| { | |
| "best_metric": 0.32597243785858154, | |
| "best_model_checkpoint": "data/Llama-31-8B_task-3_120-samples_config-1_auto/checkpoint-66", | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 143, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 4.012484550476074, | |
| "learning_rate": 1.818181818181818e-06, | |
| "loss": 3.0293, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 2.3296754360198975, | |
| "learning_rate": 3.636363636363636e-06, | |
| "loss": 2.2505, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 2.2629120349884033, | |
| "learning_rate": 7.272727272727272e-06, | |
| "loss": 2.3565, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 1.9213932752609253, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 2.2364, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 3.3934085369110107, | |
| "learning_rate": 1.4545454545454545e-05, | |
| "loss": 2.1148, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 2.808576822280884, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 2.0265, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.866349220275879, | |
| "eval_runtime": 21.1106, | |
| "eval_samples_per_second": 1.137, | |
| "eval_steps_per_second": 1.137, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 2.242879629135132, | |
| "learning_rate": 2.1818181818181818e-05, | |
| "loss": 1.928, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 2.0829696655273438, | |
| "learning_rate": 2.5454545454545454e-05, | |
| "loss": 1.5942, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 1.9415721893310547, | |
| "learning_rate": 2.909090909090909e-05, | |
| "loss": 1.6149, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 1.7251673936843872, | |
| "learning_rate": 3.272727272727273e-05, | |
| "loss": 1.2239, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 1.3666032552719116, | |
| "learning_rate": 3.6363636363636364e-05, | |
| "loss": 1.0307, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.7636884450912476, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6418, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.653621256351471, | |
| "eval_runtime": 21.1068, | |
| "eval_samples_per_second": 1.137, | |
| "eval_steps_per_second": 1.137, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.9443819522857666, | |
| "learning_rate": 4.3636363636363636e-05, | |
| "loss": 0.5085, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 1.1375559568405151, | |
| "learning_rate": 4.7272727272727275e-05, | |
| "loss": 0.5995, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.6164768934249878, | |
| "learning_rate": 5.090909090909091e-05, | |
| "loss": 0.5521, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.8946235179901123, | |
| "learning_rate": 5.4545454545454546e-05, | |
| "loss": 0.438, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.6255387663841248, | |
| "learning_rate": 5.818181818181818e-05, | |
| "loss": 0.2518, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.41362643241882324, | |
| "eval_runtime": 21.104, | |
| "eval_samples_per_second": 1.137, | |
| "eval_steps_per_second": 1.137, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.090909090909091, | |
| "grad_norm": 1.8066771030426025, | |
| "learning_rate": 6.181818181818182e-05, | |
| "loss": 0.4488, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.29350921511650085, | |
| "learning_rate": 6.545454545454546e-05, | |
| "loss": 0.1088, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.4545454545454546, | |
| "grad_norm": 0.44908615946769714, | |
| "learning_rate": 6.90909090909091e-05, | |
| "loss": 0.3992, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.3139246106147766, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 0.3445, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.8181818181818183, | |
| "grad_norm": 0.4786546528339386, | |
| "learning_rate": 7.636363636363637e-05, | |
| "loss": 0.4866, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.3976927101612091, | |
| "learning_rate": 8e-05, | |
| "loss": 0.3244, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.3522801697254181, | |
| "eval_runtime": 21.1014, | |
| "eval_samples_per_second": 1.137, | |
| "eval_steps_per_second": 1.137, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.181818181818182, | |
| "grad_norm": 1.2649972438812256, | |
| "learning_rate": 8.363636363636364e-05, | |
| "loss": 0.2522, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.44340747594833374, | |
| "learning_rate": 8.727272727272727e-05, | |
| "loss": 0.4279, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 0.573393702507019, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.2589, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.3285563588142395, | |
| "learning_rate": 9.454545454545455e-05, | |
| "loss": 0.2671, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "grad_norm": 5.318609714508057, | |
| "learning_rate": 9.818181818181818e-05, | |
| "loss": 0.3276, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.33851274847984314, | |
| "eval_runtime": 21.3331, | |
| "eval_samples_per_second": 1.125, | |
| "eval_steps_per_second": 1.125, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.385530561208725, | |
| "learning_rate": 9.999899300364532e-05, | |
| "loss": 0.3258, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.2727272727272725, | |
| "grad_norm": 0.42068204283714294, | |
| "learning_rate": 9.99909372761763e-05, | |
| "loss": 0.3124, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.14537262916564941, | |
| "learning_rate": 9.997482711915927e-05, | |
| "loss": 0.1665, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.636363636363637, | |
| "grad_norm": 0.38247016072273254, | |
| "learning_rate": 9.99506651282272e-05, | |
| "loss": 0.2435, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.3936985731124878, | |
| "learning_rate": 9.991845519630678e-05, | |
| "loss": 0.2423, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.4892041087150574, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.3029, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.32597243785858154, | |
| "eval_runtime": 21.3339, | |
| "eval_samples_per_second": 1.125, | |
| "eval_steps_per_second": 1.125, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.30393707752227783, | |
| "learning_rate": 9.982991356370404e-05, | |
| "loss": 0.1767, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 0.3415716588497162, | |
| "learning_rate": 9.977359612865423e-05, | |
| "loss": 0.2164, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.4968898892402649, | |
| "learning_rate": 9.970925928158274e-05, | |
| "loss": 0.3642, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.41892415285110474, | |
| "learning_rate": 9.963691338830044e-05, | |
| "loss": 0.2064, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.4685850739479065, | |
| "learning_rate": 9.955657010501806e-05, | |
| "loss": 0.2127, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.34463366866111755, | |
| "eval_runtime": 21.1697, | |
| "eval_samples_per_second": 1.134, | |
| "eval_steps_per_second": 1.134, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.090909090909091, | |
| "grad_norm": 0.5011253952980042, | |
| "learning_rate": 9.946824237646824e-05, | |
| "loss": 0.1221, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.8317933082580566, | |
| "learning_rate": 9.937194443381972e-05, | |
| "loss": 0.2695, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 7.454545454545454, | |
| "grad_norm": 0.47007039189338684, | |
| "learning_rate": 9.926769179238466e-05, | |
| "loss": 0.0938, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.4090757668018341, | |
| "learning_rate": 9.915550124911866e-05, | |
| "loss": 0.1174, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 7.818181818181818, | |
| "grad_norm": 0.46983206272125244, | |
| "learning_rate": 9.903539087991462e-05, | |
| "loss": 0.169, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.8168413639068604, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.1592, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.3746682405471802, | |
| "eval_runtime": 21.1418, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 1.135, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.181818181818182, | |
| "grad_norm": 0.5034617781639099, | |
| "learning_rate": 9.877148934427037e-05, | |
| "loss": 0.1256, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.7391631603240967, | |
| "learning_rate": 9.862774069706346e-05, | |
| "loss": 0.092, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 8.545454545454545, | |
| "grad_norm": 1.0084911584854126, | |
| "learning_rate": 9.847615725553456e-05, | |
| "loss": 0.0646, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.5537567734718323, | |
| "learning_rate": 9.831676344247342e-05, | |
| "loss": 0.1261, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "grad_norm": 0.8034816980361938, | |
| "learning_rate": 9.814958493905963e-05, | |
| "loss": 0.1194, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.4065481722354889, | |
| "eval_runtime": 21.3736, | |
| "eval_samples_per_second": 1.123, | |
| "eval_steps_per_second": 1.123, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.6414387822151184, | |
| "learning_rate": 9.797464868072488e-05, | |
| "loss": 0.0815, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 9.272727272727273, | |
| "grad_norm": 0.14650435745716095, | |
| "learning_rate": 9.779198285281325e-05, | |
| "loss": 0.0249, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.9511333107948303, | |
| "learning_rate": 9.760161688604008e-05, | |
| "loss": 0.0989, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 9.636363636363637, | |
| "grad_norm": 1.684899926185608, | |
| "learning_rate": 9.740358145174998e-05, | |
| "loss": 0.0617, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.6999126076698303, | |
| "learning_rate": 9.719790845697533e-05, | |
| "loss": 0.0418, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.7452811002731323, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.0429, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.49909672141075134, | |
| "eval_runtime": 21.3601, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.3761460483074188, | |
| "learning_rate": 9.676378356149734e-05, | |
| "loss": 0.0316, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 10.363636363636363, | |
| "grad_norm": 0.5788271427154541, | |
| "learning_rate": 9.653540160603956e-05, | |
| "loss": 0.0477, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.5197237133979797, | |
| "learning_rate": 9.629952196931901e-05, | |
| "loss": 0.0287, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 10.727272727272727, | |
| "grad_norm": 1.2195073366165161, | |
| "learning_rate": 9.60561826557425e-05, | |
| "loss": 0.0533, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.29727470874786377, | |
| "learning_rate": 9.580542287160348e-05, | |
| "loss": 0.0213, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 0.5292800068855286, | |
| "eval_runtime": 21.1582, | |
| "eval_samples_per_second": 1.134, | |
| "eval_steps_per_second": 1.134, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 11.090909090909092, | |
| "grad_norm": 0.37346208095550537, | |
| "learning_rate": 9.554728301876526e-05, | |
| "loss": 0.0289, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.579096257686615, | |
| "learning_rate": 9.528180468815155e-05, | |
| "loss": 0.02, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 11.454545454545455, | |
| "grad_norm": 0.2732711434364319, | |
| "learning_rate": 9.50090306530454e-05, | |
| "loss": 0.016, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.42670881748199463, | |
| "learning_rate": 9.472900486219769e-05, | |
| "loss": 0.0173, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 11.818181818181818, | |
| "grad_norm": 0.614210307598114, | |
| "learning_rate": 9.444177243274618e-05, | |
| "loss": 0.0156, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 1.006243109703064, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.0509, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.5700822472572327, | |
| "eval_runtime": 21.136, | |
| "eval_samples_per_second": 1.136, | |
| "eval_steps_per_second": 1.136, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 12.181818181818182, | |
| "grad_norm": 0.11038227379322052, | |
| "learning_rate": 9.384587392471515e-05, | |
| "loss": 0.0061, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 12.363636363636363, | |
| "grad_norm": 0.04612269997596741, | |
| "learning_rate": 9.353730385598887e-05, | |
| "loss": 0.0071, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 12.545454545454545, | |
| "grad_norm": 0.6978845596313477, | |
| "learning_rate": 9.322171915289635e-05, | |
| "loss": 0.0241, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 12.727272727272727, | |
| "grad_norm": 1.0454477071762085, | |
| "learning_rate": 9.289917066174886e-05, | |
| "loss": 0.0276, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 12.909090909090908, | |
| "grad_norm": 0.3554699122905731, | |
| "learning_rate": 9.256971035084785e-05, | |
| "loss": 0.0154, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 0.5771800875663757, | |
| "eval_runtime": 21.1492, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 1.135, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 143, | |
| "total_flos": 9.013378055090995e+16, | |
| "train_loss": 0.4252611890409168, | |
| "train_runtime": 3333.0705, | |
| "train_samples_per_second": 1.32, | |
| "train_steps_per_second": 0.165 | |
| } | |
| ], | |
| "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": 9.013378055090995e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |