| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 21.231422505307854, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.06, | |
| "eval_loss": 2.901623249053955, | |
| "eval_runtime": 78.4215, | |
| "eval_samples_per_second": 35.985, | |
| "eval_steps_per_second": 3.009, | |
| "eval_wer": 0.9995142002989537, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 9.970000000000001e-05, | |
| "loss": 2.877, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "eval_loss": 0.981178343296051, | |
| "eval_runtime": 78.2116, | |
| "eval_samples_per_second": 36.082, | |
| "eval_steps_per_second": 3.017, | |
| "eval_wer": 0.3484678624813154, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "eval_loss": 0.784228503704071, | |
| "eval_runtime": 78.8199, | |
| "eval_samples_per_second": 35.803, | |
| "eval_steps_per_second": 2.994, | |
| "eval_wer": 0.273168908819133, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "learning_rate": 8.892222222222223e-05, | |
| "loss": 0.7834, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "eval_loss": 0.6962451934814453, | |
| "eval_runtime": 78.959, | |
| "eval_samples_per_second": 35.74, | |
| "eval_steps_per_second": 2.989, | |
| "eval_wer": 0.21917040358744394, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.31, | |
| "eval_loss": 0.6526582837104797, | |
| "eval_runtime": 78.1018, | |
| "eval_samples_per_second": 36.132, | |
| "eval_steps_per_second": 3.022, | |
| "eval_wer": 0.20418535127055307, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.37, | |
| "learning_rate": 7.78111111111111e-05, | |
| "loss": 0.6084, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 6.37, | |
| "eval_loss": 0.6220057606697083, | |
| "eval_runtime": 79.4519, | |
| "eval_samples_per_second": 35.518, | |
| "eval_steps_per_second": 2.97, | |
| "eval_wer": 0.19715994020926755, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.43, | |
| "eval_loss": 0.6441554427146912, | |
| "eval_runtime": 80.6733, | |
| "eval_samples_per_second": 34.981, | |
| "eval_steps_per_second": 2.925, | |
| "eval_wer": 0.19338565022421525, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "learning_rate": 6.670000000000001e-05, | |
| "loss": 0.5147, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "eval_loss": 0.6793447732925415, | |
| "eval_runtime": 80.545, | |
| "eval_samples_per_second": 35.036, | |
| "eval_steps_per_second": 2.93, | |
| "eval_wer": 0.19495515695067264, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 9.55, | |
| "eval_loss": 0.6431596875190735, | |
| "eval_runtime": 83.1308, | |
| "eval_samples_per_second": 33.947, | |
| "eval_steps_per_second": 2.839, | |
| "eval_wer": 0.19200298953662182, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 10.62, | |
| "learning_rate": 5.558888888888889e-05, | |
| "loss": 0.4566, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 10.62, | |
| "eval_loss": 0.6605134010314941, | |
| "eval_runtime": 80.2378, | |
| "eval_samples_per_second": 35.17, | |
| "eval_steps_per_second": 2.941, | |
| "eval_wer": 0.1853139013452915, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 11.68, | |
| "eval_loss": 0.6392588019371033, | |
| "eval_runtime": 78.2211, | |
| "eval_samples_per_second": 36.077, | |
| "eval_steps_per_second": 3.017, | |
| "eval_wer": 0.18658445440956653, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 12.74, | |
| "learning_rate": 4.448888888888889e-05, | |
| "loss": 0.4155, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 12.74, | |
| "eval_loss": 0.6917754411697388, | |
| "eval_runtime": 78.1406, | |
| "eval_samples_per_second": 36.114, | |
| "eval_steps_per_second": 3.02, | |
| "eval_wer": 0.18030642750373693, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 13.8, | |
| "eval_loss": 0.6514243483543396, | |
| "eval_runtime": 78.7289, | |
| "eval_samples_per_second": 35.845, | |
| "eval_steps_per_second": 2.998, | |
| "eval_wer": 0.17911061285500748, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 14.86, | |
| "learning_rate": 3.337777777777778e-05, | |
| "loss": 0.372, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 14.86, | |
| "eval_loss": 0.7009950876235962, | |
| "eval_runtime": 78.8286, | |
| "eval_samples_per_second": 35.799, | |
| "eval_steps_per_second": 2.994, | |
| "eval_wer": 0.1851270553064275, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 15.92, | |
| "eval_loss": 0.6824212670326233, | |
| "eval_runtime": 79.8558, | |
| "eval_samples_per_second": 35.339, | |
| "eval_steps_per_second": 2.955, | |
| "eval_wer": 0.17858744394618833, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 16.99, | |
| "learning_rate": 2.2277777777777778e-05, | |
| "loss": 0.3368, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 16.99, | |
| "eval_loss": 0.6895105838775635, | |
| "eval_runtime": 81.107, | |
| "eval_samples_per_second": 34.794, | |
| "eval_steps_per_second": 2.91, | |
| "eval_wer": 0.1780269058295964, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 18.05, | |
| "eval_loss": 0.7150452136993408, | |
| "eval_runtime": 79.9454, | |
| "eval_samples_per_second": 35.299, | |
| "eval_steps_per_second": 2.952, | |
| "eval_wer": 0.17589686098654708, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 19.11, | |
| "learning_rate": 1.1177777777777779e-05, | |
| "loss": 0.3244, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 19.11, | |
| "eval_loss": 0.7141156792640686, | |
| "eval_runtime": 85.0058, | |
| "eval_samples_per_second": 33.198, | |
| "eval_steps_per_second": 2.776, | |
| "eval_wer": 0.17589686098654708, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 20.17, | |
| "eval_loss": 0.7225230932235718, | |
| "eval_runtime": 77.8748, | |
| "eval_samples_per_second": 36.238, | |
| "eval_steps_per_second": 3.031, | |
| "eval_wer": 0.17556053811659192, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 21.23, | |
| "learning_rate": 6.666666666666667e-08, | |
| "loss": 0.2981, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 21.23, | |
| "eval_loss": 0.7286972403526306, | |
| "eval_runtime": 78.6838, | |
| "eval_samples_per_second": 35.865, | |
| "eval_steps_per_second": 2.999, | |
| "eval_wer": 0.17556053811659192, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 21.23, | |
| "step": 10000, | |
| "total_flos": 2.495144460541511e+19, | |
| "train_loss": 0.6986962738037109, | |
| "train_runtime": 10488.9956, | |
| "train_samples_per_second": 22.881, | |
| "train_steps_per_second": 0.953 | |
| } | |
| ], | |
| "max_steps": 10000, | |
| "num_train_epochs": 22, | |
| "total_flos": 2.495144460541511e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |