| { | |
| "best_metric": 0.2838810384273529, | |
| "best_model_checkpoint": "data/Mistral-7B_task-3_120-samples_config-1_auto/checkpoint-44", | |
| "epoch": 11.0, | |
| "eval_steps": 500, | |
| "global_step": 121, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 26.1705265045166, | |
| "learning_rate": 1.818181818181818e-06, | |
| "loss": 5.0531, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 17.742868423461914, | |
| "learning_rate": 3.636363636363636e-06, | |
| "loss": 3.3443, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 15.82374095916748, | |
| "learning_rate": 7.272727272727272e-06, | |
| "loss": 3.5057, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 11.942649841308594, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 3.1071, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 18.01738739013672, | |
| "learning_rate": 1.4545454545454545e-05, | |
| "loss": 2.5314, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 9.661962509155273, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 2.0683, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.5827921628952026, | |
| "eval_runtime": 25.8948, | |
| "eval_samples_per_second": 0.927, | |
| "eval_steps_per_second": 0.927, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 6.589108943939209, | |
| "learning_rate": 2.1818181818181818e-05, | |
| "loss": 1.7145, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 5.2075514793396, | |
| "learning_rate": 2.5454545454545454e-05, | |
| "loss": 1.2169, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 3.836588144302368, | |
| "learning_rate": 2.909090909090909e-05, | |
| "loss": 0.9367, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 2.0355136394500732, | |
| "learning_rate": 3.272727272727273e-05, | |
| "loss": 0.7085, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 1.877949595451355, | |
| "learning_rate": 3.6363636363636364e-05, | |
| "loss": 0.5436, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.243712067604065, | |
| "learning_rate": 4e-05, | |
| "loss": 0.3118, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.41756853461265564, | |
| "eval_runtime": 25.9327, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.9926459193229675, | |
| "learning_rate": 4.3636363636363636e-05, | |
| "loss": 0.3122, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 1.3329486846923828, | |
| "learning_rate": 4.7272727272727275e-05, | |
| "loss": 0.4175, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.8565844297409058, | |
| "learning_rate": 5.090909090909091e-05, | |
| "loss": 0.3766, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 1.0575027465820312, | |
| "learning_rate": 5.4545454545454546e-05, | |
| "loss": 0.3069, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.7391346096992493, | |
| "learning_rate": 5.818181818181818e-05, | |
| "loss": 0.1835, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.30521002411842346, | |
| "eval_runtime": 25.9236, | |
| "eval_samples_per_second": 0.926, | |
| "eval_steps_per_second": 0.926, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.090909090909091, | |
| "grad_norm": 0.837920069694519, | |
| "learning_rate": 6.181818181818182e-05, | |
| "loss": 0.3316, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.4749152362346649, | |
| "learning_rate": 6.545454545454546e-05, | |
| "loss": 0.0724, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.4545454545454546, | |
| "grad_norm": 0.9835648536682129, | |
| "learning_rate": 6.90909090909091e-05, | |
| "loss": 0.2561, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 1.0047273635864258, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 0.2265, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.8181818181818183, | |
| "grad_norm": 0.945101797580719, | |
| "learning_rate": 7.636363636363637e-05, | |
| "loss": 0.3011, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.3383798599243164, | |
| "learning_rate": 8e-05, | |
| "loss": 0.2052, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.2838810384273529, | |
| "eval_runtime": 25.927, | |
| "eval_samples_per_second": 0.926, | |
| "eval_steps_per_second": 0.926, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.181818181818182, | |
| "grad_norm": 0.6298382878303528, | |
| "learning_rate": 8.363636363636364e-05, | |
| "loss": 0.1359, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 1.335944652557373, | |
| "learning_rate": 8.727272727272727e-05, | |
| "loss": 0.2115, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 0.7435305118560791, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.1205, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.9755720496177673, | |
| "learning_rate": 9.454545454545455e-05, | |
| "loss": 0.1468, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "grad_norm": 0.9170610308647156, | |
| "learning_rate": 9.818181818181818e-05, | |
| "loss": 0.136, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.29344168305397034, | |
| "eval_runtime": 25.9261, | |
| "eval_samples_per_second": 0.926, | |
| "eval_steps_per_second": 0.926, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.8854531645774841, | |
| "learning_rate": 9.999899300364532e-05, | |
| "loss": 0.1279, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.2727272727272725, | |
| "grad_norm": 0.9929145574569702, | |
| "learning_rate": 9.99909372761763e-05, | |
| "loss": 0.0941, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.3176187574863434, | |
| "learning_rate": 9.997482711915927e-05, | |
| "loss": 0.0445, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.636363636363637, | |
| "grad_norm": 1.0197572708129883, | |
| "learning_rate": 9.99506651282272e-05, | |
| "loss": 0.069, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 1.3820139169692993, | |
| "learning_rate": 9.991845519630678e-05, | |
| "loss": 0.0702, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.1240243911743164, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.0724, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.35454604029655457, | |
| "eval_runtime": 25.8871, | |
| "eval_samples_per_second": 0.927, | |
| "eval_steps_per_second": 0.927, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.45250990986824036, | |
| "learning_rate": 9.982991356370404e-05, | |
| "loss": 0.0267, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 0.7566531300544739, | |
| "learning_rate": 9.977359612865423e-05, | |
| "loss": 0.0363, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.89458167552948, | |
| "learning_rate": 9.970925928158274e-05, | |
| "loss": 0.0696, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.6780168414115906, | |
| "learning_rate": 9.963691338830044e-05, | |
| "loss": 0.0387, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.7051839828491211, | |
| "learning_rate": 9.955657010501806e-05, | |
| "loss": 0.044, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.3940676152706146, | |
| "eval_runtime": 25.8841, | |
| "eval_samples_per_second": 0.927, | |
| "eval_steps_per_second": 0.927, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.090909090909091, | |
| "grad_norm": 0.7892889976501465, | |
| "learning_rate": 9.946824237646824e-05, | |
| "loss": 0.0172, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 1.2990819215774536, | |
| "learning_rate": 9.937194443381972e-05, | |
| "loss": 0.0399, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 7.454545454545454, | |
| "grad_norm": 0.8396146297454834, | |
| "learning_rate": 9.926769179238466e-05, | |
| "loss": 0.0133, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.7266479730606079, | |
| "learning_rate": 9.915550124911866e-05, | |
| "loss": 0.0259, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 7.818181818181818, | |
| "grad_norm": 0.36522987484931946, | |
| "learning_rate": 9.903539087991462e-05, | |
| "loss": 0.0304, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.8812476992607117, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.024, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.38941457867622375, | |
| "eval_runtime": 25.8888, | |
| "eval_samples_per_second": 0.927, | |
| "eval_steps_per_second": 0.927, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.181818181818182, | |
| "grad_norm": 0.4447575807571411, | |
| "learning_rate": 9.877148934427037e-05, | |
| "loss": 0.0154, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.7535117864608765, | |
| "learning_rate": 9.862774069706346e-05, | |
| "loss": 0.0145, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 8.545454545454545, | |
| "grad_norm": 0.5318331122398376, | |
| "learning_rate": 9.847615725553456e-05, | |
| "loss": 0.0103, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.3066743016242981, | |
| "learning_rate": 9.831676344247342e-05, | |
| "loss": 0.0136, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "grad_norm": 0.6094489693641663, | |
| "learning_rate": 9.814958493905963e-05, | |
| "loss": 0.0209, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.4123961925506592, | |
| "eval_runtime": 25.8844, | |
| "eval_samples_per_second": 0.927, | |
| "eval_steps_per_second": 0.927, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.3180105686187744, | |
| "learning_rate": 9.797464868072488e-05, | |
| "loss": 0.0087, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 9.272727272727273, | |
| "grad_norm": 0.08984816819429398, | |
| "learning_rate": 9.779198285281325e-05, | |
| "loss": 0.003, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.7463743686676025, | |
| "learning_rate": 9.760161688604008e-05, | |
| "loss": 0.0182, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 9.636363636363637, | |
| "grad_norm": 0.8020153045654297, | |
| "learning_rate": 9.740358145174998e-05, | |
| "loss": 0.0152, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.25512656569480896, | |
| "learning_rate": 9.719790845697533e-05, | |
| "loss": 0.0047, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.1539861559867859, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.0046, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.4324527680873871, | |
| "eval_runtime": 25.8849, | |
| "eval_samples_per_second": 0.927, | |
| "eval_steps_per_second": 0.927, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.27275049686431885, | |
| "learning_rate": 9.676378356149734e-05, | |
| "loss": 0.0035, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 10.363636363636363, | |
| "grad_norm": 0.9052261710166931, | |
| "learning_rate": 9.653540160603956e-05, | |
| "loss": 0.0135, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.4687228798866272, | |
| "learning_rate": 9.629952196931901e-05, | |
| "loss": 0.0049, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 10.727272727272727, | |
| "grad_norm": 0.448128342628479, | |
| "learning_rate": 9.60561826557425e-05, | |
| "loss": 0.0147, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.09168056398630142, | |
| "learning_rate": 9.580542287160348e-05, | |
| "loss": 0.0046, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 0.4639182984828949, | |
| "eval_runtime": 25.8862, | |
| "eval_samples_per_second": 0.927, | |
| "eval_steps_per_second": 0.927, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "step": 121, | |
| "total_flos": 9.479167540199424e+16, | |
| "train_loss": 0.42315512989870896, | |
| "train_runtime": 3524.0466, | |
| "train_samples_per_second": 1.249, | |
| "train_steps_per_second": 0.156 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.479167540199424e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |