| { | |
| "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": 4.274783951494615, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 2.0554, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 2.961220476156441, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 1.9786, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.6180429458618164, | |
| "eval_runtime": 2.0806, | |
| "eval_samples_per_second": 5.768, | |
| "eval_steps_per_second": 0.481, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.9112171217097137, | |
| "learning_rate": 1.973044870579824e-05, | |
| "loss": 1.75, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.998845100402832, | |
| "eval_runtime": 2.0684, | |
| "eval_samples_per_second": 5.801, | |
| "eval_steps_per_second": 0.483, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.0494875968698665, | |
| "learning_rate": 1.866025403784439e-05, | |
| "loss": 1.5489, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.9909167289733887, | |
| "eval_runtime": 2.0598, | |
| "eval_samples_per_second": 5.826, | |
| "eval_steps_per_second": 0.485, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 0.7506291843468882, | |
| "learning_rate": 1.686241637868734e-05, | |
| "loss": 1.3631, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.9766260385513306, | |
| "eval_runtime": 2.0535, | |
| "eval_samples_per_second": 5.844, | |
| "eval_steps_per_second": 0.487, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 1.1207567831148806, | |
| "learning_rate": 1.4487991802004625e-05, | |
| "loss": 1.1908, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.0514743700194535, | |
| "learning_rate": 1.1736481776669307e-05, | |
| "loss": 1.0134, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 2.1003313064575195, | |
| "eval_runtime": 2.0426, | |
| "eval_samples_per_second": 5.875, | |
| "eval_steps_per_second": 0.49, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "grad_norm": 0.9419508909251836, | |
| "learning_rate": 8.839070858747697e-06, | |
| "loss": 0.8425, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.168269395828247, | |
| "eval_runtime": 2.0544, | |
| "eval_samples_per_second": 5.841, | |
| "eval_steps_per_second": 0.487, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "grad_norm": 1.8575888725444458, | |
| "learning_rate": 6.039202339608432e-06, | |
| "loss": 0.7051, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 2.373701810836792, | |
| "eval_runtime": 2.056, | |
| "eval_samples_per_second": 5.836, | |
| "eval_steps_per_second": 0.486, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 1.7968257086708292, | |
| "learning_rate": 3.5721239031346067e-06, | |
| "loss": 0.594, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.5296216011047363, | |
| "eval_runtime": 2.0509, | |
| "eval_samples_per_second": 5.851, | |
| "eval_steps_per_second": 0.488, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "grad_norm": 1.4292640479394765, | |
| "learning_rate": 1.6451218858706374e-06, | |
| "loss": 0.5107, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 2.5202696323394775, | |
| "eval_runtime": 2.0562, | |
| "eval_samples_per_second": 5.836, | |
| "eval_steps_per_second": 0.486, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 9.17, | |
| "grad_norm": 0.9522351371996648, | |
| "learning_rate": 4.2010487684511105e-07, | |
| "loss": 0.4652, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.01561769419806, | |
| "learning_rate": 0.0, | |
| "loss": 0.4594, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 2.5297091007232666, | |
| "eval_runtime": 2.053, | |
| "eval_samples_per_second": 5.845, | |
| "eval_steps_per_second": 0.487, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 60, | |
| "total_flos": 12562779340800.0, | |
| "train_loss": 1.0364325324694315, | |
| "train_runtime": 358.0118, | |
| "train_samples_per_second": 20.642, | |
| "train_steps_per_second": 0.168 | |
| } | |
| ], | |
| "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 | |
| } | |