| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "global_step": 4000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 0.00019800000000000002, | |
| "loss": 4.9661, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 0.000398, | |
| "loss": 2.9221, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "learning_rate": 0.000598, | |
| "loss": 1.6582, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.0007980000000000001, | |
| "loss": 0.4613, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 0.000998, | |
| "loss": 0.3082, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "eval_loss": 0.38708925247192383, | |
| "eval_runtime": 140.6014, | |
| "eval_samples_per_second": 18.584, | |
| "eval_steps_per_second": 0.263, | |
| "eval_wer": 0.4907113872685772, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "learning_rate": 0.0009980271764103532, | |
| "loss": 0.2368, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 17.5, | |
| "learning_rate": 0.000992044732251972, | |
| "loss": 0.202, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 0.0009821006332271156, | |
| "loss": 0.1745, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 22.5, | |
| "learning_rate": 0.0009682749433740962, | |
| "loss": 0.1594, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 0.0009506789790182364, | |
| "loss": 0.1497, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_loss": 0.41679587960243225, | |
| "eval_runtime": 138.5858, | |
| "eval_samples_per_second": 18.855, | |
| "eval_steps_per_second": 0.267, | |
| "eval_wer": 0.4278468171443064, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 27.5, | |
| "learning_rate": 0.00092945441251827, | |
| "loss": 0.1406, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "learning_rate": 0.0009047721316038118, | |
| "loss": 0.1308, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 32.5, | |
| "learning_rate": 0.0008768308634878388, | |
| "loss": 0.1285, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "learning_rate": 0.0008458555748320216, | |
| "loss": 0.1213, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 37.5, | |
| "learning_rate": 0.0008120956604474414, | |
| "loss": 0.1243, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 37.5, | |
| "eval_loss": 0.4445520341396332, | |
| "eval_runtime": 137.9299, | |
| "eval_samples_per_second": 18.944, | |
| "eval_steps_per_second": 0.268, | |
| "eval_wer": 0.4220136951559726, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "learning_rate": 0.0007758229353142152, | |
| "loss": 0.1215, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 42.5, | |
| "learning_rate": 0.0007373294460870985, | |
| "loss": 0.1185, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "learning_rate": 0.0006969251197075427, | |
| "loss": 0.1064, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 47.5, | |
| "learning_rate": 0.0006549352680541975, | |
| "loss": 0.1085, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 0.000611697968722942, | |
| "loss": 0.0954, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_loss": 0.44264093041419983, | |
| "eval_runtime": 138.1747, | |
| "eval_samples_per_second": 18.911, | |
| "eval_steps_per_second": 0.268, | |
| "eval_wer": 0.39459168146081663, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 52.5, | |
| "learning_rate": 0.0005675613430248713, | |
| "loss": 0.0923, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 55.0, | |
| "learning_rate": 0.0005228807531181908, | |
| "loss": 0.0847, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 57.5, | |
| "learning_rate": 0.00047801594084106763, | |
| "loss": 0.0907, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "learning_rate": 0.0004333281312818746, | |
| "loss": 0.0833, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 62.5, | |
| "learning_rate": 0.00038917712440717607, | |
| "loss": 0.0741, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 62.5, | |
| "eval_loss": 0.4501730799674988, | |
| "eval_runtime": 137.8839, | |
| "eval_samples_per_second": 18.951, | |
| "eval_steps_per_second": 0.268, | |
| "eval_wer": 0.3800405782399188, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 65.0, | |
| "learning_rate": 0.00034591839816395533, | |
| "loss": 0.072, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 67.5, | |
| "learning_rate": 0.00030390024638020374, | |
| "loss": 0.0684, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "learning_rate": 0.0002634609745078109, | |
| "loss": 0.062, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 72.5, | |
| "learning_rate": 0.00022492617578598646, | |
| "loss": 0.06, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "learning_rate": 0.00018860610975594382, | |
| "loss": 0.0533, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "eval_loss": 0.4618484675884247, | |
| "eval_runtime": 138.9018, | |
| "eval_samples_per_second": 18.812, | |
| "eval_steps_per_second": 0.266, | |
| "eval_wer": 0.3653309662693381, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 77.5, | |
| "learning_rate": 0.0001547932042335039, | |
| "loss": 0.0542, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "learning_rate": 0.00012375970085226701, | |
| "loss": 0.0514, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 82.5, | |
| "learning_rate": 9.575546313405425e-05, | |
| "loss": 0.0501, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 85.0, | |
| "learning_rate": 7.100596473474763e-05, | |
| "loss": 0.0466, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 87.5, | |
| "learning_rate": 4.9710474062988955e-05, | |
| "loss": 0.0447, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 87.5, | |
| "eval_loss": 0.4518083930015564, | |
| "eval_runtime": 138.0899, | |
| "eval_samples_per_second": 18.922, | |
| "eval_steps_per_second": 0.268, | |
| "eval_wer": 0.3460563023078874, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "learning_rate": 3.204044988812144e-05, | |
| "loss": 0.0414, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 92.5, | |
| "learning_rate": 1.8138160854995144e-05, | |
| "loss": 0.0415, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 95.0, | |
| "learning_rate": 8.115540020491363e-06, | |
| "loss": 0.0406, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 97.5, | |
| "learning_rate": 2.053283634363745e-06, | |
| "loss": 0.0416, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 2.0142048445803695e-10, | |
| "loss": 0.0396, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "eval_loss": 0.46233049035072327, | |
| "eval_runtime": 138.5502, | |
| "eval_samples_per_second": 18.86, | |
| "eval_steps_per_second": 0.267, | |
| "eval_wer": 0.34199847831600305, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 4000, | |
| "total_flos": 1.2065530437849294e+20, | |
| "train_loss": 0.3406678358316422, | |
| "train_runtime": 40685.8922, | |
| "train_samples_per_second": 14.049, | |
| "train_steps_per_second": 0.098 | |
| } | |
| ], | |
| "max_steps": 4000, | |
| "num_train_epochs": 100, | |
| "total_flos": 1.2065530437849294e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |