| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 35, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 29.255616813764313, | |
| "learning_rate": 4.989935734988098e-06, | |
| "loss": 1.0801, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 2.625827463029639, | |
| "learning_rate": 4.752422169756048e-06, | |
| "loss": 0.4063, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.7421250973739865, | |
| "learning_rate": 4.058724504646834e-06, | |
| "loss": 0.2072, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 0.6523449084691579, | |
| "learning_rate": 3.056302334890786e-06, | |
| "loss": 0.1583, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.7323202901972329, | |
| "learning_rate": 1.9436976651092143e-06, | |
| "loss": 0.1154, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 0.6205542406578355, | |
| "learning_rate": 9.412754953531664e-07, | |
| "loss": 0.0833, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 0.5523010023195977, | |
| "learning_rate": 2.4757783024395244e-07, | |
| "loss": 0.0699, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.5103448900296131, | |
| "learning_rate": 0.0, | |
| "loss": 0.0634, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 35, | |
| "total_flos": 2.2283279731051725e+17, | |
| "train_loss": 0.1769442856311798, | |
| "train_runtime": 3175.0371, | |
| "train_samples_per_second": 1.403, | |
| "train_steps_per_second": 0.011 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 35, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.2283279731051725e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |