| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.4367816091954024, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 2.7599999999999997e-05, | |
| "loss": 10.9066, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 5.6399999999999995e-05, | |
| "loss": 7.3263, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "eval_cer": 0.9998702550415184, | |
| "eval_loss": 5.000309467315674, | |
| "eval_runtime": 31.7393, | |
| "eval_samples_per_second": 14.399, | |
| "eval_steps_per_second": 0.914, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 8.639999999999999e-05, | |
| "loss": 3.4967, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 0.0001164, | |
| "loss": 3.1464, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "eval_cer": 0.9998702550415184, | |
| "eval_loss": 3.5314910411834717, | |
| "eval_runtime": 31.3912, | |
| "eval_samples_per_second": 14.558, | |
| "eval_steps_per_second": 0.924, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 0.00014639999999999998, | |
| "loss": 3.1238, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 0.00017639999999999998, | |
| "loss": 3.1152, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "eval_cer": 0.9998702550415184, | |
| "eval_loss": 3.6040682792663574, | |
| "eval_runtime": 31.267, | |
| "eval_samples_per_second": 14.616, | |
| "eval_steps_per_second": 0.927, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 0.00020639999999999998, | |
| "loss": 3.088, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 0.0002364, | |
| "loss": 3.0626, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_cer": 0.9743290332147093, | |
| "eval_loss": 3.385751247406006, | |
| "eval_runtime": 31.1188, | |
| "eval_samples_per_second": 14.686, | |
| "eval_steps_per_second": 0.932, | |
| "eval_wer": 0.9992471013401596, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 0.00026639999999999997, | |
| "loss": 3.032, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 0.0002964, | |
| "loss": 2.3387, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "eval_cer": 0.3497553380782918, | |
| "eval_loss": 1.502172589302063, | |
| "eval_runtime": 31.3947, | |
| "eval_samples_per_second": 14.557, | |
| "eval_steps_per_second": 0.924, | |
| "eval_wer": 1.0904984189128144, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 0.00029168765743073046, | |
| "loss": 1.1185, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 0.0002822418136020151, | |
| "loss": 0.7737, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "eval_cer": 0.21517274614472123, | |
| "eval_loss": 0.6938613057136536, | |
| "eval_runtime": 31.1722, | |
| "eval_samples_per_second": 14.661, | |
| "eval_steps_per_second": 0.93, | |
| "eval_wer": 0.872910706218943, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 0.00027279596977329974, | |
| "loss": 0.647, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 0.00026335012594458433, | |
| "loss": 0.5643, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "eval_cer": 0.15467452550415184, | |
| "eval_loss": 0.4780799448490143, | |
| "eval_runtime": 30.8612, | |
| "eval_samples_per_second": 14.808, | |
| "eval_steps_per_second": 0.94, | |
| "eval_wer": 0.7375395271796417, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 0.00025390428211586897, | |
| "loss": 0.5188, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 0.0002444584382871536, | |
| "loss": 0.4762, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "eval_cer": 0.1396055753262159, | |
| "eval_loss": 0.41990911960601807, | |
| "eval_runtime": 31.306, | |
| "eval_samples_per_second": 14.598, | |
| "eval_steps_per_second": 0.926, | |
| "eval_wer": 0.6905586508056015, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 0.00023501259445843828, | |
| "loss": 0.4614, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 0.0002255667506297229, | |
| "loss": 0.4371, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "eval_cer": 0.12713152431791222, | |
| "eval_loss": 0.382140189409256, | |
| "eval_runtime": 31.4276, | |
| "eval_samples_per_second": 14.541, | |
| "eval_steps_per_second": 0.923, | |
| "eval_wer": 0.6390603824725192, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 0.00021612090680100753, | |
| "loss": 0.4259, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 0.00020667506297229217, | |
| "loss": 0.4138, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "eval_cer": 0.12171930604982206, | |
| "eval_loss": 0.36281564831733704, | |
| "eval_runtime": 31.2629, | |
| "eval_samples_per_second": 14.618, | |
| "eval_steps_per_second": 0.928, | |
| "eval_wer": 0.6143653064297545, | |
| "step": 1000 | |
| } | |
| ], | |
| "max_steps": 2088, | |
| "num_train_epochs": 3, | |
| "total_flos": 8.198176189032398e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |