| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.999927855133107, | |
| "global_step": 55440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 0.0004954906204906205, | |
| "loss": 0.5606, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 0.000490981240981241, | |
| "loss": 0.4359, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 0.00048647186147186144, | |
| "loss": 0.3852, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 0.000481962481962482, | |
| "loss": 0.3778, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 0.00047745310245310245, | |
| "loss": 0.3523, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 0.0004729437229437229, | |
| "loss": 0.3319, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 0.00046843434343434346, | |
| "loss": 0.3153, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 0.00046392496392496394, | |
| "loss": 0.3027, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 0.0004594155844155844, | |
| "loss": 0.3063, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 0.00045490620490620494, | |
| "loss": 0.292, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 0.0004503968253968254, | |
| "loss": 0.2802, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 0.0004458874458874459, | |
| "loss": 0.2705, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 0.0004413780663780664, | |
| "loss": 0.2669, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 0.00043686868686868685, | |
| "loss": 0.2616, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 0.00043235930735930733, | |
| "loss": 0.2486, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 0.00042784992784992786, | |
| "loss": 0.2432, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 0.00042334054834054834, | |
| "loss": 0.2413, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 0.0004188311688311688, | |
| "loss": 0.2479, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 0.00041432178932178935, | |
| "loss": 0.2344, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 0.0004098124098124098, | |
| "loss": 0.2363, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 0.0004053030303030303, | |
| "loss": 0.2328, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 0.00040079365079365083, | |
| "loss": 0.237, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 0.0003962842712842713, | |
| "loss": 0.2284, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 0.0003917748917748918, | |
| "loss": 0.2286, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 0.00038726551226551226, | |
| "loss": 0.222, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 0.00038275613275613274, | |
| "loss": 0.2144, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 0.0003782467532467532, | |
| "loss": 0.2272, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 0.00037373737373737375, | |
| "loss": 0.2107, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 0.0003692279942279942, | |
| "loss": 0.2054, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 0.0003647186147186147, | |
| "loss": 0.1953, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 0.00036020923520923523, | |
| "loss": 0.1955, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 0.0003556998556998557, | |
| "loss": 0.2012, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 0.0003511904761904762, | |
| "loss": 0.1938, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 0.0003466810966810967, | |
| "loss": 0.1932, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 0.0003421717171717172, | |
| "loss": 0.1944, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 0.00033766233766233767, | |
| "loss": 0.1967, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 0.00033315295815295815, | |
| "loss": 0.1955, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 0.0003286435786435786, | |
| "loss": 0.1906, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 0.0003241341991341991, | |
| "loss": 0.1894, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 0.00031962481962481964, | |
| "loss": 0.1908, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 0.0003151154401154401, | |
| "loss": 0.1835, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 0.0003106060606060606, | |
| "loss": 0.1755, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "learning_rate": 0.0003060966810966811, | |
| "loss": 0.1725, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "learning_rate": 0.0003015873015873016, | |
| "loss": 0.172, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "learning_rate": 0.0002970779220779221, | |
| "loss": 0.172, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "learning_rate": 0.0002925685425685426, | |
| "loss": 0.1649, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "learning_rate": 0.0002880591630591631, | |
| "loss": 0.1683, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "learning_rate": 0.00028354978354978356, | |
| "loss": 0.1696, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "learning_rate": 0.00027904040404040404, | |
| "loss": 0.1695, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "learning_rate": 0.0002745310245310245, | |
| "loss": 0.1705, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "learning_rate": 0.000270021645021645, | |
| "loss": 0.1659, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 0.0002655122655122655, | |
| "loss": 0.1694, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "learning_rate": 0.000261002886002886, | |
| "loss": 0.1599, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "learning_rate": 0.0002564935064935065, | |
| "loss": 0.1638, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "learning_rate": 0.000251984126984127, | |
| "loss": 0.1662, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "learning_rate": 0.0002474747474747475, | |
| "loss": 0.1587, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "learning_rate": 0.00024296536796536796, | |
| "loss": 0.1453, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "learning_rate": 0.00023845598845598847, | |
| "loss": 0.1505, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "learning_rate": 0.00023394660894660897, | |
| "loss": 0.1498, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "learning_rate": 0.00022943722943722945, | |
| "loss": 0.145, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 0.00022492784992784992, | |
| "loss": 0.1446, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "learning_rate": 0.00022041847041847043, | |
| "loss": 0.1425, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "learning_rate": 0.0002159090909090909, | |
| "loss": 0.1465, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "learning_rate": 0.0002113997113997114, | |
| "loss": 0.1501, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "learning_rate": 0.00020689033189033191, | |
| "loss": 0.1488, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "learning_rate": 0.0002023809523809524, | |
| "loss": 0.1506, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "learning_rate": 0.00019787157287157287, | |
| "loss": 0.1457, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "learning_rate": 0.00019336219336219337, | |
| "loss": 0.1436, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "learning_rate": 0.00018885281385281385, | |
| "loss": 0.1447, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "learning_rate": 0.00018434343434343435, | |
| "loss": 0.1381, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "learning_rate": 0.00017983405483405486, | |
| "loss": 0.1297, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 5.19, | |
| "learning_rate": 0.00017532467532467534, | |
| "loss": 0.1284, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 5.27, | |
| "learning_rate": 0.0001708152958152958, | |
| "loss": 0.1375, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 5.34, | |
| "learning_rate": 0.00016630591630591632, | |
| "loss": 0.1326, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 5.41, | |
| "learning_rate": 0.0001617965367965368, | |
| "loss": 0.1351, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 5.48, | |
| "learning_rate": 0.00015728715728715727, | |
| "loss": 0.1338, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 5.56, | |
| "learning_rate": 0.0001527777777777778, | |
| "loss": 0.1303, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 5.63, | |
| "learning_rate": 0.00014826839826839828, | |
| "loss": 0.1311, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "learning_rate": 0.00014375901875901876, | |
| "loss": 0.1332, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "learning_rate": 0.00013924963924963926, | |
| "loss": 0.132, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "learning_rate": 0.00013474025974025974, | |
| "loss": 0.1299, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 5.92, | |
| "learning_rate": 0.00013023088023088021, | |
| "loss": 0.128, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 5.99, | |
| "learning_rate": 0.00012572150072150075, | |
| "loss": 0.131, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "learning_rate": 0.00012121212121212122, | |
| "loss": 0.1173, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 6.13, | |
| "learning_rate": 0.0001167027417027417, | |
| "loss": 0.1195, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "learning_rate": 0.00011219336219336219, | |
| "loss": 0.1184, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 6.28, | |
| "learning_rate": 0.0001076839826839827, | |
| "loss": 0.1226, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 6.35, | |
| "learning_rate": 0.00010317460317460317, | |
| "loss": 0.1182, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 6.42, | |
| "learning_rate": 9.866522366522366e-05, | |
| "loss": 0.1249, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 6.49, | |
| "learning_rate": 9.415584415584417e-05, | |
| "loss": 0.1187, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 6.57, | |
| "learning_rate": 8.964646464646464e-05, | |
| "loss": 0.1214, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 6.64, | |
| "learning_rate": 8.513708513708513e-05, | |
| "loss": 0.1194, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 6.71, | |
| "learning_rate": 8.062770562770564e-05, | |
| "loss": 0.1187, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 6.78, | |
| "learning_rate": 7.611832611832612e-05, | |
| "loss": 0.1168, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 6.85, | |
| "learning_rate": 7.16089466089466e-05, | |
| "loss": 0.12, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 6.93, | |
| "learning_rate": 6.709956709956711e-05, | |
| "loss": 0.1198, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 6.259018759018759e-05, | |
| "loss": 0.1189, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 7.07, | |
| "learning_rate": 5.808080808080808e-05, | |
| "loss": 0.1116, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 7.14, | |
| "learning_rate": 5.357142857142857e-05, | |
| "loss": 0.1114, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 7.21, | |
| "learning_rate": 4.9062049062049066e-05, | |
| "loss": 0.1108, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 7.29, | |
| "learning_rate": 4.455266955266955e-05, | |
| "loss": 0.1101, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 7.36, | |
| "learning_rate": 4.004329004329004e-05, | |
| "loss": 0.1062, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 7.43, | |
| "learning_rate": 3.553391053391054e-05, | |
| "loss": 0.1141, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "learning_rate": 3.102453102453102e-05, | |
| "loss": 0.1106, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 7.58, | |
| "learning_rate": 2.6515151515151516e-05, | |
| "loss": 0.1105, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 7.65, | |
| "learning_rate": 2.2005772005772003e-05, | |
| "loss": 0.11, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 7.72, | |
| "learning_rate": 1.7496392496392497e-05, | |
| "loss": 0.1117, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 7.79, | |
| "learning_rate": 1.2987012987012988e-05, | |
| "loss": 0.111, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "learning_rate": 8.477633477633478e-06, | |
| "loss": 0.1144, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 7.94, | |
| "learning_rate": 3.968253968253968e-06, | |
| "loss": 0.1126, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 55440, | |
| "total_flos": 9.377895890180506e+17, | |
| "train_loss": 0.18044453823205198, | |
| "train_runtime": 140707.6037, | |
| "train_samples_per_second": 12.609, | |
| "train_steps_per_second": 0.394 | |
| } | |
| ], | |
| "max_steps": 55440, | |
| "num_train_epochs": 8, | |
| "total_flos": 9.377895890180506e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |