| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0959752321981426, | |
| "eval_steps": 100, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15479876160990713, | |
| "eval_loss": 3.5622992515563965, | |
| "eval_runtime": 158.441, | |
| "eval_samples_per_second": 35.698, | |
| "eval_steps_per_second": 4.462, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.30959752321981426, | |
| "eval_loss": 3.2967281341552734, | |
| "eval_runtime": 153.6671, | |
| "eval_samples_per_second": 36.807, | |
| "eval_steps_per_second": 4.601, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.46439628482972134, | |
| "eval_loss": 2.648355722427368, | |
| "eval_runtime": 153.3096, | |
| "eval_samples_per_second": 36.893, | |
| "eval_steps_per_second": 4.612, | |
| "eval_wer": 0.9999518544077289, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6191950464396285, | |
| "eval_loss": 1.060172200202942, | |
| "eval_runtime": 153.5019, | |
| "eval_samples_per_second": 36.846, | |
| "eval_steps_per_second": 4.606, | |
| "eval_wer": 0.7315080804352362, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7739938080495357, | |
| "grad_norm": 2.8772997856140137, | |
| "learning_rate": 0.00029699999999999996, | |
| "loss": 3.6398, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7739938080495357, | |
| "eval_loss": 0.8942204713821411, | |
| "eval_runtime": 154.5421, | |
| "eval_samples_per_second": 36.598, | |
| "eval_steps_per_second": 4.575, | |
| "eval_wer": 0.6695607517131806, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9287925696594427, | |
| "eval_loss": 0.7115893959999084, | |
| "eval_runtime": 155.1607, | |
| "eval_samples_per_second": 36.453, | |
| "eval_steps_per_second": 4.557, | |
| "eval_wer": 0.5361172184686492, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.08359133126935, | |
| "eval_loss": 0.6647589802742004, | |
| "eval_runtime": 154.7086, | |
| "eval_samples_per_second": 36.559, | |
| "eval_steps_per_second": 4.57, | |
| "eval_wer": 0.5100704530500233, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.238390092879257, | |
| "eval_loss": 0.5869073867797852, | |
| "eval_runtime": 154.314, | |
| "eval_samples_per_second": 36.653, | |
| "eval_steps_per_second": 4.582, | |
| "eval_wer": 0.45284139237052845, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.3931888544891642, | |
| "eval_loss": 0.5697966814041138, | |
| "eval_runtime": 153.6252, | |
| "eval_samples_per_second": 36.817, | |
| "eval_steps_per_second": 4.602, | |
| "eval_wer": 0.43589414389112674, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.5479876160990713, | |
| "grad_norm": 0.41244009137153625, | |
| "learning_rate": 0.0002016, | |
| "loss": 0.5976, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5479876160990713, | |
| "eval_loss": 0.54075688123703, | |
| "eval_runtime": 154.0436, | |
| "eval_samples_per_second": 36.717, | |
| "eval_steps_per_second": 4.59, | |
| "eval_wer": 0.41122755211760365, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.7027863777089784, | |
| "eval_loss": 0.5229234099388123, | |
| "eval_runtime": 154.4025, | |
| "eval_samples_per_second": 36.632, | |
| "eval_steps_per_second": 4.579, | |
| "eval_wer": 0.41960488517276245, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.8575851393188856, | |
| "eval_loss": 0.5054619312286377, | |
| "eval_runtime": 154.4414, | |
| "eval_samples_per_second": 36.622, | |
| "eval_steps_per_second": 4.578, | |
| "eval_wer": 0.39546789491422063, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.0123839009287927, | |
| "eval_loss": 0.48077496886253357, | |
| "eval_runtime": 154.8241, | |
| "eval_samples_per_second": 36.532, | |
| "eval_steps_per_second": 4.566, | |
| "eval_wer": 0.3708815457944825, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.1671826625387, | |
| "eval_loss": 0.46668991446495056, | |
| "eval_runtime": 154.4804, | |
| "eval_samples_per_second": 36.613, | |
| "eval_steps_per_second": 4.577, | |
| "eval_wer": 0.3579785270658471, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.321981424148607, | |
| "grad_norm": 0.45199868083000183, | |
| "learning_rate": 0.00010179999999999998, | |
| "loss": 0.443, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.321981424148607, | |
| "eval_loss": 0.45730867981910706, | |
| "eval_runtime": 154.0073, | |
| "eval_samples_per_second": 36.726, | |
| "eval_steps_per_second": 4.591, | |
| "eval_wer": 0.35817110943493125, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.476780185758514, | |
| "eval_loss": 0.44747394323349, | |
| "eval_runtime": 156.828, | |
| "eval_samples_per_second": 36.065, | |
| "eval_steps_per_second": 4.508, | |
| "eval_wer": 0.3452199451140248, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "eval_loss": 0.43689531087875366, | |
| "eval_runtime": 155.5732, | |
| "eval_samples_per_second": 36.356, | |
| "eval_steps_per_second": 4.544, | |
| "eval_wer": 0.3477877100351463, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.7863777089783284, | |
| "eval_loss": 0.4227137863636017, | |
| "eval_runtime": 156.702, | |
| "eval_samples_per_second": 36.094, | |
| "eval_steps_per_second": 4.512, | |
| "eval_wer": 0.32978125852578194, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "eval_loss": 0.4168964922428131, | |
| "eval_runtime": 153.7827, | |
| "eval_samples_per_second": 36.779, | |
| "eval_steps_per_second": 4.597, | |
| "eval_wer": 0.3270690568278474, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.0959752321981426, | |
| "grad_norm": 0.4056684672832489, | |
| "learning_rate": 2.4e-06, | |
| "loss": 0.3475, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0959752321981426, | |
| "eval_loss": 0.42000359296798706, | |
| "eval_runtime": 154.6879, | |
| "eval_samples_per_second": 36.564, | |
| "eval_steps_per_second": 4.57, | |
| "eval_wer": 0.3226557108696699, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0959752321981426, | |
| "step": 2000, | |
| "total_flos": 1.8144283373471887e+19, | |
| "train_loss": 1.2569797668457032, | |
| "train_runtime": 7798.4407, | |
| "train_samples_per_second": 16.414, | |
| "train_steps_per_second": 0.256 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 400, | |
| "total_flos": 1.8144283373471887e+19, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |