| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 3000, | |
| "global_step": 88686, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.382721060821325e-05, | |
| "grad_norm": 1760.0, | |
| "learning_rate": 0.00029999661727893914, | |
| "loss": 57.5, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.10148163182463973, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 0.000289851836817536, | |
| "loss": 4.5083, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.10148163182463973, | |
| "eval_loss": 2.8375253677368164, | |
| "eval_runtime": 88.2347, | |
| "eval_samples_per_second": 1069.931, | |
| "eval_steps_per_second": 8.364, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.20296326364927947, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 0.000279703673635072, | |
| "loss": 3.6752, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.20296326364927947, | |
| "eval_loss": 2.74216628074646, | |
| "eval_runtime": 88.2236, | |
| "eval_samples_per_second": 1070.065, | |
| "eval_steps_per_second": 8.365, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.30444489547391923, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 0.00026955551045260807, | |
| "loss": 3.598, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.30444489547391923, | |
| "eval_loss": 2.70210862159729, | |
| "eval_runtime": 88.2673, | |
| "eval_samples_per_second": 1069.535, | |
| "eval_steps_per_second": 8.361, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.40592652729855894, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 0.0002594073472701441, | |
| "loss": 3.5651, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.40592652729855894, | |
| "eval_loss": 2.686271905899048, | |
| "eval_runtime": 88.2934, | |
| "eval_samples_per_second": 1069.219, | |
| "eval_steps_per_second": 8.358, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.5074081591231987, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 0.0002492591840876801, | |
| "loss": 3.5409, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.5074081591231987, | |
| "eval_loss": 2.672954797744751, | |
| "eval_runtime": 88.2673, | |
| "eval_samples_per_second": 1069.535, | |
| "eval_steps_per_second": 8.361, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.6088897909478385, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 0.0002391110209052161, | |
| "loss": 3.5353, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.6088897909478385, | |
| "eval_loss": 2.6632580757141113, | |
| "eval_runtime": 88.3039, | |
| "eval_samples_per_second": 1069.092, | |
| "eval_steps_per_second": 8.358, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.7103714227724782, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 0.00022896285772275215, | |
| "loss": 3.5268, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.7103714227724782, | |
| "eval_loss": 2.6613523960113525, | |
| "eval_runtime": 88.3108, | |
| "eval_samples_per_second": 1069.009, | |
| "eval_steps_per_second": 8.357, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.8118530545971179, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 0.0002188146945402882, | |
| "loss": 3.5223, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.8118530545971179, | |
| "eval_loss": 2.6542599201202393, | |
| "eval_runtime": 88.2938, | |
| "eval_samples_per_second": 1069.214, | |
| "eval_steps_per_second": 8.358, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.9133346864217576, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 0.00020866653135782423, | |
| "loss": 3.5185, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.9133346864217576, | |
| "eval_loss": 2.6546196937561035, | |
| "eval_runtime": 88.3801, | |
| "eval_samples_per_second": 1068.17, | |
| "eval_steps_per_second": 8.35, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.0148163182463974, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 0.00019851836817536025, | |
| "loss": 3.515, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.0148163182463974, | |
| "eval_loss": 2.6509780883789062, | |
| "eval_runtime": 88.3972, | |
| "eval_samples_per_second": 1067.963, | |
| "eval_steps_per_second": 8.349, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.116297950071037, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 0.0001883702049928963, | |
| "loss": 3.5138, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 1.116297950071037, | |
| "eval_loss": 2.6480565071105957, | |
| "eval_runtime": 88.3803, | |
| "eval_samples_per_second": 1068.168, | |
| "eval_steps_per_second": 8.35, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 1.217779581895677, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 0.0001782220418104323, | |
| "loss": 3.508, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.217779581895677, | |
| "eval_loss": 2.6456427574157715, | |
| "eval_runtime": 88.3938, | |
| "eval_samples_per_second": 1068.005, | |
| "eval_steps_per_second": 8.349, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.3192612137203166, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 0.00016807387862796832, | |
| "loss": 3.5071, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 1.3192612137203166, | |
| "eval_loss": 2.642657518386841, | |
| "eval_runtime": 88.3875, | |
| "eval_samples_per_second": 1068.081, | |
| "eval_steps_per_second": 8.35, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 1.4207428455449564, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 0.00015792571544550436, | |
| "loss": 3.5058, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 1.4207428455449564, | |
| "eval_loss": 2.6432926654815674, | |
| "eval_runtime": 88.4239, | |
| "eval_samples_per_second": 1067.641, | |
| "eval_steps_per_second": 8.346, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 1.522224477369596, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 0.00014777755226304037, | |
| "loss": 3.5063, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 1.522224477369596, | |
| "eval_loss": 2.644076108932495, | |
| "eval_runtime": 88.4016, | |
| "eval_samples_per_second": 1067.91, | |
| "eval_steps_per_second": 8.348, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 1.6237061091942357, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 0.0001376293890805764, | |
| "loss": 3.5003, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 1.6237061091942357, | |
| "eval_loss": 2.6422553062438965, | |
| "eval_runtime": 88.447, | |
| "eval_samples_per_second": 1067.362, | |
| "eval_steps_per_second": 8.344, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 1.7251877410188756, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 0.00012748122589811243, | |
| "loss": 3.5018, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 1.7251877410188756, | |
| "eval_loss": 2.643885612487793, | |
| "eval_runtime": 88.2483, | |
| "eval_samples_per_second": 1069.765, | |
| "eval_steps_per_second": 8.363, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 1.8266693728435153, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 0.00011733306271564845, | |
| "loss": 3.4992, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 1.8266693728435153, | |
| "eval_loss": 2.6422975063323975, | |
| "eval_runtime": 88.2892, | |
| "eval_samples_per_second": 1069.27, | |
| "eval_steps_per_second": 8.359, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 1.928151004668155, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 0.00010718489953318448, | |
| "loss": 3.5011, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 1.928151004668155, | |
| "eval_loss": 2.6424670219421387, | |
| "eval_runtime": 88.2616, | |
| "eval_samples_per_second": 1069.604, | |
| "eval_steps_per_second": 8.362, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 2.029632636492795, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 9.703673635072052e-05, | |
| "loss": 3.5017, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 2.029632636492795, | |
| "eval_loss": 2.6414718627929688, | |
| "eval_runtime": 88.2423, | |
| "eval_samples_per_second": 1069.838, | |
| "eval_steps_per_second": 8.363, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 2.1311142683174347, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 8.688857316825655e-05, | |
| "loss": 3.5014, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 2.1311142683174347, | |
| "eval_loss": 2.6421916484832764, | |
| "eval_runtime": 88.2249, | |
| "eval_samples_per_second": 1070.049, | |
| "eval_steps_per_second": 8.365, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 2.232595900142074, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 7.674040998579256e-05, | |
| "loss": 3.4996, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 2.232595900142074, | |
| "eval_loss": 2.6411330699920654, | |
| "eval_runtime": 88.2534, | |
| "eval_samples_per_second": 1069.703, | |
| "eval_steps_per_second": 8.362, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 2.334077531966714, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 6.659224680332859e-05, | |
| "loss": 3.4998, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 2.334077531966714, | |
| "eval_loss": 2.6422553062438965, | |
| "eval_runtime": 88.2778, | |
| "eval_samples_per_second": 1069.408, | |
| "eval_steps_per_second": 8.36, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 2.435559163791354, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 5.644408362086462e-05, | |
| "loss": 3.5002, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 2.435559163791354, | |
| "eval_loss": 2.6412177085876465, | |
| "eval_runtime": 88.2555, | |
| "eval_samples_per_second": 1069.678, | |
| "eval_steps_per_second": 8.362, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 2.5370407956159937, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 4.629592043840065e-05, | |
| "loss": 3.5011, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 2.5370407956159937, | |
| "eval_loss": 2.641514301300049, | |
| "eval_runtime": 88.2215, | |
| "eval_samples_per_second": 1070.09, | |
| "eval_steps_per_second": 8.365, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 2.638522427440633, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 3.614775725593667e-05, | |
| "loss": 3.5005, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 2.638522427440633, | |
| "eval_loss": 2.6411118507385254, | |
| "eval_runtime": 88.2883, | |
| "eval_samples_per_second": 1069.281, | |
| "eval_steps_per_second": 8.359, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 2.740004059265273, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 2.59995940734727e-05, | |
| "loss": 3.4975, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 2.740004059265273, | |
| "eval_loss": 2.6413235664367676, | |
| "eval_runtime": 88.2689, | |
| "eval_samples_per_second": 1069.516, | |
| "eval_steps_per_second": 8.361, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 2.841485691089913, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.5851430891008727e-05, | |
| "loss": 3.5013, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 2.841485691089913, | |
| "eval_loss": 2.6411330699920654, | |
| "eval_runtime": 88.2658, | |
| "eval_samples_per_second": 1069.554, | |
| "eval_steps_per_second": 8.361, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 2.9429673229145523, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 5.703267708544753e-06, | |
| "loss": 3.5006, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 2.9429673229145523, | |
| "eval_loss": 2.6412177085876465, | |
| "eval_runtime": 88.2303, | |
| "eval_samples_per_second": 1069.984, | |
| "eval_steps_per_second": 8.364, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 88686, | |
| "total_flos": 1.2409318398044897e+18, | |
| "train_loss": 3.5531258104435874, | |
| "train_runtime": 34812.4565, | |
| "train_samples_per_second": 326.075, | |
| "train_steps_per_second": 2.548 | |
| } | |
| ], | |
| "logging_steps": 3000, | |
| "max_steps": 88686, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 3000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.2409318398044897e+18, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |