| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.23076923076923, | |
| "eval_steps": 500, | |
| "global_step": 60, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 390.0, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 40.7928, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 69.0, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 37.3003, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "eval_loss": 12.323984146118164, | |
| "eval_runtime": 2.0206, | |
| "eval_samples_per_second": 4.949, | |
| "eval_steps_per_second": 0.99, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 26.0, | |
| "learning_rate": 0.00019730448705798239, | |
| "loss": 21.0453, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 7.588448524475098, | |
| "eval_runtime": 2.0384, | |
| "eval_samples_per_second": 4.906, | |
| "eval_steps_per_second": 0.981, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 16.75, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 17.809, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "eval_loss": 5.654947757720947, | |
| "eval_runtime": 2.0499, | |
| "eval_samples_per_second": 4.878, | |
| "eval_steps_per_second": 0.976, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 58.0, | |
| "learning_rate": 0.0001686241637868734, | |
| "loss": 13.3008, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 8.0, | |
| "learning_rate": 0.00014487991802004623, | |
| "loss": 3.5386, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 3.160724401473999, | |
| "eval_runtime": 2.0507, | |
| "eval_samples_per_second": 4.876, | |
| "eval_steps_per_second": 0.975, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 2.02, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "eval_loss": 2.8709278106689453, | |
| "eval_runtime": 2.0507, | |
| "eval_samples_per_second": 4.876, | |
| "eval_steps_per_second": 0.975, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 5.384615384615385, | |
| "grad_norm": 3.375, | |
| "learning_rate": 8.839070858747697e-05, | |
| "loss": 1.732, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.789313793182373, | |
| "eval_runtime": 2.0444, | |
| "eval_samples_per_second": 4.891, | |
| "eval_steps_per_second": 0.978, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "grad_norm": 4.0, | |
| "learning_rate": 6.039202339608432e-05, | |
| "loss": 1.5413, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.923076923076923, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 1.4072, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 6.923076923076923, | |
| "eval_loss": 2.797020435333252, | |
| "eval_runtime": 2.049, | |
| "eval_samples_per_second": 4.88, | |
| "eval_steps_per_second": 0.976, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 1.6451218858706374e-05, | |
| "loss": 1.3061, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.8124804496765137, | |
| "eval_runtime": 2.0441, | |
| "eval_samples_per_second": 4.892, | |
| "eval_steps_per_second": 0.978, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 8.461538461538462, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 4.20104876845111e-06, | |
| "loss": 1.2613, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 8.923076923076923, | |
| "eval_loss": 2.8097996711730957, | |
| "eval_runtime": 2.0484, | |
| "eval_samples_per_second": 4.882, | |
| "eval_steps_per_second": 0.976, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 9.23076923076923, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 0.0, | |
| "loss": 1.2304, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 9.23076923076923, | |
| "eval_loss": 2.8104407787323, | |
| "eval_runtime": 2.0425, | |
| "eval_samples_per_second": 4.896, | |
| "eval_steps_per_second": 0.979, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 9.23076923076923, | |
| "step": 60, | |
| "total_flos": 4.58661949299753e+16, | |
| "train_loss": 8.682566889127095, | |
| "train_runtime": 414.6045, | |
| "train_samples_per_second": 2.484, | |
| "train_steps_per_second": 0.145 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 60, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "total_flos": 4.58661949299753e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |