| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 6, | |
| "global_step": 23, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 6.0, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.7259, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "eval_loss": 1.913841724395752, | |
| "eval_runtime": 0.1057, | |
| "eval_samples_per_second": 18.922, | |
| "eval_steps_per_second": 18.922, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 7.0, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.8344, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 6e-06, | |
| "loss": 1.8468, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.7955, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 1e-05, | |
| "loss": 1.8767, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.8148, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_loss": 1.901075005531311, | |
| "eval_runtime": 0.1023, | |
| "eval_samples_per_second": 19.547, | |
| "eval_steps_per_second": 19.547, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 1.9312, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 6.25, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.832, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 6.375, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.8511, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9536, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 1.9282, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 6.125, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.8631, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "eval_loss": 1.8658802509307861, | |
| "eval_runtime": 0.1025, | |
| "eval_samples_per_second": 19.508, | |
| "eval_steps_per_second": 19.508, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 1.8138, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 1.7873, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 12.875, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9504, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.7767, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 1.9669, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 3.6e-05, | |
| "loss": 1.8768, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "eval_loss": 1.8019580841064453, | |
| "eval_runtime": 0.1033, | |
| "eval_samples_per_second": 19.362, | |
| "eval_steps_per_second": 19.362, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 3.8e-05, | |
| "loss": 1.6976, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 4e-05, | |
| "loss": 1.7206, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 6.5, | |
| "learning_rate": 4.2e-05, | |
| "loss": 1.884, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 2.005, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 1.7929, | |
| "step": 23 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 23, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "total_flos": 292378028015616.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |