| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.006750342790844847, | |
| "eval_steps": 5, | |
| "global_step": 20, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00033751713954224236, | |
| "grad_norm": 5.253102779388428, | |
| "learning_rate": 1e-05, | |
| "loss": 4.166, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00033751713954224236, | |
| "eval_loss": 4.157853603363037, | |
| "eval_runtime": 138.9493, | |
| "eval_samples_per_second": 35.912, | |
| "eval_steps_per_second": 17.956, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0006750342790844847, | |
| "grad_norm": 5.244392395019531, | |
| "learning_rate": 2e-05, | |
| "loss": 3.7615, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0010125514186267272, | |
| "grad_norm": 6.120645523071289, | |
| "learning_rate": 3e-05, | |
| "loss": 4.1821, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0013500685581689694, | |
| "grad_norm": 5.252912521362305, | |
| "learning_rate": 4e-05, | |
| "loss": 4.1008, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0016875856977112118, | |
| "grad_norm": 5.319488048553467, | |
| "learning_rate": 5e-05, | |
| "loss": 3.9749, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0016875856977112118, | |
| "eval_loss": 3.8743081092834473, | |
| "eval_runtime": 37.142, | |
| "eval_samples_per_second": 134.349, | |
| "eval_steps_per_second": 67.175, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0020251028372534545, | |
| "grad_norm": 5.740655899047852, | |
| "learning_rate": 6e-05, | |
| "loss": 3.7718, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0023626199767956967, | |
| "grad_norm": 5.769639015197754, | |
| "learning_rate": 7e-05, | |
| "loss": 3.8983, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.002700137116337939, | |
| "grad_norm": 5.479471683502197, | |
| "learning_rate": 8e-05, | |
| "loss": 3.2239, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0030376542558801815, | |
| "grad_norm": 5.583406448364258, | |
| "learning_rate": 9e-05, | |
| "loss": 3.1888, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0033751713954224237, | |
| "grad_norm": 5.544492721557617, | |
| "learning_rate": 0.0001, | |
| "loss": 3.2513, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0033751713954224237, | |
| "eval_loss": 2.91571044921875, | |
| "eval_runtime": 51.272, | |
| "eval_samples_per_second": 97.324, | |
| "eval_steps_per_second": 48.662, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0037126885349646663, | |
| "grad_norm": 5.027786731719971, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.6374, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.004050205674506909, | |
| "grad_norm": 5.445701599121094, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 2.6293, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.004387722814049151, | |
| "grad_norm": 4.952109336853027, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 1.8693, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.004725239953591393, | |
| "grad_norm": 5.304759502410889, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.9581, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0050627570931336355, | |
| "grad_norm": 4.6097307205200195, | |
| "learning_rate": 5e-05, | |
| "loss": 2.0891, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0050627570931336355, | |
| "eval_loss": 1.750018835067749, | |
| "eval_runtime": 58.061, | |
| "eval_samples_per_second": 85.944, | |
| "eval_steps_per_second": 42.972, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.005400274232675878, | |
| "grad_norm": 4.610104560852051, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 1.7243, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.005737791372218121, | |
| "grad_norm": 5.061768054962158, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 1.734, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.006075308511760363, | |
| "grad_norm": 4.05675745010376, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 1.4461, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.006412825651302605, | |
| "grad_norm": 3.7175137996673584, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 1.7627, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.006750342790844847, | |
| "grad_norm": 3.5504698753356934, | |
| "learning_rate": 0.0, | |
| "loss": 1.5403, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.006750342790844847, | |
| "eval_loss": 1.5085777044296265, | |
| "eval_runtime": 45.4389, | |
| "eval_samples_per_second": 109.818, | |
| "eval_steps_per_second": 54.909, | |
| "step": 20 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 5, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 615708515893248.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |