| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 20, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 173.0, | |
| "learning_rate": 0.0001, | |
| "loss": 45.5656, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 16.504554748535156, | |
| "eval_runtime": 0.2461, | |
| "eval_samples_per_second": 40.635, | |
| "eval_steps_per_second": 4.063, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 14.199999809265137, | |
| "eval_runtime": 0.2394, | |
| "eval_samples_per_second": 41.776, | |
| "eval_steps_per_second": 4.178, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 31.5, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 35.6654, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 12.994392395019531, | |
| "eval_runtime": 0.2412, | |
| "eval_samples_per_second": 41.452, | |
| "eval_steps_per_second": 4.145, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 11.569496154785156, | |
| "eval_runtime": 0.2326, | |
| "eval_samples_per_second": 42.992, | |
| "eval_steps_per_second": 4.299, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 10.5, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 22.2461, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 10.306487083435059, | |
| "eval_runtime": 0.2322, | |
| "eval_samples_per_second": 43.073, | |
| "eval_steps_per_second": 4.307, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 9.36452579498291, | |
| "eval_runtime": 0.2332, | |
| "eval_samples_per_second": 42.887, | |
| "eval_steps_per_second": 4.289, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 8.907111167907715, | |
| "eval_runtime": 0.2345, | |
| "eval_samples_per_second": 42.648, | |
| "eval_steps_per_second": 4.265, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 4.5, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 19.7508, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 8.693410873413086, | |
| "eval_runtime": 0.2425, | |
| "eval_samples_per_second": 41.229, | |
| "eval_steps_per_second": 4.123, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 8.628682136535645, | |
| "eval_runtime": 0.2394, | |
| "eval_samples_per_second": 41.765, | |
| "eval_steps_per_second": 4.177, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 4.125, | |
| "learning_rate": 0.0, | |
| "loss": 19.172, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 8.61988639831543, | |
| "eval_runtime": 0.232, | |
| "eval_samples_per_second": 43.094, | |
| "eval_steps_per_second": 4.309, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 20, | |
| "total_flos": 6.098355666236211e+16, | |
| "train_loss": 24.70357437133789, | |
| "train_runtime": 50.9898, | |
| "train_samples_per_second": 21.769, | |
| "train_steps_per_second": 0.392 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "total_flos": 6.098355666236211e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |