| { | |
| "best_metric": 0.004768455401062965, | |
| "best_model_checkpoint": "/mnt/ml_drive/kcardenas/limbxy_hands/checkpoint-1020", | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 1020, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.49019607843137253, | |
| "grad_norm": 556763.0, | |
| "learning_rate": 5e-06, | |
| "loss": 0.2302, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.9803921568627451, | |
| "grad_norm": 1358784.625, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1563, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.14388924837112427, | |
| "eval_rmse": 0.37932735681533813, | |
| "eval_runtime": 4.9465, | |
| "eval_samples_per_second": 114.83, | |
| "eval_steps_per_second": 1.819, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.4705882352941178, | |
| "grad_norm": 4929205.5, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.1396, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.9607843137254903, | |
| "grad_norm": 3301384.0, | |
| "learning_rate": 2e-05, | |
| "loss": 0.066, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.06285522878170013, | |
| "eval_rmse": 0.2507094442844391, | |
| "eval_runtime": 5.0622, | |
| "eval_samples_per_second": 112.205, | |
| "eval_steps_per_second": 1.778, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.450980392156863, | |
| "grad_norm": 1424321.875, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0443, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "grad_norm": 677306.75, | |
| "learning_rate": 3e-05, | |
| "loss": 0.031, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.024024562910199165, | |
| "eval_rmse": 0.15499858558177948, | |
| "eval_runtime": 5.0919, | |
| "eval_samples_per_second": 111.551, | |
| "eval_steps_per_second": 1.768, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 3.431372549019608, | |
| "grad_norm": 1847882.875, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.0227, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.9215686274509802, | |
| "grad_norm": 2230610.75, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0189, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.022217219695448875, | |
| "eval_rmse": 0.1490544229745865, | |
| "eval_runtime": 5.1067, | |
| "eval_samples_per_second": 111.227, | |
| "eval_steps_per_second": 1.762, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 4.411764705882353, | |
| "grad_norm": 2538623.25, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.0294, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.901960784313726, | |
| "grad_norm": 1391400.75, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0362, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.07003403455018997, | |
| "eval_rmse": 0.2646394371986389, | |
| "eval_runtime": 5.0039, | |
| "eval_samples_per_second": 113.512, | |
| "eval_steps_per_second": 1.799, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 5.392156862745098, | |
| "grad_norm": 1861511.0, | |
| "learning_rate": 4.8376623376623384e-05, | |
| "loss": 0.0243, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 591873.25, | |
| "learning_rate": 4.675324675324675e-05, | |
| "loss": 0.0106, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.011829150840640068, | |
| "eval_rmse": 0.10876189917325974, | |
| "eval_runtime": 5.0793, | |
| "eval_samples_per_second": 111.825, | |
| "eval_steps_per_second": 1.772, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 6.372549019607844, | |
| "grad_norm": 717852.0625, | |
| "learning_rate": 4.5129870129870135e-05, | |
| "loss": 0.011, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 6.862745098039216, | |
| "grad_norm": 5022134.0, | |
| "learning_rate": 4.3506493506493503e-05, | |
| "loss": 0.0217, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.04193717613816261, | |
| "eval_rmse": 0.20478568971157074, | |
| "eval_runtime": 5.0411, | |
| "eval_samples_per_second": 112.673, | |
| "eval_steps_per_second": 1.785, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 7.352941176470588, | |
| "grad_norm": 4874459.0, | |
| "learning_rate": 4.1883116883116886e-05, | |
| "loss": 0.0272, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 7.8431372549019605, | |
| "grad_norm": 2519686.0, | |
| "learning_rate": 4.025974025974026e-05, | |
| "loss": 0.0167, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.01697073131799698, | |
| "eval_rmse": 0.1302717626094818, | |
| "eval_runtime": 5.0791, | |
| "eval_samples_per_second": 111.83, | |
| "eval_steps_per_second": 1.772, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 1922916.0, | |
| "learning_rate": 3.8636363636363636e-05, | |
| "loss": 0.0125, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 8.823529411764707, | |
| "grad_norm": 1355185.75, | |
| "learning_rate": 3.701298701298702e-05, | |
| "loss": 0.0078, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.011065785773098469, | |
| "eval_rmse": 0.10519403964281082, | |
| "eval_runtime": 5.0743, | |
| "eval_samples_per_second": 111.938, | |
| "eval_steps_per_second": 1.774, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 9.313725490196079, | |
| "grad_norm": 3503622.25, | |
| "learning_rate": 3.5389610389610387e-05, | |
| "loss": 0.0128, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 9.803921568627452, | |
| "grad_norm": 2042047.625, | |
| "learning_rate": 3.376623376623377e-05, | |
| "loss": 0.0102, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.01578633300960064, | |
| "eval_rmse": 0.12564367055892944, | |
| "eval_runtime": 5.0736, | |
| "eval_samples_per_second": 111.953, | |
| "eval_steps_per_second": 1.774, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 10.294117647058824, | |
| "grad_norm": 1890998.625, | |
| "learning_rate": 3.2142857142857144e-05, | |
| "loss": 0.0067, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 10.784313725490197, | |
| "grad_norm": 1245511.75, | |
| "learning_rate": 3.051948051948052e-05, | |
| "loss": 0.009, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 0.006434537936002016, | |
| "eval_rmse": 0.08021557331085205, | |
| "eval_runtime": 4.9055, | |
| "eval_samples_per_second": 115.79, | |
| "eval_steps_per_second": 1.835, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 11.27450980392157, | |
| "grad_norm": 1712539.625, | |
| "learning_rate": 2.8896103896103898e-05, | |
| "loss": 0.0044, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 11.764705882352942, | |
| "grad_norm": 3221122.25, | |
| "learning_rate": 2.7272727272727273e-05, | |
| "loss": 0.0067, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.00743119465187192, | |
| "eval_rmse": 0.08620437234640121, | |
| "eval_runtime": 5.1361, | |
| "eval_samples_per_second": 110.589, | |
| "eval_steps_per_second": 1.752, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 12.254901960784313, | |
| "grad_norm": 862932.25, | |
| "learning_rate": 2.5649350649350652e-05, | |
| "loss": 0.0058, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 12.745098039215687, | |
| "grad_norm": 1713839.5, | |
| "learning_rate": 2.4025974025974027e-05, | |
| "loss": 0.0047, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 0.0054638562723994255, | |
| "eval_rmse": 0.07391790300607681, | |
| "eval_runtime": 5.1487, | |
| "eval_samples_per_second": 110.319, | |
| "eval_steps_per_second": 1.748, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 13.235294117647058, | |
| "grad_norm": 1322453.625, | |
| "learning_rate": 2.2402597402597402e-05, | |
| "loss": 0.0043, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 13.72549019607843, | |
| "grad_norm": 941169.0, | |
| "learning_rate": 2.077922077922078e-05, | |
| "loss": 0.0036, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.006842494942247868, | |
| "eval_rmse": 0.08271937817335129, | |
| "eval_runtime": 4.825, | |
| "eval_samples_per_second": 117.719, | |
| "eval_steps_per_second": 1.865, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 14.215686274509803, | |
| "grad_norm": 2176495.5, | |
| "learning_rate": 1.9155844155844156e-05, | |
| "loss": 0.0036, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 14.705882352941176, | |
| "grad_norm": 411098.53125, | |
| "learning_rate": 1.7532467532467535e-05, | |
| "loss": 0.0072, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 0.0075486451387405396, | |
| "eval_rmse": 0.08688294142484665, | |
| "eval_runtime": 5.1462, | |
| "eval_samples_per_second": 110.372, | |
| "eval_steps_per_second": 1.749, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 15.196078431372548, | |
| "grad_norm": 2002094.375, | |
| "learning_rate": 1.590909090909091e-05, | |
| "loss": 0.0037, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 15.686274509803921, | |
| "grad_norm": 2112328.75, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.006, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.005468303803354502, | |
| "eval_rmse": 0.073947973549366, | |
| "eval_runtime": 5.0395, | |
| "eval_samples_per_second": 112.709, | |
| "eval_steps_per_second": 1.786, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 16.176470588235293, | |
| "grad_norm": 1033486.875, | |
| "learning_rate": 1.2662337662337662e-05, | |
| "loss": 0.0036, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 16.666666666666668, | |
| "grad_norm": 321803.625, | |
| "learning_rate": 1.103896103896104e-05, | |
| "loss": 0.0022, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_loss": 0.005201099440455437, | |
| "eval_rmse": 0.07211864739656448, | |
| "eval_runtime": 5.018, | |
| "eval_samples_per_second": 113.192, | |
| "eval_steps_per_second": 1.794, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 17.15686274509804, | |
| "grad_norm": 10877.533203125, | |
| "learning_rate": 9.415584415584416e-06, | |
| "loss": 0.0016, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 17.647058823529413, | |
| "grad_norm": 1244264.875, | |
| "learning_rate": 7.792207792207792e-06, | |
| "loss": 0.0051, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 0.005439546890556812, | |
| "eval_rmse": 0.07375328242778778, | |
| "eval_runtime": 5.1056, | |
| "eval_samples_per_second": 111.251, | |
| "eval_steps_per_second": 1.763, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 18.137254901960784, | |
| "grad_norm": 395613.0, | |
| "learning_rate": 6.168831168831169e-06, | |
| "loss": 0.0033, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 18.627450980392158, | |
| "grad_norm": 1329285.375, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 0.0023, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_loss": 0.0050988770090043545, | |
| "eval_rmse": 0.07140642404556274, | |
| "eval_runtime": 5.0748, | |
| "eval_samples_per_second": 111.926, | |
| "eval_steps_per_second": 1.773, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 19.11764705882353, | |
| "grad_norm": 406520.3125, | |
| "learning_rate": 2.922077922077922e-06, | |
| "loss": 0.0019, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 19.607843137254903, | |
| "grad_norm": 275064.03125, | |
| "learning_rate": 1.2987012987012988e-06, | |
| "loss": 0.0016, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 0.004768455401062965, | |
| "eval_rmse": 0.0690540000796318, | |
| "eval_runtime": 5.0305, | |
| "eval_samples_per_second": 112.912, | |
| "eval_steps_per_second": 1.789, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 1020, | |
| "total_flos": 0.0, | |
| "train_loss": 0.02494959205973382, | |
| "train_runtime": 1869.8943, | |
| "train_samples_per_second": 34.398, | |
| "train_steps_per_second": 0.545 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 1020, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |