| { | |
| "best_metric": 166.9327798179786, | |
| "best_model_checkpoint": "./whisper-training-blog/checkpoint-396", | |
| "epoch": 2.330357142857143, | |
| "global_step": 448, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 1.6666666666666668e-07, | |
| "loss": 1.7185, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 3.3333333333333335e-07, | |
| "loss": 1.6809, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5.555555555555555e-07, | |
| "loss": 1.7119, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 7.777777777777778e-07, | |
| "loss": 1.543, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 1e-06, | |
| "loss": 1.609, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 1.2222222222222223e-06, | |
| "loss": 1.5656, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 1.4444444444444445e-06, | |
| "loss": 1.6078, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 1.5075, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.8888888888888888e-06, | |
| "loss": 1.5183, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 2.1111111111111114e-06, | |
| "loss": 1.4949, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 2.3333333333333336e-06, | |
| "loss": 1.4111, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "eval_loss": 1.491944432258606, | |
| "eval_runtime": 119.9494, | |
| "eval_samples_per_second": 2.751, | |
| "eval_steps_per_second": 0.35, | |
| "eval_wer": 245.34568098355422, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 2.5555555555555557e-06, | |
| "loss": 1.3192, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2.7777777777777775e-06, | |
| "loss": 1.3334, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 3e-06, | |
| "loss": 1.2896, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 3.2222222222222222e-06, | |
| "loss": 1.2748, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 3.4444444444444444e-06, | |
| "loss": 1.3855, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 3.6666666666666666e-06, | |
| "loss": 1.2078, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 3.888888888888889e-06, | |
| "loss": 1.2186, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.111111111111111e-06, | |
| "loss": 1.1661, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.333333333333333e-06, | |
| "loss": 1.1352, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.555555555555555e-06, | |
| "loss": 1.0729, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.777777777777778e-06, | |
| "loss": 1.0501, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 1.225460410118103, | |
| "eval_runtime": 124.4384, | |
| "eval_samples_per_second": 2.652, | |
| "eval_steps_per_second": 0.338, | |
| "eval_wer": 225.88216509659907, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.9999999999999996e-06, | |
| "loss": 1.1278, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 5.2222222222222226e-06, | |
| "loss": 1.0564, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 5.444444444444445e-06, | |
| "loss": 0.9929, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5.666666666666667e-06, | |
| "loss": 1.0878, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5.888888888888889e-06, | |
| "loss": 1.0416, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 6.111111111111111e-06, | |
| "loss": 0.9337, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 6.333333333333333e-06, | |
| "loss": 0.9686, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 6.555555555555556e-06, | |
| "loss": 0.9557, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 6.777777777777778e-06, | |
| "loss": 1.0193, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 7e-06, | |
| "loss": 0.851, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 7.222222222222222e-06, | |
| "loss": 0.9032, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "eval_loss": 1.1202839612960815, | |
| "eval_runtime": 115.008, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 0.365, | |
| "eval_wer": 211.65575602746287, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 7.444444444444444e-06, | |
| "loss": 0.917, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 7.428115015974441e-06, | |
| "loss": 0.9444, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 7.332268370607029e-06, | |
| "loss": 0.8926, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 7.236421725239617e-06, | |
| "loss": 0.9975, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 7.140575079872205e-06, | |
| "loss": 1.239, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 7.044728434504793e-06, | |
| "loss": 0.9147, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 6.948881789137381e-06, | |
| "loss": 1.0116, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 6.853035143769968e-06, | |
| "loss": 0.9081, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 6.7571884984025565e-06, | |
| "loss": 0.8435, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 6.6613418530351436e-06, | |
| "loss": 0.8671, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 6.565495207667732e-06, | |
| "loss": 0.8141, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "eval_loss": 1.0674982070922852, | |
| "eval_runtime": 102.5987, | |
| "eval_samples_per_second": 3.216, | |
| "eval_steps_per_second": 0.409, | |
| "eval_wer": 184.62398211719625, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 6.469648562300319e-06, | |
| "loss": 0.8328, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 6.373801916932908e-06, | |
| "loss": 0.8191, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 6.277955271565495e-06, | |
| "loss": 0.8659, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 6.182108626198084e-06, | |
| "loss": 0.8211, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 6.086261980830671e-06, | |
| "loss": 0.8517, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 5.990415335463259e-06, | |
| "loss": 0.7813, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 5.894568690095847e-06, | |
| "loss": 0.7541, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 5.798722044728435e-06, | |
| "loss": 0.7938, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 5.702875399361023e-06, | |
| "loss": 0.818, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 5.607028753993611e-06, | |
| "loss": 0.7461, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 5.5111821086261985e-06, | |
| "loss": 0.8029, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "eval_loss": 1.0393612384796143, | |
| "eval_runtime": 103.1889, | |
| "eval_samples_per_second": 3.198, | |
| "eval_steps_per_second": 0.407, | |
| "eval_wer": 178.4129011655756, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 5.415335463258786e-06, | |
| "loss": 0.7541, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 5.319488817891374e-06, | |
| "loss": 0.7663, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 5.223642172523962e-06, | |
| "loss": 0.7253, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 5.127795527156549e-06, | |
| "loss": 0.6943, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 5.031948881789138e-06, | |
| "loss": 0.728, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 4.936102236421725e-06, | |
| "loss": 0.784, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 4.840255591054313e-06, | |
| "loss": 0.7154, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 4.744408945686901e-06, | |
| "loss": 0.7076, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 4.648562300319489e-06, | |
| "loss": 0.7811, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 4.552715654952077e-06, | |
| "loss": 0.6817, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 4.456869009584665e-06, | |
| "loss": 0.6325, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_loss": 1.0301399230957031, | |
| "eval_runtime": 117.4437, | |
| "eval_samples_per_second": 2.81, | |
| "eval_steps_per_second": 0.358, | |
| "eval_wer": 216.63739422002237, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 4.361022364217253e-06, | |
| "loss": 0.6825, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 4.26517571884984e-06, | |
| "loss": 0.7186, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 4.169329073482428e-06, | |
| "loss": 0.6699, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 4.0734824281150155e-06, | |
| "loss": 0.5677, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 3.977635782747604e-06, | |
| "loss": 0.6832, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 3.881789137380191e-06, | |
| "loss": 0.6293, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 3.7859424920127796e-06, | |
| "loss": 0.6719, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 3.6900958466453675e-06, | |
| "loss": 0.712, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 3.5942492012779555e-06, | |
| "loss": 0.8767, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 3.4984025559105434e-06, | |
| "loss": 0.6725, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 3.4025559105431313e-06, | |
| "loss": 0.6971, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "eval_loss": 1.0135136842727661, | |
| "eval_runtime": 114.2748, | |
| "eval_samples_per_second": 2.888, | |
| "eval_steps_per_second": 0.368, | |
| "eval_wer": 184.4004470700942, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 3.306709265175719e-06, | |
| "loss": 0.7433, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 3.2108626198083067e-06, | |
| "loss": 0.6264, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 3.1150159744408946e-06, | |
| "loss": 0.6605, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 3.0191693290734825e-06, | |
| "loss": 0.6411, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 2.9233226837060704e-06, | |
| "loss": 0.654, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 2.8274760383386583e-06, | |
| "loss": 0.6111, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 2.7316293929712462e-06, | |
| "loss": 0.6827, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 2.635782747603834e-06, | |
| "loss": 0.6498, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 2.539936102236422e-06, | |
| "loss": 0.6641, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 2.44408945686901e-06, | |
| "loss": 0.6146, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 2.3482428115015974e-06, | |
| "loss": 0.6051, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "eval_loss": 1.0065311193466187, | |
| "eval_runtime": 110.2985, | |
| "eval_samples_per_second": 2.992, | |
| "eval_steps_per_second": 0.381, | |
| "eval_wer": 194.7149928149449, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 2.2523961661341854e-06, | |
| "loss": 0.6236, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 2.1565495207667733e-06, | |
| "loss": 0.5992, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 2.060702875399361e-06, | |
| "loss": 0.6777, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "learning_rate": 1.964856230031949e-06, | |
| "loss": 0.6014, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 1.8690095846645368e-06, | |
| "loss": 0.6221, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 1.7731629392971245e-06, | |
| "loss": 0.6288, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 1.6773162939297124e-06, | |
| "loss": 0.6081, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 1.5814696485623003e-06, | |
| "loss": 0.5887, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 1.4856230031948882e-06, | |
| "loss": 0.5933, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 1.389776357827476e-06, | |
| "loss": 0.6373, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 1.2939297124600638e-06, | |
| "loss": 0.6047, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "eval_loss": 1.0029499530792236, | |
| "eval_runtime": 97.7372, | |
| "eval_samples_per_second": 3.376, | |
| "eval_steps_per_second": 0.43, | |
| "eval_wer": 166.9327798179786, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 1.1980830670926517e-06, | |
| "loss": 0.5773, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 1.1022364217252397e-06, | |
| "loss": 0.6325, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 1.0063897763578274e-06, | |
| "loss": 0.6009, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 9.105431309904153e-07, | |
| "loss": 0.5616, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 8.146964856230032e-07, | |
| "loss": 0.5729, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 7.188498402555911e-07, | |
| "loss": 0.5754, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 6.230031948881789e-07, | |
| "loss": 0.6207, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 5.271565495207668e-07, | |
| "loss": 0.5041, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 4.313099041533546e-07, | |
| "loss": 0.5543, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 3.3546325878594247e-07, | |
| "loss": 0.5607, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 2.3961661341853033e-07, | |
| "loss": 0.585, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "eval_loss": 1.0049597024917603, | |
| "eval_runtime": 112.1305, | |
| "eval_samples_per_second": 2.943, | |
| "eval_steps_per_second": 0.375, | |
| "eval_wer": 191.23423279578478, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 1.437699680511182e-07, | |
| "loss": 0.5486, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 4.792332268370607e-08, | |
| "loss": 0.6385, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "step": 448, | |
| "total_flos": 1.7572960198656e+17, | |
| "train_loss": 0.8751586728862354, | |
| "train_runtime": 2260.8457, | |
| "train_samples_per_second": 3.17, | |
| "train_steps_per_second": 0.198 | |
| } | |
| ], | |
| "max_steps": 448, | |
| "num_train_epochs": 9223372036854775807, | |
| "total_flos": 1.7572960198656e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |