| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.6923076923076925, | |
| "eval_steps": 5, | |
| "global_step": 50, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 0.7337189811685318, | |
| "learning_rate": 4e-05, | |
| "loss": 1.4896, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.788344053257058, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4804, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "eval_loss": 1.1582854986190796, | |
| "eval_runtime": 2.6274, | |
| "eval_samples_per_second": 7.612, | |
| "eval_steps_per_second": 1.142, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.4371059352187654, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 0.9708, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_loss": 0.8723791241645813, | |
| "eval_runtime": 1.3731, | |
| "eval_samples_per_second": 14.565, | |
| "eval_steps_per_second": 2.185, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.25625240789847736, | |
| "learning_rate": 0.0001766044443118978, | |
| "loss": 0.7685, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "eval_loss": 0.7142642140388489, | |
| "eval_runtime": 1.3722, | |
| "eval_samples_per_second": 14.575, | |
| "eval_steps_per_second": 2.186, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.37095540648653375, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.6454, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "eval_loss": 0.660171389579773, | |
| "eval_runtime": 1.3569, | |
| "eval_samples_per_second": 14.739, | |
| "eval_steps_per_second": 2.211, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.269490348868028, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.5826, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "eval_loss": 0.6317092776298523, | |
| "eval_runtime": 1.3548, | |
| "eval_samples_per_second": 14.762, | |
| "eval_steps_per_second": 2.214, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.1367860640067106, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 0.5458, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "eval_loss": 0.6193430423736572, | |
| "eval_runtime": 1.3508, | |
| "eval_samples_per_second": 14.806, | |
| "eval_steps_per_second": 2.221, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.384615384615385, | |
| "grad_norm": 0.14497012519202485, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.4844, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 5.384615384615385, | |
| "eval_loss": 0.6151253581047058, | |
| "eval_runtime": 1.3512, | |
| "eval_samples_per_second": 14.802, | |
| "eval_steps_per_second": 2.22, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "grad_norm": 0.1438702628497647, | |
| "learning_rate": 2.339555568810221e-05, | |
| "loss": 0.506, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "eval_loss": 0.6133347749710083, | |
| "eval_runtime": 1.358, | |
| "eval_samples_per_second": 14.728, | |
| "eval_steps_per_second": 2.209, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.923076923076923, | |
| "grad_norm": 0.15810290565530724, | |
| "learning_rate": 6.030737921409169e-06, | |
| "loss": 0.4766, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 6.923076923076923, | |
| "eval_loss": 0.6125746965408325, | |
| "eval_runtime": 1.3498, | |
| "eval_samples_per_second": 14.817, | |
| "eval_steps_per_second": 2.223, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "grad_norm": 0.16553701241746885, | |
| "learning_rate": 0.0, | |
| "loss": 0.4639, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "eval_loss": 0.6129434108734131, | |
| "eval_runtime": 1.3475, | |
| "eval_samples_per_second": 14.842, | |
| "eval_steps_per_second": 2.226, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "step": 50, | |
| "total_flos": 1205290795008.0, | |
| "train_loss": 0.692623188495636, | |
| "train_runtime": 177.9895, | |
| "train_samples_per_second": 4.495, | |
| "train_steps_per_second": 0.281 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 50, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1205290795008.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |