| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.0028050638414998676, | |
| "eval_steps": 25, | |
| "global_step": 50, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 5.610127682999735e-05, | |
| "eval_loss": NaN, | |
| "eval_runtime": 22644.069, | |
| "eval_samples_per_second": 2.799, | |
| "eval_steps_per_second": 2.799, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00028050638414998675, | |
| "grad_norm": 8.75216007232666, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 8.2406, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0005610127682999735, | |
| "grad_norm": 6.417360305786133, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 4.3853, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0008415191524499603, | |
| "grad_norm": 7.757654190063477, | |
| "learning_rate": 8.296729075500344e-05, | |
| "loss": 2.1783, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.001122025536599947, | |
| "grad_norm": 4.7035722732543945, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 1.887, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0014025319207499338, | |
| "grad_norm": 3.3673248291015625, | |
| "learning_rate": 5.327015646150716e-05, | |
| "loss": 1.5397, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0014025319207499338, | |
| "eval_loss": NaN, | |
| "eval_runtime": 22696.0747, | |
| "eval_samples_per_second": 2.792, | |
| "eval_steps_per_second": 2.792, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0016830383048999206, | |
| "grad_norm": 3.5548179149627686, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 1.3131, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.001963544689049907, | |
| "grad_norm": 3.1754086017608643, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 1.738, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.002244051073199894, | |
| "grad_norm": 3.659442663192749, | |
| "learning_rate": 1.0332332985438248e-05, | |
| "loss": 1.8102, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.002524557457349881, | |
| "grad_norm": 3.112903356552124, | |
| "learning_rate": 2.653493525244721e-06, | |
| "loss": 1.3295, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0028050638414998676, | |
| "grad_norm": 4.095558166503906, | |
| "learning_rate": 0.0, | |
| "loss": 1.6155, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0028050638414998676, | |
| "eval_loss": NaN, | |
| "eval_runtime": 22725.7487, | |
| "eval_samples_per_second": 2.789, | |
| "eval_steps_per_second": 2.789, | |
| "step": 50 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 50, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.483774567120896e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |