| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.719665271966527, | |
| "eval_steps": 100, | |
| "global_step": 2600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.10460251046025104, | |
| "eval_loss": 3.505904197692871, | |
| "eval_runtime": 155.3669, | |
| "eval_samples_per_second": 36.404, | |
| "eval_steps_per_second": 4.551, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.20920502092050208, | |
| "eval_loss": 3.2686526775360107, | |
| "eval_runtime": 151.6355, | |
| "eval_samples_per_second": 37.3, | |
| "eval_steps_per_second": 4.662, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3138075313807531, | |
| "eval_loss": 2.7369139194488525, | |
| "eval_runtime": 151.6605, | |
| "eval_samples_per_second": 37.294, | |
| "eval_steps_per_second": 4.662, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.41841004184100417, | |
| "eval_loss": 1.303921103477478, | |
| "eval_runtime": 151.8516, | |
| "eval_samples_per_second": 37.247, | |
| "eval_steps_per_second": 4.656, | |
| "eval_wer": 0.8544558745646836, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5230125523012552, | |
| "grad_norm": 2.7574479579925537, | |
| "learning_rate": 0.0002958, | |
| "loss": 3.5426, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5230125523012552, | |
| "eval_loss": 1.051817536354065, | |
| "eval_runtime": 152.6837, | |
| "eval_samples_per_second": 37.044, | |
| "eval_steps_per_second": 4.63, | |
| "eval_wer": 0.7738120075107124, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6276150627615062, | |
| "eval_loss": 0.839580774307251, | |
| "eval_runtime": 153.5079, | |
| "eval_samples_per_second": 36.845, | |
| "eval_steps_per_second": 4.606, | |
| "eval_wer": 0.6148994559548073, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7322175732217573, | |
| "eval_loss": 0.7548021674156189, | |
| "eval_runtime": 152.4899, | |
| "eval_samples_per_second": 37.091, | |
| "eval_steps_per_second": 4.636, | |
| "eval_wer": 0.5686636388438638, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.8368200836820083, | |
| "eval_loss": 0.6868751645088196, | |
| "eval_runtime": 153.2403, | |
| "eval_samples_per_second": 36.909, | |
| "eval_steps_per_second": 4.614, | |
| "eval_wer": 0.5185119802282101, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.9414225941422594, | |
| "eval_loss": 0.6501098871231079, | |
| "eval_runtime": 152.6877, | |
| "eval_samples_per_second": 37.043, | |
| "eval_steps_per_second": 4.63, | |
| "eval_wer": 0.4895283336810515, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.0460251046025104, | |
| "grad_norm": 1.2487365007400513, | |
| "learning_rate": 0.00022957142857142856, | |
| "loss": 0.7785, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0460251046025104, | |
| "eval_loss": 0.5817079544067383, | |
| "eval_runtime": 153.4094, | |
| "eval_samples_per_second": 36.869, | |
| "eval_steps_per_second": 4.609, | |
| "eval_wer": 0.4500649965495659, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.1506276150627615, | |
| "eval_loss": 0.5646732449531555, | |
| "eval_runtime": 152.9118, | |
| "eval_samples_per_second": 36.989, | |
| "eval_steps_per_second": 4.624, | |
| "eval_wer": 0.4288006932965287, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.2552301255230125, | |
| "eval_loss": 0.5424243807792664, | |
| "eval_runtime": 152.6245, | |
| "eval_samples_per_second": 37.058, | |
| "eval_steps_per_second": 4.632, | |
| "eval_wer": 0.42786987851262215, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.3598326359832635, | |
| "eval_loss": 0.5264406800270081, | |
| "eval_runtime": 153.2158, | |
| "eval_samples_per_second": 36.915, | |
| "eval_steps_per_second": 4.614, | |
| "eval_wer": 0.4050809648376691, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.4644351464435146, | |
| "eval_loss": 0.5098912715911865, | |
| "eval_runtime": 153.9853, | |
| "eval_samples_per_second": 36.731, | |
| "eval_steps_per_second": 4.591, | |
| "eval_wer": 0.3977467862817159, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.5690376569037658, | |
| "grad_norm": 0.9644107222557068, | |
| "learning_rate": 0.0001582857142857143, | |
| "loss": 0.5795, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.5690376569037658, | |
| "eval_loss": 0.5057937502861023, | |
| "eval_runtime": 153.7283, | |
| "eval_samples_per_second": 36.792, | |
| "eval_steps_per_second": 4.599, | |
| "eval_wer": 0.3952913610758935, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.6736401673640167, | |
| "eval_loss": 0.48043006658554077, | |
| "eval_runtime": 153.7864, | |
| "eval_samples_per_second": 36.778, | |
| "eval_steps_per_second": 4.597, | |
| "eval_wer": 0.37892185970374415, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.778242677824268, | |
| "eval_loss": 0.46723026037216187, | |
| "eval_runtime": 153.9064, | |
| "eval_samples_per_second": 36.75, | |
| "eval_steps_per_second": 4.594, | |
| "eval_wer": 0.3698062942337629, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.8828451882845187, | |
| "eval_loss": 0.46052873134613037, | |
| "eval_runtime": 153.9609, | |
| "eval_samples_per_second": 36.737, | |
| "eval_steps_per_second": 4.592, | |
| "eval_wer": 0.3712185649403797, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.98744769874477, | |
| "eval_loss": 0.4490562677383423, | |
| "eval_runtime": 153.6441, | |
| "eval_samples_per_second": 36.812, | |
| "eval_steps_per_second": 4.602, | |
| "eval_wer": 0.35561939304456675, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.092050209205021, | |
| "grad_norm": 1.1409815549850464, | |
| "learning_rate": 8.714285714285714e-05, | |
| "loss": 0.5057, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.092050209205021, | |
| "eval_loss": 0.44942042231559753, | |
| "eval_runtime": 154.6428, | |
| "eval_samples_per_second": 36.575, | |
| "eval_steps_per_second": 4.572, | |
| "eval_wer": 0.34534833336008086, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.196652719665272, | |
| "eval_loss": 0.4431760907173157, | |
| "eval_runtime": 154.0218, | |
| "eval_samples_per_second": 36.722, | |
| "eval_steps_per_second": 4.59, | |
| "eval_wer": 0.3396832020028566, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.301255230125523, | |
| "eval_loss": 0.43776625394821167, | |
| "eval_runtime": 154.8082, | |
| "eval_samples_per_second": 36.536, | |
| "eval_steps_per_second": 4.567, | |
| "eval_wer": 0.335061225144838, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.405857740585774, | |
| "eval_loss": 0.42906874418258667, | |
| "eval_runtime": 154.6416, | |
| "eval_samples_per_second": 36.575, | |
| "eval_steps_per_second": 4.572, | |
| "eval_wer": 0.33104909245558567, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.510460251046025, | |
| "eval_loss": 0.4279017746448517, | |
| "eval_runtime": 154.5726, | |
| "eval_samples_per_second": 36.591, | |
| "eval_steps_per_second": 4.574, | |
| "eval_wer": 0.3294281908491278, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.6150627615062763, | |
| "grad_norm": 0.6568267345428467, | |
| "learning_rate": 1.6e-05, | |
| "loss": 0.3986, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.6150627615062763, | |
| "eval_loss": 0.4233950078487396, | |
| "eval_runtime": 154.8004, | |
| "eval_samples_per_second": 36.537, | |
| "eval_steps_per_second": 4.567, | |
| "eval_wer": 0.3259456596748568, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.719665271966527, | |
| "eval_loss": 0.41978269815444946, | |
| "eval_runtime": 153.7416, | |
| "eval_samples_per_second": 36.789, | |
| "eval_steps_per_second": 4.599, | |
| "eval_wer": 0.32485435958338016, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.719665271966527, | |
| "step": 2600, | |
| "total_flos": 1.0779565758702488e+19, | |
| "train_loss": 1.1314451643136831, | |
| "train_runtime": 7270.6329, | |
| "train_samples_per_second": 11.443, | |
| "train_steps_per_second": 0.358 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2600, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 400, | |
| "total_flos": 1.0779565758702488e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |