| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 11.148272017837236, | |
| "eval_steps": 500, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 4.136017322540283, | |
| "learning_rate": 4.75e-05, | |
| "loss": 1.4862, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 3.5059304237365723, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.2182, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 3.1735312938690186, | |
| "learning_rate": 4.25e-05, | |
| "loss": 1.0549, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 3.1064469814300537, | |
| "learning_rate": 4e-05, | |
| "loss": 0.9527, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 2.8415098190307617, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.8717, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "grad_norm": 2.4496607780456543, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.783, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 2.9042794704437256, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.7372, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 2.628190517425537, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6509, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 2.9309325218200684, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.6398, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "grad_norm": 2.357445240020752, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.56, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 6.13, | |
| "grad_norm": 2.1420226097106934, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.5543, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 6.69, | |
| "grad_norm": 2.6394078731536865, | |
| "learning_rate": 2e-05, | |
| "loss": 0.5104, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 7.25, | |
| "grad_norm": 2.395305633544922, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.4832, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 2.048461675643921, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.4705, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 8.36, | |
| "grad_norm": 2.379493236541748, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.442, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 8.92, | |
| "grad_norm": 2.170802116394043, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4264, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 9.48, | |
| "grad_norm": 2.277756690979004, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.405, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 10.03, | |
| "grad_norm": 2.3412973880767822, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4069, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 10.59, | |
| "grad_norm": 2.220499038696289, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.3937, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 11.15, | |
| "grad_norm": 2.1124320030212402, | |
| "learning_rate": 0.0, | |
| "loss": 0.3872, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 11.15, | |
| "step": 10000, | |
| "total_flos": 5225840640000000.0, | |
| "train_loss": 0.6717062408447265, | |
| "train_runtime": 2130.1373, | |
| "train_samples_per_second": 4.695, | |
| "train_steps_per_second": 4.695 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 12, | |
| "save_steps": 1000, | |
| "total_flos": 5225840640000000.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |