| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.0, |
| "eval_steps": 6, |
| "global_step": 12, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.25, |
| "grad_norm": 88.1607934988436, |
| "learning_rate": 2e-05, |
| "loss": 2.335, |
| "mean_token_accuracy": 0.648490309715271, |
| "num_tokens": 21926.0, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 22.802312407394005, |
| "learning_rate": 1.8333333333333333e-05, |
| "loss": 1.1659, |
| "mean_token_accuracy": 0.7188374400138855, |
| "num_tokens": 46341.0, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.75, |
| "grad_norm": 15.376065491761672, |
| "learning_rate": 1.6666666666666667e-05, |
| "loss": 0.9474, |
| "mean_token_accuracy": 0.7389319539070129, |
| "num_tokens": 68818.0, |
| "step": 3 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 8.136678583492891, |
| "learning_rate": 1.5000000000000002e-05, |
| "loss": 0.7871, |
| "mean_token_accuracy": 0.7817161083221436, |
| "num_tokens": 90481.0, |
| "step": 4 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 6.055337469344572, |
| "learning_rate": 1.3333333333333333e-05, |
| "loss": 0.6484, |
| "mean_token_accuracy": 0.8068888187408447, |
| "num_tokens": 112083.0, |
| "step": 5 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 2.4346589397744043, |
| "learning_rate": 1.1666666666666668e-05, |
| "loss": 0.6135, |
| "mean_token_accuracy": 0.8205102682113647, |
| "num_tokens": 135399.0, |
| "step": 6 |
| }, |
| { |
| "epoch": 1.5, |
| "eval_loss": 0.6730726361274719, |
| "eval_mean_token_accuracy": 0.8092885613441467, |
| "eval_num_tokens": 135399.0, |
| "eval_runtime": 0.1928, |
| "eval_samples_per_second": 20.749, |
| "eval_steps_per_second": 5.187, |
| "step": 6 |
| }, |
| { |
| "epoch": 1.75, |
| "grad_norm": 1.9800824362829181, |
| "learning_rate": 1e-05, |
| "loss": 0.5734, |
| "mean_token_accuracy": 0.8318648934364319, |
| "num_tokens": 158941.0, |
| "step": 7 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 1.7869036745320295, |
| "learning_rate": 8.333333333333334e-06, |
| "loss": 0.4918, |
| "mean_token_accuracy": 0.8499757647514343, |
| "num_tokens": 180442.0, |
| "step": 8 |
| }, |
| { |
| "epoch": 2.25, |
| "grad_norm": 1.6183509198161528, |
| "learning_rate": 6.666666666666667e-06, |
| "loss": 0.4559, |
| "mean_token_accuracy": 0.8571715354919434, |
| "num_tokens": 203358.0, |
| "step": 9 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 1.4356865952246156, |
| "learning_rate": 5e-06, |
| "loss": 0.432, |
| "mean_token_accuracy": 0.8671599626541138, |
| "num_tokens": 226202.0, |
| "step": 10 |
| }, |
| { |
| "epoch": 2.75, |
| "grad_norm": 1.4409611264021458, |
| "learning_rate": 3.3333333333333333e-06, |
| "loss": 0.4264, |
| "mean_token_accuracy": 0.8663450479507446, |
| "num_tokens": 248675.0, |
| "step": 11 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 1.4198612779446906, |
| "learning_rate": 1.6666666666666667e-06, |
| "loss": 0.4066, |
| "mean_token_accuracy": 0.872953474521637, |
| "num_tokens": 271758.0, |
| "step": 12 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_loss": 0.6228852272033691, |
| "eval_mean_token_accuracy": 0.8320158123970032, |
| "eval_num_tokens": 271758.0, |
| "eval_runtime": 0.2559, |
| "eval_samples_per_second": 15.634, |
| "eval_steps_per_second": 3.908, |
| "step": 12 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 12, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 640950927360.0, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
