{ "best_metric": null, "best_model_checkpoint": null, "epoch": 4.668374426299499, "global_step": 200000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 5e-06, "loss": 9.1117, "step": 500 }, { "epoch": 0.03, "learning_rate": 1e-05, "loss": 7.3389, "step": 1000 }, { "epoch": 0.04, "learning_rate": 1.5e-05, "loss": 6.8415, "step": 1500 }, { "epoch": 0.05, "learning_rate": 2e-05, "loss": 6.6819, "step": 2000 }, { "epoch": 0.07, "learning_rate": 2.5e-05, "loss": 6.5753, "step": 2500 }, { "epoch": 0.08, "learning_rate": 3e-05, "loss": 6.4883, "step": 3000 }, { "epoch": 0.09, "learning_rate": 3.5e-05, "loss": 6.421, "step": 3500 }, { "epoch": 0.11, "learning_rate": 4e-05, "loss": 6.3655, "step": 4000 }, { "epoch": 0.12, "learning_rate": 4.5e-05, "loss": 6.3167, "step": 4500 }, { "epoch": 0.13, "learning_rate": 5e-05, "loss": 6.2735, "step": 5000 }, { "epoch": 0.15, "learning_rate": 5.500000000000001e-05, "loss": 6.2371, "step": 5500 }, { "epoch": 0.16, "learning_rate": 6e-05, "loss": 6.2028, "step": 6000 }, { "epoch": 0.17, "learning_rate": 6.500000000000001e-05, "loss": 6.1733, "step": 6500 }, { "epoch": 0.19, "learning_rate": 7e-05, "loss": 6.1502, "step": 7000 }, { "epoch": 0.2, "learning_rate": 7.500000000000001e-05, "loss": 6.1308, "step": 7500 }, { "epoch": 0.21, "learning_rate": 8e-05, "loss": 6.1092, "step": 8000 }, { "epoch": 0.23, "learning_rate": 8.5e-05, "loss": 6.0945, "step": 8500 }, { "epoch": 0.24, "learning_rate": 8.999000000000001e-05, "loss": 6.0738, "step": 9000 }, { "epoch": 0.25, "learning_rate": 9.499e-05, "loss": 6.0627, "step": 9500 }, { "epoch": 0.27, "learning_rate": 9.999000000000001e-05, "loss": 6.0486, "step": 10000 }, { "epoch": 0.28, "learning_rate": 9.994959595959596e-05, "loss": 6.0388, "step": 10500 }, { "epoch": 0.29, "learning_rate": 9.989919191919192e-05, "loss": 6.0244, "step": 11000 }, { "epoch": 0.31, "learning_rate": 9.984868686868687e-05, "loss": 6.015, "step": 11500 }, { "epoch": 0.32, "learning_rate": 9.979818181818183e-05, "loss": 6.0047, "step": 12000 }, { "epoch": 0.33, "learning_rate": 9.974767676767676e-05, "loss": 5.9944, "step": 12500 }, { "epoch": 0.35, "learning_rate": 9.969727272727272e-05, "loss": 5.9915, "step": 13000 }, { "epoch": 0.36, "learning_rate": 9.964676767676768e-05, "loss": 5.9777, "step": 13500 }, { "epoch": 0.37, "learning_rate": 9.959626262626263e-05, "loss": 5.976, "step": 14000 }, { "epoch": 0.39, "learning_rate": 9.954575757575758e-05, "loss": 5.9659, "step": 14500 }, { "epoch": 0.4, "learning_rate": 9.949535353535355e-05, "loss": 5.9644, "step": 15000 }, { "epoch": 0.01, "learning_rate": 9.944484848484848e-05, "loss": 5.954, "step": 15500 }, { "epoch": 0.03, "learning_rate": 9.939434343434344e-05, "loss": 5.9498, "step": 16000 }, { "epoch": 0.04, "learning_rate": 9.934383838383839e-05, "loss": 5.946, "step": 16500 }, { "epoch": 0.05, "learning_rate": 9.929343434343435e-05, "loss": 5.941, "step": 17000 }, { "epoch": 0.07, "learning_rate": 9.924292929292929e-05, "loss": 5.9365, "step": 17500 }, { "epoch": 0.08, "learning_rate": 9.919242424242424e-05, "loss": 5.9317, "step": 18000 }, { "epoch": 0.09, "learning_rate": 9.91420202020202e-05, "loss": 5.9262, "step": 18500 }, { "epoch": 0.11, "learning_rate": 9.909151515151516e-05, "loss": 5.9223, "step": 19000 }, { "epoch": 0.12, "learning_rate": 9.90410101010101e-05, "loss": 5.918, "step": 19500 }, { "epoch": 0.13, "learning_rate": 9.899050505050505e-05, "loss": 5.9124, "step": 20000 }, { "epoch": 0.15, "learning_rate": 9.894e-05, "loss": 5.9087, "step": 20500 }, { "epoch": 0.16, "learning_rate": 9.888959595959596e-05, "loss": 5.9039, "step": 21000 }, { "epoch": 0.17, "learning_rate": 9.883909090909092e-05, "loss": 5.8978, "step": 21500 }, { "epoch": 0.19, "learning_rate": 9.878858585858586e-05, "loss": 5.8957, "step": 22000 }, { "epoch": 0.2, "learning_rate": 9.873808080808081e-05, "loss": 5.7647, "step": 22500 }, { "epoch": 0.21, "learning_rate": 9.868757575757577e-05, "loss": 5.5977, "step": 23000 }, { "epoch": 0.23, "learning_rate": 9.863717171717172e-05, "loss": 5.5037, "step": 23500 }, { "epoch": 0.24, "learning_rate": 9.858666666666668e-05, "loss": 5.2742, "step": 24000 }, { "epoch": 0.25, "learning_rate": 9.853616161616162e-05, "loss": 4.8529, "step": 24500 }, { "epoch": 0.27, "learning_rate": 9.848565656565657e-05, "loss": 4.462, "step": 25000 }, { "epoch": 0.28, "learning_rate": 9.843515151515153e-05, "loss": 3.8976, "step": 25500 }, { "epoch": 0.29, "learning_rate": 9.838464646464647e-05, "loss": 3.5281, "step": 26000 }, { "epoch": 0.31, "learning_rate": 9.833424242424243e-05, "loss": 3.3119, "step": 26500 }, { "epoch": 0.32, "learning_rate": 9.82838383838384e-05, "loss": 3.1643, "step": 27000 }, { "epoch": 0.33, "learning_rate": 9.823333333333333e-05, "loss": 3.0554, "step": 27500 }, { "epoch": 0.35, "learning_rate": 9.81829292929293e-05, "loss": 2.9745, "step": 28000 }, { "epoch": 0.36, "learning_rate": 9.813242424242425e-05, "loss": 2.9042, "step": 28500 }, { "epoch": 0.37, "learning_rate": 9.808191919191919e-05, "loss": 2.8514, "step": 29000 }, { "epoch": 0.39, "learning_rate": 9.803141414141415e-05, "loss": 2.7973, "step": 29500 }, { "epoch": 0.4, "learning_rate": 9.79809090909091e-05, "loss": 2.7575, "step": 30000 }, { "epoch": 0.41, "learning_rate": 9.793040404040405e-05, "loss": 2.7213, "step": 30500 }, { "epoch": 0.43, "learning_rate": 9.7879898989899e-05, "loss": 2.6879, "step": 31000 }, { "epoch": 0.44, "learning_rate": 9.782939393939394e-05, "loss": 2.6595, "step": 31500 }, { "epoch": 0.45, "learning_rate": 9.77788888888889e-05, "loss": 2.6326, "step": 32000 }, { "epoch": 0.47, "learning_rate": 9.772848484848486e-05, "loss": 2.6054, "step": 32500 }, { "epoch": 0.48, "learning_rate": 9.76779797979798e-05, "loss": 2.5839, "step": 33000 }, { "epoch": 0.49, "learning_rate": 9.762747474747475e-05, "loss": 2.5596, "step": 33500 }, { "epoch": 0.51, "learning_rate": 9.757707070707072e-05, "loss": 2.5399, "step": 34000 }, { "epoch": 0.52, "learning_rate": 9.752656565656566e-05, "loss": 2.5139, "step": 34500 }, { "epoch": 0.53, "learning_rate": 9.747606060606062e-05, "loss": 2.4959, "step": 35000 }, { "epoch": 0.55, "learning_rate": 9.742555555555556e-05, "loss": 2.4758, "step": 35500 }, { "epoch": 0.56, "learning_rate": 9.737505050505051e-05, "loss": 2.4582, "step": 36000 }, { "epoch": 0.57, "learning_rate": 9.732454545454545e-05, "loss": 2.443, "step": 36500 }, { "epoch": 0.59, "learning_rate": 9.727404040404041e-05, "loss": 2.4262, "step": 37000 }, { "epoch": 0.6, "learning_rate": 9.722353535353536e-05, "loss": 2.4139, "step": 37500 }, { "epoch": 0.61, "learning_rate": 9.717303030303031e-05, "loss": 2.3976, "step": 38000 }, { "epoch": 0.63, "learning_rate": 9.712252525252525e-05, "loss": 2.3856, "step": 38500 }, { "epoch": 0.64, "learning_rate": 9.707202020202021e-05, "loss": 2.3699, "step": 39000 }, { "epoch": 0.65, "learning_rate": 9.702151515151515e-05, "loss": 2.3563, "step": 39500 }, { "epoch": 0.67, "learning_rate": 9.69710101010101e-05, "loss": 2.3455, "step": 40000 }, { "epoch": 0.68, "learning_rate": 9.692050505050505e-05, "loss": 2.3341, "step": 40500 }, { "epoch": 0.69, "learning_rate": 9.687e-05, "loss": 2.3209, "step": 41000 }, { "epoch": 0.71, "learning_rate": 9.681949494949495e-05, "loss": 2.3087, "step": 41500 }, { "epoch": 0.72, "learning_rate": 9.67689898989899e-05, "loss": 2.3016, "step": 42000 }, { "epoch": 0.73, "learning_rate": 9.671848484848485e-05, "loss": 2.2899, "step": 42500 }, { "epoch": 0.75, "learning_rate": 9.66679797979798e-05, "loss": 2.2808, "step": 43000 }, { "epoch": 0.76, "learning_rate": 9.661757575757576e-05, "loss": 2.2732, "step": 43500 }, { "epoch": 0.77, "learning_rate": 9.656707070707072e-05, "loss": 2.2635, "step": 44000 }, { "epoch": 0.79, "learning_rate": 9.651656565656566e-05, "loss": 2.2496, "step": 44500 }, { "epoch": 0.8, "learning_rate": 9.646606060606061e-05, "loss": 2.247, "step": 45000 }, { "epoch": 0.81, "learning_rate": 9.641555555555556e-05, "loss": 2.2398, "step": 45500 }, { "epoch": 0.83, "learning_rate": 9.636505050505052e-05, "loss": 2.2295, "step": 46000 }, { "epoch": 0.84, "learning_rate": 9.631454545454546e-05, "loss": 2.2218, "step": 46500 }, { "epoch": 0.85, "learning_rate": 9.626404040404041e-05, "loss": 2.2126, "step": 47000 }, { "epoch": 0.87, "learning_rate": 9.621373737373738e-05, "loss": 2.2047, "step": 47500 }, { "epoch": 0.88, "learning_rate": 9.616323232323233e-05, "loss": 2.1967, "step": 48000 }, { "epoch": 0.89, "learning_rate": 9.611313131313133e-05, "loss": 2.1927, "step": 48500 }, { "epoch": 0.91, "learning_rate": 9.606262626262626e-05, "loss": 2.187, "step": 49000 }, { "epoch": 0.92, "learning_rate": 9.601212121212122e-05, "loss": 2.1775, "step": 49500 }, { "epoch": 0.93, "learning_rate": 9.596161616161617e-05, "loss": 2.1728, "step": 50000 }, { "epoch": 0.95, "learning_rate": 9.591111111111111e-05, "loss": 2.1627, "step": 50500 }, { "epoch": 0.96, "learning_rate": 9.586060606060606e-05, "loss": 2.1593, "step": 51000 }, { "epoch": 0.97, "learning_rate": 9.581010101010102e-05, "loss": 2.1523, "step": 51500 }, { "epoch": 0.99, "learning_rate": 9.575959595959597e-05, "loss": 2.1481, "step": 52000 }, { "epoch": 1.0, "learning_rate": 9.570909090909091e-05, "loss": 2.1405, "step": 52500 }, { "epoch": 1.01, "learning_rate": 9.565858585858586e-05, "loss": 2.1305, "step": 53000 }, { "epoch": 1.03, "learning_rate": 9.560808080808082e-05, "loss": 2.1286, "step": 53500 }, { "epoch": 1.04, "learning_rate": 9.555757575757575e-05, "loss": 2.1233, "step": 54000 }, { "epoch": 1.05, "learning_rate": 9.550707070707071e-05, "loss": 2.1175, "step": 54500 }, { "epoch": 1.07, "learning_rate": 9.545656565656567e-05, "loss": 2.1127, "step": 55000 }, { "epoch": 1.08, "learning_rate": 9.54060606060606e-05, "loss": 2.1042, "step": 55500 }, { "epoch": 1.09, "learning_rate": 9.535555555555556e-05, "loss": 2.1054, "step": 56000 }, { "epoch": 1.11, "learning_rate": 9.530505050505051e-05, "loss": 2.099, "step": 56500 }, { "epoch": 1.12, "learning_rate": 9.525454545454547e-05, "loss": 2.096, "step": 57000 }, { "epoch": 1.13, "learning_rate": 9.52040404040404e-05, "loss": 2.0863, "step": 57500 }, { "epoch": 1.15, "learning_rate": 9.515353535353536e-05, "loss": 2.0835, "step": 58000 }, { "epoch": 1.16, "learning_rate": 9.510303030303031e-05, "loss": 2.0819, "step": 58500 }, { "epoch": 1.17, "learning_rate": 9.505252525252526e-05, "loss": 2.0762, "step": 59000 }, { "epoch": 1.19, "learning_rate": 9.500212121212121e-05, "loss": 2.0713, "step": 59500 }, { "epoch": 1.2, "learning_rate": 9.495161616161616e-05, "loss": 2.068, "step": 60000 }, { "epoch": 1.01, "learning_rate": 9.490111111111112e-05, "loss": 2.0617, "step": 60500 }, { "epoch": 1.03, "learning_rate": 9.485060606060607e-05, "loss": 2.062, "step": 61000 }, { "epoch": 1.04, "learning_rate": 9.480010101010101e-05, "loss": 2.0536, "step": 61500 }, { "epoch": 1.05, "learning_rate": 9.474959595959596e-05, "loss": 2.0479, "step": 62000 }, { "epoch": 1.07, "learning_rate": 9.469909090909092e-05, "loss": 2.0468, "step": 62500 }, { "epoch": 1.08, "learning_rate": 9.464858585858585e-05, "loss": 2.0424, "step": 63000 }, { "epoch": 1.09, "learning_rate": 9.459818181818183e-05, "loss": 2.0416, "step": 63500 }, { "epoch": 1.11, "learning_rate": 9.454777777777779e-05, "loss": 2.0356, "step": 64000 }, { "epoch": 1.12, "learning_rate": 9.449727272727273e-05, "loss": 2.0325, "step": 64500 }, { "epoch": 1.13, "learning_rate": 9.444676767676768e-05, "loss": 2.0273, "step": 65000 }, { "epoch": 1.15, "learning_rate": 9.439626262626264e-05, "loss": 2.0243, "step": 65500 }, { "epoch": 1.16, "learning_rate": 9.434575757575757e-05, "loss": 2.0211, "step": 66000 }, { "epoch": 1.17, "learning_rate": 9.429525252525253e-05, "loss": 2.0163, "step": 66500 }, { "epoch": 1.19, "learning_rate": 9.424484848484849e-05, "loss": 2.0159, "step": 67000 }, { "epoch": 1.2, "learning_rate": 9.419434343434344e-05, "loss": 2.0124, "step": 67500 }, { "epoch": 1.21, "learning_rate": 9.41438383838384e-05, "loss": 2.0099, "step": 68000 }, { "epoch": 1.23, "learning_rate": 9.409333333333333e-05, "loss": 2.0052, "step": 68500 }, { "epoch": 1.24, "learning_rate": 9.404282828282829e-05, "loss": 2.0026, "step": 69000 }, { "epoch": 1.25, "learning_rate": 9.399232323232324e-05, "loss": 1.9975, "step": 69500 }, { "epoch": 1.27, "learning_rate": 9.394181818181818e-05, "loss": 1.9983, "step": 70000 }, { "epoch": 1.28, "learning_rate": 9.389131313131314e-05, "loss": 1.9933, "step": 70500 }, { "epoch": 1.29, "learning_rate": 9.384090909090909e-05, "loss": 1.9904, "step": 71000 }, { "epoch": 1.31, "learning_rate": 9.379040404040405e-05, "loss": 1.9854, "step": 71500 }, { "epoch": 1.32, "learning_rate": 9.3739898989899e-05, "loss": 1.9836, "step": 72000 }, { "epoch": 1.33, "learning_rate": 9.368939393939394e-05, "loss": 1.9802, "step": 72500 }, { "epoch": 1.35, "learning_rate": 9.36389898989899e-05, "loss": 1.9806, "step": 73000 }, { "epoch": 1.36, "learning_rate": 9.358858585858586e-05, "loss": 1.9775, "step": 73500 }, { "epoch": 1.37, "learning_rate": 9.35380808080808e-05, "loss": 1.9737, "step": 74000 }, { "epoch": 1.39, "learning_rate": 9.348757575757577e-05, "loss": 1.9708, "step": 74500 }, { "epoch": 1.4, "learning_rate": 9.343707070707071e-05, "loss": 1.9689, "step": 75000 }, { "epoch": 1.41, "learning_rate": 9.338656565656566e-05, "loss": 1.9649, "step": 75500 }, { "epoch": 1.43, "learning_rate": 9.333606060606062e-05, "loss": 1.9629, "step": 76000 }, { "epoch": 1.44, "learning_rate": 9.328555555555556e-05, "loss": 1.9583, "step": 76500 }, { "epoch": 1.45, "learning_rate": 9.323505050505051e-05, "loss": 1.957, "step": 77000 }, { "epoch": 1.47, "learning_rate": 9.318464646464647e-05, "loss": 1.9549, "step": 77500 }, { "epoch": 1.48, "learning_rate": 9.313414141414142e-05, "loss": 1.9576, "step": 78000 }, { "epoch": 1.49, "learning_rate": 9.308363636363638e-05, "loss": 1.9471, "step": 78500 }, { "epoch": 1.51, "learning_rate": 9.303313131313132e-05, "loss": 1.9478, "step": 79000 }, { "epoch": 1.52, "learning_rate": 9.298262626262627e-05, "loss": 1.9437, "step": 79500 }, { "epoch": 1.53, "learning_rate": 9.293212121212122e-05, "loss": 1.9433, "step": 80000 }, { "epoch": 1.55, "learning_rate": 9.288161616161618e-05, "loss": 1.9401, "step": 80500 }, { "epoch": 1.56, "learning_rate": 9.283111111111111e-05, "loss": 1.9401, "step": 81000 }, { "epoch": 1.57, "learning_rate": 9.278070707070707e-05, "loss": 1.9385, "step": 81500 }, { "epoch": 1.59, "learning_rate": 9.273020202020203e-05, "loss": 1.9356, "step": 82000 }, { "epoch": 1.6, "learning_rate": 9.267969696969697e-05, "loss": 1.9305, "step": 82500 }, { "epoch": 1.61, "learning_rate": 9.262919191919192e-05, "loss": 1.9306, "step": 83000 }, { "epoch": 1.63, "learning_rate": 9.257868686868687e-05, "loss": 1.9265, "step": 83500 }, { "epoch": 1.64, "learning_rate": 9.252818181818183e-05, "loss": 1.9266, "step": 84000 }, { "epoch": 1.65, "learning_rate": 9.247777777777779e-05, "loss": 1.922, "step": 84500 }, { "epoch": 1.67, "learning_rate": 9.242727272727273e-05, "loss": 1.924, "step": 85000 }, { "epoch": 1.68, "learning_rate": 9.237676767676768e-05, "loss": 1.9211, "step": 85500 }, { "epoch": 1.69, "learning_rate": 9.232626262626263e-05, "loss": 1.9181, "step": 86000 }, { "epoch": 1.71, "learning_rate": 9.227585858585859e-05, "loss": 1.9191, "step": 86500 }, { "epoch": 1.72, "learning_rate": 9.222535353535354e-05, "loss": 1.9154, "step": 87000 }, { "epoch": 1.73, "learning_rate": 9.217484848484849e-05, "loss": 1.9104, "step": 87500 }, { "epoch": 1.75, "learning_rate": 9.212434343434344e-05, "loss": 1.9092, "step": 88000 }, { "epoch": 1.76, "learning_rate": 9.20739393939394e-05, "loss": 1.9071, "step": 88500 }, { "epoch": 1.77, "learning_rate": 9.202343434343434e-05, "loss": 1.9066, "step": 89000 }, { "epoch": 1.79, "learning_rate": 9.19729292929293e-05, "loss": 1.9048, "step": 89500 }, { "epoch": 1.8, "learning_rate": 9.192242424242425e-05, "loss": 1.9006, "step": 90000 }, { "epoch": 1.81, "learning_rate": 9.18719191919192e-05, "loss": 1.9012, "step": 90500 }, { "epoch": 1.83, "learning_rate": 9.182151515151516e-05, "loss": 1.8953, "step": 91000 }, { "epoch": 1.84, "learning_rate": 9.17710101010101e-05, "loss": 1.8962, "step": 91500 }, { "epoch": 1.85, "learning_rate": 9.172050505050506e-05, "loss": 1.8991, "step": 92000 }, { "epoch": 1.87, "learning_rate": 9.167e-05, "loss": 1.8911, "step": 92500 }, { "epoch": 1.88, "learning_rate": 9.161949494949495e-05, "loss": 1.8945, "step": 93000 }, { "epoch": 1.89, "learning_rate": 9.15689898989899e-05, "loss": 1.8904, "step": 93500 }, { "epoch": 1.91, "learning_rate": 9.151848484848485e-05, "loss": 1.8859, "step": 94000 }, { "epoch": 1.92, "learning_rate": 9.14679797979798e-05, "loss": 1.8836, "step": 94500 }, { "epoch": 1.93, "learning_rate": 9.141747474747475e-05, "loss": 1.88, "step": 95000 }, { "epoch": 1.95, "learning_rate": 9.13669696969697e-05, "loss": 1.8834, "step": 95500 }, { "epoch": 1.96, "learning_rate": 9.131656565656566e-05, "loss": 1.8813, "step": 96000 }, { "epoch": 1.97, "learning_rate": 9.126616161616162e-05, "loss": 1.881, "step": 96500 }, { "epoch": 1.99, "learning_rate": 9.121575757575758e-05, "loss": 1.8785, "step": 97000 }, { "epoch": 2.0, "learning_rate": 9.116525252525252e-05, "loss": 1.8735, "step": 97500 }, { "epoch": 2.01, "learning_rate": 9.111474747474747e-05, "loss": 1.8707, "step": 98000 }, { "epoch": 2.03, "learning_rate": 9.106424242424243e-05, "loss": 1.8702, "step": 98500 }, { "epoch": 2.04, "learning_rate": 9.101373737373738e-05, "loss": 1.8696, "step": 99000 }, { "epoch": 2.05, "learning_rate": 9.096323232323232e-05, "loss": 1.8667, "step": 99500 }, { "epoch": 2.07, "learning_rate": 9.091272727272727e-05, "loss": 1.8645, "step": 100000 }, { "epoch": 2.01, "learning_rate": 9.086222222222223e-05, "loss": 1.8615, "step": 100500 }, { "epoch": 2.03, "learning_rate": 9.081171717171718e-05, "loss": 1.8605, "step": 101000 }, { "epoch": 2.04, "learning_rate": 9.076131313131314e-05, "loss": 1.8579, "step": 101500 }, { "epoch": 2.05, "learning_rate": 9.071080808080808e-05, "loss": 1.8513, "step": 102000 }, { "epoch": 2.07, "learning_rate": 9.066030303030304e-05, "loss": 1.854, "step": 102500 }, { "epoch": 2.08, "learning_rate": 9.060979797979799e-05, "loss": 1.8577, "step": 103000 }, { "epoch": 2.09, "learning_rate": 9.055929292929294e-05, "loss": 1.8557, "step": 103500 }, { "epoch": 2.11, "learning_rate": 9.05088888888889e-05, "loss": 1.854, "step": 104000 }, { "epoch": 2.12, "learning_rate": 9.045838383838384e-05, "loss": 1.8532, "step": 104500 }, { "epoch": 2.13, "learning_rate": 9.04078787878788e-05, "loss": 1.8507, "step": 105000 }, { "epoch": 2.15, "learning_rate": 9.035737373737375e-05, "loss": 1.8485, "step": 105500 }, { "epoch": 2.16, "learning_rate": 9.03069696969697e-05, "loss": 1.8498, "step": 106000 }, { "epoch": 2.17, "learning_rate": 9.025646464646465e-05, "loss": 1.8447, "step": 106500 }, { "epoch": 2.19, "learning_rate": 9.02059595959596e-05, "loss": 1.8434, "step": 107000 }, { "epoch": 2.2, "learning_rate": 9.015545454545456e-05, "loss": 1.8453, "step": 107500 }, { "epoch": 2.21, "learning_rate": 9.010505050505052e-05, "loss": 1.8461, "step": 108000 }, { "epoch": 2.23, "learning_rate": 9.005454545454545e-05, "loss": 1.8435, "step": 108500 }, { "epoch": 2.24, "learning_rate": 9.000404040404041e-05, "loss": 1.8408, "step": 109000 }, { "epoch": 2.25, "learning_rate": 8.995353535353536e-05, "loss": 1.8387, "step": 109500 }, { "epoch": 2.27, "learning_rate": 8.990303030303032e-05, "loss": 1.8399, "step": 110000 }, { "epoch": 2.28, "learning_rate": 8.985252525252525e-05, "loss": 1.838, "step": 110500 }, { "epoch": 2.29, "learning_rate": 8.980212121212121e-05, "loss": 1.8371, "step": 111000 }, { "epoch": 2.31, "learning_rate": 8.975161616161617e-05, "loss": 1.8337, "step": 111500 }, { "epoch": 2.32, "learning_rate": 8.970111111111112e-05, "loss": 1.8354, "step": 112000 }, { "epoch": 2.33, "learning_rate": 8.965060606060606e-05, "loss": 1.8291, "step": 112500 }, { "epoch": 2.35, "learning_rate": 8.960010101010101e-05, "loss": 1.8313, "step": 113000 }, { "epoch": 2.36, "learning_rate": 8.954959595959597e-05, "loss": 1.8295, "step": 113500 }, { "epoch": 2.37, "learning_rate": 8.949909090909092e-05, "loss": 1.8289, "step": 114000 }, { "epoch": 2.39, "learning_rate": 8.944858585858586e-05, "loss": 1.8248, "step": 114500 }, { "epoch": 2.4, "learning_rate": 8.939818181818182e-05, "loss": 1.8248, "step": 115000 }, { "epoch": 2.41, "learning_rate": 8.934767676767677e-05, "loss": 1.8244, "step": 115500 }, { "epoch": 2.43, "learning_rate": 8.929717171717173e-05, "loss": 1.8228, "step": 116000 }, { "epoch": 2.44, "learning_rate": 8.924666666666667e-05, "loss": 1.8262, "step": 116500 }, { "epoch": 2.45, "learning_rate": 8.919616161616162e-05, "loss": 1.8237, "step": 117000 }, { "epoch": 2.47, "learning_rate": 8.914565656565657e-05, "loss": 1.8182, "step": 117500 }, { "epoch": 2.48, "learning_rate": 8.909515151515153e-05, "loss": 1.8198, "step": 118000 }, { "epoch": 2.49, "learning_rate": 8.904464646464646e-05, "loss": 1.8176, "step": 118500 }, { "epoch": 2.51, "learning_rate": 8.899424242424242e-05, "loss": 1.8169, "step": 119000 }, { "epoch": 2.52, "learning_rate": 8.894373737373738e-05, "loss": 1.8131, "step": 119500 }, { "epoch": 2.53, "learning_rate": 8.889323232323233e-05, "loss": 1.812, "step": 120000 }, { "epoch": 2.55, "learning_rate": 8.884272727272727e-05, "loss": 1.8113, "step": 120500 }, { "epoch": 2.56, "learning_rate": 8.879222222222222e-05, "loss": 1.8122, "step": 121000 }, { "epoch": 2.57, "learning_rate": 8.87420202020202e-05, "loss": 1.8105, "step": 121500 }, { "epoch": 2.59, "learning_rate": 8.869151515151516e-05, "loss": 1.8107, "step": 122000 }, { "epoch": 2.6, "learning_rate": 8.86410101010101e-05, "loss": 1.8084, "step": 122500 }, { "epoch": 2.61, "learning_rate": 8.859050505050506e-05, "loss": 1.8092, "step": 123000 }, { "epoch": 2.63, "learning_rate": 8.854e-05, "loss": 1.8055, "step": 123500 }, { "epoch": 2.64, "learning_rate": 8.848949494949495e-05, "loss": 1.8041, "step": 124000 }, { "epoch": 2.65, "learning_rate": 8.84389898989899e-05, "loss": 1.8036, "step": 124500 }, { "epoch": 2.67, "learning_rate": 8.838858585858585e-05, "loss": 1.8059, "step": 125000 }, { "epoch": 2.68, "learning_rate": 8.833808080808081e-05, "loss": 1.8004, "step": 125500 }, { "epoch": 2.69, "learning_rate": 8.828757575757576e-05, "loss": 1.8008, "step": 126000 }, { "epoch": 2.71, "learning_rate": 8.823707070707071e-05, "loss": 1.8048, "step": 126500 }, { "epoch": 2.72, "learning_rate": 8.818656565656565e-05, "loss": 1.7992, "step": 127000 }, { "epoch": 2.73, "learning_rate": 8.813606060606061e-05, "loss": 1.797, "step": 127500 }, { "epoch": 2.75, "learning_rate": 8.808555555555556e-05, "loss": 1.7949, "step": 128000 }, { "epoch": 2.76, "learning_rate": 8.80350505050505e-05, "loss": 1.7984, "step": 128500 }, { "epoch": 2.77, "learning_rate": 8.798454545454547e-05, "loss": 1.7954, "step": 129000 }, { "epoch": 2.79, "learning_rate": 8.793414141414141e-05, "loss": 1.7979, "step": 129500 }, { "epoch": 2.8, "learning_rate": 8.788363636363637e-05, "loss": 1.795, "step": 130000 }, { "epoch": 2.81, "learning_rate": 8.783313131313132e-05, "loss": 1.7944, "step": 130500 }, { "epoch": 2.83, "learning_rate": 8.778262626262627e-05, "loss": 1.7907, "step": 131000 }, { "epoch": 2.84, "learning_rate": 8.773212121212122e-05, "loss": 1.7912, "step": 131500 }, { "epoch": 2.85, "learning_rate": 8.768171717171717e-05, "loss": 1.7935, "step": 132000 }, { "epoch": 2.87, "learning_rate": 8.763121212121213e-05, "loss": 1.79, "step": 132500 }, { "epoch": 2.88, "learning_rate": 8.758070707070708e-05, "loss": 1.7892, "step": 133000 }, { "epoch": 2.89, "learning_rate": 8.753020202020202e-05, "loss": 1.787, "step": 133500 }, { "epoch": 2.91, "learning_rate": 8.747969696969698e-05, "loss": 1.7887, "step": 134000 }, { "epoch": 2.92, "learning_rate": 8.742919191919192e-05, "loss": 1.7821, "step": 134500 }, { "epoch": 2.93, "learning_rate": 8.737878787878788e-05, "loss": 1.7862, "step": 135000 }, { "epoch": 2.95, "learning_rate": 8.732838383838385e-05, "loss": 1.7833, "step": 135500 }, { "epoch": 2.96, "learning_rate": 8.72778787878788e-05, "loss": 1.7795, "step": 136000 }, { "epoch": 2.97, "learning_rate": 8.722737373737374e-05, "loss": 1.7825, "step": 136500 }, { "epoch": 2.99, "learning_rate": 8.71768686868687e-05, "loss": 1.7804, "step": 137000 }, { "epoch": 3.0, "learning_rate": 8.712636363636363e-05, "loss": 1.7818, "step": 137500 }, { "epoch": 3.01, "learning_rate": 8.70758585858586e-05, "loss": 1.7751, "step": 138000 }, { "epoch": 3.03, "learning_rate": 8.702535353535354e-05, "loss": 1.7759, "step": 138500 }, { "epoch": 3.04, "learning_rate": 8.697484848484849e-05, "loss": 1.7723, "step": 139000 }, { "epoch": 3.05, "learning_rate": 8.692434343434343e-05, "loss": 1.7776, "step": 139500 }, { "epoch": 3.07, "learning_rate": 8.687393939393939e-05, "loss": 1.772, "step": 140000 }, { "epoch": 3.08, "learning_rate": 8.682343434343435e-05, "loss": 1.7723, "step": 140500 }, { "epoch": 3.09, "learning_rate": 8.67729292929293e-05, "loss": 1.773, "step": 141000 }, { "epoch": 3.11, "learning_rate": 8.672242424242425e-05, "loss": 1.7728, "step": 141500 }, { "epoch": 3.12, "learning_rate": 8.667191919191919e-05, "loss": 1.7723, "step": 142000 }, { "epoch": 3.13, "learning_rate": 8.662151515151515e-05, "loss": 1.7704, "step": 142500 }, { "epoch": 3.15, "learning_rate": 8.657101010101011e-05, "loss": 1.7684, "step": 143000 }, { "epoch": 3.16, "learning_rate": 8.652050505050506e-05, "loss": 1.7666, "step": 143500 }, { "epoch": 3.17, "learning_rate": 8.647e-05, "loss": 1.7679, "step": 144000 }, { "epoch": 3.19, "learning_rate": 8.641949494949495e-05, "loss": 1.7665, "step": 144500 }, { "epoch": 3.2, "learning_rate": 8.636909090909091e-05, "loss": 1.7651, "step": 145000 }, { "epoch": 3.21, "learning_rate": 8.631858585858587e-05, "loss": 1.7629, "step": 145500 }, { "epoch": 3.23, "learning_rate": 8.62680808080808e-05, "loss": 1.7663, "step": 146000 }, { "epoch": 3.24, "learning_rate": 8.621757575757576e-05, "loss": 1.762, "step": 146500 }, { "epoch": 3.25, "learning_rate": 8.616717171717172e-05, "loss": 1.7659, "step": 147000 }, { "epoch": 3.27, "learning_rate": 8.611666666666667e-05, "loss": 1.763, "step": 147500 }, { "epoch": 3.28, "learning_rate": 8.606626262626263e-05, "loss": 1.7618, "step": 148000 }, { "epoch": 3.29, "learning_rate": 8.601575757575759e-05, "loss": 1.7611, "step": 148500 }, { "epoch": 3.31, "learning_rate": 8.596525252525252e-05, "loss": 1.7598, "step": 149000 }, { "epoch": 3.32, "learning_rate": 8.591474747474748e-05, "loss": 1.7596, "step": 149500 }, { "epoch": 3.33, "learning_rate": 8.586424242424243e-05, "loss": 1.7591, "step": 150000 }, { "epoch": 3.35, "learning_rate": 8.581373737373737e-05, "loss": 1.7579, "step": 150500 }, { "epoch": 3.36, "learning_rate": 8.576323232323232e-05, "loss": 1.7576, "step": 151000 }, { "epoch": 3.37, "learning_rate": 8.571272727272728e-05, "loss": 1.7565, "step": 151500 }, { "epoch": 3.39, "learning_rate": 8.566222222222223e-05, "loss": 1.7553, "step": 152000 }, { "epoch": 3.4, "learning_rate": 8.561171717171717e-05, "loss": 1.7523, "step": 152500 }, { "epoch": 3.41, "learning_rate": 8.556131313131313e-05, "loss": 1.7522, "step": 153000 }, { "epoch": 3.43, "learning_rate": 8.551080808080808e-05, "loss": 1.7551, "step": 153500 }, { "epoch": 3.44, "learning_rate": 8.546030303030304e-05, "loss": 1.7525, "step": 154000 }, { "epoch": 3.45, "learning_rate": 8.540979797979798e-05, "loss": 1.7523, "step": 154500 }, { "epoch": 3.47, "learning_rate": 8.535929292929293e-05, "loss": 1.7543, "step": 155000 }, { "epoch": 3.48, "learning_rate": 8.530888888888889e-05, "loss": 1.7491, "step": 155500 }, { "epoch": 3.49, "learning_rate": 8.525848484848485e-05, "loss": 1.7486, "step": 156000 }, { "epoch": 3.51, "learning_rate": 8.52079797979798e-05, "loss": 1.7521, "step": 156500 }, { "epoch": 3.52, "learning_rate": 8.515747474747476e-05, "loss": 1.7481, "step": 157000 }, { "epoch": 3.53, "learning_rate": 8.51069696969697e-05, "loss": 1.7482, "step": 157500 }, { "epoch": 3.55, "learning_rate": 8.505646464646465e-05, "loss": 1.7481, "step": 158000 }, { "epoch": 3.56, "learning_rate": 8.500595959595961e-05, "loss": 1.7466, "step": 158500 }, { "epoch": 3.57, "learning_rate": 8.495545454545455e-05, "loss": 1.7479, "step": 159000 }, { "epoch": 3.59, "learning_rate": 8.49049494949495e-05, "loss": 1.7431, "step": 159500 }, { "epoch": 3.6, "learning_rate": 8.485444444444445e-05, "loss": 1.7464, "step": 160000 }, { "epoch": 3.61, "learning_rate": 8.480404040404041e-05, "loss": 1.7433, "step": 160500 }, { "epoch": 3.63, "learning_rate": 8.475363636363637e-05, "loss": 1.7421, "step": 161000 }, { "epoch": 3.64, "learning_rate": 8.470313131313131e-05, "loss": 1.7432, "step": 161500 }, { "epoch": 3.65, "learning_rate": 8.465262626262627e-05, "loss": 1.7413, "step": 162000 }, { "epoch": 3.67, "learning_rate": 8.460212121212122e-05, "loss": 1.7415, "step": 162500 }, { "epoch": 3.68, "learning_rate": 8.455161616161617e-05, "loss": 1.7388, "step": 163000 }, { "epoch": 3.69, "learning_rate": 8.450111111111113e-05, "loss": 1.74, "step": 163500 }, { "epoch": 3.71, "learning_rate": 8.445070707070709e-05, "loss": 1.7387, "step": 164000 }, { "epoch": 3.72, "learning_rate": 8.440020202020202e-05, "loss": 1.7392, "step": 164500 }, { "epoch": 3.73, "learning_rate": 8.434969696969698e-05, "loss": 1.7391, "step": 165000 }, { "epoch": 3.75, "learning_rate": 8.429919191919192e-05, "loss": 1.7389, "step": 165500 }, { "epoch": 3.76, "learning_rate": 8.424868686868687e-05, "loss": 1.7355, "step": 166000 }, { "epoch": 3.77, "learning_rate": 8.419818181818182e-05, "loss": 1.736, "step": 166500 }, { "epoch": 3.79, "learning_rate": 8.414767676767678e-05, "loss": 1.734, "step": 167000 }, { "epoch": 3.8, "learning_rate": 8.409717171717172e-05, "loss": 1.7356, "step": 167500 }, { "epoch": 3.81, "learning_rate": 8.404676767676768e-05, "loss": 1.7363, "step": 168000 }, { "epoch": 3.83, "learning_rate": 8.399626262626263e-05, "loss": 1.7342, "step": 168500 }, { "epoch": 3.84, "learning_rate": 8.394575757575758e-05, "loss": 1.7305, "step": 169000 }, { "epoch": 3.85, "learning_rate": 8.389525252525254e-05, "loss": 1.7314, "step": 169500 }, { "epoch": 3.87, "learning_rate": 8.384474747474748e-05, "loss": 1.729, "step": 170000 }, { "epoch": 3.88, "learning_rate": 8.379424242424243e-05, "loss": 1.7323, "step": 170500 }, { "epoch": 3.89, "learning_rate": 8.374383838383839e-05, "loss": 1.7309, "step": 171000 }, { "epoch": 3.91, "learning_rate": 8.369333333333333e-05, "loss": 1.7317, "step": 171500 }, { "epoch": 3.92, "learning_rate": 8.36428282828283e-05, "loss": 1.729, "step": 172000 }, { "epoch": 3.93, "learning_rate": 8.359232323232323e-05, "loss": 1.7302, "step": 172500 }, { "epoch": 3.95, "learning_rate": 8.354181818181819e-05, "loss": 1.7301, "step": 173000 }, { "epoch": 3.96, "learning_rate": 8.349131313131313e-05, "loss": 1.7267, "step": 173500 }, { "epoch": 3.97, "learning_rate": 8.344080808080809e-05, "loss": 1.7279, "step": 174000 }, { "epoch": 3.99, "learning_rate": 8.339030303030303e-05, "loss": 1.7258, "step": 174500 }, { "epoch": 4.0, "learning_rate": 8.333979797979799e-05, "loss": 1.7249, "step": 175000 }, { "epoch": 4.01, "learning_rate": 8.328939393939394e-05, "loss": 1.7266, "step": 175500 }, { "epoch": 4.03, "learning_rate": 8.32389898989899e-05, "loss": 1.7213, "step": 176000 }, { "epoch": 4.04, "learning_rate": 8.318848484848485e-05, "loss": 1.7227, "step": 176500 }, { "epoch": 4.05, "learning_rate": 8.31379797979798e-05, "loss": 1.7208, "step": 177000 }, { "epoch": 4.07, "learning_rate": 8.308747474747474e-05, "loss": 1.7235, "step": 177500 }, { "epoch": 4.08, "learning_rate": 8.30370707070707e-05, "loss": 1.7204, "step": 178000 }, { "epoch": 4.09, "learning_rate": 8.298656565656566e-05, "loss": 1.7214, "step": 178500 }, { "epoch": 4.11, "learning_rate": 8.293606060606061e-05, "loss": 1.7182, "step": 179000 }, { "epoch": 4.12, "learning_rate": 8.288555555555556e-05, "loss": 1.7178, "step": 179500 }, { "epoch": 4.13, "learning_rate": 8.28350505050505e-05, "loss": 1.7176, "step": 180000 }, { "epoch": 4.15, "learning_rate": 8.278454545454546e-05, "loss": 1.7194, "step": 180500 }, { "epoch": 4.16, "learning_rate": 8.273414141414142e-05, "loss": 1.7195, "step": 181000 }, { "epoch": 4.17, "learning_rate": 8.268373737373738e-05, "loss": 1.7167, "step": 181500 }, { "epoch": 4.19, "learning_rate": 8.263323232323233e-05, "loss": 1.7161, "step": 182000 }, { "epoch": 4.2, "learning_rate": 8.258272727272727e-05, "loss": 1.7162, "step": 182500 }, { "epoch": 4.21, "learning_rate": 8.253222222222222e-05, "loss": 1.716, "step": 183000 }, { "epoch": 4.23, "learning_rate": 8.248171717171718e-05, "loss": 1.7145, "step": 183500 }, { "epoch": 4.24, "learning_rate": 8.243131313131314e-05, "loss": 1.7113, "step": 184000 }, { "epoch": 4.25, "learning_rate": 8.238080808080809e-05, "loss": 1.7151, "step": 184500 }, { "epoch": 4.27, "learning_rate": 8.233030303030303e-05, "loss": 1.7118, "step": 185000 }, { "epoch": 4.28, "learning_rate": 8.227979797979798e-05, "loss": 1.7147, "step": 185500 }, { "epoch": 4.29, "learning_rate": 8.222929292929294e-05, "loss": 1.7144, "step": 186000 }, { "epoch": 4.31, "learning_rate": 8.217878787878787e-05, "loss": 1.7127, "step": 186500 }, { "epoch": 4.32, "learning_rate": 8.212828282828283e-05, "loss": 1.7103, "step": 187000 }, { "epoch": 4.33, "learning_rate": 8.207777777777779e-05, "loss": 1.7132, "step": 187500 }, { "epoch": 4.35, "learning_rate": 8.202737373737374e-05, "loss": 1.7088, "step": 188000 }, { "epoch": 4.36, "learning_rate": 8.19768686868687e-05, "loss": 1.7088, "step": 188500 }, { "epoch": 4.37, "learning_rate": 8.192636363636364e-05, "loss": 1.7094, "step": 189000 }, { "epoch": 4.39, "learning_rate": 8.187585858585859e-05, "loss": 1.709, "step": 189500 }, { "epoch": 4.4, "learning_rate": 8.182535353535355e-05, "loss": 1.7098, "step": 190000 }, { "epoch": 4.41, "learning_rate": 8.177505050505051e-05, "loss": 1.7083, "step": 190500 }, { "epoch": 4.43, "learning_rate": 8.172454545454546e-05, "loss": 1.7063, "step": 191000 }, { "epoch": 4.44, "learning_rate": 8.167404040404042e-05, "loss": 1.7083, "step": 191500 }, { "epoch": 4.45, "learning_rate": 8.162353535353535e-05, "loss": 1.7079, "step": 192000 }, { "epoch": 4.47, "learning_rate": 8.157303030303031e-05, "loss": 1.7069, "step": 192500 }, { "epoch": 4.48, "learning_rate": 8.152252525252527e-05, "loss": 1.7064, "step": 193000 }, { "epoch": 4.49, "learning_rate": 8.14720202020202e-05, "loss": 1.7037, "step": 193500 }, { "epoch": 4.51, "learning_rate": 8.142151515151516e-05, "loss": 1.7043, "step": 194000 }, { "epoch": 4.52, "learning_rate": 8.13710101010101e-05, "loss": 1.7044, "step": 194500 }, { "epoch": 4.53, "learning_rate": 8.132050505050505e-05, "loss": 1.7019, "step": 195000 }, { "epoch": 4.55, "learning_rate": 8.127010101010101e-05, "loss": 1.7043, "step": 195500 }, { "epoch": 4.56, "learning_rate": 8.121959595959596e-05, "loss": 1.7025, "step": 196000 }, { "epoch": 4.57, "learning_rate": 8.116909090909092e-05, "loss": 1.7028, "step": 196500 }, { "epoch": 4.59, "learning_rate": 8.111858585858587e-05, "loss": 1.6991, "step": 197000 }, { "epoch": 4.6, "learning_rate": 8.106808080808081e-05, "loss": 1.7023, "step": 197500 }, { "epoch": 4.62, "learning_rate": 8.101757575757576e-05, "loss": 1.7004, "step": 198000 }, { "epoch": 4.63, "learning_rate": 8.096707070707072e-05, "loss": 1.7012, "step": 198500 }, { "epoch": 4.64, "learning_rate": 8.091656565656565e-05, "loss": 1.6983, "step": 199000 }, { "epoch": 4.66, "learning_rate": 8.086616161616162e-05, "loss": 1.7011, "step": 199500 }, { "epoch": 4.67, "learning_rate": 8.081565656565657e-05, "loss": 1.6972, "step": 200000 } ], "max_steps": 200000, "num_train_epochs": 6, "total_flos": 1.3476043989572387e+19, "trial_name": null, "trial_params": null }