| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 49.994535519125684, | |
| "global_step": 4550, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 3.7499999999999997e-06, | |
| "loss": 14.8405, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 7.499999999999999e-06, | |
| "loss": 6.8597, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "learning_rate": 1.1249999999999999e-05, | |
| "loss": 4.4132, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "learning_rate": 1.4999999999999999e-05, | |
| "loss": 3.7795, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 5.49, | |
| "learning_rate": 1.875e-05, | |
| "loss": 3.4129, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.49, | |
| "eval_loss": 3.3224499225616455, | |
| "eval_runtime": 168.3653, | |
| "eval_samples_per_second": 28.765, | |
| "eval_steps_per_second": 0.903, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 6.59, | |
| "learning_rate": 2.2499999999999998e-05, | |
| "loss": 3.2045, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 2.6249999999999998e-05, | |
| "loss": 3.142, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "learning_rate": 2.9999999999999997e-05, | |
| "loss": 3.0872, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 9.89, | |
| "learning_rate": 3.375e-05, | |
| "loss": 2.9848, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 10.98, | |
| "learning_rate": 3.75e-05, | |
| "loss": 2.9323, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 10.98, | |
| "eval_loss": 2.9127891063690186, | |
| "eval_runtime": 166.7051, | |
| "eval_samples_per_second": 29.051, | |
| "eval_steps_per_second": 0.912, | |
| "eval_wer": 1.0000283848992335, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 12.09, | |
| "learning_rate": 4.125e-05, | |
| "loss": 2.9013, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 13.19, | |
| "learning_rate": 4.4999999999999996e-05, | |
| "loss": 2.729, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 14.28, | |
| "learning_rate": 4.875e-05, | |
| "loss": 2.1939, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "learning_rate": 5.2499999999999995e-05, | |
| "loss": 1.8125, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 16.48, | |
| "learning_rate": 5.625e-05, | |
| "loss": 1.6839, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 16.48, | |
| "eval_loss": 0.7740097641944885, | |
| "eval_runtime": 167.3266, | |
| "eval_samples_per_second": 28.943, | |
| "eval_steps_per_second": 0.908, | |
| "eval_wer": 0.685381776894692, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 17.58, | |
| "learning_rate": 5.9999999999999995e-05, | |
| "loss": 1.6194, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 18.68, | |
| "learning_rate": 6.374999999999999e-05, | |
| "loss": 1.5613, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 19.78, | |
| "learning_rate": 6.75e-05, | |
| "loss": 1.5197, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 20.87, | |
| "learning_rate": 7.1175e-05, | |
| "loss": 1.5009, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 21.97, | |
| "learning_rate": 7.492499999999999e-05, | |
| "loss": 1.485, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 21.97, | |
| "eval_loss": 0.5829736590385437, | |
| "eval_runtime": 166.3956, | |
| "eval_samples_per_second": 29.105, | |
| "eval_steps_per_second": 0.913, | |
| "eval_wer": 0.597615668464377, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 23.08, | |
| "learning_rate": 7.211764705882351e-05, | |
| "loss": 1.4606, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 24.17, | |
| "learning_rate": 6.920588235294117e-05, | |
| "loss": 1.4257, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 25.27, | |
| "learning_rate": 6.626470588235294e-05, | |
| "loss": 1.4002, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 26.37, | |
| "learning_rate": 6.33235294117647e-05, | |
| "loss": 1.3856, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 27.47, | |
| "learning_rate": 6.038235294117646e-05, | |
| "loss": 1.362, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 27.47, | |
| "eval_loss": 0.48657190799713135, | |
| "eval_runtime": 166.2387, | |
| "eval_samples_per_second": 29.133, | |
| "eval_steps_per_second": 0.914, | |
| "eval_wer": 0.490519443655975, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 28.57, | |
| "learning_rate": 5.744117647058823e-05, | |
| "loss": 1.3342, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 29.67, | |
| "learning_rate": 5.4499999999999997e-05, | |
| "loss": 1.318, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 30.77, | |
| "learning_rate": 5.155882352941176e-05, | |
| "loss": 1.3013, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 31.86, | |
| "learning_rate": 4.861764705882352e-05, | |
| "loss": 1.2849, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 32.96, | |
| "learning_rate": 4.567647058823529e-05, | |
| "loss": 1.2752, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 32.96, | |
| "eval_loss": 0.42398178577423096, | |
| "eval_runtime": 167.0219, | |
| "eval_samples_per_second": 28.996, | |
| "eval_steps_per_second": 0.91, | |
| "eval_wer": 0.4966789667896679, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 34.07, | |
| "learning_rate": 4.2735294117647056e-05, | |
| "loss": 1.2682, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 35.16, | |
| "learning_rate": 3.979411764705882e-05, | |
| "loss": 1.2428, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 36.26, | |
| "learning_rate": 3.6852941176470586e-05, | |
| "loss": 1.2282, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 37.36, | |
| "learning_rate": 3.391176470588235e-05, | |
| "loss": 1.2198, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "learning_rate": 3.0970588235294116e-05, | |
| "loss": 1.1957, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "eval_loss": 0.38985687494277954, | |
| "eval_runtime": 166.2154, | |
| "eval_samples_per_second": 29.137, | |
| "eval_steps_per_second": 0.914, | |
| "eval_wer": 0.42577348850411584, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 39.56, | |
| "learning_rate": 2.8029411764705878e-05, | |
| "loss": 1.1891, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 40.66, | |
| "learning_rate": 2.5088235294117646e-05, | |
| "loss": 1.1847, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 41.75, | |
| "learning_rate": 2.2147058823529408e-05, | |
| "loss": 1.1779, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 42.85, | |
| "learning_rate": 1.9205882352941176e-05, | |
| "loss": 1.1654, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 43.95, | |
| "learning_rate": 1.626470588235294e-05, | |
| "loss": 1.1646, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 43.95, | |
| "eval_loss": 0.3597247898578644, | |
| "eval_runtime": 165.7161, | |
| "eval_samples_per_second": 29.225, | |
| "eval_steps_per_second": 0.917, | |
| "eval_wer": 0.401447629860914, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 45.05, | |
| "learning_rate": 1.3323529411764704e-05, | |
| "loss": 1.1574, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 46.15, | |
| "learning_rate": 1.0411764705882353e-05, | |
| "loss": 1.1513, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 47.25, | |
| "learning_rate": 7.470588235294117e-06, | |
| "loss": 1.1412, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 48.35, | |
| "learning_rate": 4.529411764705882e-06, | |
| "loss": 1.1383, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 49.45, | |
| "learning_rate": 1.5882352941176468e-06, | |
| "loss": 1.1265, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 49.45, | |
| "eval_loss": 0.35590532422065735, | |
| "eval_runtime": 165.6811, | |
| "eval_samples_per_second": 29.231, | |
| "eval_steps_per_second": 0.917, | |
| "eval_wer": 0.38288390576213455, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 49.99, | |
| "step": 4550, | |
| "total_flos": 7.032770514539837e+19, | |
| "train_loss": 2.1873197584885817, | |
| "train_runtime": 245048.5131, | |
| "train_samples_per_second": 2.385, | |
| "train_steps_per_second": 0.019 | |
| } | |
| ], | |
| "max_steps": 4550, | |
| "num_train_epochs": 50, | |
| "total_flos": 7.032770514539837e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |