| { | |
| "best_metric": 0.39128661155700684, | |
| "best_model_checkpoint": "data/Mistral-7B_task-3_120-samples_config-1/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": 26.581195831298828, | |
| "learning_rate": 1.818181818181818e-06, | |
| "loss": 5.0961, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 17.186176300048828, | |
| "learning_rate": 3.636363636363636e-06, | |
| "loss": 3.7834, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 16.22742462158203, | |
| "learning_rate": 7.272727272727272e-06, | |
| "loss": 3.7902, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 13.555403709411621, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 3.3208, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 19.296531677246094, | |
| "learning_rate": 1.4545454545454545e-05, | |
| "loss": 2.8761, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 10.272661209106445, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 2.2415, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.829300880432129, | |
| "eval_runtime": 25.7663, | |
| "eval_samples_per_second": 0.931, | |
| "eval_steps_per_second": 0.931, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 7.439228057861328, | |
| "learning_rate": 2.1818181818181818e-05, | |
| "loss": 1.8772, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 5.770608901977539, | |
| "learning_rate": 2.5454545454545454e-05, | |
| "loss": 1.4832, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 4.287290573120117, | |
| "learning_rate": 2.909090909090909e-05, | |
| "loss": 1.0633, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 2.3973379135131836, | |
| "learning_rate": 3.272727272727273e-05, | |
| "loss": 0.8615, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 2.0681209564208984, | |
| "learning_rate": 3.6363636363636364e-05, | |
| "loss": 0.6223, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.4477177858352661, | |
| "learning_rate": 4e-05, | |
| "loss": 0.3793, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.5601645112037659, | |
| "eval_runtime": 25.7673, | |
| "eval_samples_per_second": 0.931, | |
| "eval_steps_per_second": 0.931, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 1.0610381364822388, | |
| "learning_rate": 4.3636363636363636e-05, | |
| "loss": 0.4043, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 1.2741479873657227, | |
| "learning_rate": 4.7272727272727275e-05, | |
| "loss": 0.5729, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.8593302369117737, | |
| "learning_rate": 5.090909090909091e-05, | |
| "loss": 0.3782, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 1.1046785116195679, | |
| "learning_rate": 5.4545454545454546e-05, | |
| "loss": 0.3238, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.7316509485244751, | |
| "learning_rate": 5.818181818181818e-05, | |
| "loss": 0.2382, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.42310047149658203, | |
| "eval_runtime": 25.7592, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.090909090909091, | |
| "grad_norm": 1.131074070930481, | |
| "learning_rate": 6.181818181818182e-05, | |
| "loss": 0.4555, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.38236701488494873, | |
| "learning_rate": 6.545454545454546e-05, | |
| "loss": 0.1055, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.4545454545454546, | |
| "grad_norm": 4.889406681060791, | |
| "learning_rate": 6.90909090909091e-05, | |
| "loss": 0.2955, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.8969003558158875, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 0.281, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.8181818181818183, | |
| "grad_norm": 1.034072995185852, | |
| "learning_rate": 7.636363636363637e-05, | |
| "loss": 0.3261, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.0129034519195557, | |
| "learning_rate": 8e-05, | |
| "loss": 0.2918, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.39155831933021545, | |
| "eval_runtime": 25.7607, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.181818181818182, | |
| "grad_norm": 0.6423981785774231, | |
| "learning_rate": 8.363636363636364e-05, | |
| "loss": 0.1505, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 1.1609551906585693, | |
| "learning_rate": 8.727272727272727e-05, | |
| "loss": 0.2499, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 0.9071617722511292, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.1717, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 2.986572027206421, | |
| "learning_rate": 9.454545454545455e-05, | |
| "loss": 0.1974, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "grad_norm": 1.1126000881195068, | |
| "learning_rate": 9.818181818181818e-05, | |
| "loss": 0.1887, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.39128661155700684, | |
| "eval_runtime": 25.7604, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.8015605807304382, | |
| "learning_rate": 9.999899300364532e-05, | |
| "loss": 0.1774, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.2727272727272725, | |
| "grad_norm": 0.818981409072876, | |
| "learning_rate": 9.99909372761763e-05, | |
| "loss": 0.1105, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.3991180658340454, | |
| "learning_rate": 9.997482711915927e-05, | |
| "loss": 0.0658, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.636363636363637, | |
| "grad_norm": 1.6436625719070435, | |
| "learning_rate": 9.99506651282272e-05, | |
| "loss": 0.1566, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 1.0331753492355347, | |
| "learning_rate": 9.991845519630678e-05, | |
| "loss": 0.0605, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.9555497169494629, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.0868, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.45599040389060974, | |
| "eval_runtime": 25.7595, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.46245571970939636, | |
| "learning_rate": 9.982991356370404e-05, | |
| "loss": 0.0335, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 0.4522000253200531, | |
| "learning_rate": 9.977359612865423e-05, | |
| "loss": 0.0224, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 1.2461261749267578, | |
| "learning_rate": 9.970925928158274e-05, | |
| "loss": 0.0728, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.866970419883728, | |
| "learning_rate": 9.963691338830044e-05, | |
| "loss": 0.0479, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 1.0041601657867432, | |
| "learning_rate": 9.955657010501806e-05, | |
| "loss": 0.0333, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.5413926243782043, | |
| "eval_runtime": 25.7595, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.090909090909091, | |
| "grad_norm": 0.7442712783813477, | |
| "learning_rate": 9.946824237646824e-05, | |
| "loss": 0.0094, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 1.4385738372802734, | |
| "learning_rate": 9.937194443381972e-05, | |
| "loss": 0.023, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 7.454545454545454, | |
| "grad_norm": 0.7076303958892822, | |
| "learning_rate": 9.926769179238466e-05, | |
| "loss": 0.0098, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.49934569001197815, | |
| "learning_rate": 9.915550124911866e-05, | |
| "loss": 0.018, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 7.818181818181818, | |
| "grad_norm": 0.6519784331321716, | |
| "learning_rate": 9.903539087991462e-05, | |
| "loss": 0.032, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.025206446647644, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.0184, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.5703032612800598, | |
| "eval_runtime": 25.7606, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.181818181818182, | |
| "grad_norm": 0.32475417852401733, | |
| "learning_rate": 9.877148934427037e-05, | |
| "loss": 0.0117, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.3134045898914337, | |
| "learning_rate": 9.862774069706346e-05, | |
| "loss": 0.0058, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 8.545454545454545, | |
| "grad_norm": 0.2756688892841339, | |
| "learning_rate": 9.847615725553456e-05, | |
| "loss": 0.0025, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.28128984570503235, | |
| "learning_rate": 9.831676344247342e-05, | |
| "loss": 0.0189, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "grad_norm": 1.5065057277679443, | |
| "learning_rate": 9.814958493905963e-05, | |
| "loss": 0.0298, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.5940575003623962, | |
| "eval_runtime": 25.767, | |
| "eval_samples_per_second": 0.931, | |
| "eval_steps_per_second": 0.931, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.26251956820487976, | |
| "learning_rate": 9.797464868072488e-05, | |
| "loss": 0.0124, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 9.272727272727273, | |
| "grad_norm": 0.011677834205329418, | |
| "learning_rate": 9.779198285281325e-05, | |
| "loss": 0.0009, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.3504819869995117, | |
| "learning_rate": 9.760161688604008e-05, | |
| "loss": 0.01, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 9.636363636363637, | |
| "grad_norm": 0.32781001925468445, | |
| "learning_rate": 9.740358145174998e-05, | |
| "loss": 0.0029, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 1.1787352561950684, | |
| "learning_rate": 9.719790845697533e-05, | |
| "loss": 0.0088, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.2411821335554123, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.0026, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.632849395275116, | |
| "eval_runtime": 25.7634, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.5745018124580383, | |
| "learning_rate": 9.676378356149734e-05, | |
| "loss": 0.0022, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 10.363636363636363, | |
| "grad_norm": 0.12452679872512817, | |
| "learning_rate": 9.653540160603956e-05, | |
| "loss": 0.002, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.06492336839437485, | |
| "learning_rate": 9.629952196931901e-05, | |
| "loss": 0.0015, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 10.727272727272727, | |
| "grad_norm": 0.6480187773704529, | |
| "learning_rate": 9.60561826557425e-05, | |
| "loss": 0.0057, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.02387073077261448, | |
| "learning_rate": 9.580542287160348e-05, | |
| "loss": 0.0007, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 0.6887062191963196, | |
| "eval_runtime": 25.7643, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 11.090909090909092, | |
| "grad_norm": 0.021298233419656754, | |
| "learning_rate": 9.554728301876526e-05, | |
| "loss": 0.0037, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.160748690366745, | |
| "learning_rate": 9.528180468815155e-05, | |
| "loss": 0.0009, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 11.454545454545455, | |
| "grad_norm": 0.018241755664348602, | |
| "learning_rate": 9.50090306530454e-05, | |
| "loss": 0.0009, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.5487923622131348, | |
| "learning_rate": 9.472900486219769e-05, | |
| "loss": 0.0031, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 11.818181818181818, | |
| "grad_norm": 0.021489502862095833, | |
| "learning_rate": 9.444177243274618e-05, | |
| "loss": 0.0008, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.4083148241043091, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.0086, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.7278346419334412, | |
| "eval_runtime": 25.7635, | |
| "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": 0.4348593218530754, | |
| "train_runtime": 3817.9089, | |
| "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 | |
| } | |