| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 256.0, | |
| "eval_steps": 500, | |
| "global_step": 256, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.485157862469004, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5302, | |
| "mean_token_accuracy": 0.8446678519248962, | |
| "num_tokens": 39860.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.4913851914815907, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1203, | |
| "mean_token_accuracy": 0.9620425701141357, | |
| "num_tokens": 398600.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.09925451841277072, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0011, | |
| "mean_token_accuracy": 0.9998640537261962, | |
| "num_tokens": 797200.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.027909330308405398, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0001, | |
| "mean_token_accuracy": 0.9999841988086701, | |
| "num_tokens": 1195800.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 0.004164190395265697, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 1594400.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 0.0006083710274524334, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 1993000.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 0.00037327028596591483, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 2391600.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "grad_norm": 0.0002836318539897977, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 2790200.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "grad_norm": 0.00023454512604283856, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 3188800.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "grad_norm": 0.00020209688966584525, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 3587400.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "grad_norm": 0.00018017548608497045, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 3986000.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 110.0, | |
| "grad_norm": 0.00016141799457905403, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 4384600.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 120.0, | |
| "grad_norm": 0.0001468881382632941, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 4783200.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 130.0, | |
| "grad_norm": 0.00013386155677500393, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 5181800.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 140.0, | |
| "grad_norm": 0.00012394062245490835, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 5580400.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "grad_norm": 0.00011400167970745115, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 5979000.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 160.0, | |
| "grad_norm": 0.00010600962535540816, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 6377600.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 170.0, | |
| "grad_norm": 9.868668824939607e-05, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 6776200.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 180.0, | |
| "grad_norm": 9.3079991238424e-05, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 7174800.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 190.0, | |
| "grad_norm": 8.760954102112786e-05, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 7573400.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "grad_norm": 8.181827299982745e-05, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 7972000.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 210.0, | |
| "grad_norm": 7.786118641109334e-05, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 8370600.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 220.0, | |
| "grad_norm": 7.325181941887582e-05, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 8769200.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 230.0, | |
| "grad_norm": 6.92188534097631e-05, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 9167800.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 240.0, | |
| "grad_norm": 6.594493440123658e-05, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 9566400.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 250.0, | |
| "grad_norm": 6.291430923179703e-05, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "mean_token_accuracy": 1.0, | |
| "num_tokens": 9965000.0, | |
| "step": 250 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 256, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 256, | |
| "save_steps": 64, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 23071161384960.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |