| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 70, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 197.0, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 51.1707, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 104.0, | |
| "learning_rate": 0.00014285714285714287, | |
| "loss": 45.0263, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 13.753079414367676, | |
| "eval_runtime": 0.2534, | |
| "eval_samples_per_second": 39.465, | |
| "eval_steps_per_second": 3.946, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 20.25, | |
| "learning_rate": 0.00019888308262251285, | |
| "loss": 26.2686, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 8.776823997497559, | |
| "eval_runtime": 0.2365, | |
| "eval_samples_per_second": 42.288, | |
| "eval_steps_per_second": 4.229, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 4.875, | |
| "learning_rate": 0.00019214762118704076, | |
| "loss": 20.9714, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 5.0, | |
| "learning_rate": 0.00017971325072229226, | |
| "loss": 19.4522, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 7.711602210998535, | |
| "eval_runtime": 0.2355, | |
| "eval_samples_per_second": 42.47, | |
| "eval_steps_per_second": 4.247, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 0.00016234898018587337, | |
| "loss": 17.9837, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 7.1805419921875, | |
| "eval_runtime": 0.2358, | |
| "eval_samples_per_second": 42.415, | |
| "eval_steps_per_second": 4.242, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 13.75, | |
| "learning_rate": 0.00014112871031306119, | |
| "loss": 16.9838, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 18.375, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 14.7187, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 7.078407287597656, | |
| "eval_runtime": 0.2345, | |
| "eval_samples_per_second": 42.639, | |
| "eval_steps_per_second": 4.264, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 23.25, | |
| "learning_rate": 9.252699064135758e-05, | |
| "loss": 12.1795, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 7.036722660064697, | |
| "eval_runtime": 0.2361, | |
| "eval_samples_per_second": 42.352, | |
| "eval_steps_per_second": 4.235, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 6.428571428571429, | |
| "grad_norm": 27.125, | |
| "learning_rate": 6.815133497483157e-05, | |
| "loss": 9.1407, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 6.5706281661987305, | |
| "eval_runtime": 0.236, | |
| "eval_samples_per_second": 42.371, | |
| "eval_steps_per_second": 4.237, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 26.25, | |
| "learning_rate": 4.574537361342407e-05, | |
| "loss": 6.7417, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 7.857142857142857, | |
| "grad_norm": 22.25, | |
| "learning_rate": 2.669481281701739e-05, | |
| "loss": 5.133, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 6.185364723205566, | |
| "eval_runtime": 0.2416, | |
| "eval_samples_per_second": 41.387, | |
| "eval_steps_per_second": 4.139, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 22.125, | |
| "learning_rate": 1.2177842662977135e-05, | |
| "loss": 4.2385, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 6.027551174163818, | |
| "eval_runtime": 0.2379, | |
| "eval_samples_per_second": 42.042, | |
| "eval_steps_per_second": 4.204, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 9.285714285714286, | |
| "grad_norm": 20.5, | |
| "learning_rate": 3.092271377092215e-06, | |
| "loss": 3.8744, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 20.25, | |
| "learning_rate": 0.0, | |
| "loss": 3.8163, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 6.005556583404541, | |
| "eval_runtime": 0.2357, | |
| "eval_samples_per_second": 42.424, | |
| "eval_steps_per_second": 4.242, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 70, | |
| "total_flos": 2.1344245100262195e+17, | |
| "train_loss": 14.839824295043945, | |
| "train_runtime": 171.5695, | |
| "train_samples_per_second": 24.655, | |
| "train_steps_per_second": 0.408 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 70, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "total_flos": 2.1344245100262195e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |