{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.893939393939394, "global_step": 1500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 1.9873737373737378e-05, "loss": 8.938, "step": 10 }, { "epoch": 0.03, "learning_rate": 1.9747474747474747e-05, "loss": 0.3506, "step": 20 }, { "epoch": 0.04, "learning_rate": 1.9621212121212123e-05, "loss": 0.1109, "step": 30 }, { "epoch": 0.05, "learning_rate": 1.9494949494949496e-05, "loss": 0.1104, "step": 40 }, { "epoch": 0.06, "learning_rate": 1.936868686868687e-05, "loss": 0.1353, "step": 50 }, { "epoch": 0.08, "learning_rate": 1.9242424242424244e-05, "loss": 0.1138, "step": 60 }, { "epoch": 0.09, "learning_rate": 1.9116161616161617e-05, "loss": 0.1159, "step": 70 }, { "epoch": 0.1, "learning_rate": 1.8989898989898993e-05, "loss": 0.1157, "step": 80 }, { "epoch": 0.11, "learning_rate": 1.8863636363636366e-05, "loss": 0.1018, "step": 90 }, { "epoch": 0.13, "learning_rate": 1.873737373737374e-05, "loss": 0.1111, "step": 100 }, { "epoch": 0.14, "learning_rate": 1.8611111111111114e-05, "loss": 0.1432, "step": 110 }, { "epoch": 0.15, "learning_rate": 1.8484848484848487e-05, "loss": 0.1059, "step": 120 }, { "epoch": 0.16, "learning_rate": 1.835858585858586e-05, "loss": 0.1213, "step": 130 }, { "epoch": 0.18, "learning_rate": 1.8232323232323236e-05, "loss": 0.1344, "step": 140 }, { "epoch": 0.19, "learning_rate": 1.810606060606061e-05, "loss": 0.1174, "step": 150 }, { "epoch": 0.2, "learning_rate": 1.797979797979798e-05, "loss": 0.1579, "step": 160 }, { "epoch": 0.21, "learning_rate": 1.7853535353535354e-05, "loss": 0.109, "step": 170 }, { "epoch": 0.23, "learning_rate": 1.772727272727273e-05, "loss": 0.1852, "step": 180 }, { "epoch": 0.24, "learning_rate": 1.7601010101010102e-05, "loss": 0.0766, "step": 190 }, { "epoch": 0.25, "learning_rate": 1.7474747474747475e-05, "loss": 0.1577, "step": 200 }, { "epoch": 0.27, "learning_rate": 1.734848484848485e-05, "loss": 0.1063, "step": 210 }, { "epoch": 0.28, "learning_rate": 1.7222222222222224e-05, "loss": 0.1002, "step": 220 }, { "epoch": 0.29, "learning_rate": 1.7095959595959596e-05, "loss": 0.1122, "step": 230 }, { "epoch": 0.3, "learning_rate": 1.6969696969696972e-05, "loss": 0.1283, "step": 240 }, { "epoch": 0.32, "learning_rate": 1.6843434343434345e-05, "loss": 0.1163, "step": 250 }, { "epoch": 0.33, "learning_rate": 1.6717171717171718e-05, "loss": 0.1092, "step": 260 }, { "epoch": 0.34, "learning_rate": 1.6590909090909094e-05, "loss": 0.114, "step": 270 }, { "epoch": 0.35, "learning_rate": 1.6464646464646466e-05, "loss": 0.0944, "step": 280 }, { "epoch": 0.37, "learning_rate": 1.633838383838384e-05, "loss": 0.1325, "step": 290 }, { "epoch": 0.38, "learning_rate": 1.6212121212121212e-05, "loss": 0.0885, "step": 300 }, { "epoch": 0.39, "learning_rate": 1.6085858585858588e-05, "loss": 0.1479, "step": 310 }, { "epoch": 0.4, "learning_rate": 1.595959595959596e-05, "loss": 0.135, "step": 320 }, { "epoch": 0.42, "learning_rate": 1.5833333333333333e-05, "loss": 0.1261, "step": 330 }, { "epoch": 0.43, "learning_rate": 1.570707070707071e-05, "loss": 0.0947, "step": 340 }, { "epoch": 0.44, "learning_rate": 1.5580808080808082e-05, "loss": 0.1241, "step": 350 }, { "epoch": 0.45, "learning_rate": 1.5454545454545454e-05, "loss": 0.1229, "step": 360 }, { "epoch": 0.47, "learning_rate": 1.532828282828283e-05, "loss": 0.1149, "step": 370 }, { "epoch": 0.48, "learning_rate": 1.5202020202020203e-05, "loss": 0.133, "step": 380 }, { "epoch": 0.49, "learning_rate": 1.5075757575757577e-05, "loss": 0.1303, "step": 390 }, { "epoch": 0.51, "learning_rate": 1.4949494949494952e-05, "loss": 0.0846, "step": 400 }, { "epoch": 0.52, "learning_rate": 1.4823232323232324e-05, "loss": 0.0893, "step": 410 }, { "epoch": 0.53, "learning_rate": 1.4696969696969699e-05, "loss": 0.0996, "step": 420 }, { "epoch": 0.54, "learning_rate": 1.4570707070707071e-05, "loss": 0.0987, "step": 430 }, { "epoch": 0.56, "learning_rate": 1.4444444444444446e-05, "loss": 0.1045, "step": 440 }, { "epoch": 0.57, "learning_rate": 1.431818181818182e-05, "loss": 0.1004, "step": 450 }, { "epoch": 0.58, "learning_rate": 1.4191919191919193e-05, "loss": 0.0917, "step": 460 }, { "epoch": 0.59, "learning_rate": 1.4065656565656567e-05, "loss": 0.15, "step": 470 }, { "epoch": 0.61, "learning_rate": 1.3939393939393942e-05, "loss": 0.0974, "step": 480 }, { "epoch": 0.62, "learning_rate": 1.3813131313131314e-05, "loss": 0.1036, "step": 490 }, { "epoch": 0.63, "learning_rate": 1.3686868686868689e-05, "loss": 0.1123, "step": 500 }, { "epoch": 0.64, "learning_rate": 1.3560606060606061e-05, "loss": 0.0806, "step": 510 }, { "epoch": 0.66, "learning_rate": 1.3434343434343436e-05, "loss": 0.1059, "step": 520 }, { "epoch": 0.67, "learning_rate": 1.330808080808081e-05, "loss": 0.0984, "step": 530 }, { "epoch": 0.68, "learning_rate": 1.3181818181818183e-05, "loss": 0.1285, "step": 540 }, { "epoch": 0.69, "learning_rate": 1.3055555555555557e-05, "loss": 0.1089, "step": 550 }, { "epoch": 0.71, "learning_rate": 1.2929292929292931e-05, "loss": 0.0928, "step": 560 }, { "epoch": 0.72, "learning_rate": 1.2803030303030304e-05, "loss": 0.0948, "step": 570 }, { "epoch": 0.73, "learning_rate": 1.2676767676767678e-05, "loss": 0.0934, "step": 580 }, { "epoch": 0.74, "learning_rate": 1.2550505050505051e-05, "loss": 0.094, "step": 590 }, { "epoch": 0.76, "learning_rate": 1.2424242424242425e-05, "loss": 0.0862, "step": 600 }, { "epoch": 0.77, "learning_rate": 1.22979797979798e-05, "loss": 0.0951, "step": 610 }, { "epoch": 0.78, "learning_rate": 1.2171717171717172e-05, "loss": 0.1, "step": 620 }, { "epoch": 0.8, "learning_rate": 1.2045454545454547e-05, "loss": 0.0789, "step": 630 }, { "epoch": 0.81, "learning_rate": 1.191919191919192e-05, "loss": 0.0726, "step": 640 }, { "epoch": 0.82, "learning_rate": 1.1792929292929294e-05, "loss": 0.1093, "step": 650 }, { "epoch": 0.83, "learning_rate": 1.1666666666666668e-05, "loss": 0.1078, "step": 660 }, { "epoch": 0.85, "learning_rate": 1.154040404040404e-05, "loss": 0.1226, "step": 670 }, { "epoch": 0.86, "learning_rate": 1.1414141414141415e-05, "loss": 0.0753, "step": 680 }, { "epoch": 0.87, "learning_rate": 1.128787878787879e-05, "loss": 0.0905, "step": 690 }, { "epoch": 0.88, "learning_rate": 1.1161616161616162e-05, "loss": 0.0887, "step": 700 }, { "epoch": 0.9, "learning_rate": 1.1035353535353536e-05, "loss": 0.1082, "step": 710 }, { "epoch": 0.91, "learning_rate": 1.0909090909090909e-05, "loss": 0.0809, "step": 720 }, { "epoch": 0.92, "learning_rate": 1.0782828282828283e-05, "loss": 0.1048, "step": 730 }, { "epoch": 0.93, "learning_rate": 1.0656565656565658e-05, "loss": 0.1034, "step": 740 }, { "epoch": 0.95, "learning_rate": 1.053030303030303e-05, "loss": 0.1425, "step": 750 }, { "epoch": 0.96, "learning_rate": 1.0404040404040405e-05, "loss": 0.0966, "step": 760 }, { "epoch": 0.97, "learning_rate": 1.0277777777777777e-05, "loss": 0.118, "step": 770 }, { "epoch": 0.98, "learning_rate": 1.0151515151515152e-05, "loss": 0.1258, "step": 780 }, { "epoch": 1.0, "learning_rate": 1.0025252525252526e-05, "loss": 0.1023, "step": 790 }, { "epoch": 1.01, "learning_rate": 9.8989898989899e-06, "loss": 0.0791, "step": 800 }, { "epoch": 1.02, "learning_rate": 9.772727272727273e-06, "loss": 0.0656, "step": 810 }, { "epoch": 1.04, "learning_rate": 9.646464646464647e-06, "loss": 0.0656, "step": 820 }, { "epoch": 1.05, "learning_rate": 9.52020202020202e-06, "loss": 0.0741, "step": 830 }, { "epoch": 1.06, "learning_rate": 9.393939393939396e-06, "loss": 0.0925, "step": 840 }, { "epoch": 1.07, "learning_rate": 9.267676767676769e-06, "loss": 0.0753, "step": 850 }, { "epoch": 1.09, "learning_rate": 9.141414141414143e-06, "loss": 0.0735, "step": 860 }, { "epoch": 1.1, "learning_rate": 9.015151515151516e-06, "loss": 0.0843, "step": 870 }, { "epoch": 1.11, "learning_rate": 8.888888888888888e-06, "loss": 0.0641, "step": 880 }, { "epoch": 1.12, "learning_rate": 8.762626262626264e-06, "loss": 0.078, "step": 890 }, { "epoch": 1.14, "learning_rate": 8.636363636363637e-06, "loss": 0.0819, "step": 900 }, { "epoch": 1.15, "learning_rate": 8.510101010101011e-06, "loss": 0.079, "step": 910 }, { "epoch": 1.16, "learning_rate": 8.383838383838384e-06, "loss": 0.0794, "step": 920 }, { "epoch": 1.17, "learning_rate": 8.257575757575758e-06, "loss": 0.0684, "step": 930 }, { "epoch": 1.19, "learning_rate": 8.131313131313133e-06, "loss": 0.0833, "step": 940 }, { "epoch": 1.2, "learning_rate": 8.005050505050505e-06, "loss": 0.079, "step": 950 }, { "epoch": 1.21, "learning_rate": 7.87878787878788e-06, "loss": 0.0675, "step": 960 }, { "epoch": 1.22, "learning_rate": 7.752525252525254e-06, "loss": 0.1101, "step": 970 }, { "epoch": 1.24, "learning_rate": 7.6262626262626275e-06, "loss": 0.072, "step": 980 }, { "epoch": 1.25, "learning_rate": 7.500000000000001e-06, "loss": 0.0754, "step": 990 }, { "epoch": 1.26, "learning_rate": 7.373737373737374e-06, "loss": 0.0807, "step": 1000 }, { "epoch": 1.28, "learning_rate": 7.247474747474747e-06, "loss": 0.0794, "step": 1010 }, { "epoch": 1.29, "learning_rate": 7.121212121212122e-06, "loss": 0.0866, "step": 1020 }, { "epoch": 1.3, "learning_rate": 6.994949494949496e-06, "loss": 0.0901, "step": 1030 }, { "epoch": 1.31, "learning_rate": 6.868686868686869e-06, "loss": 0.0676, "step": 1040 }, { "epoch": 1.33, "learning_rate": 6.742424242424243e-06, "loss": 0.0816, "step": 1050 }, { "epoch": 1.34, "learning_rate": 6.616161616161617e-06, "loss": 0.089, "step": 1060 }, { "epoch": 1.35, "learning_rate": 6.489898989898991e-06, "loss": 0.0831, "step": 1070 }, { "epoch": 1.36, "learning_rate": 6.363636363636364e-06, "loss": 0.0721, "step": 1080 }, { "epoch": 1.38, "learning_rate": 6.237373737373738e-06, "loss": 0.0743, "step": 1090 }, { "epoch": 1.39, "learning_rate": 6.111111111111112e-06, "loss": 0.0833, "step": 1100 }, { "epoch": 1.4, "learning_rate": 5.984848484848486e-06, "loss": 0.1008, "step": 1110 }, { "epoch": 1.41, "learning_rate": 5.858585858585859e-06, "loss": 0.1181, "step": 1120 }, { "epoch": 1.43, "learning_rate": 5.732323232323233e-06, "loss": 0.0634, "step": 1130 }, { "epoch": 1.44, "learning_rate": 5.606060606060606e-06, "loss": 0.0656, "step": 1140 }, { "epoch": 1.45, "learning_rate": 5.4797979797979804e-06, "loss": 0.0797, "step": 1150 }, { "epoch": 1.46, "learning_rate": 5.353535353535354e-06, "loss": 0.0644, "step": 1160 }, { "epoch": 1.48, "learning_rate": 5.2272727272727274e-06, "loss": 0.0605, "step": 1170 }, { "epoch": 1.49, "learning_rate": 5.101010101010101e-06, "loss": 0.0668, "step": 1180 }, { "epoch": 1.5, "learning_rate": 4.974747474747475e-06, "loss": 0.0638, "step": 1190 }, { "epoch": 1.52, "learning_rate": 4.848484848484849e-06, "loss": 0.0797, "step": 1200 }, { "epoch": 1.53, "learning_rate": 4.722222222222222e-06, "loss": 0.0941, "step": 1210 }, { "epoch": 1.54, "learning_rate": 4.595959595959597e-06, "loss": 0.0784, "step": 1220 }, { "epoch": 1.55, "learning_rate": 4.46969696969697e-06, "loss": 0.0561, "step": 1230 }, { "epoch": 1.57, "learning_rate": 4.343434343434344e-06, "loss": 0.0661, "step": 1240 }, { "epoch": 1.58, "learning_rate": 4.217171717171717e-06, "loss": 0.071, "step": 1250 }, { "epoch": 1.59, "learning_rate": 4.0909090909090915e-06, "loss": 0.0702, "step": 1260 }, { "epoch": 1.6, "learning_rate": 3.964646464646465e-06, "loss": 0.0702, "step": 1270 }, { "epoch": 1.62, "learning_rate": 3.8383838383838385e-06, "loss": 0.0795, "step": 1280 }, { "epoch": 1.63, "learning_rate": 3.7121212121212124e-06, "loss": 0.0627, "step": 1290 }, { "epoch": 1.64, "learning_rate": 3.585858585858586e-06, "loss": 0.086, "step": 1300 }, { "epoch": 1.65, "learning_rate": 3.45959595959596e-06, "loss": 0.0567, "step": 1310 }, { "epoch": 1.67, "learning_rate": 3.3333333333333333e-06, "loss": 0.0764, "step": 1320 }, { "epoch": 1.68, "learning_rate": 3.2070707070707072e-06, "loss": 0.0576, "step": 1330 }, { "epoch": 1.69, "learning_rate": 3.0808080808080807e-06, "loss": 0.0783, "step": 1340 }, { "epoch": 1.7, "learning_rate": 2.954545454545455e-06, "loss": 0.0649, "step": 1350 }, { "epoch": 1.72, "learning_rate": 2.8282828282828286e-06, "loss": 0.0756, "step": 1360 }, { "epoch": 1.73, "learning_rate": 2.7020202020202025e-06, "loss": 0.0655, "step": 1370 }, { "epoch": 1.74, "learning_rate": 2.575757575757576e-06, "loss": 0.0664, "step": 1380 }, { "epoch": 1.76, "learning_rate": 2.4494949494949495e-06, "loss": 0.0814, "step": 1390 }, { "epoch": 1.77, "learning_rate": 2.3232323232323234e-06, "loss": 0.0612, "step": 1400 }, { "epoch": 1.78, "learning_rate": 2.196969696969697e-06, "loss": 0.0664, "step": 1410 }, { "epoch": 1.79, "learning_rate": 2.070707070707071e-06, "loss": 0.0484, "step": 1420 }, { "epoch": 1.81, "learning_rate": 1.944444444444445e-06, "loss": 0.0733, "step": 1430 }, { "epoch": 1.82, "learning_rate": 1.8181818181818183e-06, "loss": 0.0644, "step": 1440 }, { "epoch": 1.83, "learning_rate": 1.6919191919191922e-06, "loss": 0.0675, "step": 1450 }, { "epoch": 1.84, "learning_rate": 1.565656565656566e-06, "loss": 0.0749, "step": 1460 }, { "epoch": 1.86, "learning_rate": 1.4393939393939396e-06, "loss": 0.0604, "step": 1470 }, { "epoch": 1.87, "learning_rate": 1.3131313131313134e-06, "loss": 0.074, "step": 1480 }, { "epoch": 1.88, "learning_rate": 1.186868686868687e-06, "loss": 0.0808, "step": 1490 }, { "epoch": 1.89, "learning_rate": 1.0606060606060608e-06, "loss": 0.0812, "step": 1500 } ], "max_steps": 1584, "num_train_epochs": 2, "total_flos": 2785192250966016.0, "trial_name": null, "trial_params": null }