{ | |
"best_metric": null, | |
"best_model_checkpoint": null, | |
"epoch": 2.9555555555555557, | |
"global_step": 332500, | |
"is_hyper_param_search": false, | |
"is_local_process_zero": true, | |
"is_world_process_zero": true, | |
"log_history": [ | |
{ | |
"epoch": 0.0, | |
"learning_rate": 0.001, | |
"loss": 1.693, | |
"step": 500 | |
}, | |
{ | |
"epoch": 0.01, | |
"learning_rate": 0.001, | |
"loss": 1.5462, | |
"step": 1000 | |
}, | |
{ | |
"epoch": 0.01, | |
"learning_rate": 0.001, | |
"loss": 1.5029, | |
"step": 1500 | |
}, | |
{ | |
"epoch": 0.02, | |
"learning_rate": 0.001, | |
"loss": 1.4654, | |
"step": 2000 | |
}, | |
{ | |
"epoch": 0.02, | |
"learning_rate": 0.001, | |
"loss": 1.4407, | |
"step": 2500 | |
}, | |
{ | |
"epoch": 0.03, | |
"learning_rate": 0.001, | |
"loss": 1.3838, | |
"step": 3000 | |
}, | |
{ | |
"epoch": 0.03, | |
"learning_rate": 0.001, | |
"loss": 1.37, | |
"step": 3500 | |
}, | |
{ | |
"epoch": 0.04, | |
"learning_rate": 0.001, | |
"loss": 1.3818, | |
"step": 4000 | |
}, | |
{ | |
"epoch": 0.04, | |
"learning_rate": 0.001, | |
"loss": 1.3539, | |
"step": 4500 | |
}, | |
{ | |
"epoch": 0.04, | |
"learning_rate": 0.001, | |
"loss": 1.3605, | |
"step": 5000 | |
}, | |
{ | |
"epoch": 0.05, | |
"learning_rate": 0.001, | |
"loss": 1.3485, | |
"step": 5500 | |
}, | |
{ | |
"epoch": 0.05, | |
"learning_rate": 0.001, | |
"loss": 1.3284, | |
"step": 6000 | |
}, | |
{ | |
"epoch": 0.06, | |
"learning_rate": 0.001, | |
"loss": 1.3229, | |
"step": 6500 | |
}, | |
{ | |
"epoch": 0.06, | |
"learning_rate": 0.001, | |
"loss": 1.3203, | |
"step": 7000 | |
}, | |
{ | |
"epoch": 0.07, | |
"learning_rate": 0.001, | |
"loss": 1.3039, | |
"step": 7500 | |
}, | |
{ | |
"epoch": 0.07, | |
"learning_rate": 0.001, | |
"loss": 1.2939, | |
"step": 8000 | |
}, | |
{ | |
"epoch": 0.08, | |
"learning_rate": 0.001, | |
"loss": 1.2585, | |
"step": 8500 | |
}, | |
{ | |
"epoch": 0.08, | |
"learning_rate": 0.001, | |
"loss": 1.2824, | |
"step": 9000 | |
}, | |
{ | |
"epoch": 0.08, | |
"learning_rate": 0.001, | |
"loss": 1.2786, | |
"step": 9500 | |
}, | |
{ | |
"epoch": 0.09, | |
"learning_rate": 0.001, | |
"loss": 1.2889, | |
"step": 10000 | |
}, | |
{ | |
"epoch": 0.09, | |
"learning_rate": 0.001, | |
"loss": 1.2513, | |
"step": 10500 | |
}, | |
{ | |
"epoch": 0.1, | |
"learning_rate": 0.001, | |
"loss": 1.2813, | |
"step": 11000 | |
}, | |
{ | |
"epoch": 0.1, | |
"learning_rate": 0.001, | |
"loss": 1.2363, | |
"step": 11500 | |
}, | |
{ | |
"epoch": 0.11, | |
"learning_rate": 0.001, | |
"loss": 1.2591, | |
"step": 12000 | |
}, | |
{ | |
"epoch": 0.11, | |
"learning_rate": 0.001, | |
"loss": 1.238, | |
"step": 12500 | |
}, | |
{ | |
"epoch": 0.12, | |
"learning_rate": 0.001, | |
"loss": 1.2293, | |
"step": 13000 | |
}, | |
{ | |
"epoch": 0.12, | |
"learning_rate": 0.001, | |
"loss": 1.2416, | |
"step": 13500 | |
}, | |
{ | |
"epoch": 0.12, | |
"learning_rate": 0.001, | |
"loss": 1.2246, | |
"step": 14000 | |
}, | |
{ | |
"epoch": 0.13, | |
"learning_rate": 0.001, | |
"loss": 1.2235, | |
"step": 14500 | |
}, | |
{ | |
"epoch": 0.13, | |
"learning_rate": 0.001, | |
"loss": 1.2157, | |
"step": 15000 | |
}, | |
{ | |
"epoch": 0.14, | |
"learning_rate": 0.001, | |
"loss": 1.2175, | |
"step": 15500 | |
}, | |
{ | |
"epoch": 0.14, | |
"learning_rate": 0.001, | |
"loss": 1.2135, | |
"step": 16000 | |
}, | |
{ | |
"epoch": 0.15, | |
"learning_rate": 0.001, | |
"loss": 1.2295, | |
"step": 16500 | |
}, | |
{ | |
"epoch": 0.15, | |
"learning_rate": 0.001, | |
"loss": 1.2025, | |
"step": 17000 | |
}, | |
{ | |
"epoch": 0.16, | |
"learning_rate": 0.001, | |
"loss": 1.1972, | |
"step": 17500 | |
}, | |
{ | |
"epoch": 0.16, | |
"learning_rate": 0.001, | |
"loss": 1.2001, | |
"step": 18000 | |
}, | |
{ | |
"epoch": 0.16, | |
"learning_rate": 0.001, | |
"loss": 1.1981, | |
"step": 18500 | |
}, | |
{ | |
"epoch": 0.17, | |
"learning_rate": 0.001, | |
"loss": 1.1821, | |
"step": 19000 | |
}, | |
{ | |
"epoch": 0.17, | |
"learning_rate": 0.001, | |
"loss": 1.1872, | |
"step": 19500 | |
}, | |
{ | |
"epoch": 0.18, | |
"learning_rate": 0.001, | |
"loss": 1.1764, | |
"step": 20000 | |
}, | |
{ | |
"epoch": 0.18, | |
"learning_rate": 0.001, | |
"loss": 1.1847, | |
"step": 20500 | |
}, | |
{ | |
"epoch": 0.19, | |
"learning_rate": 0.001, | |
"loss": 1.1703, | |
"step": 21000 | |
}, | |
{ | |
"epoch": 0.19, | |
"learning_rate": 0.001, | |
"loss": 1.1738, | |
"step": 21500 | |
}, | |
{ | |
"epoch": 0.2, | |
"learning_rate": 0.001, | |
"loss": 1.1653, | |
"step": 22000 | |
}, | |
{ | |
"epoch": 0.2, | |
"learning_rate": 0.001, | |
"loss": 1.1625, | |
"step": 22500 | |
}, | |
{ | |
"epoch": 0.2, | |
"learning_rate": 0.001, | |
"loss": 1.188, | |
"step": 23000 | |
}, | |
{ | |
"epoch": 0.21, | |
"learning_rate": 0.001, | |
"loss": 1.1871, | |
"step": 23500 | |
}, | |
{ | |
"epoch": 0.21, | |
"learning_rate": 0.001, | |
"loss": 1.1614, | |
"step": 24000 | |
}, | |
{ | |
"epoch": 0.22, | |
"learning_rate": 0.001, | |
"loss": 1.1665, | |
"step": 24500 | |
}, | |
{ | |
"epoch": 0.22, | |
"learning_rate": 0.001, | |
"loss": 1.1696, | |
"step": 25000 | |
}, | |
{ | |
"epoch": 0.23, | |
"learning_rate": 0.001, | |
"loss": 1.1465, | |
"step": 25500 | |
}, | |
{ | |
"epoch": 0.23, | |
"learning_rate": 0.001, | |
"loss": 1.1538, | |
"step": 26000 | |
}, | |
{ | |
"epoch": 0.24, | |
"learning_rate": 0.001, | |
"loss": 1.1523, | |
"step": 26500 | |
}, | |
{ | |
"epoch": 0.24, | |
"learning_rate": 0.001, | |
"loss": 1.1376, | |
"step": 27000 | |
}, | |
{ | |
"epoch": 0.24, | |
"learning_rate": 0.001, | |
"loss": 1.1364, | |
"step": 27500 | |
}, | |
{ | |
"epoch": 0.25, | |
"learning_rate": 0.001, | |
"loss": 1.1567, | |
"step": 28000 | |
}, | |
{ | |
"epoch": 0.25, | |
"learning_rate": 0.001, | |
"loss": 1.1393, | |
"step": 28500 | |
}, | |
{ | |
"epoch": 0.26, | |
"learning_rate": 0.001, | |
"loss": 1.1399, | |
"step": 29000 | |
}, | |
{ | |
"epoch": 0.26, | |
"learning_rate": 0.001, | |
"loss": 1.1328, | |
"step": 29500 | |
}, | |
{ | |
"epoch": 0.27, | |
"learning_rate": 0.001, | |
"loss": 1.1456, | |
"step": 30000 | |
}, | |
{ | |
"epoch": 0.27, | |
"learning_rate": 0.001, | |
"loss": 1.1311, | |
"step": 30500 | |
}, | |
{ | |
"epoch": 0.28, | |
"learning_rate": 0.001, | |
"loss": 1.1246, | |
"step": 31000 | |
}, | |
{ | |
"epoch": 0.28, | |
"learning_rate": 0.001, | |
"loss": 1.1358, | |
"step": 31500 | |
}, | |
{ | |
"epoch": 0.28, | |
"learning_rate": 0.001, | |
"loss": 1.1271, | |
"step": 32000 | |
}, | |
{ | |
"epoch": 0.29, | |
"learning_rate": 0.001, | |
"loss": 1.1082, | |
"step": 32500 | |
}, | |
{ | |
"epoch": 0.29, | |
"learning_rate": 0.001, | |
"loss": 1.1064, | |
"step": 33000 | |
}, | |
{ | |
"epoch": 0.3, | |
"learning_rate": 0.001, | |
"loss": 1.122, | |
"step": 33500 | |
}, | |
{ | |
"epoch": 0.3, | |
"learning_rate": 0.001, | |
"loss": 1.1282, | |
"step": 34000 | |
}, | |
{ | |
"epoch": 0.31, | |
"learning_rate": 0.001, | |
"loss": 1.1088, | |
"step": 34500 | |
}, | |
{ | |
"epoch": 0.31, | |
"learning_rate": 0.001, | |
"loss": 1.1167, | |
"step": 35000 | |
}, | |
{ | |
"epoch": 0.32, | |
"learning_rate": 0.001, | |
"loss": 1.0914, | |
"step": 35500 | |
}, | |
{ | |
"epoch": 0.32, | |
"learning_rate": 0.001, | |
"loss": 1.1197, | |
"step": 36000 | |
}, | |
{ | |
"epoch": 0.32, | |
"learning_rate": 0.001, | |
"loss": 1.1085, | |
"step": 36500 | |
}, | |
{ | |
"epoch": 0.33, | |
"learning_rate": 0.001, | |
"loss": 1.0976, | |
"step": 37000 | |
}, | |
{ | |
"epoch": 0.33, | |
"learning_rate": 0.001, | |
"loss": 1.1057, | |
"step": 37500 | |
}, | |
{ | |
"epoch": 0.34, | |
"learning_rate": 0.001, | |
"loss": 1.0936, | |
"step": 38000 | |
}, | |
{ | |
"epoch": 0.34, | |
"learning_rate": 0.001, | |
"loss": 1.0813, | |
"step": 38500 | |
}, | |
{ | |
"epoch": 0.35, | |
"learning_rate": 0.001, | |
"loss": 1.0897, | |
"step": 39000 | |
}, | |
{ | |
"epoch": 0.35, | |
"learning_rate": 0.001, | |
"loss": 1.1009, | |
"step": 39500 | |
}, | |
{ | |
"epoch": 0.36, | |
"learning_rate": 0.001, | |
"loss": 1.0939, | |
"step": 40000 | |
}, | |
{ | |
"epoch": 0.36, | |
"learning_rate": 0.001, | |
"loss": 1.0905, | |
"step": 40500 | |
}, | |
{ | |
"epoch": 0.36, | |
"learning_rate": 0.001, | |
"loss": 1.0965, | |
"step": 41000 | |
}, | |
{ | |
"epoch": 0.37, | |
"learning_rate": 0.001, | |
"loss": 1.0823, | |
"step": 41500 | |
}, | |
{ | |
"epoch": 0.37, | |
"learning_rate": 0.001, | |
"loss": 1.0951, | |
"step": 42000 | |
}, | |
{ | |
"epoch": 0.38, | |
"learning_rate": 0.001, | |
"loss": 1.0823, | |
"step": 42500 | |
}, | |
{ | |
"epoch": 0.38, | |
"learning_rate": 0.001, | |
"loss": 1.0741, | |
"step": 43000 | |
}, | |
{ | |
"epoch": 0.39, | |
"learning_rate": 0.001, | |
"loss": 1.0622, | |
"step": 43500 | |
}, | |
{ | |
"epoch": 0.39, | |
"learning_rate": 0.001, | |
"loss": 1.0879, | |
"step": 44000 | |
}, | |
{ | |
"epoch": 0.4, | |
"learning_rate": 0.001, | |
"loss": 1.0559, | |
"step": 44500 | |
}, | |
{ | |
"epoch": 0.4, | |
"learning_rate": 0.001, | |
"loss": 1.0766, | |
"step": 45000 | |
}, | |
{ | |
"epoch": 0.4, | |
"learning_rate": 0.001, | |
"loss": 1.0609, | |
"step": 45500 | |
}, | |
{ | |
"epoch": 0.41, | |
"learning_rate": 0.001, | |
"loss": 1.063, | |
"step": 46000 | |
}, | |
{ | |
"epoch": 0.41, | |
"learning_rate": 0.001, | |
"loss": 1.0618, | |
"step": 46500 | |
}, | |
{ | |
"epoch": 0.42, | |
"learning_rate": 0.001, | |
"loss": 1.0618, | |
"step": 47000 | |
}, | |
{ | |
"epoch": 0.42, | |
"learning_rate": 0.001, | |
"loss": 1.0488, | |
"step": 47500 | |
}, | |
{ | |
"epoch": 0.43, | |
"learning_rate": 0.001, | |
"loss": 1.0686, | |
"step": 48000 | |
}, | |
{ | |
"epoch": 0.43, | |
"learning_rate": 0.001, | |
"loss": 1.0457, | |
"step": 48500 | |
}, | |
{ | |
"epoch": 0.44, | |
"learning_rate": 0.001, | |
"loss": 1.0494, | |
"step": 49000 | |
}, | |
{ | |
"epoch": 0.44, | |
"learning_rate": 0.001, | |
"loss": 1.0498, | |
"step": 49500 | |
}, | |
{ | |
"epoch": 0.44, | |
"learning_rate": 0.001, | |
"loss": 1.0677, | |
"step": 50000 | |
}, | |
{ | |
"epoch": 0.45, | |
"learning_rate": 0.001, | |
"loss": 1.0548, | |
"step": 50500 | |
}, | |
{ | |
"epoch": 0.45, | |
"learning_rate": 0.001, | |
"loss": 1.0721, | |
"step": 51000 | |
}, | |
{ | |
"epoch": 0.46, | |
"learning_rate": 0.001, | |
"loss": 1.0442, | |
"step": 51500 | |
}, | |
{ | |
"epoch": 0.46, | |
"learning_rate": 0.001, | |
"loss": 1.0502, | |
"step": 52000 | |
}, | |
{ | |
"epoch": 0.47, | |
"learning_rate": 0.001, | |
"loss": 1.0457, | |
"step": 52500 | |
}, | |
{ | |
"epoch": 0.47, | |
"learning_rate": 0.001, | |
"loss": 1.0459, | |
"step": 53000 | |
}, | |
{ | |
"epoch": 0.48, | |
"learning_rate": 0.001, | |
"loss": 1.0419, | |
"step": 53500 | |
}, | |
{ | |
"epoch": 0.48, | |
"learning_rate": 0.001, | |
"loss": 1.0533, | |
"step": 54000 | |
}, | |
{ | |
"epoch": 0.48, | |
"learning_rate": 0.001, | |
"loss": 1.0498, | |
"step": 54500 | |
}, | |
{ | |
"epoch": 0.49, | |
"learning_rate": 0.001, | |
"loss": 1.0347, | |
"step": 55000 | |
}, | |
{ | |
"epoch": 0.49, | |
"learning_rate": 0.001, | |
"loss": 1.0422, | |
"step": 55500 | |
}, | |
{ | |
"epoch": 0.5, | |
"learning_rate": 0.001, | |
"loss": 1.0431, | |
"step": 56000 | |
}, | |
{ | |
"epoch": 0.5, | |
"learning_rate": 0.001, | |
"loss": 1.0365, | |
"step": 56500 | |
}, | |
{ | |
"epoch": 0.51, | |
"learning_rate": 0.001, | |
"loss": 1.0383, | |
"step": 57000 | |
}, | |
{ | |
"epoch": 0.51, | |
"learning_rate": 0.001, | |
"loss": 1.0291, | |
"step": 57500 | |
}, | |
{ | |
"epoch": 0.52, | |
"learning_rate": 0.001, | |
"loss": 1.037, | |
"step": 58000 | |
}, | |
{ | |
"epoch": 0.52, | |
"learning_rate": 0.001, | |
"loss": 1.0293, | |
"step": 58500 | |
}, | |
{ | |
"epoch": 0.52, | |
"learning_rate": 0.001, | |
"loss": 1.0243, | |
"step": 59000 | |
}, | |
{ | |
"epoch": 0.53, | |
"learning_rate": 0.001, | |
"loss": 1.0319, | |
"step": 59500 | |
}, | |
{ | |
"epoch": 0.53, | |
"learning_rate": 0.001, | |
"loss": 1.0237, | |
"step": 60000 | |
}, | |
{ | |
"epoch": 0.54, | |
"learning_rate": 0.001, | |
"loss": 1.0394, | |
"step": 60500 | |
}, | |
{ | |
"epoch": 0.54, | |
"learning_rate": 0.001, | |
"loss": 1.0306, | |
"step": 61000 | |
}, | |
{ | |
"epoch": 0.55, | |
"learning_rate": 0.001, | |
"loss": 1.0184, | |
"step": 61500 | |
}, | |
{ | |
"epoch": 0.55, | |
"learning_rate": 0.001, | |
"loss": 1.0149, | |
"step": 62000 | |
}, | |
{ | |
"epoch": 0.56, | |
"learning_rate": 0.001, | |
"loss": 1.024, | |
"step": 62500 | |
}, | |
{ | |
"epoch": 0.56, | |
"learning_rate": 0.001, | |
"loss": 1.0095, | |
"step": 63000 | |
}, | |
{ | |
"epoch": 0.56, | |
"learning_rate": 0.001, | |
"loss": 1.008, | |
"step": 63500 | |
}, | |
{ | |
"epoch": 0.57, | |
"learning_rate": 0.001, | |
"loss": 1.0057, | |
"step": 64000 | |
}, | |
{ | |
"epoch": 0.57, | |
"learning_rate": 0.001, | |
"loss": 1.0232, | |
"step": 64500 | |
}, | |
{ | |
"epoch": 0.58, | |
"learning_rate": 0.001, | |
"loss": 1.02, | |
"step": 65000 | |
}, | |
{ | |
"epoch": 0.58, | |
"learning_rate": 0.001, | |
"loss": 1.0141, | |
"step": 65500 | |
}, | |
{ | |
"epoch": 0.59, | |
"learning_rate": 0.001, | |
"loss": 1.0132, | |
"step": 66000 | |
}, | |
{ | |
"epoch": 0.59, | |
"learning_rate": 0.001, | |
"loss": 1.0229, | |
"step": 66500 | |
}, | |
{ | |
"epoch": 0.6, | |
"learning_rate": 0.001, | |
"loss": 1.0211, | |
"step": 67000 | |
}, | |
{ | |
"epoch": 0.6, | |
"learning_rate": 0.001, | |
"loss": 1.0062, | |
"step": 67500 | |
}, | |
{ | |
"epoch": 0.6, | |
"learning_rate": 0.001, | |
"loss": 1.0062, | |
"step": 68000 | |
}, | |
{ | |
"epoch": 0.61, | |
"learning_rate": 0.001, | |
"loss": 1.0044, | |
"step": 68500 | |
}, | |
{ | |
"epoch": 0.61, | |
"learning_rate": 0.001, | |
"loss": 1.0211, | |
"step": 69000 | |
}, | |
{ | |
"epoch": 0.62, | |
"learning_rate": 0.001, | |
"loss": 0.9982, | |
"step": 69500 | |
}, | |
{ | |
"epoch": 0.62, | |
"learning_rate": 0.001, | |
"loss": 1.011, | |
"step": 70000 | |
}, | |
{ | |
"epoch": 0.63, | |
"learning_rate": 0.001, | |
"loss": 0.9963, | |
"step": 70500 | |
}, | |
{ | |
"epoch": 0.63, | |
"learning_rate": 0.001, | |
"loss": 1.0009, | |
"step": 71000 | |
}, | |
{ | |
"epoch": 0.64, | |
"learning_rate": 0.001, | |
"loss": 0.9925, | |
"step": 71500 | |
}, | |
{ | |
"epoch": 0.64, | |
"learning_rate": 0.001, | |
"loss": 0.9819, | |
"step": 72000 | |
}, | |
{ | |
"epoch": 0.64, | |
"learning_rate": 0.001, | |
"loss": 1.012, | |
"step": 72500 | |
}, | |
{ | |
"epoch": 0.65, | |
"learning_rate": 0.001, | |
"loss": 0.9856, | |
"step": 73000 | |
}, | |
{ | |
"epoch": 0.65, | |
"learning_rate": 0.001, | |
"loss": 0.9991, | |
"step": 73500 | |
}, | |
{ | |
"epoch": 0.66, | |
"learning_rate": 0.001, | |
"loss": 0.9929, | |
"step": 74000 | |
}, | |
{ | |
"epoch": 0.66, | |
"learning_rate": 0.001, | |
"loss": 0.9949, | |
"step": 74500 | |
}, | |
{ | |
"epoch": 0.67, | |
"learning_rate": 0.001, | |
"loss": 0.9874, | |
"step": 75000 | |
}, | |
{ | |
"epoch": 0.67, | |
"learning_rate": 0.001, | |
"loss": 0.9843, | |
"step": 75500 | |
}, | |
{ | |
"epoch": 0.68, | |
"learning_rate": 0.001, | |
"loss": 1.0074, | |
"step": 76000 | |
}, | |
{ | |
"epoch": 0.68, | |
"learning_rate": 0.001, | |
"loss": 0.9937, | |
"step": 76500 | |
}, | |
{ | |
"epoch": 0.68, | |
"learning_rate": 0.001, | |
"loss": 0.9898, | |
"step": 77000 | |
}, | |
{ | |
"epoch": 0.69, | |
"learning_rate": 0.001, | |
"loss": 0.9799, | |
"step": 77500 | |
}, | |
{ | |
"epoch": 0.69, | |
"learning_rate": 0.001, | |
"loss": 0.9855, | |
"step": 78000 | |
}, | |
{ | |
"epoch": 0.7, | |
"learning_rate": 0.001, | |
"loss": 0.9797, | |
"step": 78500 | |
}, | |
{ | |
"epoch": 0.7, | |
"learning_rate": 0.001, | |
"loss": 0.9774, | |
"step": 79000 | |
}, | |
{ | |
"epoch": 0.71, | |
"learning_rate": 0.001, | |
"loss": 0.981, | |
"step": 79500 | |
}, | |
{ | |
"epoch": 0.71, | |
"learning_rate": 0.001, | |
"loss": 0.9909, | |
"step": 80000 | |
}, | |
{ | |
"epoch": 0.72, | |
"learning_rate": 0.001, | |
"loss": 0.9734, | |
"step": 80500 | |
}, | |
{ | |
"epoch": 0.72, | |
"learning_rate": 0.001, | |
"loss": 0.9857, | |
"step": 81000 | |
}, | |
{ | |
"epoch": 0.72, | |
"learning_rate": 0.001, | |
"loss": 0.976, | |
"step": 81500 | |
}, | |
{ | |
"epoch": 0.73, | |
"learning_rate": 0.001, | |
"loss": 0.9842, | |
"step": 82000 | |
}, | |
{ | |
"epoch": 0.73, | |
"learning_rate": 0.001, | |
"loss": 0.9798, | |
"step": 82500 | |
}, | |
{ | |
"epoch": 0.74, | |
"learning_rate": 0.001, | |
"loss": 0.9816, | |
"step": 83000 | |
}, | |
{ | |
"epoch": 0.74, | |
"learning_rate": 0.001, | |
"loss": 0.9686, | |
"step": 83500 | |
}, | |
{ | |
"epoch": 0.75, | |
"learning_rate": 0.001, | |
"loss": 0.9741, | |
"step": 84000 | |
}, | |
{ | |
"epoch": 0.75, | |
"learning_rate": 0.001, | |
"loss": 1.0041, | |
"step": 84500 | |
}, | |
{ | |
"epoch": 0.76, | |
"learning_rate": 0.001, | |
"loss": 0.9844, | |
"step": 85000 | |
}, | |
{ | |
"epoch": 0.76, | |
"learning_rate": 0.001, | |
"loss": 0.9761, | |
"step": 85500 | |
}, | |
{ | |
"epoch": 0.76, | |
"learning_rate": 0.001, | |
"loss": 0.9775, | |
"step": 86000 | |
}, | |
{ | |
"epoch": 0.77, | |
"learning_rate": 0.001, | |
"loss": 0.9667, | |
"step": 86500 | |
}, | |
{ | |
"epoch": 0.77, | |
"learning_rate": 0.001, | |
"loss": 0.9434, | |
"step": 87000 | |
}, | |
{ | |
"epoch": 0.78, | |
"learning_rate": 0.001, | |
"loss": 0.976, | |
"step": 87500 | |
}, | |
{ | |
"epoch": 0.78, | |
"learning_rate": 0.001, | |
"loss": 0.962, | |
"step": 88000 | |
}, | |
{ | |
"epoch": 0.79, | |
"learning_rate": 0.001, | |
"loss": 0.9782, | |
"step": 88500 | |
}, | |
{ | |
"epoch": 0.79, | |
"learning_rate": 0.001, | |
"loss": 0.9739, | |
"step": 89000 | |
}, | |
{ | |
"epoch": 0.8, | |
"learning_rate": 0.001, | |
"loss": 0.9858, | |
"step": 89500 | |
}, | |
{ | |
"epoch": 0.8, | |
"learning_rate": 0.001, | |
"loss": 0.9573, | |
"step": 90000 | |
}, | |
{ | |
"epoch": 0.8, | |
"learning_rate": 0.001, | |
"loss": 0.9776, | |
"step": 90500 | |
}, | |
{ | |
"epoch": 0.81, | |
"learning_rate": 0.001, | |
"loss": 0.9729, | |
"step": 91000 | |
}, | |
{ | |
"epoch": 0.81, | |
"learning_rate": 0.001, | |
"loss": 0.9878, | |
"step": 91500 | |
}, | |
{ | |
"epoch": 0.82, | |
"learning_rate": 0.001, | |
"loss": 0.9735, | |
"step": 92000 | |
}, | |
{ | |
"epoch": 0.82, | |
"learning_rate": 0.001, | |
"loss": 0.9581, | |
"step": 92500 | |
}, | |
{ | |
"epoch": 0.83, | |
"learning_rate": 0.001, | |
"loss": 0.9593, | |
"step": 93000 | |
}, | |
{ | |
"epoch": 0.83, | |
"learning_rate": 0.001, | |
"loss": 0.9611, | |
"step": 93500 | |
}, | |
{ | |
"epoch": 0.84, | |
"learning_rate": 0.001, | |
"loss": 0.9602, | |
"step": 94000 | |
}, | |
{ | |
"epoch": 0.84, | |
"learning_rate": 0.001, | |
"loss": 0.9516, | |
"step": 94500 | |
}, | |
{ | |
"epoch": 0.84, | |
"learning_rate": 0.001, | |
"loss": 0.934, | |
"step": 95000 | |
}, | |
{ | |
"epoch": 0.85, | |
"learning_rate": 0.001, | |
"loss": 0.9451, | |
"step": 95500 | |
}, | |
{ | |
"epoch": 0.85, | |
"learning_rate": 0.001, | |
"loss": 0.956, | |
"step": 96000 | |
}, | |
{ | |
"epoch": 0.86, | |
"learning_rate": 0.001, | |
"loss": 0.9612, | |
"step": 96500 | |
}, | |
{ | |
"epoch": 0.86, | |
"learning_rate": 0.001, | |
"loss": 0.946, | |
"step": 97000 | |
}, | |
{ | |
"epoch": 0.87, | |
"learning_rate": 0.001, | |
"loss": 0.9406, | |
"step": 97500 | |
}, | |
{ | |
"epoch": 0.87, | |
"learning_rate": 0.001, | |
"loss": 0.9641, | |
"step": 98000 | |
}, | |
{ | |
"epoch": 0.88, | |
"learning_rate": 0.001, | |
"loss": 0.9696, | |
"step": 98500 | |
}, | |
{ | |
"epoch": 0.88, | |
"learning_rate": 0.001, | |
"loss": 0.9599, | |
"step": 99000 | |
}, | |
{ | |
"epoch": 0.88, | |
"learning_rate": 0.001, | |
"loss": 0.949, | |
"step": 99500 | |
}, | |
{ | |
"epoch": 0.89, | |
"learning_rate": 0.001, | |
"loss": 0.9491, | |
"step": 100000 | |
}, | |
{ | |
"epoch": 0.89, | |
"learning_rate": 0.001, | |
"loss": 0.9468, | |
"step": 100500 | |
}, | |
{ | |
"epoch": 0.9, | |
"learning_rate": 0.001, | |
"loss": 0.9301, | |
"step": 101000 | |
}, | |
{ | |
"epoch": 0.9, | |
"learning_rate": 0.001, | |
"loss": 0.9481, | |
"step": 101500 | |
}, | |
{ | |
"epoch": 0.91, | |
"learning_rate": 0.001, | |
"loss": 0.946, | |
"step": 102000 | |
}, | |
{ | |
"epoch": 0.91, | |
"learning_rate": 0.001, | |
"loss": 0.94, | |
"step": 102500 | |
}, | |
{ | |
"epoch": 0.92, | |
"learning_rate": 0.001, | |
"loss": 0.9288, | |
"step": 103000 | |
}, | |
{ | |
"epoch": 0.92, | |
"learning_rate": 0.001, | |
"loss": 0.9514, | |
"step": 103500 | |
}, | |
{ | |
"epoch": 0.92, | |
"learning_rate": 0.001, | |
"loss": 0.9371, | |
"step": 104000 | |
}, | |
{ | |
"epoch": 0.93, | |
"learning_rate": 0.001, | |
"loss": 0.9269, | |
"step": 104500 | |
}, | |
{ | |
"epoch": 0.93, | |
"learning_rate": 0.001, | |
"loss": 0.9405, | |
"step": 105000 | |
}, | |
{ | |
"epoch": 0.94, | |
"learning_rate": 0.001, | |
"loss": 0.9306, | |
"step": 105500 | |
}, | |
{ | |
"epoch": 0.94, | |
"learning_rate": 0.001, | |
"loss": 0.9397, | |
"step": 106000 | |
}, | |
{ | |
"epoch": 0.95, | |
"learning_rate": 0.001, | |
"loss": 0.9247, | |
"step": 106500 | |
}, | |
{ | |
"epoch": 0.95, | |
"learning_rate": 0.001, | |
"loss": 0.9122, | |
"step": 107000 | |
}, | |
{ | |
"epoch": 0.96, | |
"learning_rate": 0.001, | |
"loss": 0.9361, | |
"step": 107500 | |
}, | |
{ | |
"epoch": 0.96, | |
"learning_rate": 0.001, | |
"loss": 0.9255, | |
"step": 108000 | |
}, | |
{ | |
"epoch": 0.96, | |
"learning_rate": 0.001, | |
"loss": 0.9434, | |
"step": 108500 | |
}, | |
{ | |
"epoch": 0.97, | |
"learning_rate": 0.001, | |
"loss": 0.9383, | |
"step": 109000 | |
}, | |
{ | |
"epoch": 0.97, | |
"learning_rate": 0.001, | |
"loss": 0.9371, | |
"step": 109500 | |
}, | |
{ | |
"epoch": 0.98, | |
"learning_rate": 0.001, | |
"loss": 0.9272, | |
"step": 110000 | |
}, | |
{ | |
"epoch": 0.98, | |
"learning_rate": 0.001, | |
"loss": 0.9288, | |
"step": 110500 | |
}, | |
{ | |
"epoch": 0.99, | |
"learning_rate": 0.001, | |
"loss": 0.9248, | |
"step": 111000 | |
}, | |
{ | |
"epoch": 0.99, | |
"learning_rate": 0.001, | |
"loss": 0.924, | |
"step": 111500 | |
}, | |
{ | |
"epoch": 1.0, | |
"learning_rate": 0.001, | |
"loss": 0.9319, | |
"step": 112000 | |
}, | |
{ | |
"epoch": 1.0, | |
"learning_rate": 0.001, | |
"loss": 0.9324, | |
"step": 112500 | |
}, | |
{ | |
"epoch": 1.0, | |
"eval_bleu": 25.3594, | |
"eval_gen_len": 11.3081, | |
"eval_loss": 1.1617918014526367, | |
"eval_runtime": 419.1845, | |
"eval_samples_per_second": 33.398, | |
"step": 112500 | |
}, | |
{ | |
"epoch": 1.0, | |
"learning_rate": 0.001, | |
"loss": 0.8688, | |
"step": 113000 | |
}, | |
{ | |
"epoch": 1.01, | |
"learning_rate": 0.001, | |
"loss": 0.8644, | |
"step": 113500 | |
}, | |
{ | |
"epoch": 1.01, | |
"learning_rate": 0.001, | |
"loss": 0.8799, | |
"step": 114000 | |
}, | |
{ | |
"epoch": 1.02, | |
"learning_rate": 0.001, | |
"loss": 0.8541, | |
"step": 114500 | |
}, | |
{ | |
"epoch": 1.02, | |
"learning_rate": 0.001, | |
"loss": 0.8661, | |
"step": 115000 | |
}, | |
{ | |
"epoch": 1.03, | |
"learning_rate": 0.001, | |
"loss": 0.8715, | |
"step": 115500 | |
}, | |
{ | |
"epoch": 1.03, | |
"learning_rate": 0.001, | |
"loss": 0.8783, | |
"step": 116000 | |
}, | |
{ | |
"epoch": 1.04, | |
"learning_rate": 0.001, | |
"loss": 0.8769, | |
"step": 116500 | |
}, | |
{ | |
"epoch": 1.04, | |
"learning_rate": 0.001, | |
"loss": 0.8697, | |
"step": 117000 | |
}, | |
{ | |
"epoch": 1.04, | |
"learning_rate": 0.001, | |
"loss": 0.8621, | |
"step": 117500 | |
}, | |
{ | |
"epoch": 1.05, | |
"learning_rate": 0.001, | |
"loss": 0.86, | |
"step": 118000 | |
}, | |
{ | |
"epoch": 1.05, | |
"learning_rate": 0.001, | |
"loss": 0.8785, | |
"step": 118500 | |
}, | |
{ | |
"epoch": 1.06, | |
"learning_rate": 0.001, | |
"loss": 0.8818, | |
"step": 119000 | |
}, | |
{ | |
"epoch": 1.06, | |
"learning_rate": 0.001, | |
"loss": 0.8571, | |
"step": 119500 | |
}, | |
{ | |
"epoch": 1.07, | |
"learning_rate": 0.001, | |
"loss": 0.8614, | |
"step": 120000 | |
}, | |
{ | |
"epoch": 1.07, | |
"learning_rate": 0.001, | |
"loss": 0.8627, | |
"step": 120500 | |
}, | |
{ | |
"epoch": 1.08, | |
"learning_rate": 0.001, | |
"loss": 0.8623, | |
"step": 121000 | |
}, | |
{ | |
"epoch": 1.08, | |
"learning_rate": 0.001, | |
"loss": 0.8851, | |
"step": 121500 | |
}, | |
{ | |
"epoch": 1.08, | |
"learning_rate": 0.001, | |
"loss": 0.8638, | |
"step": 122000 | |
}, | |
{ | |
"epoch": 1.09, | |
"learning_rate": 0.001, | |
"loss": 0.8637, | |
"step": 122500 | |
}, | |
{ | |
"epoch": 1.09, | |
"learning_rate": 0.001, | |
"loss": 0.8661, | |
"step": 123000 | |
}, | |
{ | |
"epoch": 1.1, | |
"learning_rate": 0.001, | |
"loss": 0.852, | |
"step": 123500 | |
}, | |
{ | |
"epoch": 1.1, | |
"learning_rate": 0.001, | |
"loss": 0.8738, | |
"step": 124000 | |
}, | |
{ | |
"epoch": 1.11, | |
"learning_rate": 0.001, | |
"loss": 0.871, | |
"step": 124500 | |
}, | |
{ | |
"epoch": 1.11, | |
"learning_rate": 0.001, | |
"loss": 0.8642, | |
"step": 125000 | |
}, | |
{ | |
"epoch": 1.12, | |
"learning_rate": 0.001, | |
"loss": 0.8721, | |
"step": 125500 | |
}, | |
{ | |
"epoch": 1.12, | |
"learning_rate": 0.001, | |
"loss": 0.8672, | |
"step": 126000 | |
}, | |
{ | |
"epoch": 1.12, | |
"learning_rate": 0.001, | |
"loss": 0.8628, | |
"step": 126500 | |
}, | |
{ | |
"epoch": 1.13, | |
"learning_rate": 0.001, | |
"loss": 0.8731, | |
"step": 127000 | |
}, | |
{ | |
"epoch": 1.13, | |
"learning_rate": 0.001, | |
"loss": 0.8753, | |
"step": 127500 | |
}, | |
{ | |
"epoch": 1.14, | |
"learning_rate": 0.001, | |
"loss": 0.8566, | |
"step": 128000 | |
}, | |
{ | |
"epoch": 1.14, | |
"learning_rate": 0.001, | |
"loss": 0.8601, | |
"step": 128500 | |
}, | |
{ | |
"epoch": 1.15, | |
"learning_rate": 0.001, | |
"loss": 0.8622, | |
"step": 129000 | |
}, | |
{ | |
"epoch": 1.15, | |
"learning_rate": 0.001, | |
"loss": 0.8706, | |
"step": 129500 | |
}, | |
{ | |
"epoch": 1.16, | |
"learning_rate": 0.001, | |
"loss": 0.8695, | |
"step": 130000 | |
}, | |
{ | |
"epoch": 1.16, | |
"learning_rate": 0.001, | |
"loss": 0.8633, | |
"step": 130500 | |
}, | |
{ | |
"epoch": 1.16, | |
"learning_rate": 0.001, | |
"loss": 0.8651, | |
"step": 131000 | |
}, | |
{ | |
"epoch": 1.17, | |
"learning_rate": 0.001, | |
"loss": 0.8501, | |
"step": 131500 | |
}, | |
{ | |
"epoch": 1.17, | |
"learning_rate": 0.001, | |
"loss": 0.86, | |
"step": 132000 | |
}, | |
{ | |
"epoch": 1.18, | |
"learning_rate": 0.001, | |
"loss": 0.8713, | |
"step": 132500 | |
}, | |
{ | |
"epoch": 1.18, | |
"learning_rate": 0.001, | |
"loss": 0.8549, | |
"step": 133000 | |
}, | |
{ | |
"epoch": 1.19, | |
"learning_rate": 0.001, | |
"loss": 0.8755, | |
"step": 133500 | |
}, | |
{ | |
"epoch": 1.19, | |
"learning_rate": 0.001, | |
"loss": 0.8511, | |
"step": 134000 | |
}, | |
{ | |
"epoch": 1.2, | |
"learning_rate": 0.001, | |
"loss": 0.8587, | |
"step": 134500 | |
}, | |
{ | |
"epoch": 1.2, | |
"learning_rate": 0.001, | |
"loss": 0.8637, | |
"step": 135000 | |
}, | |
{ | |
"epoch": 1.2, | |
"learning_rate": 0.001, | |
"loss": 0.8625, | |
"step": 135500 | |
}, | |
{ | |
"epoch": 1.21, | |
"learning_rate": 0.001, | |
"loss": 0.8718, | |
"step": 136000 | |
}, | |
{ | |
"epoch": 1.21, | |
"learning_rate": 0.001, | |
"loss": 0.8617, | |
"step": 136500 | |
}, | |
{ | |
"epoch": 1.22, | |
"learning_rate": 0.001, | |
"loss": 0.8564, | |
"step": 137000 | |
}, | |
{ | |
"epoch": 1.22, | |
"learning_rate": 0.001, | |
"loss": 0.859, | |
"step": 137500 | |
}, | |
{ | |
"epoch": 1.23, | |
"learning_rate": 0.001, | |
"loss": 0.8669, | |
"step": 138000 | |
}, | |
{ | |
"epoch": 1.23, | |
"learning_rate": 0.001, | |
"loss": 0.8575, | |
"step": 138500 | |
}, | |
{ | |
"epoch": 1.24, | |
"learning_rate": 0.001, | |
"loss": 0.856, | |
"step": 139000 | |
}, | |
{ | |
"epoch": 1.24, | |
"learning_rate": 0.001, | |
"loss": 0.847, | |
"step": 139500 | |
}, | |
{ | |
"epoch": 1.24, | |
"learning_rate": 0.001, | |
"loss": 0.8527, | |
"step": 140000 | |
}, | |
{ | |
"epoch": 1.25, | |
"learning_rate": 0.001, | |
"loss": 0.8722, | |
"step": 140500 | |
}, | |
{ | |
"epoch": 1.25, | |
"learning_rate": 0.001, | |
"loss": 0.8562, | |
"step": 141000 | |
}, | |
{ | |
"epoch": 1.26, | |
"learning_rate": 0.001, | |
"loss": 0.8549, | |
"step": 141500 | |
}, | |
{ | |
"epoch": 1.26, | |
"learning_rate": 0.001, | |
"loss": 0.8518, | |
"step": 142000 | |
}, | |
{ | |
"epoch": 1.27, | |
"learning_rate": 0.001, | |
"loss": 0.8492, | |
"step": 142500 | |
}, | |
{ | |
"epoch": 1.27, | |
"learning_rate": 0.001, | |
"loss": 0.8715, | |
"step": 143000 | |
}, | |
{ | |
"epoch": 1.28, | |
"learning_rate": 0.001, | |
"loss": 0.8469, | |
"step": 143500 | |
}, | |
{ | |
"epoch": 1.28, | |
"learning_rate": 0.001, | |
"loss": 0.8494, | |
"step": 144000 | |
}, | |
{ | |
"epoch": 1.28, | |
"learning_rate": 0.001, | |
"loss": 0.8518, | |
"step": 144500 | |
}, | |
{ | |
"epoch": 1.29, | |
"learning_rate": 0.001, | |
"loss": 0.8491, | |
"step": 145000 | |
}, | |
{ | |
"epoch": 1.29, | |
"learning_rate": 0.001, | |
"loss": 0.861, | |
"step": 145500 | |
}, | |
{ | |
"epoch": 1.3, | |
"learning_rate": 0.001, | |
"loss": 0.8455, | |
"step": 146000 | |
}, | |
{ | |
"epoch": 1.3, | |
"learning_rate": 0.001, | |
"loss": 0.8601, | |
"step": 146500 | |
}, | |
{ | |
"epoch": 1.31, | |
"learning_rate": 0.001, | |
"loss": 0.8354, | |
"step": 147000 | |
}, | |
{ | |
"epoch": 1.31, | |
"learning_rate": 0.001, | |
"loss": 0.8499, | |
"step": 147500 | |
}, | |
{ | |
"epoch": 1.32, | |
"learning_rate": 0.001, | |
"loss": 0.8503, | |
"step": 148000 | |
}, | |
{ | |
"epoch": 1.32, | |
"learning_rate": 0.001, | |
"loss": 0.8558, | |
"step": 148500 | |
}, | |
{ | |
"epoch": 1.32, | |
"learning_rate": 0.001, | |
"loss": 0.8644, | |
"step": 149000 | |
}, | |
{ | |
"epoch": 1.33, | |
"learning_rate": 0.001, | |
"loss": 0.8425, | |
"step": 149500 | |
}, | |
{ | |
"epoch": 1.33, | |
"learning_rate": 0.001, | |
"loss": 0.8499, | |
"step": 150000 | |
}, | |
{ | |
"epoch": 1.34, | |
"learning_rate": 0.001, | |
"loss": 0.8523, | |
"step": 150500 | |
}, | |
{ | |
"epoch": 1.34, | |
"learning_rate": 0.001, | |
"loss": 0.8515, | |
"step": 151000 | |
}, | |
{ | |
"epoch": 1.35, | |
"learning_rate": 0.001, | |
"loss": 0.8603, | |
"step": 151500 | |
}, | |
{ | |
"epoch": 1.35, | |
"learning_rate": 0.001, | |
"loss": 0.8536, | |
"step": 152000 | |
}, | |
{ | |
"epoch": 1.36, | |
"learning_rate": 0.001, | |
"loss": 0.8479, | |
"step": 152500 | |
}, | |
{ | |
"epoch": 1.36, | |
"learning_rate": 0.001, | |
"loss": 0.8473, | |
"step": 153000 | |
}, | |
{ | |
"epoch": 1.36, | |
"learning_rate": 0.001, | |
"loss": 0.8405, | |
"step": 153500 | |
}, | |
{ | |
"epoch": 1.37, | |
"learning_rate": 0.001, | |
"loss": 0.8304, | |
"step": 154000 | |
}, | |
{ | |
"epoch": 1.37, | |
"learning_rate": 0.001, | |
"loss": 0.8331, | |
"step": 154500 | |
}, | |
{ | |
"epoch": 1.38, | |
"learning_rate": 0.001, | |
"loss": 0.8403, | |
"step": 155000 | |
}, | |
{ | |
"epoch": 1.38, | |
"learning_rate": 0.001, | |
"loss": 0.8535, | |
"step": 155500 | |
}, | |
{ | |
"epoch": 1.39, | |
"learning_rate": 0.001, | |
"loss": 0.8434, | |
"step": 156000 | |
}, | |
{ | |
"epoch": 1.39, | |
"learning_rate": 0.001, | |
"loss": 0.8458, | |
"step": 156500 | |
}, | |
{ | |
"epoch": 1.4, | |
"learning_rate": 0.001, | |
"loss": 0.8428, | |
"step": 157000 | |
}, | |
{ | |
"epoch": 1.4, | |
"learning_rate": 0.001, | |
"loss": 0.8446, | |
"step": 157500 | |
}, | |
{ | |
"epoch": 1.4, | |
"learning_rate": 0.001, | |
"loss": 0.8419, | |
"step": 158000 | |
}, | |
{ | |
"epoch": 1.41, | |
"learning_rate": 0.001, | |
"loss": 0.8543, | |
"step": 158500 | |
}, | |
{ | |
"epoch": 1.41, | |
"learning_rate": 0.001, | |
"loss": 0.8475, | |
"step": 159000 | |
}, | |
{ | |
"epoch": 1.42, | |
"learning_rate": 0.001, | |
"loss": 0.8414, | |
"step": 159500 | |
}, | |
{ | |
"epoch": 1.42, | |
"learning_rate": 0.001, | |
"loss": 0.8504, | |
"step": 160000 | |
}, | |
{ | |
"epoch": 1.43, | |
"learning_rate": 0.001, | |
"loss": 0.8609, | |
"step": 160500 | |
}, | |
{ | |
"epoch": 1.43, | |
"learning_rate": 0.001, | |
"loss": 0.82, | |
"step": 161000 | |
}, | |
{ | |
"epoch": 1.44, | |
"learning_rate": 0.001, | |
"loss": 0.835, | |
"step": 161500 | |
}, | |
{ | |
"epoch": 1.44, | |
"learning_rate": 0.001, | |
"loss": 0.8393, | |
"step": 162000 | |
}, | |
{ | |
"epoch": 1.44, | |
"learning_rate": 0.001, | |
"loss": 0.8514, | |
"step": 162500 | |
}, | |
{ | |
"epoch": 1.45, | |
"learning_rate": 0.001, | |
"loss": 0.8474, | |
"step": 163000 | |
}, | |
{ | |
"epoch": 1.45, | |
"learning_rate": 0.001, | |
"loss": 0.8449, | |
"step": 163500 | |
}, | |
{ | |
"epoch": 1.46, | |
"learning_rate": 0.001, | |
"loss": 0.8425, | |
"step": 164000 | |
}, | |
{ | |
"epoch": 1.46, | |
"learning_rate": 0.001, | |
"loss": 0.8369, | |
"step": 164500 | |
}, | |
{ | |
"epoch": 1.47, | |
"learning_rate": 0.001, | |
"loss": 0.8265, | |
"step": 165000 | |
}, | |
{ | |
"epoch": 1.47, | |
"learning_rate": 0.001, | |
"loss": 0.8347, | |
"step": 165500 | |
}, | |
{ | |
"epoch": 1.48, | |
"learning_rate": 0.001, | |
"loss": 0.8335, | |
"step": 166000 | |
}, | |
{ | |
"epoch": 1.48, | |
"learning_rate": 0.001, | |
"loss": 0.8301, | |
"step": 166500 | |
}, | |
{ | |
"epoch": 1.48, | |
"learning_rate": 0.001, | |
"loss": 0.8308, | |
"step": 167000 | |
}, | |
{ | |
"epoch": 1.49, | |
"learning_rate": 0.001, | |
"loss": 0.8444, | |
"step": 167500 | |
}, | |
{ | |
"epoch": 1.49, | |
"learning_rate": 0.001, | |
"loss": 0.831, | |
"step": 168000 | |
}, | |
{ | |
"epoch": 1.5, | |
"learning_rate": 0.001, | |
"loss": 0.8535, | |
"step": 168500 | |
}, | |
{ | |
"epoch": 1.5, | |
"learning_rate": 0.001, | |
"loss": 0.8472, | |
"step": 169000 | |
}, | |
{ | |
"epoch": 1.51, | |
"learning_rate": 0.001, | |
"loss": 0.8475, | |
"step": 169500 | |
}, | |
{ | |
"epoch": 1.51, | |
"learning_rate": 0.001, | |
"loss": 0.8209, | |
"step": 170000 | |
}, | |
{ | |
"epoch": 1.52, | |
"learning_rate": 0.001, | |
"loss": 0.8422, | |
"step": 170500 | |
}, | |
{ | |
"epoch": 1.52, | |
"learning_rate": 0.001, | |
"loss": 0.8297, | |
"step": 171000 | |
}, | |
{ | |
"epoch": 1.52, | |
"learning_rate": 0.001, | |
"loss": 0.8287, | |
"step": 171500 | |
}, | |
{ | |
"epoch": 1.53, | |
"learning_rate": 0.001, | |
"loss": 0.8265, | |
"step": 172000 | |
}, | |
{ | |
"epoch": 1.53, | |
"learning_rate": 0.001, | |
"loss": 0.8265, | |
"step": 172500 | |
}, | |
{ | |
"epoch": 1.54, | |
"learning_rate": 0.001, | |
"loss": 0.8394, | |
"step": 173000 | |
}, | |
{ | |
"epoch": 1.54, | |
"learning_rate": 0.001, | |
"loss": 0.8364, | |
"step": 173500 | |
}, | |
{ | |
"epoch": 1.55, | |
"learning_rate": 0.001, | |
"loss": 0.8242, | |
"step": 174000 | |
}, | |
{ | |
"epoch": 1.55, | |
"learning_rate": 0.001, | |
"loss": 0.8311, | |
"step": 174500 | |
}, | |
{ | |
"epoch": 1.56, | |
"learning_rate": 0.001, | |
"loss": 0.822, | |
"step": 175000 | |
}, | |
{ | |
"epoch": 1.56, | |
"learning_rate": 0.001, | |
"loss": 0.8245, | |
"step": 175500 | |
}, | |
{ | |
"epoch": 1.56, | |
"learning_rate": 0.001, | |
"loss": 0.8281, | |
"step": 176000 | |
}, | |
{ | |
"epoch": 1.57, | |
"learning_rate": 0.001, | |
"loss": 0.8208, | |
"step": 176500 | |
}, | |
{ | |
"epoch": 1.57, | |
"learning_rate": 0.001, | |
"loss": 0.8328, | |
"step": 177000 | |
}, | |
{ | |
"epoch": 1.58, | |
"learning_rate": 0.001, | |
"loss": 0.83, | |
"step": 177500 | |
}, | |
{ | |
"epoch": 1.58, | |
"learning_rate": 0.001, | |
"loss": 0.8314, | |
"step": 178000 | |
}, | |
{ | |
"epoch": 1.59, | |
"learning_rate": 0.001, | |
"loss": 0.822, | |
"step": 178500 | |
}, | |
{ | |
"epoch": 1.59, | |
"learning_rate": 0.001, | |
"loss": 0.8189, | |
"step": 179000 | |
}, | |
{ | |
"epoch": 1.6, | |
"learning_rate": 0.001, | |
"loss": 0.8221, | |
"step": 179500 | |
}, | |
{ | |
"epoch": 1.6, | |
"learning_rate": 0.001, | |
"loss": 0.8187, | |
"step": 180000 | |
}, | |
{ | |
"epoch": 1.6, | |
"learning_rate": 0.001, | |
"loss": 0.8311, | |
"step": 180500 | |
}, | |
{ | |
"epoch": 1.61, | |
"learning_rate": 0.001, | |
"loss": 0.8193, | |
"step": 181000 | |
}, | |
{ | |
"epoch": 1.61, | |
"learning_rate": 0.001, | |
"loss": 0.8208, | |
"step": 181500 | |
}, | |
{ | |
"epoch": 1.62, | |
"learning_rate": 0.001, | |
"loss": 0.8305, | |
"step": 182000 | |
}, | |
{ | |
"epoch": 1.62, | |
"learning_rate": 0.001, | |
"loss": 0.8302, | |
"step": 182500 | |
}, | |
{ | |
"epoch": 1.63, | |
"learning_rate": 0.001, | |
"loss": 0.8354, | |
"step": 183000 | |
}, | |
{ | |
"epoch": 1.63, | |
"learning_rate": 0.001, | |
"loss": 0.8353, | |
"step": 183500 | |
}, | |
{ | |
"epoch": 1.64, | |
"learning_rate": 0.001, | |
"loss": 0.8353, | |
"step": 184000 | |
}, | |
{ | |
"epoch": 1.64, | |
"learning_rate": 0.001, | |
"loss": 0.8325, | |
"step": 184500 | |
}, | |
{ | |
"epoch": 1.64, | |
"learning_rate": 0.001, | |
"loss": 0.8132, | |
"step": 185000 | |
}, | |
{ | |
"epoch": 1.65, | |
"learning_rate": 0.001, | |
"loss": 0.8275, | |
"step": 185500 | |
}, | |
{ | |
"epoch": 1.65, | |
"learning_rate": 0.001, | |
"loss": 0.8165, | |
"step": 186000 | |
}, | |
{ | |
"epoch": 1.66, | |
"learning_rate": 0.001, | |
"loss": 0.8165, | |
"step": 186500 | |
}, | |
{ | |
"epoch": 1.66, | |
"learning_rate": 0.001, | |
"loss": 0.8246, | |
"step": 187000 | |
}, | |
{ | |
"epoch": 1.67, | |
"learning_rate": 0.001, | |
"loss": 0.8358, | |
"step": 187500 | |
}, | |
{ | |
"epoch": 1.67, | |
"learning_rate": 0.001, | |
"loss": 0.8265, | |
"step": 188000 | |
}, | |
{ | |
"epoch": 1.68, | |
"learning_rate": 0.001, | |
"loss": 0.8096, | |
"step": 188500 | |
}, | |
{ | |
"epoch": 1.68, | |
"learning_rate": 0.001, | |
"loss": 0.813, | |
"step": 189000 | |
}, | |
{ | |
"epoch": 1.68, | |
"learning_rate": 0.001, | |
"loss": 0.8203, | |
"step": 189500 | |
}, | |
{ | |
"epoch": 1.69, | |
"learning_rate": 0.001, | |
"loss": 0.8272, | |
"step": 190000 | |
}, | |
{ | |
"epoch": 1.69, | |
"learning_rate": 0.001, | |
"loss": 0.8114, | |
"step": 190500 | |
}, | |
{ | |
"epoch": 1.7, | |
"learning_rate": 0.001, | |
"loss": 0.8045, | |
"step": 191000 | |
}, | |
{ | |
"epoch": 1.7, | |
"learning_rate": 0.001, | |
"loss": 0.8295, | |
"step": 191500 | |
}, | |
{ | |
"epoch": 1.71, | |
"learning_rate": 0.001, | |
"loss": 0.8153, | |
"step": 192000 | |
}, | |
{ | |
"epoch": 1.71, | |
"learning_rate": 0.001, | |
"loss": 0.8181, | |
"step": 192500 | |
}, | |
{ | |
"epoch": 1.72, | |
"learning_rate": 0.001, | |
"loss": 0.8101, | |
"step": 193000 | |
}, | |
{ | |
"epoch": 1.72, | |
"learning_rate": 0.001, | |
"loss": 0.816, | |
"step": 193500 | |
}, | |
{ | |
"epoch": 1.72, | |
"learning_rate": 0.001, | |
"loss": 0.8139, | |
"step": 194000 | |
}, | |
{ | |
"epoch": 1.73, | |
"learning_rate": 0.001, | |
"loss": 0.8126, | |
"step": 194500 | |
}, | |
{ | |
"epoch": 1.73, | |
"learning_rate": 0.001, | |
"loss": 0.8168, | |
"step": 195000 | |
}, | |
{ | |
"epoch": 1.74, | |
"learning_rate": 0.001, | |
"loss": 0.8348, | |
"step": 195500 | |
}, | |
{ | |
"epoch": 1.74, | |
"learning_rate": 0.001, | |
"loss": 0.8248, | |
"step": 196000 | |
}, | |
{ | |
"epoch": 1.75, | |
"learning_rate": 0.001, | |
"loss": 0.8201, | |
"step": 196500 | |
}, | |
{ | |
"epoch": 1.75, | |
"learning_rate": 0.001, | |
"loss": 0.8142, | |
"step": 197000 | |
}, | |
{ | |
"epoch": 1.76, | |
"learning_rate": 0.001, | |
"loss": 0.8046, | |
"step": 197500 | |
}, | |
{ | |
"epoch": 1.76, | |
"learning_rate": 0.001, | |
"loss": 0.8027, | |
"step": 198000 | |
}, | |
{ | |
"epoch": 1.76, | |
"learning_rate": 0.001, | |
"loss": 0.8127, | |
"step": 198500 | |
}, | |
{ | |
"epoch": 1.77, | |
"learning_rate": 0.001, | |
"loss": 0.8314, | |
"step": 199000 | |
}, | |
{ | |
"epoch": 1.77, | |
"learning_rate": 0.001, | |
"loss": 0.8192, | |
"step": 199500 | |
}, | |
{ | |
"epoch": 1.78, | |
"learning_rate": 0.001, | |
"loss": 0.8005, | |
"step": 200000 | |
}, | |
{ | |
"epoch": 1.78, | |
"learning_rate": 0.001, | |
"loss": 0.8173, | |
"step": 200500 | |
}, | |
{ | |
"epoch": 1.79, | |
"learning_rate": 0.001, | |
"loss": 0.8133, | |
"step": 201000 | |
}, | |
{ | |
"epoch": 1.79, | |
"learning_rate": 0.001, | |
"loss": 0.8144, | |
"step": 201500 | |
}, | |
{ | |
"epoch": 1.8, | |
"learning_rate": 0.001, | |
"loss": 0.8199, | |
"step": 202000 | |
}, | |
{ | |
"epoch": 1.8, | |
"learning_rate": 0.001, | |
"loss": 0.7955, | |
"step": 202500 | |
}, | |
{ | |
"epoch": 1.8, | |
"learning_rate": 0.001, | |
"loss": 0.8012, | |
"step": 203000 | |
}, | |
{ | |
"epoch": 1.81, | |
"learning_rate": 0.001, | |
"loss": 0.807, | |
"step": 203500 | |
}, | |
{ | |
"epoch": 1.81, | |
"learning_rate": 0.001, | |
"loss": 0.8053, | |
"step": 204000 | |
}, | |
{ | |
"epoch": 1.82, | |
"learning_rate": 0.001, | |
"loss": 0.8077, | |
"step": 204500 | |
}, | |
{ | |
"epoch": 1.82, | |
"learning_rate": 0.001, | |
"loss": 0.8047, | |
"step": 205000 | |
}, | |
{ | |
"epoch": 1.83, | |
"learning_rate": 0.001, | |
"loss": 0.7873, | |
"step": 205500 | |
}, | |
{ | |
"epoch": 1.83, | |
"learning_rate": 0.001, | |
"loss": 0.8178, | |
"step": 206000 | |
}, | |
{ | |
"epoch": 1.84, | |
"learning_rate": 0.001, | |
"loss": 0.8041, | |
"step": 206500 | |
}, | |
{ | |
"epoch": 1.84, | |
"learning_rate": 0.001, | |
"loss": 0.8003, | |
"step": 207000 | |
}, | |
{ | |
"epoch": 1.84, | |
"learning_rate": 0.001, | |
"loss": 0.799, | |
"step": 207500 | |
}, | |
{ | |
"epoch": 1.85, | |
"learning_rate": 0.001, | |
"loss": 0.8095, | |
"step": 208000 | |
}, | |
{ | |
"epoch": 1.85, | |
"learning_rate": 0.001, | |
"loss": 0.8133, | |
"step": 208500 | |
}, | |
{ | |
"epoch": 1.86, | |
"learning_rate": 0.001, | |
"loss": 0.8091, | |
"step": 209000 | |
}, | |
{ | |
"epoch": 1.86, | |
"learning_rate": 0.001, | |
"loss": 0.8016, | |
"step": 209500 | |
}, | |
{ | |
"epoch": 1.87, | |
"learning_rate": 0.001, | |
"loss": 0.796, | |
"step": 210000 | |
}, | |
{ | |
"epoch": 1.87, | |
"learning_rate": 0.001, | |
"loss": 0.7999, | |
"step": 210500 | |
}, | |
{ | |
"epoch": 1.88, | |
"learning_rate": 0.001, | |
"loss": 0.8037, | |
"step": 211000 | |
}, | |
{ | |
"epoch": 1.88, | |
"learning_rate": 0.001, | |
"loss": 0.8096, | |
"step": 211500 | |
}, | |
{ | |
"epoch": 1.88, | |
"learning_rate": 0.001, | |
"loss": 0.7965, | |
"step": 212000 | |
}, | |
{ | |
"epoch": 1.89, | |
"learning_rate": 0.001, | |
"loss": 0.8053, | |
"step": 212500 | |
}, | |
{ | |
"epoch": 1.89, | |
"learning_rate": 0.001, | |
"loss": 0.7845, | |
"step": 213000 | |
}, | |
{ | |
"epoch": 1.9, | |
"learning_rate": 0.001, | |
"loss": 0.812, | |
"step": 213500 | |
}, | |
{ | |
"epoch": 1.9, | |
"learning_rate": 0.001, | |
"loss": 0.7981, | |
"step": 214000 | |
}, | |
{ | |
"epoch": 1.91, | |
"learning_rate": 0.001, | |
"loss": 0.7889, | |
"step": 214500 | |
}, | |
{ | |
"epoch": 1.91, | |
"learning_rate": 0.001, | |
"loss": 0.7928, | |
"step": 215000 | |
}, | |
{ | |
"epoch": 1.92, | |
"learning_rate": 0.001, | |
"loss": 0.7977, | |
"step": 215500 | |
}, | |
{ | |
"epoch": 1.92, | |
"learning_rate": 0.001, | |
"loss": 0.8108, | |
"step": 216000 | |
}, | |
{ | |
"epoch": 1.92, | |
"learning_rate": 0.001, | |
"loss": 0.7999, | |
"step": 216500 | |
}, | |
{ | |
"epoch": 1.93, | |
"learning_rate": 0.001, | |
"loss": 0.8031, | |
"step": 217000 | |
}, | |
{ | |
"epoch": 1.93, | |
"learning_rate": 0.001, | |
"loss": 0.8014, | |
"step": 217500 | |
}, | |
{ | |
"epoch": 1.94, | |
"learning_rate": 0.001, | |
"loss": 0.7943, | |
"step": 218000 | |
}, | |
{ | |
"epoch": 1.94, | |
"learning_rate": 0.001, | |
"loss": 0.7998, | |
"step": 218500 | |
}, | |
{ | |
"epoch": 1.95, | |
"learning_rate": 0.001, | |
"loss": 0.8035, | |
"step": 219000 | |
}, | |
{ | |
"epoch": 1.95, | |
"learning_rate": 0.001, | |
"loss": 0.8102, | |
"step": 219500 | |
}, | |
{ | |
"epoch": 1.96, | |
"learning_rate": 0.001, | |
"loss": 0.7945, | |
"step": 220000 | |
}, | |
{ | |
"epoch": 1.96, | |
"learning_rate": 0.001, | |
"loss": 0.7827, | |
"step": 220500 | |
}, | |
{ | |
"epoch": 1.96, | |
"learning_rate": 0.001, | |
"loss": 0.8058, | |
"step": 221000 | |
}, | |
{ | |
"epoch": 1.97, | |
"learning_rate": 0.001, | |
"loss": 0.7898, | |
"step": 221500 | |
}, | |
{ | |
"epoch": 1.97, | |
"learning_rate": 0.001, | |
"loss": 0.8023, | |
"step": 222000 | |
}, | |
{ | |
"epoch": 1.98, | |
"learning_rate": 0.001, | |
"loss": 0.7975, | |
"step": 222500 | |
}, | |
{ | |
"epoch": 1.98, | |
"learning_rate": 0.001, | |
"loss": 0.7971, | |
"step": 223000 | |
}, | |
{ | |
"epoch": 1.99, | |
"learning_rate": 0.001, | |
"loss": 0.7937, | |
"step": 223500 | |
}, | |
{ | |
"epoch": 1.99, | |
"learning_rate": 0.001, | |
"loss": 0.8034, | |
"step": 224000 | |
}, | |
{ | |
"epoch": 2.0, | |
"learning_rate": 0.001, | |
"loss": 0.7894, | |
"step": 224500 | |
}, | |
{ | |
"epoch": 2.0, | |
"learning_rate": 0.001, | |
"loss": 0.7931, | |
"step": 225000 | |
}, | |
{ | |
"epoch": 2.0, | |
"eval_bleu": 25.3566, | |
"eval_gen_len": 11.3335, | |
"eval_loss": 1.1507222652435303, | |
"eval_runtime": 423.4735, | |
"eval_samples_per_second": 33.06, | |
"step": 225000 | |
}, | |
{ | |
"epoch": 2.0, | |
"learning_rate": 0.001, | |
"loss": 0.7253, | |
"step": 225500 | |
}, | |
{ | |
"epoch": 2.01, | |
"learning_rate": 0.001, | |
"loss": 0.7256, | |
"step": 226000 | |
}, | |
{ | |
"epoch": 2.01, | |
"learning_rate": 0.001, | |
"loss": 0.737, | |
"step": 226500 | |
}, | |
{ | |
"epoch": 2.02, | |
"learning_rate": 0.001, | |
"loss": 0.7269, | |
"step": 227000 | |
}, | |
{ | |
"epoch": 2.02, | |
"learning_rate": 0.001, | |
"loss": 0.7416, | |
"step": 227500 | |
}, | |
{ | |
"epoch": 2.03, | |
"learning_rate": 0.001, | |
"loss": 0.7401, | |
"step": 228000 | |
}, | |
{ | |
"epoch": 2.03, | |
"learning_rate": 0.001, | |
"loss": 0.7468, | |
"step": 228500 | |
}, | |
{ | |
"epoch": 2.04, | |
"learning_rate": 0.001, | |
"loss": 0.7392, | |
"step": 229000 | |
}, | |
{ | |
"epoch": 2.04, | |
"learning_rate": 0.001, | |
"loss": 0.728, | |
"step": 229500 | |
}, | |
{ | |
"epoch": 2.04, | |
"learning_rate": 0.001, | |
"loss": 0.741, | |
"step": 230000 | |
}, | |
{ | |
"epoch": 2.05, | |
"learning_rate": 0.001, | |
"loss": 0.7415, | |
"step": 230500 | |
}, | |
{ | |
"epoch": 2.05, | |
"learning_rate": 0.001, | |
"loss": 0.721, | |
"step": 231000 | |
}, | |
{ | |
"epoch": 2.06, | |
"learning_rate": 0.001, | |
"loss": 0.7308, | |
"step": 231500 | |
}, | |
{ | |
"epoch": 2.06, | |
"learning_rate": 0.001, | |
"loss": 0.7378, | |
"step": 232000 | |
}, | |
{ | |
"epoch": 2.07, | |
"learning_rate": 0.001, | |
"loss": 0.7433, | |
"step": 232500 | |
}, | |
{ | |
"epoch": 2.07, | |
"learning_rate": 0.001, | |
"loss": 0.7267, | |
"step": 233000 | |
}, | |
{ | |
"epoch": 2.08, | |
"learning_rate": 0.001, | |
"loss": 0.7335, | |
"step": 233500 | |
}, | |
{ | |
"epoch": 2.08, | |
"learning_rate": 0.001, | |
"loss": 0.7513, | |
"step": 234000 | |
}, | |
{ | |
"epoch": 2.08, | |
"learning_rate": 0.001, | |
"loss": 0.7453, | |
"step": 234500 | |
}, | |
{ | |
"epoch": 2.09, | |
"learning_rate": 0.001, | |
"loss": 0.7346, | |
"step": 235000 | |
}, | |
{ | |
"epoch": 2.09, | |
"learning_rate": 0.001, | |
"loss": 0.7454, | |
"step": 235500 | |
}, | |
{ | |
"epoch": 2.1, | |
"learning_rate": 0.001, | |
"loss": 0.7376, | |
"step": 236000 | |
}, | |
{ | |
"epoch": 2.1, | |
"learning_rate": 0.001, | |
"loss": 0.7269, | |
"step": 236500 | |
}, | |
{ | |
"epoch": 2.11, | |
"learning_rate": 0.001, | |
"loss": 0.7299, | |
"step": 237000 | |
}, | |
{ | |
"epoch": 2.11, | |
"learning_rate": 0.001, | |
"loss": 0.7482, | |
"step": 237500 | |
}, | |
{ | |
"epoch": 2.12, | |
"learning_rate": 0.001, | |
"loss": 0.7489, | |
"step": 238000 | |
}, | |
{ | |
"epoch": 2.12, | |
"learning_rate": 0.001, | |
"loss": 0.7369, | |
"step": 238500 | |
}, | |
{ | |
"epoch": 2.12, | |
"learning_rate": 0.001, | |
"loss": 0.737, | |
"step": 239000 | |
}, | |
{ | |
"epoch": 2.13, | |
"learning_rate": 0.001, | |
"loss": 0.738, | |
"step": 239500 | |
}, | |
{ | |
"epoch": 2.13, | |
"learning_rate": 0.001, | |
"loss": 0.7427, | |
"step": 240000 | |
}, | |
{ | |
"epoch": 2.14, | |
"learning_rate": 0.001, | |
"loss": 0.7427, | |
"step": 240500 | |
}, | |
{ | |
"epoch": 2.14, | |
"learning_rate": 0.001, | |
"loss": 0.7381, | |
"step": 241000 | |
}, | |
{ | |
"epoch": 2.15, | |
"learning_rate": 0.001, | |
"loss": 0.7344, | |
"step": 241500 | |
}, | |
{ | |
"epoch": 2.15, | |
"learning_rate": 0.001, | |
"loss": 0.7378, | |
"step": 242000 | |
}, | |
{ | |
"epoch": 2.16, | |
"learning_rate": 0.001, | |
"loss": 0.7354, | |
"step": 242500 | |
}, | |
{ | |
"epoch": 2.16, | |
"learning_rate": 0.001, | |
"loss": 0.7284, | |
"step": 243000 | |
}, | |
{ | |
"epoch": 2.16, | |
"learning_rate": 0.001, | |
"loss": 0.7332, | |
"step": 243500 | |
}, | |
{ | |
"epoch": 2.17, | |
"learning_rate": 0.001, | |
"loss": 0.7295, | |
"step": 244000 | |
}, | |
{ | |
"epoch": 2.17, | |
"learning_rate": 0.001, | |
"loss": 0.7441, | |
"step": 244500 | |
}, | |
{ | |
"epoch": 2.18, | |
"learning_rate": 0.001, | |
"loss": 0.7482, | |
"step": 245000 | |
}, | |
{ | |
"epoch": 2.18, | |
"learning_rate": 0.001, | |
"loss": 0.7387, | |
"step": 245500 | |
}, | |
{ | |
"epoch": 2.19, | |
"learning_rate": 0.001, | |
"loss": 0.7443, | |
"step": 246000 | |
}, | |
{ | |
"epoch": 2.19, | |
"learning_rate": 0.001, | |
"loss": 0.7458, | |
"step": 246500 | |
}, | |
{ | |
"epoch": 2.2, | |
"learning_rate": 0.001, | |
"loss": 0.7347, | |
"step": 247000 | |
}, | |
{ | |
"epoch": 2.2, | |
"learning_rate": 0.001, | |
"loss": 0.7474, | |
"step": 247500 | |
}, | |
{ | |
"epoch": 2.2, | |
"learning_rate": 0.001, | |
"loss": 0.7404, | |
"step": 248000 | |
}, | |
{ | |
"epoch": 2.21, | |
"learning_rate": 0.001, | |
"loss": 0.7364, | |
"step": 248500 | |
}, | |
{ | |
"epoch": 2.21, | |
"learning_rate": 0.001, | |
"loss": 0.7472, | |
"step": 249000 | |
}, | |
{ | |
"epoch": 2.22, | |
"learning_rate": 0.001, | |
"loss": 0.7474, | |
"step": 249500 | |
}, | |
{ | |
"epoch": 2.22, | |
"learning_rate": 0.001, | |
"loss": 0.7337, | |
"step": 250000 | |
}, | |
{ | |
"epoch": 2.23, | |
"learning_rate": 0.001, | |
"loss": 0.7369, | |
"step": 250500 | |
}, | |
{ | |
"epoch": 2.23, | |
"learning_rate": 0.001, | |
"loss": 0.7431, | |
"step": 251000 | |
}, | |
{ | |
"epoch": 2.24, | |
"learning_rate": 0.001, | |
"loss": 0.7295, | |
"step": 251500 | |
}, | |
{ | |
"epoch": 2.24, | |
"learning_rate": 0.001, | |
"loss": 0.7479, | |
"step": 252000 | |
}, | |
{ | |
"epoch": 2.24, | |
"learning_rate": 0.001, | |
"loss": 0.7457, | |
"step": 252500 | |
}, | |
{ | |
"epoch": 2.25, | |
"learning_rate": 0.001, | |
"loss": 0.741, | |
"step": 253000 | |
}, | |
{ | |
"epoch": 2.25, | |
"learning_rate": 0.001, | |
"loss": 0.7333, | |
"step": 253500 | |
}, | |
{ | |
"epoch": 2.26, | |
"learning_rate": 0.001, | |
"loss": 0.7445, | |
"step": 254000 | |
}, | |
{ | |
"epoch": 2.26, | |
"learning_rate": 0.001, | |
"loss": 0.7334, | |
"step": 254500 | |
}, | |
{ | |
"epoch": 2.27, | |
"learning_rate": 0.001, | |
"loss": 0.7405, | |
"step": 255000 | |
}, | |
{ | |
"epoch": 2.27, | |
"learning_rate": 0.001, | |
"loss": 0.7382, | |
"step": 255500 | |
}, | |
{ | |
"epoch": 2.28, | |
"learning_rate": 0.001, | |
"loss": 0.7421, | |
"step": 256000 | |
}, | |
{ | |
"epoch": 2.28, | |
"learning_rate": 0.001, | |
"loss": 0.7422, | |
"step": 256500 | |
}, | |
{ | |
"epoch": 2.28, | |
"learning_rate": 0.001, | |
"loss": 0.7331, | |
"step": 257000 | |
}, | |
{ | |
"epoch": 2.29, | |
"learning_rate": 0.001, | |
"loss": 0.7303, | |
"step": 257500 | |
}, | |
{ | |
"epoch": 2.29, | |
"learning_rate": 0.001, | |
"loss": 0.7348, | |
"step": 258000 | |
}, | |
{ | |
"epoch": 2.3, | |
"learning_rate": 0.001, | |
"loss": 0.7532, | |
"step": 258500 | |
}, | |
{ | |
"epoch": 2.3, | |
"learning_rate": 0.001, | |
"loss": 0.7396, | |
"step": 259000 | |
}, | |
{ | |
"epoch": 2.31, | |
"learning_rate": 0.001, | |
"loss": 0.7383, | |
"step": 259500 | |
}, | |
{ | |
"epoch": 2.31, | |
"learning_rate": 0.001, | |
"loss": 0.7474, | |
"step": 260000 | |
}, | |
{ | |
"epoch": 2.32, | |
"learning_rate": 0.001, | |
"loss": 0.733, | |
"step": 260500 | |
}, | |
{ | |
"epoch": 2.32, | |
"learning_rate": 0.001, | |
"loss": 0.7368, | |
"step": 261000 | |
}, | |
{ | |
"epoch": 2.32, | |
"learning_rate": 0.001, | |
"loss": 0.7301, | |
"step": 261500 | |
}, | |
{ | |
"epoch": 2.33, | |
"learning_rate": 0.001, | |
"loss": 0.7414, | |
"step": 262000 | |
}, | |
{ | |
"epoch": 2.33, | |
"learning_rate": 0.001, | |
"loss": 0.7374, | |
"step": 262500 | |
}, | |
{ | |
"epoch": 2.34, | |
"learning_rate": 0.001, | |
"loss": 0.7529, | |
"step": 263000 | |
}, | |
{ | |
"epoch": 2.34, | |
"learning_rate": 0.001, | |
"loss": 0.7328, | |
"step": 263500 | |
}, | |
{ | |
"epoch": 2.35, | |
"learning_rate": 0.001, | |
"loss": 0.7442, | |
"step": 264000 | |
}, | |
{ | |
"epoch": 2.35, | |
"learning_rate": 0.001, | |
"loss": 0.7357, | |
"step": 264500 | |
}, | |
{ | |
"epoch": 2.36, | |
"learning_rate": 0.001, | |
"loss": 0.7313, | |
"step": 265000 | |
}, | |
{ | |
"epoch": 2.36, | |
"learning_rate": 0.001, | |
"loss": 0.7404, | |
"step": 265500 | |
}, | |
{ | |
"epoch": 2.36, | |
"learning_rate": 0.001, | |
"loss": 0.7372, | |
"step": 266000 | |
}, | |
{ | |
"epoch": 2.37, | |
"learning_rate": 0.001, | |
"loss": 0.7419, | |
"step": 266500 | |
}, | |
{ | |
"epoch": 2.37, | |
"learning_rate": 0.001, | |
"loss": 0.7305, | |
"step": 267000 | |
}, | |
{ | |
"epoch": 2.38, | |
"learning_rate": 0.001, | |
"loss": 0.7376, | |
"step": 267500 | |
}, | |
{ | |
"epoch": 2.38, | |
"learning_rate": 0.001, | |
"loss": 0.722, | |
"step": 268000 | |
}, | |
{ | |
"epoch": 2.39, | |
"learning_rate": 0.001, | |
"loss": 0.7442, | |
"step": 268500 | |
}, | |
{ | |
"epoch": 2.39, | |
"learning_rate": 0.001, | |
"loss": 0.7323, | |
"step": 269000 | |
}, | |
{ | |
"epoch": 2.4, | |
"learning_rate": 0.001, | |
"loss": 0.7262, | |
"step": 269500 | |
}, | |
{ | |
"epoch": 2.4, | |
"learning_rate": 0.001, | |
"loss": 0.7313, | |
"step": 270000 | |
}, | |
{ | |
"epoch": 2.4, | |
"learning_rate": 0.001, | |
"loss": 0.7305, | |
"step": 270500 | |
}, | |
{ | |
"epoch": 2.41, | |
"learning_rate": 0.001, | |
"loss": 0.7271, | |
"step": 271000 | |
}, | |
{ | |
"epoch": 2.41, | |
"learning_rate": 0.001, | |
"loss": 0.7391, | |
"step": 271500 | |
}, | |
{ | |
"epoch": 2.42, | |
"learning_rate": 0.001, | |
"loss": 0.7361, | |
"step": 272000 | |
}, | |
{ | |
"epoch": 2.42, | |
"learning_rate": 0.001, | |
"loss": 0.7339, | |
"step": 272500 | |
}, | |
{ | |
"epoch": 2.43, | |
"learning_rate": 0.001, | |
"loss": 0.7318, | |
"step": 273000 | |
}, | |
{ | |
"epoch": 2.43, | |
"learning_rate": 0.001, | |
"loss": 0.7278, | |
"step": 273500 | |
}, | |
{ | |
"epoch": 2.44, | |
"learning_rate": 0.001, | |
"loss": 0.7438, | |
"step": 274000 | |
}, | |
{ | |
"epoch": 2.44, | |
"learning_rate": 0.001, | |
"loss": 0.7411, | |
"step": 274500 | |
}, | |
{ | |
"epoch": 2.44, | |
"learning_rate": 0.001, | |
"loss": 0.7447, | |
"step": 275000 | |
}, | |
{ | |
"epoch": 2.45, | |
"learning_rate": 0.001, | |
"loss": 0.7326, | |
"step": 275500 | |
}, | |
{ | |
"epoch": 2.45, | |
"learning_rate": 0.001, | |
"loss": 0.7282, | |
"step": 276000 | |
}, | |
{ | |
"epoch": 2.46, | |
"learning_rate": 0.001, | |
"loss": 0.7312, | |
"step": 276500 | |
}, | |
{ | |
"epoch": 2.46, | |
"learning_rate": 0.001, | |
"loss": 0.7387, | |
"step": 277000 | |
}, | |
{ | |
"epoch": 2.47, | |
"learning_rate": 0.001, | |
"loss": 0.7263, | |
"step": 277500 | |
}, | |
{ | |
"epoch": 2.47, | |
"learning_rate": 0.001, | |
"loss": 0.715, | |
"step": 278000 | |
}, | |
{ | |
"epoch": 2.48, | |
"learning_rate": 0.001, | |
"loss": 0.722, | |
"step": 278500 | |
}, | |
{ | |
"epoch": 2.48, | |
"learning_rate": 0.001, | |
"loss": 0.7268, | |
"step": 279000 | |
}, | |
{ | |
"epoch": 2.48, | |
"learning_rate": 0.001, | |
"loss": 0.7386, | |
"step": 279500 | |
}, | |
{ | |
"epoch": 2.49, | |
"learning_rate": 0.001, | |
"loss": 0.7435, | |
"step": 280000 | |
}, | |
{ | |
"epoch": 2.49, | |
"learning_rate": 0.001, | |
"loss": 0.7273, | |
"step": 280500 | |
}, | |
{ | |
"epoch": 2.5, | |
"learning_rate": 0.001, | |
"loss": 0.7378, | |
"step": 281000 | |
}, | |
{ | |
"epoch": 2.5, | |
"learning_rate": 0.001, | |
"loss": 0.7376, | |
"step": 281500 | |
}, | |
{ | |
"epoch": 2.51, | |
"learning_rate": 0.001, | |
"loss": 0.7397, | |
"step": 282000 | |
}, | |
{ | |
"epoch": 2.51, | |
"learning_rate": 0.001, | |
"loss": 0.7364, | |
"step": 282500 | |
}, | |
{ | |
"epoch": 2.52, | |
"learning_rate": 0.001, | |
"loss": 0.7197, | |
"step": 283000 | |
}, | |
{ | |
"epoch": 2.52, | |
"learning_rate": 0.001, | |
"loss": 0.7421, | |
"step": 283500 | |
}, | |
{ | |
"epoch": 2.52, | |
"learning_rate": 0.001, | |
"loss": 0.7335, | |
"step": 284000 | |
}, | |
{ | |
"epoch": 2.53, | |
"learning_rate": 0.001, | |
"loss": 0.7281, | |
"step": 284500 | |
}, | |
{ | |
"epoch": 2.53, | |
"learning_rate": 0.001, | |
"loss": 0.7263, | |
"step": 285000 | |
}, | |
{ | |
"epoch": 2.54, | |
"learning_rate": 0.001, | |
"loss": 0.7195, | |
"step": 285500 | |
}, | |
{ | |
"epoch": 2.54, | |
"learning_rate": 0.001, | |
"loss": 0.7336, | |
"step": 286000 | |
}, | |
{ | |
"epoch": 2.55, | |
"learning_rate": 0.001, | |
"loss": 0.7307, | |
"step": 286500 | |
}, | |
{ | |
"epoch": 2.55, | |
"learning_rate": 0.001, | |
"loss": 0.7255, | |
"step": 287000 | |
}, | |
{ | |
"epoch": 2.56, | |
"learning_rate": 0.001, | |
"loss": 0.7471, | |
"step": 287500 | |
}, | |
{ | |
"epoch": 2.56, | |
"learning_rate": 0.001, | |
"loss": 0.7307, | |
"step": 288000 | |
}, | |
{ | |
"epoch": 2.56, | |
"learning_rate": 0.001, | |
"loss": 0.7251, | |
"step": 288500 | |
}, | |
{ | |
"epoch": 2.57, | |
"learning_rate": 0.001, | |
"loss": 0.7269, | |
"step": 289000 | |
}, | |
{ | |
"epoch": 2.57, | |
"learning_rate": 0.001, | |
"loss": 0.7324, | |
"step": 289500 | |
}, | |
{ | |
"epoch": 2.58, | |
"learning_rate": 0.001, | |
"loss": 0.7142, | |
"step": 290000 | |
}, | |
{ | |
"epoch": 2.58, | |
"learning_rate": 0.001, | |
"loss": 0.7318, | |
"step": 290500 | |
}, | |
{ | |
"epoch": 2.59, | |
"learning_rate": 0.001, | |
"loss": 0.7236, | |
"step": 291000 | |
}, | |
{ | |
"epoch": 2.59, | |
"learning_rate": 0.001, | |
"loss": 0.7267, | |
"step": 291500 | |
}, | |
{ | |
"epoch": 2.6, | |
"learning_rate": 0.001, | |
"loss": 0.7387, | |
"step": 292000 | |
}, | |
{ | |
"epoch": 2.6, | |
"learning_rate": 0.001, | |
"loss": 0.7299, | |
"step": 292500 | |
}, | |
{ | |
"epoch": 2.6, | |
"learning_rate": 0.001, | |
"loss": 0.7518, | |
"step": 293000 | |
}, | |
{ | |
"epoch": 2.61, | |
"learning_rate": 0.001, | |
"loss": 0.7227, | |
"step": 293500 | |
}, | |
{ | |
"epoch": 2.61, | |
"learning_rate": 0.001, | |
"loss": 0.7261, | |
"step": 294000 | |
}, | |
{ | |
"epoch": 2.62, | |
"learning_rate": 0.001, | |
"loss": 0.7325, | |
"step": 294500 | |
}, | |
{ | |
"epoch": 2.62, | |
"learning_rate": 0.001, | |
"loss": 0.7262, | |
"step": 295000 | |
}, | |
{ | |
"epoch": 2.63, | |
"learning_rate": 0.001, | |
"loss": 0.7226, | |
"step": 295500 | |
}, | |
{ | |
"epoch": 2.63, | |
"learning_rate": 0.001, | |
"loss": 0.7218, | |
"step": 296000 | |
}, | |
{ | |
"epoch": 2.64, | |
"learning_rate": 0.001, | |
"loss": 0.7392, | |
"step": 296500 | |
}, | |
{ | |
"epoch": 2.64, | |
"learning_rate": 0.001, | |
"loss": 0.7262, | |
"step": 297000 | |
}, | |
{ | |
"epoch": 2.64, | |
"learning_rate": 0.001, | |
"loss": 0.7234, | |
"step": 297500 | |
}, | |
{ | |
"epoch": 2.65, | |
"learning_rate": 0.001, | |
"loss": 0.7265, | |
"step": 298000 | |
}, | |
{ | |
"epoch": 2.65, | |
"learning_rate": 0.001, | |
"loss": 0.721, | |
"step": 298500 | |
}, | |
{ | |
"epoch": 2.66, | |
"learning_rate": 0.001, | |
"loss": 0.7163, | |
"step": 299000 | |
}, | |
{ | |
"epoch": 2.66, | |
"learning_rate": 0.001, | |
"loss": 0.7361, | |
"step": 299500 | |
}, | |
{ | |
"epoch": 2.67, | |
"learning_rate": 0.001, | |
"loss": 0.7102, | |
"step": 300000 | |
}, | |
{ | |
"epoch": 2.67, | |
"learning_rate": 0.001, | |
"loss": 0.727, | |
"step": 300500 | |
}, | |
{ | |
"epoch": 2.68, | |
"learning_rate": 0.001, | |
"loss": 0.7309, | |
"step": 301000 | |
}, | |
{ | |
"epoch": 2.68, | |
"learning_rate": 0.001, | |
"loss": 0.7185, | |
"step": 301500 | |
}, | |
{ | |
"epoch": 2.68, | |
"learning_rate": 0.001, | |
"loss": 0.7242, | |
"step": 302000 | |
}, | |
{ | |
"epoch": 2.69, | |
"learning_rate": 0.001, | |
"loss": 0.7308, | |
"step": 302500 | |
}, | |
{ | |
"epoch": 2.69, | |
"learning_rate": 0.001, | |
"loss": 0.7323, | |
"step": 303000 | |
}, | |
{ | |
"epoch": 2.7, | |
"learning_rate": 0.001, | |
"loss": 0.7293, | |
"step": 303500 | |
}, | |
{ | |
"epoch": 2.7, | |
"learning_rate": 0.001, | |
"loss": 0.7078, | |
"step": 304000 | |
}, | |
{ | |
"epoch": 2.71, | |
"learning_rate": 0.001, | |
"loss": 0.719, | |
"step": 304500 | |
}, | |
{ | |
"epoch": 2.71, | |
"learning_rate": 0.001, | |
"loss": 0.7252, | |
"step": 305000 | |
}, | |
{ | |
"epoch": 2.72, | |
"learning_rate": 0.001, | |
"loss": 0.7252, | |
"step": 305500 | |
}, | |
{ | |
"epoch": 2.72, | |
"learning_rate": 0.001, | |
"loss": 0.7243, | |
"step": 306000 | |
}, | |
{ | |
"epoch": 2.72, | |
"learning_rate": 0.001, | |
"loss": 0.7154, | |
"step": 306500 | |
}, | |
{ | |
"epoch": 2.73, | |
"learning_rate": 0.001, | |
"loss": 0.718, | |
"step": 307000 | |
}, | |
{ | |
"epoch": 2.73, | |
"learning_rate": 0.001, | |
"loss": 0.7238, | |
"step": 307500 | |
}, | |
{ | |
"epoch": 2.74, | |
"learning_rate": 0.001, | |
"loss": 0.7181, | |
"step": 308000 | |
}, | |
{ | |
"epoch": 2.74, | |
"learning_rate": 0.001, | |
"loss": 0.7288, | |
"step": 308500 | |
}, | |
{ | |
"epoch": 2.75, | |
"learning_rate": 0.001, | |
"loss": 0.717, | |
"step": 309000 | |
}, | |
{ | |
"epoch": 2.75, | |
"learning_rate": 0.001, | |
"loss": 0.7242, | |
"step": 309500 | |
}, | |
{ | |
"epoch": 2.76, | |
"learning_rate": 0.001, | |
"loss": 0.7265, | |
"step": 310000 | |
}, | |
{ | |
"epoch": 2.76, | |
"learning_rate": 0.001, | |
"loss": 0.7091, | |
"step": 310500 | |
}, | |
{ | |
"epoch": 2.76, | |
"learning_rate": 0.001, | |
"loss": 0.7156, | |
"step": 311000 | |
}, | |
{ | |
"epoch": 2.77, | |
"learning_rate": 0.001, | |
"loss": 0.7215, | |
"step": 311500 | |
}, | |
{ | |
"epoch": 2.77, | |
"learning_rate": 0.001, | |
"loss": 0.719, | |
"step": 312000 | |
}, | |
{ | |
"epoch": 2.78, | |
"learning_rate": 0.001, | |
"loss": 0.7224, | |
"step": 312500 | |
}, | |
{ | |
"epoch": 2.78, | |
"learning_rate": 0.001, | |
"loss": 0.726, | |
"step": 313000 | |
}, | |
{ | |
"epoch": 2.79, | |
"learning_rate": 0.001, | |
"loss": 0.7081, | |
"step": 313500 | |
}, | |
{ | |
"epoch": 2.79, | |
"learning_rate": 0.001, | |
"loss": 0.714, | |
"step": 314000 | |
}, | |
{ | |
"epoch": 2.8, | |
"learning_rate": 0.001, | |
"loss": 0.7191, | |
"step": 314500 | |
}, | |
{ | |
"epoch": 2.8, | |
"learning_rate": 0.001, | |
"loss": 0.7115, | |
"step": 315000 | |
}, | |
{ | |
"epoch": 2.8, | |
"learning_rate": 0.001, | |
"loss": 0.711, | |
"step": 315500 | |
}, | |
{ | |
"epoch": 2.81, | |
"learning_rate": 0.001, | |
"loss": 0.7064, | |
"step": 316000 | |
}, | |
{ | |
"epoch": 2.81, | |
"learning_rate": 0.001, | |
"loss": 0.7096, | |
"step": 316500 | |
}, | |
{ | |
"epoch": 2.82, | |
"learning_rate": 0.001, | |
"loss": 0.7159, | |
"step": 317000 | |
}, | |
{ | |
"epoch": 2.82, | |
"learning_rate": 0.001, | |
"loss": 0.7251, | |
"step": 317500 | |
}, | |
{ | |
"epoch": 2.83, | |
"learning_rate": 0.001, | |
"loss": 0.7117, | |
"step": 318000 | |
}, | |
{ | |
"epoch": 2.83, | |
"learning_rate": 0.001, | |
"loss": 0.7108, | |
"step": 318500 | |
}, | |
{ | |
"epoch": 2.84, | |
"learning_rate": 0.001, | |
"loss": 0.7104, | |
"step": 319000 | |
}, | |
{ | |
"epoch": 2.84, | |
"learning_rate": 0.001, | |
"loss": 0.7087, | |
"step": 319500 | |
}, | |
{ | |
"epoch": 2.84, | |
"learning_rate": 0.001, | |
"loss": 0.7163, | |
"step": 320000 | |
}, | |
{ | |
"epoch": 2.85, | |
"learning_rate": 0.001, | |
"loss": 0.7341, | |
"step": 320500 | |
}, | |
{ | |
"epoch": 2.85, | |
"learning_rate": 0.001, | |
"loss": 0.7153, | |
"step": 321000 | |
}, | |
{ | |
"epoch": 2.86, | |
"learning_rate": 0.001, | |
"loss": 0.717, | |
"step": 321500 | |
}, | |
{ | |
"epoch": 2.86, | |
"learning_rate": 0.001, | |
"loss": 0.7155, | |
"step": 322000 | |
}, | |
{ | |
"epoch": 2.87, | |
"learning_rate": 0.001, | |
"loss": 0.7144, | |
"step": 322500 | |
}, | |
{ | |
"epoch": 2.87, | |
"learning_rate": 0.001, | |
"loss": 0.7093, | |
"step": 323000 | |
}, | |
{ | |
"epoch": 2.88, | |
"learning_rate": 0.001, | |
"loss": 0.7071, | |
"step": 323500 | |
}, | |
{ | |
"epoch": 2.88, | |
"learning_rate": 0.001, | |
"loss": 0.7087, | |
"step": 324000 | |
}, | |
{ | |
"epoch": 2.88, | |
"learning_rate": 0.001, | |
"loss": 0.7374, | |
"step": 324500 | |
}, | |
{ | |
"epoch": 2.89, | |
"learning_rate": 0.001, | |
"loss": 0.7276, | |
"step": 325000 | |
}, | |
{ | |
"epoch": 2.89, | |
"learning_rate": 0.001, | |
"loss": 0.7269, | |
"step": 325500 | |
}, | |
{ | |
"epoch": 2.9, | |
"learning_rate": 0.001, | |
"loss": 0.7252, | |
"step": 326000 | |
}, | |
{ | |
"epoch": 2.9, | |
"learning_rate": 0.001, | |
"loss": 0.7061, | |
"step": 326500 | |
}, | |
{ | |
"epoch": 2.91, | |
"learning_rate": 0.001, | |
"loss": 0.7138, | |
"step": 327000 | |
}, | |
{ | |
"epoch": 2.91, | |
"learning_rate": 0.001, | |
"loss": 0.7033, | |
"step": 327500 | |
}, | |
{ | |
"epoch": 2.92, | |
"learning_rate": 0.001, | |
"loss": 0.7057, | |
"step": 328000 | |
}, | |
{ | |
"epoch": 2.92, | |
"learning_rate": 0.001, | |
"loss": 0.7082, | |
"step": 328500 | |
}, | |
{ | |
"epoch": 2.92, | |
"learning_rate": 0.001, | |
"loss": 0.7056, | |
"step": 329000 | |
}, | |
{ | |
"epoch": 2.93, | |
"learning_rate": 0.001, | |
"loss": 0.7125, | |
"step": 329500 | |
}, | |
{ | |
"epoch": 2.93, | |
"learning_rate": 0.001, | |
"loss": 0.7022, | |
"step": 330000 | |
}, | |
{ | |
"epoch": 2.94, | |
"learning_rate": 0.001, | |
"loss": 0.7132, | |
"step": 330500 | |
}, | |
{ | |
"epoch": 2.94, | |
"learning_rate": 0.001, | |
"loss": 0.7175, | |
"step": 331000 | |
}, | |
{ | |
"epoch": 2.95, | |
"learning_rate": 0.001, | |
"loss": 0.7104, | |
"step": 331500 | |
}, | |
{ | |
"epoch": 2.95, | |
"learning_rate": 0.001, | |
"loss": 0.7158, | |
"step": 332000 | |
}, | |
{ | |
"epoch": 2.96, | |
"learning_rate": 0.001, | |
"loss": 0.7102, | |
"step": 332500 | |
} | |
], | |
"max_steps": 337500, | |
"num_train_epochs": 3, | |
"total_flos": 369594455408640.0, | |
"trial_name": null, | |
"trial_params": null | |
} | |