| { | |
| "best_metric": 0.8607511520385742, | |
| "best_model_checkpoint": "data/Mistral-7B_task-3_120-samples_config-1_full/checkpoint-55", | |
| "epoch": 12.0, | |
| "eval_steps": 500, | |
| "global_step": 132, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 1.014518141746521, | |
| "learning_rate": 1.818181818181818e-06, | |
| "loss": 1.1726, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 1.3257009983062744, | |
| "learning_rate": 3.636363636363636e-06, | |
| "loss": 1.3941, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 1.0239942073822021, | |
| "learning_rate": 7.272727272727272e-06, | |
| "loss": 1.2286, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 0.9980838894844055, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 1.2355, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.9276570677757263, | |
| "learning_rate": 1.4545454545454545e-05, | |
| "loss": 1.2816, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.667737603187561, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 1.1682, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.1587175130844116, | |
| "eval_runtime": 25.7563, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.6125983595848083, | |
| "learning_rate": 2.1818181818181818e-05, | |
| "loss": 1.2098, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 0.4199548661708832, | |
| "learning_rate": 2.5454545454545454e-05, | |
| "loss": 1.108, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.5052300095558167, | |
| "learning_rate": 2.909090909090909e-05, | |
| "loss": 1.117, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 0.4365622401237488, | |
| "learning_rate": 3.272727272727273e-05, | |
| "loss": 1.1298, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.4207037389278412, | |
| "learning_rate": 3.6363636363636364e-05, | |
| "loss": 1.0669, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.38810834288597107, | |
| "learning_rate": 4e-05, | |
| "loss": 0.9812, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.0305534601211548, | |
| "eval_runtime": 25.761, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.3205520510673523, | |
| "learning_rate": 4.3636363636363636e-05, | |
| "loss": 1.0135, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 0.5428040027618408, | |
| "learning_rate": 4.7272727272727275e-05, | |
| "loss": 0.979, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.4363285005092621, | |
| "learning_rate": 5.090909090909091e-05, | |
| "loss": 0.9162, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.39783522486686707, | |
| "learning_rate": 5.4545454545454546e-05, | |
| "loss": 0.9397, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.3382175862789154, | |
| "learning_rate": 5.818181818181818e-05, | |
| "loss": 0.8834, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.9015331864356995, | |
| "eval_runtime": 25.756, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.090909090909091, | |
| "grad_norm": 0.29583463072776794, | |
| "learning_rate": 6.181818181818182e-05, | |
| "loss": 0.8834, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.2984974980354309, | |
| "learning_rate": 6.545454545454546e-05, | |
| "loss": 0.8559, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.4545454545454546, | |
| "grad_norm": 0.2952021360397339, | |
| "learning_rate": 6.90909090909091e-05, | |
| "loss": 0.8545, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.25758010149002075, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 0.8011, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.8181818181818183, | |
| "grad_norm": 0.36220088601112366, | |
| "learning_rate": 7.636363636363637e-05, | |
| "loss": 0.7679, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.27651169896125793, | |
| "learning_rate": 8e-05, | |
| "loss": 0.8826, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.8727057576179504, | |
| "eval_runtime": 25.7542, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.181818181818182, | |
| "grad_norm": 0.31009164452552795, | |
| "learning_rate": 8.363636363636364e-05, | |
| "loss": 0.7733, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.2957480549812317, | |
| "learning_rate": 8.727272727272727e-05, | |
| "loss": 0.8207, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 0.26560986042022705, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.7795, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.35368090867996216, | |
| "learning_rate": 9.454545454545455e-05, | |
| "loss": 0.7814, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "grad_norm": 0.2750256359577179, | |
| "learning_rate": 9.818181818181818e-05, | |
| "loss": 0.8207, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.8607511520385742, | |
| "eval_runtime": 25.7586, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.2918890118598938, | |
| "learning_rate": 9.999899300364532e-05, | |
| "loss": 0.7662, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.2727272727272725, | |
| "grad_norm": 0.332451730966568, | |
| "learning_rate": 9.99909372761763e-05, | |
| "loss": 0.7682, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.35697320103645325, | |
| "learning_rate": 9.997482711915927e-05, | |
| "loss": 0.7124, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.636363636363637, | |
| "grad_norm": 0.3677005171775818, | |
| "learning_rate": 9.99506651282272e-05, | |
| "loss": 0.7288, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.4263714551925659, | |
| "learning_rate": 9.991845519630678e-05, | |
| "loss": 0.7155, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.42828065156936646, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.7611, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.8609657287597656, | |
| "eval_runtime": 25.7705, | |
| "eval_samples_per_second": 0.931, | |
| "eval_steps_per_second": 0.931, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.4956981837749481, | |
| "learning_rate": 9.982991356370404e-05, | |
| "loss": 0.6481, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 0.6026014089584351, | |
| "learning_rate": 9.977359612865423e-05, | |
| "loss": 0.686, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.4963383376598358, | |
| "learning_rate": 9.970925928158274e-05, | |
| "loss": 0.621, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.5334177017211914, | |
| "learning_rate": 9.963691338830044e-05, | |
| "loss": 0.6601, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.5171763300895691, | |
| "learning_rate": 9.955657010501806e-05, | |
| "loss": 0.6922, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.8744406700134277, | |
| "eval_runtime": 25.765, | |
| "eval_samples_per_second": 0.931, | |
| "eval_steps_per_second": 0.931, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.090909090909091, | |
| "grad_norm": 0.5202062129974365, | |
| "learning_rate": 9.946824237646824e-05, | |
| "loss": 0.7068, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.5927935838699341, | |
| "learning_rate": 9.937194443381972e-05, | |
| "loss": 0.5278, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 7.454545454545454, | |
| "grad_norm": 0.777925431728363, | |
| "learning_rate": 9.926769179238466e-05, | |
| "loss": 0.5446, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.7315847873687744, | |
| "learning_rate": 9.915550124911866e-05, | |
| "loss": 0.5829, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 7.818181818181818, | |
| "grad_norm": 0.7044589519500732, | |
| "learning_rate": 9.903539087991462e-05, | |
| "loss": 0.5694, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.7512570023536682, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.6264, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.914975643157959, | |
| "eval_runtime": 25.7547, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.181818181818182, | |
| "grad_norm": 0.7690101265907288, | |
| "learning_rate": 9.877148934427037e-05, | |
| "loss": 0.5288, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 1.411813497543335, | |
| "learning_rate": 9.862774069706346e-05, | |
| "loss": 0.4844, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 8.545454545454545, | |
| "grad_norm": 0.875257134437561, | |
| "learning_rate": 9.847615725553456e-05, | |
| "loss": 0.509, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.7304671406745911, | |
| "learning_rate": 9.831676344247342e-05, | |
| "loss": 0.4303, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "grad_norm": 0.8690882921218872, | |
| "learning_rate": 9.814958493905963e-05, | |
| "loss": 0.4961, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.976428210735321, | |
| "eval_runtime": 25.7555, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 1.0365498065948486, | |
| "learning_rate": 9.797464868072488e-05, | |
| "loss": 0.4861, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 9.272727272727273, | |
| "grad_norm": 1.3250523805618286, | |
| "learning_rate": 9.779198285281325e-05, | |
| "loss": 0.3733, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 1.2837258577346802, | |
| "learning_rate": 9.760161688604008e-05, | |
| "loss": 0.3837, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 9.636363636363637, | |
| "grad_norm": 0.9932382702827454, | |
| "learning_rate": 9.740358145174998e-05, | |
| "loss": 0.4408, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.8722501993179321, | |
| "learning_rate": 9.719790845697533e-05, | |
| "loss": 0.3925, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.1440495252609253, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.3722, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 1.0250847339630127, | |
| "eval_runtime": 25.756, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.9267691969871521, | |
| "learning_rate": 9.676378356149734e-05, | |
| "loss": 0.3004, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 10.363636363636363, | |
| "grad_norm": 1.0404953956604004, | |
| "learning_rate": 9.653540160603956e-05, | |
| "loss": 0.3047, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 1.0245343446731567, | |
| "learning_rate": 9.629952196931901e-05, | |
| "loss": 0.3029, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 10.727272727272727, | |
| "grad_norm": 1.0685052871704102, | |
| "learning_rate": 9.60561826557425e-05, | |
| "loss": 0.3236, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 1.1701563596725464, | |
| "learning_rate": 9.580542287160348e-05, | |
| "loss": 0.3418, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 1.1046110391616821, | |
| "eval_runtime": 25.7573, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 11.090909090909092, | |
| "grad_norm": 1.055772066116333, | |
| "learning_rate": 9.554728301876526e-05, | |
| "loss": 0.2806, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 1.1970058679580688, | |
| "learning_rate": 9.528180468815155e-05, | |
| "loss": 0.2435, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 11.454545454545455, | |
| "grad_norm": 1.167232871055603, | |
| "learning_rate": 9.50090306530454e-05, | |
| "loss": 0.2556, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 1.0482594966888428, | |
| "learning_rate": 9.472900486219769e-05, | |
| "loss": 0.2869, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 11.818181818181818, | |
| "grad_norm": 1.254442572593689, | |
| "learning_rate": 9.444177243274618e-05, | |
| "loss": 0.245, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 1.1545424461364746, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.2556, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 1.1178990602493286, | |
| "eval_runtime": 25.756, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "step": 132, | |
| "total_flos": 1.0256775620486758e+17, | |
| "train_loss": 0.6957281243168947, | |
| "train_runtime": 3818.2199, | |
| "train_samples_per_second": 1.152, | |
| "train_steps_per_second": 0.144 | |
| } | |
| ], | |
| "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": 1.0256775620486758e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |