| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 252, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 114440.2421875, | |
| "learning_rate": 0.0002988358809900258, | |
| "loss": 1.5517, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 86295.7265625, | |
| "learning_rate": 0.00029536159293436166, | |
| "loss": 1.0216, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 93327.2890625, | |
| "learning_rate": 0.00028963106229663063, | |
| "loss": 0.938, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 90729.6796875, | |
| "learning_rate": 0.0002817332360055343, | |
| "loss": 1.0426, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 77492.8203125, | |
| "learning_rate": 0.0002717907008573785, | |
| "loss": 0.9682, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 71191.8359375, | |
| "learning_rate": 0.0002599577807744739, | |
| "loss": 0.9422, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 81620.265625, | |
| "learning_rate": 0.0002464181414529809, | |
| "loss": 1.0024, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 68561.46875, | |
| "learning_rate": 0.0002313819395798639, | |
| "loss": 0.9483, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.9816237688064575, | |
| "eval_runtime": 80.5903, | |
| "eval_samples_per_second": 1.377, | |
| "eval_steps_per_second": 0.347, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 79334.9375, | |
| "learning_rate": 0.00021508256086763368, | |
| "loss": 0.8421, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 100762.734375, | |
| "learning_rate": 0.00019777299753775265, | |
| "loss": 0.6904, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 88126.5390625, | |
| "learning_rate": 0.0001797219214799096, | |
| "loss": 0.7075, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 98623.3359375, | |
| "learning_rate": 0.00016120951403796364, | |
| "loss": 0.765, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 93532.2890625, | |
| "learning_rate": 0.0001425231171508954, | |
| "loss": 0.7921, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 92075.6484375, | |
| "learning_rate": 0.00012395277334996044, | |
| "loss": 0.6924, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 93960.328125, | |
| "learning_rate": 0.00010578672383836435, | |
| "loss": 0.677, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 71849.8125, | |
| "learning_rate": 8.830693453040829e-05, | |
| "loss": 0.7375, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.9629082679748535, | |
| "eval_runtime": 80.2418, | |
| "eval_samples_per_second": 1.383, | |
| "eval_steps_per_second": 0.349, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 68938.421875, | |
| "learning_rate": 7.17847194930753e-05, | |
| "loss": 0.659, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 141907.765625, | |
| "learning_rate": 5.6476529721189974e-05, | |
| "loss": 0.425, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 96978.578125, | |
| "learning_rate": 4.261997261104223e-05, | |
| "loss": 0.4418, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 142711.125, | |
| "learning_rate": 3.0430123916561672e-05, | |
| "loss": 0.415, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 90807.7421875, | |
| "learning_rate": 2.009618943233419e-05, | |
| "loss": 0.4457, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 105124.25, | |
| "learning_rate": 1.1778568219438839e-05, | |
| "loss": 0.4444, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 92137.609375, | |
| "learning_rate": 5.606362957498195e-06, | |
| "loss": 0.4239, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 96611.28125, | |
| "learning_rate": 1.6753760662307215e-06, | |
| "loss": 0.4227, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 100055.1953125, | |
| "learning_rate": 4.662269987756317e-08, | |
| "loss": 0.4589, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.0281459093093872, | |
| "eval_runtime": 80.2237, | |
| "eval_samples_per_second": 1.384, | |
| "eval_steps_per_second": 0.349, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 252, | |
| "total_flos": 2.458929889338163e+16, | |
| "train_loss": 0.735804313705081, | |
| "train_runtime": 2220.224, | |
| "train_samples_per_second": 0.45, | |
| "train_steps_per_second": 0.114 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 252, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 10, | |
| "total_flos": 2.458929889338163e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |