| { | |
| "best_metric": 1.894714117050171, | |
| "best_model_checkpoint": "gen_test/checkpoint-6252", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 6252, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 167490.5625, | |
| "learning_rate": 5.520153550863724e-05, | |
| "loss": 2.0195, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 149414.109375, | |
| "learning_rate": 5.040307101727447e-05, | |
| "loss": 1.8734, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 117651.6953125, | |
| "learning_rate": 4.560460652591171e-05, | |
| "loss": 1.8297, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.9658416509628296, | |
| "eval_runtime": 56.3883, | |
| "eval_samples_per_second": 177.342, | |
| "eval_steps_per_second": 2.784, | |
| "step": 1563 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 112915.5, | |
| "learning_rate": 4.0806142034548946e-05, | |
| "loss": 1.7959, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 105164.921875, | |
| "learning_rate": 3.6007677543186184e-05, | |
| "loss": 1.7747, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 94653.9765625, | |
| "learning_rate": 3.1209213051823415e-05, | |
| "loss": 1.7648, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.9188886880874634, | |
| "eval_runtime": 56.4098, | |
| "eval_samples_per_second": 177.274, | |
| "eval_steps_per_second": 2.783, | |
| "step": 3126 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 110468.5703125, | |
| "learning_rate": 2.6410748560460653e-05, | |
| "loss": 1.7449, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 97404.4921875, | |
| "learning_rate": 2.161228406909789e-05, | |
| "loss": 1.7349, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 89132.0234375, | |
| "learning_rate": 1.6813819577735126e-05, | |
| "loss": 1.7292, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.9047945737838745, | |
| "eval_runtime": 56.5724, | |
| "eval_samples_per_second": 176.765, | |
| "eval_steps_per_second": 2.775, | |
| "step": 4689 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 101373.7578125, | |
| "learning_rate": 1.2015355086372362e-05, | |
| "loss": 1.7234, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 96313.7734375, | |
| "learning_rate": 7.216890595009597e-06, | |
| "loss": 1.7105, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 103498.109375, | |
| "learning_rate": 2.4184261036468333e-06, | |
| "loss": 1.7146, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.894714117050171, | |
| "eval_runtime": 56.3957, | |
| "eval_samples_per_second": 177.318, | |
| "eval_steps_per_second": 2.784, | |
| "step": 6252 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 6252, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 2.090336256e+17, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |