| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9973799126637557, | |
| "eval_steps": 500, | |
| "global_step": 429, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06986899563318777, | |
| "grad_norm": 5.182594166300789, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6932, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.13973799126637554, | |
| "grad_norm": 0.8105274865614129, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6288, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2096069868995633, | |
| "grad_norm": 0.8904752466777895, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6036, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2794759825327511, | |
| "grad_norm": 0.6207860876300546, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5902, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.34934497816593885, | |
| "grad_norm": 0.7152500537832366, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5769, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4192139737991266, | |
| "grad_norm": 0.7013464450514418, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5766, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4890829694323144, | |
| "grad_norm": 0.6585965690127796, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5608, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5589519650655022, | |
| "grad_norm": 0.5056238657103318, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5628, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.62882096069869, | |
| "grad_norm": 0.5015857792164096, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5568, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6986899563318777, | |
| "grad_norm": 0.5250734927951002, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5527, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7685589519650655, | |
| "grad_norm": 0.5330395330357691, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5538, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.8384279475982532, | |
| "grad_norm": 0.5383302692811954, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5421, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.9082969432314411, | |
| "grad_norm": 0.4580590442515656, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5438, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.9781659388646288, | |
| "grad_norm": 0.5111086405959315, | |
| "learning_rate": 5e-06, | |
| "loss": 0.547, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9991266375545852, | |
| "eval_loss": 0.5411319136619568, | |
| "eval_runtime": 97.15, | |
| "eval_samples_per_second": 39.702, | |
| "eval_steps_per_second": 0.628, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.0480349344978166, | |
| "grad_norm": 0.5986185773031188, | |
| "learning_rate": 5e-06, | |
| "loss": 0.561, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.1179039301310043, | |
| "grad_norm": 0.6935147463024353, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4992, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.1877729257641922, | |
| "grad_norm": 0.828701854577101, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5037, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.25764192139738, | |
| "grad_norm": 0.47892479220169987, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4977, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.3275109170305677, | |
| "grad_norm": 0.4854767069935743, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4972, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.3973799126637554, | |
| "grad_norm": 0.4636030652817092, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4939, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.467248908296943, | |
| "grad_norm": 0.46460367066805386, | |
| "learning_rate": 5e-06, | |
| "loss": 0.497, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.537117903930131, | |
| "grad_norm": 0.5541857158168692, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4967, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.6069868995633187, | |
| "grad_norm": 0.43929258201041527, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4972, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.6768558951965065, | |
| "grad_norm": 0.5673809631668943, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5003, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.7467248908296944, | |
| "grad_norm": 0.5312435781280381, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5029, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.8165938864628821, | |
| "grad_norm": 0.42426976715635795, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4881, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.8864628820960698, | |
| "grad_norm": 0.4900651685024478, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4914, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.9563318777292578, | |
| "grad_norm": 0.5082119447902608, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4989, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.9982532751091702, | |
| "eval_loss": 0.5304298996925354, | |
| "eval_runtime": 100.0792, | |
| "eval_samples_per_second": 38.539, | |
| "eval_steps_per_second": 0.61, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.0262008733624453, | |
| "grad_norm": 0.5750485199278518, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5172, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.096069868995633, | |
| "grad_norm": 0.5082128041211101, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4502, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.165938864628821, | |
| "grad_norm": 0.45659427403567093, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4449, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.2358078602620086, | |
| "grad_norm": 0.5382190170872588, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4475, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.3056768558951966, | |
| "grad_norm": 0.5824902306199307, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4511, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.3755458515283845, | |
| "grad_norm": 0.49390187220978227, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4527, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.445414847161572, | |
| "grad_norm": 0.5075624377832567, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4544, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.51528384279476, | |
| "grad_norm": 0.49339275826514356, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4483, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.5851528384279474, | |
| "grad_norm": 0.46662802106266, | |
| "learning_rate": 5e-06, | |
| "loss": 0.454, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.6550218340611353, | |
| "grad_norm": 0.55123614316667, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4531, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.7248908296943233, | |
| "grad_norm": 0.5404203217788516, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4479, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.7947598253275108, | |
| "grad_norm": 0.45881743373968936, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4518, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.8646288209606987, | |
| "grad_norm": 0.44281752763961424, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4532, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.934497816593886, | |
| "grad_norm": 0.48358299433650903, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4588, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.9973799126637557, | |
| "eval_loss": 0.5314457416534424, | |
| "eval_runtime": 97.6794, | |
| "eval_samples_per_second": 39.486, | |
| "eval_steps_per_second": 0.624, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 2.9973799126637557, | |
| "step": 429, | |
| "total_flos": 718381598638080.0, | |
| "train_loss": 0.5104286659569729, | |
| "train_runtime": 14367.279, | |
| "train_samples_per_second": 15.301, | |
| "train_steps_per_second": 0.03 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 429, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 718381598638080.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |