_norm": 323550.15625, | |
| "learning_rate": 4.53514739229025e-06, | |
| "loss": 0.816, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.8469387755102042, | |
| "grad_norm": 588055.625, | |
| "learning_rate": 4.529478458049887e-06, | |
| "loss": 0.9765, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 1386326.375, | |
| "learning_rate": 4.523809523809524e-06, | |
| "loss": 0.8051, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.8673469387755102, | |
| "grad_norm": 351753.28125, | |
| "learning_rate": 4.518140589569162e-06, | |
| "loss": 0.8377, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.8775510204081631, | |
| "grad_norm": 547828.25, | |
| "learning_rate": 4.5124716553287985e-06, | |
| "loss": 0.6776, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.8877551020408163, | |
| "grad_norm": 332551.40625, | |
| "learning_rate": 4.506802721088436e-06, | |
| "loss": 0.9434, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.8979591836734695, | |
| "grad_norm": 597800.125, | |
| "learning_rate": 4.5011337868480735e-06, | |
| "loss": 0.6165, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.9081632653061225, | |
| "grad_norm": 598426.1875, | |
| "learning_rate": 4.49546485260771e-06, | |
| "loss": 0.9668, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.9183673469387754, | |
| "grad_norm": 508502.5, | |
| "learning_rate": 4.489795918367348e-06, | |
| "loss": 0.8707, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.9285714285714286, | |
| "grad_norm": 825221.0625, | |
| "learning_rate": 4.484126984126984e-06, | |
| "loss": 0.8437, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.9387755102040818, | |
| "grad_norm": 674555.8125, | |
| "learning_rate": 4.478458049886622e-06, | |
| "loss": 0.7054, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.9489795918367347, | |
| "grad_norm": 547481.0, | |
| "learning_rate": 4.472789115646259e-06, | |
| "loss": 0.7896, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.9591836734693877, | |
| "grad_norm": 869727.6875, | |
| "learning_rate": 4.467120181405896e-06, | |
| "loss": 0.8161, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.9693877551020407, | |
| "grad_norm": 517315.71875, | |
| "learning_rate": 4.461451247165533e-06, | |
| "loss": 0.8131, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.9795918367346939, | |
| "grad_norm": 651501.0625, | |
| "learning_rate": 4.45578231292517e-06, | |
| "loss": 0.6217, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.989795918367347, | |
| "grad_norm": 559789.5625, | |
| "learning_rate": 4.4501133786848075e-06, | |
| "loss": 0.7849, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 586872.375, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.7778, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8779342723004695, | |
| "eval_loss": 0.5649504661560059, | |
| "eval_runtime": 182.7428, | |
| "eval_samples_per_second": 16.318, | |
| "eval_steps_per_second": 0.684, | |
| "step": 980 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 4900, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.99358138673152e+18, | |
| "train_batch_size": 24, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |