| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9335978527249388, | |
| "eval_steps": 250, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2333994631812347, | |
| "grad_norm": 0.5453733205795288, | |
| "learning_rate": 8.959847237666421e-05, | |
| "loss": 0.4322, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2333994631812347, | |
| "eval_loss": 0.22301338613033295, | |
| "eval_runtime": 15.7009, | |
| "eval_samples_per_second": 5.541, | |
| "eval_steps_per_second": 5.541, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4667989263624694, | |
| "grad_norm": 0.5138410329818726, | |
| "learning_rate": 5.783666949486448e-05, | |
| "loss": 0.1503, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4667989263624694, | |
| "eval_loss": 0.13980120420455933, | |
| "eval_runtime": 15.7934, | |
| "eval_samples_per_second": 5.509, | |
| "eval_steps_per_second": 5.509, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7001983895437041, | |
| "grad_norm": 0.5098557472229004, | |
| "learning_rate": 2.179828363456768e-05, | |
| "loss": 0.1084, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7001983895437041, | |
| "eval_loss": 0.11332736909389496, | |
| "eval_runtime": 15.7373, | |
| "eval_samples_per_second": 5.528, | |
| "eval_steps_per_second": 5.528, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.9335978527249388, | |
| "grad_norm": 0.36698898673057556, | |
| "learning_rate": 1.149979013737007e-06, | |
| "loss": 0.0932, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9335978527249388, | |
| "eval_loss": 0.10717575252056122, | |
| "eval_runtime": 15.7709, | |
| "eval_samples_per_second": 5.517, | |
| "eval_steps_per_second": 5.517, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 250, | |
| "max_steps": 1071, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 250, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.32949637553152e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |