| { | |
| "best_metric": 21.278825995807125, | |
| "best_model_checkpoint": "./checkpoint-700", | |
| "epoch": 66.66666666666667, | |
| "global_step": 800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 5.081432784565141e-06, | |
| "loss": 0.9461, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 6.3901594783484795e-06, | |
| "loss": 0.1654, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "learning_rate": 7.1145874994768475e-06, | |
| "loss": 0.0159, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "learning_rate": 7.618084285704403e-06, | |
| "loss": 0.007, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "eval_loss": 0.57275390625, | |
| "eval_runtime": 17.5098, | |
| "eval_samples_per_second": 3.198, | |
| "eval_steps_per_second": 0.228, | |
| "eval_wer": 21.48846960167715, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.42, | |
| "learning_rate": 8.004371064686715e-06, | |
| "loss": 0.0035, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 8.317843796829062e-06, | |
| "loss": 0.0014, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 14.58, | |
| "learning_rate": 8.581647301884627e-06, | |
| "loss": 0.0009, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 8.809389787307027e-06, | |
| "loss": 0.0007, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "eval_loss": 0.70166015625, | |
| "eval_runtime": 17.8855, | |
| "eval_samples_per_second": 3.131, | |
| "eval_steps_per_second": 0.224, | |
| "eval_wer": 22.11740041928721, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 18.75, | |
| "learning_rate": 9.009754892271524e-06, | |
| "loss": 0.0006, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 20.83, | |
| "learning_rate": 9.18862407120431e-06, | |
| "loss": 0.0005, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 22.92, | |
| "learning_rate": 9.350166018396598e-06, | |
| "loss": 0.0004, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 9.497443253787987e-06, | |
| "loss": 0.0003, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_loss": 0.73583984375, | |
| "eval_runtime": 17.8146, | |
| "eval_samples_per_second": 3.143, | |
| "eval_steps_per_second": 0.225, | |
| "eval_wer": 21.59329140461216, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 27.08, | |
| "learning_rate": 9.632771883655444e-06, | |
| "loss": 0.0003, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 29.17, | |
| "learning_rate": 9.75794603988726e-06, | |
| "loss": 0.0003, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 31.25, | |
| "learning_rate": 9.87438382734506e-06, | |
| "loss": 0.0002, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 9.983225577361795e-06, | |
| "loss": 0.0002, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "eval_loss": 0.759765625, | |
| "eval_runtime": 17.9019, | |
| "eval_samples_per_second": 3.128, | |
| "eval_steps_per_second": 0.223, | |
| "eval_wer": 21.59329140461216, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 35.42, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.0002, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 37.5, | |
| "learning_rate": 8.875e-06, | |
| "loss": 0.0002, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 39.58, | |
| "learning_rate": 8.25e-06, | |
| "loss": 0.0002, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 41.67, | |
| "learning_rate": 7.625e-06, | |
| "loss": 0.0002, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 41.67, | |
| "eval_loss": 0.779296875, | |
| "eval_runtime": 17.9031, | |
| "eval_samples_per_second": 3.128, | |
| "eval_steps_per_second": 0.223, | |
| "eval_wer": 22.0125786163522, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 43.75, | |
| "learning_rate": 7e-06, | |
| "loss": 0.0001, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 45.83, | |
| "learning_rate": 6.375e-06, | |
| "loss": 0.0001, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 47.92, | |
| "learning_rate": 5.75e-06, | |
| "loss": 0.0001, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 5.125e-06, | |
| "loss": 0.0001, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_loss": 0.78955078125, | |
| "eval_runtime": 17.7601, | |
| "eval_samples_per_second": 3.153, | |
| "eval_steps_per_second": 0.225, | |
| "eval_wer": 22.0125786163522, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 52.08, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.0001, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 54.17, | |
| "learning_rate": 3.875e-06, | |
| "loss": 0.0001, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 56.25, | |
| "learning_rate": 3.2500000000000002e-06, | |
| "loss": 0.0001, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 58.33, | |
| "learning_rate": 2.6250000000000003e-06, | |
| "loss": 0.0001, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 58.33, | |
| "eval_loss": 0.796875, | |
| "eval_runtime": 17.8587, | |
| "eval_samples_per_second": 3.136, | |
| "eval_steps_per_second": 0.224, | |
| "eval_wer": 21.278825995807125, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 60.42, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.0001, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 62.5, | |
| "learning_rate": 1.3750000000000002e-06, | |
| "loss": 0.0001, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 64.58, | |
| "learning_rate": 7.5e-07, | |
| "loss": 0.0001, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 66.67, | |
| "learning_rate": 1.2500000000000002e-07, | |
| "loss": 0.0001, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 66.67, | |
| "eval_loss": 0.79931640625, | |
| "eval_runtime": 17.7869, | |
| "eval_samples_per_second": 3.148, | |
| "eval_steps_per_second": 0.225, | |
| "eval_wer": 21.278825995807125, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 66.67, | |
| "step": 800, | |
| "total_flos": 2.140496152649437e+20, | |
| "train_loss": 0.035802456135861575, | |
| "train_runtime": 14846.7625, | |
| "train_samples_per_second": 6.897, | |
| "train_steps_per_second": 0.054 | |
| } | |
| ], | |
| "max_steps": 800, | |
| "num_train_epochs": 67, | |
| "total_flos": 2.140496152649437e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |