| { | |
| "best_metric": 0.5890884399414062, | |
| "best_model_checkpoint": "./checkpoint-1000", | |
| "epoch": 200.0, | |
| "global_step": 1200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 8.185000000000001e-05, | |
| "loss": 9.167, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "eval_loss": 3.5598716735839844, | |
| "eval_runtime": 9.2571, | |
| "eval_samples_per_second": 36.188, | |
| "eval_steps_per_second": 0.648, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 0.0001, | |
| "loss": 3.2645, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "eval_loss": 3.1771419048309326, | |
| "eval_runtime": 9.1654, | |
| "eval_samples_per_second": 36.55, | |
| "eval_steps_per_second": 0.655, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 0.0001, | |
| "loss": 3.1509, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_loss": 3.132053852081299, | |
| "eval_runtime": 9.2483, | |
| "eval_samples_per_second": 36.223, | |
| "eval_steps_per_second": 0.649, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 66.67, | |
| "learning_rate": 0.0001, | |
| "loss": 3.0757, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 66.67, | |
| "eval_loss": 2.8593921661376953, | |
| "eval_runtime": 9.1599, | |
| "eval_samples_per_second": 36.572, | |
| "eval_steps_per_second": 0.655, | |
| "eval_wer": 1.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 83.33, | |
| "learning_rate": 0.0001, | |
| "loss": 2.5274, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 83.33, | |
| "eval_loss": 1.5286245346069336, | |
| "eval_runtime": 9.3823, | |
| "eval_samples_per_second": 35.705, | |
| "eval_steps_per_second": 0.639, | |
| "eval_wer": 0.9797033567525371, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 0.0001, | |
| "loss": 1.6826, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "eval_loss": 0.805767834186554, | |
| "eval_runtime": 10.0486, | |
| "eval_samples_per_second": 33.338, | |
| "eval_steps_per_second": 0.597, | |
| "eval_wer": 0.797423887587822, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 116.67, | |
| "learning_rate": 8.448333333333333e-05, | |
| "loss": 1.2868, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 116.67, | |
| "eval_loss": 0.6712858080863953, | |
| "eval_runtime": 9.5622, | |
| "eval_samples_per_second": 35.034, | |
| "eval_steps_per_second": 0.627, | |
| "eval_wer": 0.7279469164715067, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 133.33, | |
| "learning_rate": 6.865e-05, | |
| "loss": 1.1262, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 133.33, | |
| "eval_loss": 0.630766749382019, | |
| "eval_runtime": 9.2712, | |
| "eval_samples_per_second": 36.133, | |
| "eval_steps_per_second": 0.647, | |
| "eval_wer": 0.7033567525370804, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "learning_rate": 5.281666666666667e-05, | |
| "loss": 1.0408, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "eval_loss": 0.6055771708488464, | |
| "eval_runtime": 9.2433, | |
| "eval_samples_per_second": 36.242, | |
| "eval_steps_per_second": 0.649, | |
| "eval_wer": 0.6744730679156908, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 166.67, | |
| "learning_rate": 3.6983333333333336e-05, | |
| "loss": 0.9617, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 166.67, | |
| "eval_loss": 0.5890884399414062, | |
| "eval_runtime": 9.1774, | |
| "eval_samples_per_second": 36.503, | |
| "eval_steps_per_second": 0.654, | |
| "eval_wer": 0.6569086651053864, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 183.33, | |
| "learning_rate": 2.1150000000000002e-05, | |
| "loss": 0.9196, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 183.33, | |
| "eval_loss": 0.5912845730781555, | |
| "eval_runtime": 9.2459, | |
| "eval_samples_per_second": 36.232, | |
| "eval_steps_per_second": 0.649, | |
| "eval_wer": 0.6432474629195941, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "learning_rate": 5.316666666666669e-06, | |
| "loss": 0.8853, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "eval_loss": 0.5924161672592163, | |
| "eval_runtime": 9.2278, | |
| "eval_samples_per_second": 36.303, | |
| "eval_steps_per_second": 0.65, | |
| "eval_wer": 0.6346604215456675, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "step": 1200, | |
| "total_flos": 3.638865982927399e+19, | |
| "train_loss": 2.424034964243571, | |
| "train_runtime": 7514.3416, | |
| "train_samples_per_second": 20.441, | |
| "train_steps_per_second": 0.16 | |
| } | |
| ], | |
| "max_steps": 1200, | |
| "num_train_epochs": 200, | |
| "total_flos": 3.638865982927399e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |