| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 23.529411764705884, | |
| "eval_steps": 200, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.058823529411764705, | |
| "eval_loss": 2.0603227615356445, | |
| "eval_runtime": 6.3215, | |
| "eval_samples_per_second": 237.445, | |
| "eval_steps_per_second": 5.062, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 1.125, | |
| "learning_rate": 0.0001, | |
| "loss": 1.8865, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 0.0002, | |
| "loss": 1.7253, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.7647058823529411, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 0.000199658449300667, | |
| "loss": 1.2996, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 0.00019863613034027224, | |
| "loss": 0.9578, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "grad_norm": 1.5, | |
| "learning_rate": 0.00019694002659393305, | |
| "loss": 0.6771, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 0.00019458172417006347, | |
| "loss": 0.3386, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 4.117647058823529, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 0.00019157733266550575, | |
| "loss": 0.2562, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 0.0001879473751206489, | |
| "loss": 0.1393, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 5.294117647058823, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 0.00018371664782625287, | |
| "loss": 0.1163, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 0.87109375, | |
| "learning_rate": 0.00017891405093963938, | |
| "loss": 0.0997, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.470588235294118, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 0.00017357239106731317, | |
| "loss": 0.0658, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 7.0588235294117645, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.00016772815716257412, | |
| "loss": 0.0672, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 7.647058823529412, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.0001614212712689668, | |
| "loss": 0.0342, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 8.235294117647058, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00015469481581224272, | |
| "loss": 0.0308, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 8.823529411764707, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 0.00014759473930370736, | |
| "loss": 0.0192, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 9.411764705882353, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00014016954246529696, | |
| "loss": 0.0116, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.00013246994692046836, | |
| "loss": 0.0079, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 10.588235294117647, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.00012454854871407994, | |
| "loss": 0.0037, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 11.176470588235293, | |
| "grad_norm": 0.0595703125, | |
| "learning_rate": 0.00011645945902807341, | |
| "loss": 0.0031, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 11.764705882352942, | |
| "grad_norm": 0.062255859375, | |
| "learning_rate": 0.00010825793454723325, | |
| "loss": 0.0024, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 11.764705882352942, | |
| "eval_loss": 3.5474228858947754, | |
| "eval_runtime": 6.0534, | |
| "eval_samples_per_second": 247.958, | |
| "eval_steps_per_second": 5.286, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 12.352941176470589, | |
| "grad_norm": 0.03076171875, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0022, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 12.941176470588236, | |
| "grad_norm": 0.026123046875, | |
| "learning_rate": 9.174206545276677e-05, | |
| "loss": 0.0021, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 13.529411764705882, | |
| "grad_norm": 0.0230712890625, | |
| "learning_rate": 8.35405409719266e-05, | |
| "loss": 0.002, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 14.117647058823529, | |
| "grad_norm": 0.0223388671875, | |
| "learning_rate": 7.54514512859201e-05, | |
| "loss": 0.0021, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 14.705882352941176, | |
| "grad_norm": 0.0225830078125, | |
| "learning_rate": 6.753005307953167e-05, | |
| "loss": 0.002, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 15.294117647058824, | |
| "grad_norm": 0.0191650390625, | |
| "learning_rate": 5.983045753470308e-05, | |
| "loss": 0.002, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 15.882352941176471, | |
| "grad_norm": 0.0191650390625, | |
| "learning_rate": 5.240526069629265e-05, | |
| "loss": 0.002, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 16.470588235294116, | |
| "grad_norm": 0.01953125, | |
| "learning_rate": 4.530518418775733e-05, | |
| "loss": 0.0019, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 17.058823529411764, | |
| "grad_norm": 0.0213623046875, | |
| "learning_rate": 3.857872873103322e-05, | |
| "loss": 0.002, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 17.647058823529413, | |
| "grad_norm": 0.0230712890625, | |
| "learning_rate": 3.227184283742591e-05, | |
| "loss": 0.0019, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 18.235294117647058, | |
| "grad_norm": 0.0208740234375, | |
| "learning_rate": 2.6427608932686843e-05, | |
| "loss": 0.0019, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 18.823529411764707, | |
| "grad_norm": 0.0228271484375, | |
| "learning_rate": 2.1085949060360654e-05, | |
| "loss": 0.0019, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 19.41176470588235, | |
| "grad_norm": 0.02587890625, | |
| "learning_rate": 1.6283352173747145e-05, | |
| "loss": 0.002, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.021484375, | |
| "learning_rate": 1.2052624879351104e-05, | |
| "loss": 0.0019, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 20.58823529411765, | |
| "grad_norm": 0.0208740234375, | |
| "learning_rate": 8.422667334494249e-06, | |
| "loss": 0.002, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 21.176470588235293, | |
| "grad_norm": 0.02294921875, | |
| "learning_rate": 5.418275829936537e-06, | |
| "loss": 0.0019, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 21.764705882352942, | |
| "grad_norm": 0.020751953125, | |
| "learning_rate": 3.059973406066963e-06, | |
| "loss": 0.002, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 22.352941176470587, | |
| "grad_norm": 0.0203857421875, | |
| "learning_rate": 1.3638696597277679e-06, | |
| "loss": 0.0019, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 22.941176470588236, | |
| "grad_norm": 0.0240478515625, | |
| "learning_rate": 3.415506993330153e-07, | |
| "loss": 0.0019, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 23.529411764705884, | |
| "grad_norm": 0.0218505859375, | |
| "learning_rate": 0.0, | |
| "loss": 0.0019, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 23.529411764705884, | |
| "eval_loss": 3.6231682300567627, | |
| "eval_runtime": 6.0013, | |
| "eval_samples_per_second": 250.112, | |
| "eval_steps_per_second": 5.332, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 24, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.4407208221474816e+17, | |
| "train_batch_size": 12, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |