| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 100, | |
| "global_step": 1872, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.21367521367521367, | |
| "eval_loss": 0.2269495576620102, | |
| "eval_runtime": 301.5344, | |
| "eval_samples_per_second": 19.421, | |
| "eval_steps_per_second": 2.428, | |
| "eval_wer": 0.3076194253481403, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.42735042735042733, | |
| "eval_loss": 0.19151096045970917, | |
| "eval_runtime": 302.1327, | |
| "eval_samples_per_second": 19.382, | |
| "eval_steps_per_second": 2.423, | |
| "eval_wer": 0.2824343380927199, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6410256410256411, | |
| "eval_loss": 0.1993144452571869, | |
| "eval_runtime": 302.3659, | |
| "eval_samples_per_second": 19.367, | |
| "eval_steps_per_second": 2.421, | |
| "eval_wer": 0.296448087431694, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8547008547008547, | |
| "eval_loss": 0.18319852650165558, | |
| "eval_runtime": 303.1213, | |
| "eval_samples_per_second": 19.319, | |
| "eval_steps_per_second": 2.415, | |
| "eval_wer": 0.2734884540807333, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0683760683760684, | |
| "grad_norm": 0.8244175314903259, | |
| "learning_rate": 0.0007748306997742663, | |
| "loss": 0.9965, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0683760683760684, | |
| "eval_loss": 0.1763978898525238, | |
| "eval_runtime": 304.6466, | |
| "eval_samples_per_second": 19.222, | |
| "eval_steps_per_second": 2.403, | |
| "eval_wer": 0.26493918561607616, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.282051282051282, | |
| "eval_loss": 0.17325642704963684, | |
| "eval_runtime": 304.5973, | |
| "eval_samples_per_second": 19.225, | |
| "eval_steps_per_second": 2.403, | |
| "eval_wer": 0.2625154239379517, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.4957264957264957, | |
| "eval_loss": 0.17252753674983978, | |
| "eval_runtime": 305.7799, | |
| "eval_samples_per_second": 19.151, | |
| "eval_steps_per_second": 2.394, | |
| "eval_wer": 0.25918826017979907, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.7094017094017095, | |
| "eval_loss": 0.17061175405979156, | |
| "eval_runtime": 306.9826, | |
| "eval_samples_per_second": 19.076, | |
| "eval_steps_per_second": 2.385, | |
| "eval_wer": 0.25806451612903225, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "eval_loss": 0.16809020936489105, | |
| "eval_runtime": 306.5557, | |
| "eval_samples_per_second": 19.103, | |
| "eval_steps_per_second": 2.388, | |
| "eval_wer": 0.25854926846465714, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.1367521367521367, | |
| "grad_norm": 0.886053204536438, | |
| "learning_rate": 0.0004926636568848759, | |
| "loss": 0.2922, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.1367521367521367, | |
| "eval_loss": 0.16939426958560944, | |
| "eval_runtime": 308.102, | |
| "eval_samples_per_second": 19.007, | |
| "eval_steps_per_second": 2.376, | |
| "eval_wer": 0.2590560549973559, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.3504273504273505, | |
| "eval_loss": 0.17009997367858887, | |
| "eval_runtime": 307.4896, | |
| "eval_samples_per_second": 19.045, | |
| "eval_steps_per_second": 2.381, | |
| "eval_wer": 0.25753569539925963, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.564102564102564, | |
| "eval_loss": 0.17012837529182434, | |
| "eval_runtime": 308.404, | |
| "eval_samples_per_second": 18.988, | |
| "eval_steps_per_second": 2.374, | |
| "eval_wer": 0.2613916798871849, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "eval_loss": 0.16536183655261993, | |
| "eval_runtime": 307.6932, | |
| "eval_samples_per_second": 19.032, | |
| "eval_steps_per_second": 2.379, | |
| "eval_wer": 0.25354750572889123, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.9914529914529915, | |
| "eval_loss": 0.16437660157680511, | |
| "eval_runtime": 307.9587, | |
| "eval_samples_per_second": 19.016, | |
| "eval_steps_per_second": 2.377, | |
| "eval_wer": 0.2516525647805394, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.2051282051282053, | |
| "grad_norm": 0.9470628499984741, | |
| "learning_rate": 0.00021049661399548533, | |
| "loss": 0.2788, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.2051282051282053, | |
| "eval_loss": 0.1636321097612381, | |
| "eval_runtime": 308.3318, | |
| "eval_samples_per_second": 18.993, | |
| "eval_steps_per_second": 2.374, | |
| "eval_wer": 0.2539881896703684, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.4188034188034186, | |
| "eval_loss": 0.16157202422618866, | |
| "eval_runtime": 307.8619, | |
| "eval_samples_per_second": 19.022, | |
| "eval_steps_per_second": 2.378, | |
| "eval_wer": 0.2511678124449145, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.6324786324786325, | |
| "eval_loss": 0.1600475162267685, | |
| "eval_runtime": 308.3358, | |
| "eval_samples_per_second": 18.992, | |
| "eval_steps_per_second": 2.374, | |
| "eval_wer": 0.24702538339502908, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "eval_loss": 0.15865576267242432, | |
| "eval_runtime": 309.7473, | |
| "eval_samples_per_second": 18.906, | |
| "eval_steps_per_second": 2.363, | |
| "eval_wer": 0.2464304600740349, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 1872, | |
| "total_flos": 2.9809868531482165e+19, | |
| "train_loss": 0.4726643766093458, | |
| "train_runtime": 14997.7554, | |
| "train_samples_per_second": 3.994, | |
| "train_steps_per_second": 0.125 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1872, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.9809868531482165e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |