| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "global_step": 3900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 1.3439999999999998e-05, | |
| "loss": 21.7509, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "learning_rate": 2.7439999999999998e-05, | |
| "loss": 5.8287, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 4.1439999999999996e-05, | |
| "loss": 3.6855, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 10.26, | |
| "learning_rate": 5.544e-05, | |
| "loss": 3.0684, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 12.82, | |
| "learning_rate": 6.944e-05, | |
| "loss": 2.9421, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 12.82, | |
| "eval_loss": 2.889408826828003, | |
| "eval_runtime": 17.3993, | |
| "eval_samples_per_second": 25.288, | |
| "eval_steps_per_second": 1.609, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "learning_rate": 6.802352941176471e-05, | |
| "loss": 2.847, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 17.95, | |
| "learning_rate": 6.596470588235294e-05, | |
| "loss": 2.5861, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 20.51, | |
| "learning_rate": 6.390588235294118e-05, | |
| "loss": 1.8427, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 23.08, | |
| "learning_rate": 6.184705882352941e-05, | |
| "loss": 1.3659, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 25.64, | |
| "learning_rate": 5.978823529411764e-05, | |
| "loss": 1.1872, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 25.64, | |
| "eval_loss": 0.6688352227210999, | |
| "eval_runtime": 17.312, | |
| "eval_samples_per_second": 25.416, | |
| "eval_steps_per_second": 1.617, | |
| "eval_wer": 0.7459733087896917, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 28.21, | |
| "learning_rate": 5.772941176470588e-05, | |
| "loss": 1.087, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 30.77, | |
| "learning_rate": 5.5670588235294114e-05, | |
| "loss": 1.0304, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 5.361176470588235e-05, | |
| "loss": 0.9762, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 35.9, | |
| "learning_rate": 5.1552941176470585e-05, | |
| "loss": 0.9043, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "learning_rate": 4.949411764705882e-05, | |
| "loss": 0.8894, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "eval_loss": 0.4868045449256897, | |
| "eval_runtime": 16.7955, | |
| "eval_samples_per_second": 26.197, | |
| "eval_steps_per_second": 1.667, | |
| "eval_wer": 0.6516336861481823, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 41.03, | |
| "learning_rate": 4.7435294117647055e-05, | |
| "loss": 0.861, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 43.59, | |
| "learning_rate": 4.5376470588235284e-05, | |
| "loss": 0.8411, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 46.15, | |
| "learning_rate": 4.331764705882352e-05, | |
| "loss": 0.8349, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 48.72, | |
| "learning_rate": 4.1258823529411755e-05, | |
| "loss": 0.7912, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 51.28, | |
| "learning_rate": 3.92e-05, | |
| "loss": 0.769, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 51.28, | |
| "eval_loss": 0.495985209941864, | |
| "eval_runtime": 17.6095, | |
| "eval_samples_per_second": 24.987, | |
| "eval_steps_per_second": 1.59, | |
| "eval_wer": 0.650713299585826, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 53.85, | |
| "learning_rate": 3.714117647058823e-05, | |
| "loss": 0.7639, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 56.41, | |
| "learning_rate": 3.508235294117647e-05, | |
| "loss": 0.7421, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 58.97, | |
| "learning_rate": 3.3023529411764704e-05, | |
| "loss": 0.728, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 61.54, | |
| "learning_rate": 3.096470588235294e-05, | |
| "loss": 0.7105, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 64.1, | |
| "learning_rate": 2.8905882352941175e-05, | |
| "loss": 0.6936, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 64.1, | |
| "eval_loss": 0.47812411189079285, | |
| "eval_runtime": 17.7238, | |
| "eval_samples_per_second": 24.825, | |
| "eval_steps_per_second": 1.58, | |
| "eval_wer": 0.5384261389783709, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 66.67, | |
| "learning_rate": 2.684705882352941e-05, | |
| "loss": 0.6813, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 69.23, | |
| "learning_rate": 2.4788235294117646e-05, | |
| "loss": 0.6514, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 71.79, | |
| "learning_rate": 2.272941176470588e-05, | |
| "loss": 0.6539, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 74.36, | |
| "learning_rate": 2.0670588235294117e-05, | |
| "loss": 0.6423, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 76.92, | |
| "learning_rate": 1.8611764705882352e-05, | |
| "loss": 0.624, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 76.92, | |
| "eval_loss": 0.4642501175403595, | |
| "eval_runtime": 16.6285, | |
| "eval_samples_per_second": 26.461, | |
| "eval_steps_per_second": 1.684, | |
| "eval_wer": 0.5430280717901519, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 79.49, | |
| "learning_rate": 1.6552941176470588e-05, | |
| "loss": 0.624, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 82.05, | |
| "learning_rate": 1.4494117647058823e-05, | |
| "loss": 0.6071, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 84.62, | |
| "learning_rate": 1.2435294117647057e-05, | |
| "loss": 0.6175, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 87.18, | |
| "learning_rate": 1.0376470588235292e-05, | |
| "loss": 0.6023, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 89.74, | |
| "learning_rate": 8.317647058823528e-06, | |
| "loss": 0.5966, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 89.74, | |
| "eval_loss": 0.4529646337032318, | |
| "eval_runtime": 16.5674, | |
| "eval_samples_per_second": 26.558, | |
| "eval_steps_per_second": 1.69, | |
| "eval_wer": 0.5591348366313852, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 92.31, | |
| "learning_rate": 6.258823529411764e-06, | |
| "loss": 0.5868, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 94.87, | |
| "learning_rate": 4.2e-06, | |
| "loss": 0.5817, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 97.44, | |
| "learning_rate": 2.141176470588235e-06, | |
| "loss": 0.5687, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 8.235294117647057e-08, | |
| "loss": 0.5709, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 3900, | |
| "total_flos": 1.7166559536300583e+19, | |
| "train_loss": 1.7521888146033653, | |
| "train_runtime": 6573.2529, | |
| "train_samples_per_second": 18.56, | |
| "train_steps_per_second": 0.593 | |
| } | |
| ], | |
| "max_steps": 3900, | |
| "num_train_epochs": 100, | |
| "total_flos": 1.7166559536300583e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |