| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.999983800508664, | |
| "eval_steps": 500, | |
| "global_step": 46297, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0025, | |
| "loss": 6.1249, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.005, | |
| "loss": 4.8254, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.004944808706978387, | |
| "loss": 4.2925, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.004889617413956775, | |
| "loss": 4.0454, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.004834426120935161, | |
| "loss": 3.906, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.004779234827913549, | |
| "loss": 3.8162, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.004724043534891935, | |
| "loss": 3.7444, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 0.004668852241870323, | |
| "loss": 3.696, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 0.00461366094884871, | |
| "loss": 3.651, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 0.0045584696558270965, | |
| "loss": 3.6204, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 0.004503278362805484, | |
| "loss": 3.5844, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 0.004448087069783871, | |
| "loss": 3.5632, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 0.004392895776762258, | |
| "loss": 3.5396, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 0.004337704483740645, | |
| "loss": 3.5185, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 0.0042825131907190325, | |
| "loss": 3.4975, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 0.004227321897697419, | |
| "loss": 3.4855, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 0.004172130604675806, | |
| "loss": 3.4744, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 0.004116939311654194, | |
| "loss": 3.4505, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 0.00406174801863258, | |
| "loss": 3.4417, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 0.004006556725610968, | |
| "loss": 3.4298, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 0.003951365432589355, | |
| "loss": 3.4162, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 0.003896174139567742, | |
| "loss": 3.4018, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 0.003840982846546129, | |
| "loss": 3.3938, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 0.0037857915535245163, | |
| "loss": 3.3894, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 0.0037306002605029033, | |
| "loss": 3.3755, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 0.00367540896748129, | |
| "loss": 3.3671, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 0.003620217674459677, | |
| "loss": 3.3617, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 0.0035650263814380645, | |
| "loss": 3.356, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 0.0035098350884164515, | |
| "loss": 3.3497, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 0.0034546437953948384, | |
| "loss": 3.3382, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 0.0033994525023732258, | |
| "loss": 3.3315, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 0.0033442612093516127, | |
| "loss": 3.3179, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 0.0032890699163299997, | |
| "loss": 3.317, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 0.0032338786233083866, | |
| "loss": 3.3162, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 0.003178687330286774, | |
| "loss": 3.3065, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 0.003123496037265161, | |
| "loss": 3.3011, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 0.003068304744243548, | |
| "loss": 3.2966, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 0.0030131134512219357, | |
| "loss": 3.2864, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 0.0029579221582003226, | |
| "loss": 3.2818, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 0.002902730865178709, | |
| "loss": 3.2774, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 0.002847539572157097, | |
| "loss": 3.268, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 0.002792348279135484, | |
| "loss": 3.2678, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 0.002737156986113871, | |
| "loss": 3.2658, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 0.0026819656930922578, | |
| "loss": 3.2574, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 0.002626774400070645, | |
| "loss": 3.251, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 0.002571583107049032, | |
| "loss": 3.2492, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 0.002516391814027419, | |
| "loss": 3.2418, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 0.002461200521005806, | |
| "loss": 3.2417, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 0.0024060092279841933, | |
| "loss": 3.2284, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 0.0023508179349625803, | |
| "loss": 3.2319, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 0.0022956266419409672, | |
| "loss": 3.2256, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 0.0022404353489193546, | |
| "loss": 3.2174, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 0.0021852440558977415, | |
| "loss": 3.2171, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 0.002130052762876129, | |
| "loss": 3.2085, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 0.002074861469854516, | |
| "loss": 3.2026, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 0.002019670176832903, | |
| "loss": 3.1985, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 0.00196447888381129, | |
| "loss": 3.1999, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 0.001909287590789677, | |
| "loss": 3.1934, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 0.0018540962977680643, | |
| "loss": 3.1862, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 0.0017989050047464512, | |
| "loss": 3.181, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 0.0017437137117248384, | |
| "loss": 3.1773, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 0.0016885224187032255, | |
| "loss": 3.1727, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 0.0016333311256816125, | |
| "loss": 3.1697, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 0.0015781398326599996, | |
| "loss": 3.1615, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 0.0015229485396383866, | |
| "loss": 3.1629, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 0.0014677572466167737, | |
| "loss": 3.1587, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 0.001412565953595161, | |
| "loss": 3.1499, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 0.0013573746605735479, | |
| "loss": 3.147, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 0.001302183367551935, | |
| "loss": 3.1447, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 0.0012469920745303222, | |
| "loss": 3.1374, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 0.0011918007815087093, | |
| "loss": 3.1397, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 0.0011366094884870963, | |
| "loss": 3.1281, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 0.0010814181954654834, | |
| "loss": 3.1262, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 0.0010262269024438704, | |
| "loss": 3.1198, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 0.0009710356094222576, | |
| "loss": 3.1126, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 0.0009158443164006447, | |
| "loss": 3.1091, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 0.0008606530233790317, | |
| "loss": 3.1086, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 0.0008054617303574188, | |
| "loss": 3.1048, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 0.0007502704373358058, | |
| "loss": 3.0977, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 0.0006950791443141929, | |
| "loss": 3.0983, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 0.0006398878512925802, | |
| "loss": 3.0903, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 0.0005846965582709672, | |
| "loss": 3.0851, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 0.0005295052652493543, | |
| "loss": 3.0847, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 0.0004743139722277413, | |
| "loss": 3.0722, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 0.0004191226792061285, | |
| "loss": 3.0718, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 0.00036393138618451553, | |
| "loss": 3.0705, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 0.00030874009316290263, | |
| "loss": 3.0653, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 0.00025354880014128974, | |
| "loss": 3.064, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 0.0001983575071196768, | |
| "loss": 3.0514, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 0.0001431662140980639, | |
| "loss": 3.0493, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.797492107645099e-05, | |
| "loss": 3.0482, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 3.278362805483807e-05, | |
| "loss": 3.051, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 46297, | |
| "total_flos": 1.5484207870676828e+18, | |
| "train_loss": 3.3410520185483805, | |
| "train_runtime": 176422.7402, | |
| "train_samples_per_second": 16.795, | |
| "train_steps_per_second": 0.262 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 46297, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 5000, | |
| "total_flos": 1.5484207870676828e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |