| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.029961549345007242, | |
| "eval_steps": 1000, | |
| "global_step": 3600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 8.322652595835345e-06, | |
| "grad_norm": 14.9375, | |
| "learning_rate": 2e-06, | |
| "loss": 0.6484, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0008322652595835345, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.0002, | |
| "loss": 0.4345, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.001664530519167069, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.0004, | |
| "loss": 0.2537, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.0024967957787506035, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 0.0006, | |
| "loss": 0.2344, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.003329061038334138, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.0008, | |
| "loss": 0.222, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.004161326297917672, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 0.001, | |
| "loss": 0.2142, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.004993591557501207, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 0.0012, | |
| "loss": 0.2011, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.005825856817084741, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 0.0014, | |
| "loss": 0.1897, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.006658122076668276, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 0.0016, | |
| "loss": 0.1775, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.0074903873362518105, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 0.0018000000000000002, | |
| "loss": 0.1615, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.008322652595835344, | |
| "grad_norm": 0.083984375, | |
| "learning_rate": 0.002, | |
| "loss": 0.1444, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.008322652595835344, | |
| "eval_peoplespeech-clean-transcription_loss": 2.2110702991485596, | |
| "eval_peoplespeech-clean-transcription_model_preparation_time": 0.0065, | |
| "eval_peoplespeech-clean-transcription_runtime": 9.5405, | |
| "eval_peoplespeech-clean-transcription_samples_per_second": 6.708, | |
| "eval_peoplespeech-clean-transcription_steps_per_second": 0.839, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.009154917855418878, | |
| "grad_norm": 0.0771484375, | |
| "learning_rate": 0.001999725185109816, | |
| "loss": 0.1323, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.009987183115002414, | |
| "grad_norm": 0.083984375, | |
| "learning_rate": 0.0019989008914857113, | |
| "loss": 0.1242, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.010819448374585948, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 0.00199752757218401, | |
| "loss": 0.1162, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.011651713634169482, | |
| "grad_norm": 0.07275390625, | |
| "learning_rate": 0.001995605982021898, | |
| "loss": 0.1128, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.012483978893753018, | |
| "grad_norm": 0.07666015625, | |
| "learning_rate": 0.0019931371771625545, | |
| "loss": 0.1094, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.013316244153336551, | |
| "grad_norm": 0.06201171875, | |
| "learning_rate": 0.001990122514534651, | |
| "loss": 0.1052, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.014148509412920085, | |
| "grad_norm": 0.058837890625, | |
| "learning_rate": 0.0019865636510865464, | |
| "loss": 0.1022, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.014980774672503621, | |
| "grad_norm": 0.0625, | |
| "learning_rate": 0.001982462542875576, | |
| "loss": 0.1011, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.015813039932087155, | |
| "grad_norm": 0.06787109375, | |
| "learning_rate": 0.001977821443992945, | |
| "loss": 0.0983, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.01664530519167069, | |
| "grad_norm": 0.0537109375, | |
| "learning_rate": 0.001972642905324813, | |
| "loss": 0.0975, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.01664530519167069, | |
| "eval_peoplespeech-clean-transcription_loss": 1.6914767026901245, | |
| "eval_peoplespeech-clean-transcription_model_preparation_time": 0.0065, | |
| "eval_peoplespeech-clean-transcription_runtime": 9.5301, | |
| "eval_peoplespeech-clean-transcription_samples_per_second": 6.716, | |
| "eval_peoplespeech-clean-transcription_steps_per_second": 0.839, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.017477570451254223, | |
| "grad_norm": 0.05859375, | |
| "learning_rate": 0.0019669297731502505, | |
| "loss": 0.0947, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.018309835710837757, | |
| "grad_norm": 0.062255859375, | |
| "learning_rate": 0.00196068518757684, | |
| "loss": 0.0935, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.019142100970421294, | |
| "grad_norm": 0.059326171875, | |
| "learning_rate": 0.001953912580814779, | |
| "loss": 0.0911, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.019974366230004828, | |
| "grad_norm": 0.060302734375, | |
| "learning_rate": 0.0019466156752904343, | |
| "loss": 0.0904, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.020806631489588362, | |
| "grad_norm": 0.0615234375, | |
| "learning_rate": 0.0019387984816003866, | |
| "loss": 0.0882, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.021638896749171896, | |
| "grad_norm": 0.056884765625, | |
| "learning_rate": 0.0019304652963070869, | |
| "loss": 0.0874, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.02247116200875543, | |
| "grad_norm": 0.049072265625, | |
| "learning_rate": 0.0019216206995773372, | |
| "loss": 0.0871, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.023303427268338964, | |
| "grad_norm": 0.053955078125, | |
| "learning_rate": 0.0019122695526648968, | |
| "loss": 0.0856, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.0241356925279225, | |
| "grad_norm": 0.05517578125, | |
| "learning_rate": 0.0019024169952385887, | |
| "loss": 0.0845, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.024967957787506035, | |
| "grad_norm": 0.0546875, | |
| "learning_rate": 0.0018920684425573864, | |
| "loss": 0.0852, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.024967957787506035, | |
| "eval_peoplespeech-clean-transcription_loss": 1.6414048671722412, | |
| "eval_peoplespeech-clean-transcription_model_preparation_time": 0.0065, | |
| "eval_peoplespeech-clean-transcription_runtime": 9.8151, | |
| "eval_peoplespeech-clean-transcription_samples_per_second": 6.521, | |
| "eval_peoplespeech-clean-transcription_steps_per_second": 0.815, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.02580022304708957, | |
| "grad_norm": 0.052978515625, | |
| "learning_rate": 0.0018812295824940284, | |
| "loss": 0.0836, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.026632488306673103, | |
| "grad_norm": 0.05126953125, | |
| "learning_rate": 0.0018699063724087904, | |
| "loss": 0.0821, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.027464753566256637, | |
| "grad_norm": 0.056884765625, | |
| "learning_rate": 0.0018581050358751443, | |
| "loss": 0.0816, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.02829701882584017, | |
| "grad_norm": 0.045654296875, | |
| "learning_rate": 0.0018458320592590974, | |
| "loss": 0.0814, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.029129284085423705, | |
| "grad_norm": 0.04638671875, | |
| "learning_rate": 0.0018330941881540914, | |
| "loss": 0.0791, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.029961549345007242, | |
| "grad_norm": 0.043212890625, | |
| "learning_rate": 0.0018198984236734246, | |
| "loss": 0.0804, | |
| "step": 3600 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 14400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 3600, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.3112799838724096e+17, | |
| "train_batch_size": 24, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |