| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 60, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 10.550161397700021, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.8491, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 8.369449500494765, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 1.8197, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.1416637897491455, | |
| "eval_runtime": 2.3253, | |
| "eval_samples_per_second": 5.161, | |
| "eval_steps_per_second": 0.43, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 6.008460274326708, | |
| "learning_rate": 1.973044870579824e-05, | |
| "loss": 1.3836, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.7507458925247192, | |
| "eval_runtime": 2.1866, | |
| "eval_samples_per_second": 5.488, | |
| "eval_steps_per_second": 0.457, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 7.347536654652698, | |
| "learning_rate": 1.866025403784439e-05, | |
| "loss": 0.9111, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 2.1440508365631104, | |
| "eval_runtime": 2.1965, | |
| "eval_samples_per_second": 5.463, | |
| "eval_steps_per_second": 0.455, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 8.924017881530585, | |
| "learning_rate": 1.686241637868734e-05, | |
| "loss": 0.4865, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.5978963375091553, | |
| "eval_runtime": 2.1946, | |
| "eval_samples_per_second": 5.468, | |
| "eval_steps_per_second": 0.456, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 3.8795868824604973, | |
| "learning_rate": 1.4487991802004625e-05, | |
| "loss": 0.2128, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.3312651987420017, | |
| "learning_rate": 1.1736481776669307e-05, | |
| "loss": 0.0827, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 2.817784547805786, | |
| "eval_runtime": 2.1833, | |
| "eval_samples_per_second": 5.496, | |
| "eval_steps_per_second": 0.458, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "grad_norm": 1.668395846222473, | |
| "learning_rate": 8.839070858747697e-06, | |
| "loss": 0.0365, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 3.2582452297210693, | |
| "eval_runtime": 2.1864, | |
| "eval_samples_per_second": 5.489, | |
| "eval_steps_per_second": 0.457, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "grad_norm": 1.082911988071084, | |
| "learning_rate": 6.039202339608432e-06, | |
| "loss": 0.019, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 3.6268527507781982, | |
| "eval_runtime": 2.176, | |
| "eval_samples_per_second": 5.515, | |
| "eval_steps_per_second": 0.46, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.6033031953021625, | |
| "learning_rate": 3.5721239031346067e-06, | |
| "loss": 0.0127, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 3.825711250305176, | |
| "eval_runtime": 2.1746, | |
| "eval_samples_per_second": 5.518, | |
| "eval_steps_per_second": 0.46, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "grad_norm": 0.39765335521647993, | |
| "learning_rate": 1.6451218858706374e-06, | |
| "loss": 0.0092, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 3.902661085128784, | |
| "eval_runtime": 2.1697, | |
| "eval_samples_per_second": 5.531, | |
| "eval_steps_per_second": 0.461, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 9.17, | |
| "grad_norm": 0.42322353991647504, | |
| "learning_rate": 4.2010487684511105e-07, | |
| "loss": 0.0077, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.39922154152767036, | |
| "learning_rate": 0.0, | |
| "loss": 0.0071, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 3.917660713195801, | |
| "eval_runtime": 2.1754, | |
| "eval_samples_per_second": 5.516, | |
| "eval_steps_per_second": 0.46, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 60, | |
| "total_flos": 12562779340800.0, | |
| "train_loss": 0.41619996732721726, | |
| "train_runtime": 395.7925, | |
| "train_samples_per_second": 18.09, | |
| "train_steps_per_second": 0.152 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 60, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 1000000000, | |
| "total_flos": 12562779340800.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |