| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 10, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.5965750217437744, | |
| "learning_rate": 0.00018, | |
| "loss": 0.7517, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8888888888888888, | |
| "eval_loss": 0.4700092077255249, | |
| "eval_runtime": 0.0309, | |
| "eval_samples_per_second": 582.425, | |
| "eval_steps_per_second": 64.714, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.535392701625824, | |
| "learning_rate": 0.00016, | |
| "loss": 0.2834, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8888888888888888, | |
| "eval_loss": 0.3312886953353882, | |
| "eval_runtime": 0.0245, | |
| "eval_samples_per_second": 735.112, | |
| "eval_steps_per_second": 81.679, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.14222599565982819, | |
| "learning_rate": 0.00014, | |
| "loss": 0.0488, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8888888888888888, | |
| "eval_loss": 0.35279521346092224, | |
| "eval_runtime": 0.0265, | |
| "eval_samples_per_second": 678.105, | |
| "eval_steps_per_second": 75.345, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.09212491661310196, | |
| "learning_rate": 0.00012, | |
| "loss": 0.0181, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8888888888888888, | |
| "eval_loss": 0.6354701519012451, | |
| "eval_runtime": 0.0268, | |
| "eval_samples_per_second": 672.847, | |
| "eval_steps_per_second": 74.761, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.05534437298774719, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0079, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8888888888888888, | |
| "eval_loss": 0.6675615310668945, | |
| "eval_runtime": 0.0382, | |
| "eval_samples_per_second": 471.491, | |
| "eval_steps_per_second": 52.388, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.04630253463983536, | |
| "learning_rate": 8e-05, | |
| "loss": 0.0437, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8888888888888888, | |
| "eval_loss": 0.5816599130630493, | |
| "eval_runtime": 0.0247, | |
| "eval_samples_per_second": 728.213, | |
| "eval_steps_per_second": 80.913, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.04637857526540756, | |
| "learning_rate": 6e-05, | |
| "loss": 0.0049, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.8888888888888888, | |
| "eval_loss": 0.4498683512210846, | |
| "eval_runtime": 0.0252, | |
| "eval_samples_per_second": 713.378, | |
| "eval_steps_per_second": 79.264, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.1237090453505516, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0192, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8888888888888888, | |
| "eval_loss": 0.516182541847229, | |
| "eval_runtime": 0.0258, | |
| "eval_samples_per_second": 698.146, | |
| "eval_steps_per_second": 77.572, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.03347891941666603, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0045, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.8888888888888888, | |
| "eval_loss": 0.5420387983322144, | |
| "eval_runtime": 0.0267, | |
| "eval_samples_per_second": 673.549, | |
| "eval_steps_per_second": 74.839, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.03203904256224632, | |
| "learning_rate": 0.0, | |
| "loss": 0.0042, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.8888888888888888, | |
| "eval_loss": 0.5463488101959229, | |
| "eval_runtime": 0.0257, | |
| "eval_samples_per_second": 699.731, | |
| "eval_steps_per_second": 77.748, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 100, | |
| "total_flos": 16662479407056.0, | |
| "train_loss": 0.11865015212446452, | |
| "train_runtime": 23.7587, | |
| "train_samples_per_second": 67.344, | |
| "train_steps_per_second": 4.209 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 20, | |
| "total_flos": 16662479407056.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |