{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 205, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 2.857142857142857e-08, "loss": 1.2934, "step": 1 }, { "epoch": 0.01, "learning_rate": 5.714285714285714e-08, "loss": 1.2798, "step": 2 }, { "epoch": 0.01, "learning_rate": 8.57142857142857e-08, "loss": 1.2641, "step": 3 }, { "epoch": 0.02, "learning_rate": 1.1428571428571427e-07, "loss": 1.2162, "step": 4 }, { "epoch": 0.02, "learning_rate": 1.4285714285714285e-07, "loss": 1.3073, "step": 5 }, { "epoch": 0.03, "learning_rate": 1.714285714285714e-07, "loss": 1.3019, "step": 6 }, { "epoch": 0.03, "learning_rate": 2e-07, "loss": 1.2939, "step": 7 }, { "epoch": 0.04, "learning_rate": 1.999874127673875e-07, "loss": 1.2734, "step": 8 }, { "epoch": 0.04, "learning_rate": 1.9994965423831852e-07, "loss": 1.3112, "step": 9 }, { "epoch": 0.05, "learning_rate": 1.998867339183008e-07, "loss": 1.3059, "step": 10 }, { "epoch": 0.05, "learning_rate": 1.9979866764718843e-07, "loss": 1.3516, "step": 11 }, { "epoch": 0.06, "learning_rate": 1.9968547759519424e-07, "loss": 1.3082, "step": 12 }, { "epoch": 0.06, "learning_rate": 1.9954719225730845e-07, "loss": 1.255, "step": 13 }, { "epoch": 0.07, "learning_rate": 1.993838464461254e-07, "loss": 1.2618, "step": 14 }, { "epoch": 0.07, "learning_rate": 1.9919548128307953e-07, "loss": 1.3239, "step": 15 }, { "epoch": 0.08, "learning_rate": 1.9898214418809327e-07, "loss": 1.285, "step": 16 }, { "epoch": 0.08, "learning_rate": 1.9874388886763942e-07, "loss": 1.2076, "step": 17 }, { "epoch": 0.09, "learning_rate": 1.984807753012208e-07, "loss": 1.3292, "step": 18 }, { "epoch": 0.09, "learning_rate": 1.9819286972627065e-07, "loss": 1.2239, "step": 19 }, { "epoch": 0.1, "learning_rate": 1.9788024462147787e-07, "loss": 1.2907, "step": 20 }, { "epoch": 0.1, "learning_rate": 1.975429786885407e-07, "loss": 1.2698, "step": 21 }, { "epoch": 0.11, "learning_rate": 1.9718115683235416e-07, "loss": 1.3123, "step": 22 }, { "epoch": 0.11, "learning_rate": 1.9679487013963562e-07, "loss": 1.2604, "step": 23 }, { "epoch": 0.12, "learning_rate": 1.9638421585599422e-07, "loss": 1.3007, "step": 24 }, { "epoch": 0.12, "learning_rate": 1.9594929736144973e-07, "loss": 1.3232, "step": 25 }, { "epoch": 0.13, "learning_rate": 1.9549022414440737e-07, "loss": 1.2522, "step": 26 }, { "epoch": 0.13, "learning_rate": 1.9500711177409452e-07, "loss": 1.2735, "step": 27 }, { "epoch": 0.14, "learning_rate": 1.9450008187146683e-07, "loss": 1.2489, "step": 28 }, { "epoch": 0.14, "learning_rate": 1.9396926207859085e-07, "loss": 1.2745, "step": 29 }, { "epoch": 0.15, "learning_rate": 1.9341478602651066e-07, "loss": 1.3546, "step": 30 }, { "epoch": 0.15, "learning_rate": 1.9283679330160724e-07, "loss": 1.3067, "step": 31 }, { "epoch": 0.16, "learning_rate": 1.9223542941045813e-07, "loss": 1.2281, "step": 32 }, { "epoch": 0.16, "learning_rate": 1.9161084574320693e-07, "loss": 1.2836, "step": 33 }, { "epoch": 0.17, "learning_rate": 1.9096319953545185e-07, "loss": 1.2204, "step": 34 }, { "epoch": 0.17, "learning_rate": 1.9029265382866213e-07, "loss": 1.2464, "step": 35 }, { "epoch": 0.18, "learning_rate": 1.8959937742913357e-07, "loss": 1.2542, "step": 36 }, { "epoch": 0.18, "learning_rate": 1.8888354486549234e-07, "loss": 1.3529, "step": 37 }, { "epoch": 0.19, "learning_rate": 1.881453363447582e-07, "loss": 1.2395, "step": 38 }, { "epoch": 0.19, "learning_rate": 1.873849377069785e-07, "loss": 1.2126, "step": 39 }, { "epoch": 0.2, "learning_rate": 1.8660254037844388e-07, "loss": 1.206, "step": 40 }, { "epoch": 0.2, "learning_rate": 1.8579834132349769e-07, "loss": 1.2899, "step": 41 }, { "epoch": 0.2, "learning_rate": 1.8497254299495145e-07, "loss": 1.2112, "step": 42 }, { "epoch": 0.21, "learning_rate": 1.8412535328311812e-07, "loss": 1.2797, "step": 43 }, { "epoch": 0.21, "learning_rate": 1.8325698546347712e-07, "loss": 1.2513, "step": 44 }, { "epoch": 0.22, "learning_rate": 1.8236765814298328e-07, "loss": 1.2499, "step": 45 }, { "epoch": 0.22, "learning_rate": 1.8145759520503356e-07, "loss": 1.2332, "step": 46 }, { "epoch": 0.23, "learning_rate": 1.8052702575310586e-07, "loss": 1.1517, "step": 47 }, { "epoch": 0.23, "learning_rate": 1.795761840530832e-07, "loss": 1.2733, "step": 48 }, { "epoch": 0.24, "learning_rate": 1.7860530947427875e-07, "loss": 1.199, "step": 49 }, { "epoch": 0.24, "learning_rate": 1.7761464642917567e-07, "loss": 1.2419, "step": 50 }, { "epoch": 0.25, "learning_rate": 1.766044443118978e-07, "loss": 1.2258, "step": 51 }, { "epoch": 0.25, "learning_rate": 1.7557495743542582e-07, "loss": 1.2609, "step": 52 }, { "epoch": 0.26, "learning_rate": 1.7452644496757548e-07, "loss": 1.192, "step": 53 }, { "epoch": 0.26, "learning_rate": 1.734591708657533e-07, "loss": 1.1758, "step": 54 }, { "epoch": 0.27, "learning_rate": 1.72373403810507e-07, "loss": 1.2181, "step": 55 }, { "epoch": 0.27, "learning_rate": 1.712694171378863e-07, "loss": 1.2666, "step": 56 }, { "epoch": 0.28, "learning_rate": 1.7014748877063213e-07, "loss": 1.2084, "step": 57 }, { "epoch": 0.28, "learning_rate": 1.690079011482112e-07, "loss": 1.2985, "step": 58 }, { "epoch": 0.29, "learning_rate": 1.678509411557132e-07, "loss": 1.2617, "step": 59 }, { "epoch": 0.29, "learning_rate": 1.6667690005162916e-07, "loss": 1.1919, "step": 60 }, { "epoch": 0.3, "learning_rate": 1.6548607339452852e-07, "loss": 1.1732, "step": 61 }, { "epoch": 0.3, "learning_rate": 1.6427876096865392e-07, "loss": 1.2219, "step": 62 }, { "epoch": 0.31, "learning_rate": 1.6305526670845225e-07, "loss": 1.2498, "step": 63 }, { "epoch": 0.31, "learning_rate": 1.6181589862206052e-07, "loss": 1.2018, "step": 64 }, { "epoch": 0.32, "learning_rate": 1.6056096871376665e-07, "loss": 1.2268, "step": 65 }, { "epoch": 0.32, "learning_rate": 1.5929079290546407e-07, "loss": 1.2108, "step": 66 }, { "epoch": 0.33, "learning_rate": 1.5800569095711981e-07, "loss": 1.2669, "step": 67 }, { "epoch": 0.33, "learning_rate": 1.5670598638627704e-07, "loss": 1.2009, "step": 68 }, { "epoch": 0.34, "learning_rate": 1.5539200638661102e-07, "loss": 1.249, "step": 69 }, { "epoch": 0.34, "learning_rate": 1.5406408174555975e-07, "loss": 1.2238, "step": 70 }, { "epoch": 0.35, "learning_rate": 1.5272254676105024e-07, "loss": 1.1932, "step": 71 }, { "epoch": 0.35, "learning_rate": 1.5136773915734065e-07, "loss": 1.146, "step": 72 }, { "epoch": 0.36, "learning_rate": 1.5e-07, "loss": 1.1604, "step": 73 }, { "epoch": 0.36, "learning_rate": 1.4861967361004686e-07, "loss": 1.152, "step": 74 }, { "epoch": 0.37, "learning_rate": 1.4722710747726827e-07, "loss": 1.2761, "step": 75 }, { "epoch": 0.37, "learning_rate": 1.4582265217274104e-07, "loss": 1.2246, "step": 76 }, { "epoch": 0.38, "learning_rate": 1.4440666126057742e-07, "loss": 1.1919, "step": 77 }, { "epoch": 0.38, "learning_rate": 1.4297949120891716e-07, "loss": 1.1962, "step": 78 }, { "epoch": 0.39, "learning_rate": 1.4154150130018864e-07, "loss": 1.145, "step": 79 }, { "epoch": 0.39, "learning_rate": 1.4009305354066136e-07, "loss": 1.2599, "step": 80 }, { "epoch": 0.4, "learning_rate": 1.3863451256931284e-07, "loss": 1.1728, "step": 81 }, { "epoch": 0.4, "learning_rate": 1.3716624556603275e-07, "loss": 1.2814, "step": 82 }, { "epoch": 0.4, "learning_rate": 1.3568862215918716e-07, "loss": 1.1492, "step": 83 }, { "epoch": 0.41, "learning_rate": 1.342020143325669e-07, "loss": 1.1769, "step": 84 }, { "epoch": 0.41, "learning_rate": 1.3270679633174217e-07, "loss": 1.1795, "step": 85 }, { "epoch": 0.42, "learning_rate": 1.3120334456984869e-07, "loss": 1.187, "step": 86 }, { "epoch": 0.42, "learning_rate": 1.296920375328275e-07, "loss": 1.2158, "step": 87 }, { "epoch": 0.43, "learning_rate": 1.2817325568414297e-07, "loss": 1.1903, "step": 88 }, { "epoch": 0.43, "learning_rate": 1.2664738136900349e-07, "loss": 1.2518, "step": 89 }, { "epoch": 0.44, "learning_rate": 1.2511479871810792e-07, "loss": 1.1396, "step": 90 }, { "epoch": 0.44, "learning_rate": 1.2357589355094274e-07, "loss": 1.2396, "step": 91 }, { "epoch": 0.45, "learning_rate": 1.2203105327865406e-07, "loss": 1.2634, "step": 92 }, { "epoch": 0.45, "learning_rate": 1.2048066680651908e-07, "loss": 1.1883, "step": 93 }, { "epoch": 0.46, "learning_rate": 1.1892512443604101e-07, "loss": 1.2214, "step": 94 }, { "epoch": 0.46, "learning_rate": 1.1736481776669305e-07, "loss": 1.2324, "step": 95 }, { "epoch": 0.47, "learning_rate": 1.1580013959733499e-07, "loss": 1.165, "step": 96 }, { "epoch": 0.47, "learning_rate": 1.1423148382732852e-07, "loss": 1.1884, "step": 97 }, { "epoch": 0.48, "learning_rate": 1.1265924535737492e-07, "loss": 1.1774, "step": 98 }, { "epoch": 0.48, "learning_rate": 1.110838199901011e-07, "loss": 1.1927, "step": 99 }, { "epoch": 0.49, "learning_rate": 1.0950560433041824e-07, "loss": 1.2058, "step": 100 }, { "epoch": 0.49, "learning_rate": 1.0792499568567884e-07, "loss": 1.1488, "step": 101 }, { "epoch": 0.5, "learning_rate": 1.0634239196565645e-07, "loss": 1.1537, "step": 102 }, { "epoch": 0.5, "learning_rate": 1.0475819158237424e-07, "loss": 1.2555, "step": 103 }, { "epoch": 0.51, "learning_rate": 1.0317279334980678e-07, "loss": 1.1353, "step": 104 }, { "epoch": 0.51, "learning_rate": 1.015865963834808e-07, "loss": 1.1506, "step": 105 }, { "epoch": 0.52, "learning_rate": 1e-07, "loss": 1.2785, "step": 106 }, { "epoch": 0.52, "learning_rate": 9.841340361651919e-08, "loss": 1.221, "step": 107 }, { "epoch": 0.53, "learning_rate": 9.682720665019325e-08, "loss": 1.2045, "step": 108 }, { "epoch": 0.53, "learning_rate": 9.524180841762577e-08, "loss": 1.2851, "step": 109 }, { "epoch": 0.54, "learning_rate": 9.365760803434354e-08, "loss": 1.2263, "step": 110 }, { "epoch": 0.54, "learning_rate": 9.207500431432114e-08, "loss": 1.2084, "step": 111 }, { "epoch": 0.55, "learning_rate": 9.049439566958175e-08, "loss": 1.2495, "step": 112 }, { "epoch": 0.55, "learning_rate": 8.89161800098989e-08, "loss": 1.1412, "step": 113 }, { "epoch": 0.56, "learning_rate": 8.734075464262506e-08, "loss": 1.2215, "step": 114 }, { "epoch": 0.56, "learning_rate": 8.576851617267149e-08, "loss": 1.1579, "step": 115 }, { "epoch": 0.57, "learning_rate": 8.4199860402665e-08, "loss": 1.1943, "step": 116 }, { "epoch": 0.57, "learning_rate": 8.263518223330696e-08, "loss": 1.1636, "step": 117 }, { "epoch": 0.58, "learning_rate": 8.107487556395901e-08, "loss": 1.2279, "step": 118 }, { "epoch": 0.58, "learning_rate": 7.951933319348094e-08, "loss": 1.1739, "step": 119 }, { "epoch": 0.59, "learning_rate": 7.796894672134593e-08, "loss": 1.1788, "step": 120 }, { "epoch": 0.59, "learning_rate": 7.642410644905725e-08, "loss": 1.2144, "step": 121 }, { "epoch": 0.6, "learning_rate": 7.488520128189209e-08, "loss": 1.1632, "step": 122 }, { "epoch": 0.6, "learning_rate": 7.33526186309965e-08, "loss": 1.189, "step": 123 }, { "epoch": 0.6, "learning_rate": 7.182674431585702e-08, "loss": 1.1962, "step": 124 }, { "epoch": 0.61, "learning_rate": 7.030796246717254e-08, "loss": 1.1785, "step": 125 }, { "epoch": 0.61, "learning_rate": 6.879665543015129e-08, "loss": 1.1997, "step": 126 }, { "epoch": 0.62, "learning_rate": 6.729320366825783e-08, "loss": 1.1685, "step": 127 }, { "epoch": 0.62, "learning_rate": 6.579798566743313e-08, "loss": 1.2235, "step": 128 }, { "epoch": 0.63, "learning_rate": 6.431137784081282e-08, "loss": 1.253, "step": 129 }, { "epoch": 0.63, "learning_rate": 6.283375443396726e-08, "loss": 1.1822, "step": 130 }, { "epoch": 0.64, "learning_rate": 6.136548743068712e-08, "loss": 1.1801, "step": 131 }, { "epoch": 0.64, "learning_rate": 5.990694645933865e-08, "loss": 1.2556, "step": 132 }, { "epoch": 0.65, "learning_rate": 5.845849869981137e-08, "loss": 1.186, "step": 133 }, { "epoch": 0.65, "learning_rate": 5.7020508791082835e-08, "loss": 1.1407, "step": 134 }, { "epoch": 0.66, "learning_rate": 5.559333873942258e-08, "loss": 1.1429, "step": 135 }, { "epoch": 0.66, "learning_rate": 5.4177347827258954e-08, "loss": 1.1542, "step": 136 }, { "epoch": 0.67, "learning_rate": 5.2772892522731736e-08, "loss": 1.1555, "step": 137 }, { "epoch": 0.67, "learning_rate": 5.138032638995314e-08, "loss": 1.1717, "step": 138 }, { "epoch": 0.68, "learning_rate": 5.000000000000002e-08, "loss": 1.207, "step": 139 }, { "epoch": 0.68, "learning_rate": 4.863226084265939e-08, "loss": 1.2138, "step": 140 }, { "epoch": 0.69, "learning_rate": 4.727745323894975e-08, "loss": 1.1169, "step": 141 }, { "epoch": 0.69, "learning_rate": 4.5935918254440276e-08, "loss": 1.1844, "step": 142 }, { "epoch": 0.7, "learning_rate": 4.460799361338897e-08, "loss": 1.1074, "step": 143 }, { "epoch": 0.7, "learning_rate": 4.3294013613722934e-08, "loss": 1.2708, "step": 144 }, { "epoch": 0.71, "learning_rate": 4.199430904288019e-08, "loss": 1.1677, "step": 145 }, { "epoch": 0.71, "learning_rate": 4.070920709453597e-08, "loss": 1.1804, "step": 146 }, { "epoch": 0.72, "learning_rate": 3.943903128623335e-08, "loss": 1.2236, "step": 147 }, { "epoch": 0.72, "learning_rate": 3.818410137793947e-08, "loss": 1.2135, "step": 148 }, { "epoch": 0.73, "learning_rate": 3.694473329154778e-08, "loss": 1.1679, "step": 149 }, { "epoch": 0.73, "learning_rate": 3.572123903134606e-08, "loss": 1.1737, "step": 150 }, { "epoch": 0.74, "learning_rate": 3.45139266054715e-08, "loss": 1.2123, "step": 151 }, { "epoch": 0.74, "learning_rate": 3.3323099948370846e-08, "loss": 1.1855, "step": 152 }, { "epoch": 0.75, "learning_rate": 3.214905884428679e-08, "loss": 1.1712, "step": 153 }, { "epoch": 0.75, "learning_rate": 3.0992098851788814e-08, "loss": 1.0934, "step": 154 }, { "epoch": 0.76, "learning_rate": 2.985251122936786e-08, "loss": 1.2342, "step": 155 }, { "epoch": 0.76, "learning_rate": 2.873058286211374e-08, "loss": 1.1785, "step": 156 }, { "epoch": 0.77, "learning_rate": 2.762659618949298e-08, "loss": 1.1129, "step": 157 }, { "epoch": 0.77, "learning_rate": 2.654082913424668e-08, "loss": 1.2042, "step": 158 }, { "epoch": 0.78, "learning_rate": 2.547355503242453e-08, "loss": 1.1776, "step": 159 }, { "epoch": 0.78, "learning_rate": 2.4425042564574183e-08, "loss": 1.2015, "step": 160 }, { "epoch": 0.79, "learning_rate": 2.339555568810221e-08, "loss": 1.2125, "step": 161 }, { "epoch": 0.79, "learning_rate": 2.2385353570824306e-08, "loss": 1.1405, "step": 162 }, { "epoch": 0.8, "learning_rate": 2.139469052572127e-08, "loss": 1.1515, "step": 163 }, { "epoch": 0.8, "learning_rate": 2.042381594691678e-08, "loss": 1.0014, "step": 164 }, { "epoch": 0.8, "learning_rate": 1.9472974246894136e-08, "loss": 1.145, "step": 165 }, { "epoch": 0.81, "learning_rate": 1.8542404794966426e-08, "loss": 1.2162, "step": 166 }, { "epoch": 0.81, "learning_rate": 1.763234185701673e-08, "loss": 1.2195, "step": 167 }, { "epoch": 0.82, "learning_rate": 1.674301453652287e-08, "loss": 1.1321, "step": 168 }, { "epoch": 0.82, "learning_rate": 1.5874646716881866e-08, "loss": 1.1922, "step": 169 }, { "epoch": 0.83, "learning_rate": 1.502745700504857e-08, "loss": 1.1713, "step": 170 }, { "epoch": 0.83, "learning_rate": 1.4201658676502293e-08, "loss": 1.2084, "step": 171 }, { "epoch": 0.84, "learning_rate": 1.3397459621556128e-08, "loss": 1.1722, "step": 172 }, { "epoch": 0.84, "learning_rate": 1.2615062293021505e-08, "loss": 1.1945, "step": 173 }, { "epoch": 0.85, "learning_rate": 1.1854663655241803e-08, "loss": 1.1389, "step": 174 }, { "epoch": 0.85, "learning_rate": 1.1116455134507664e-08, "loss": 1.1721, "step": 175 }, { "epoch": 0.86, "learning_rate": 1.0400622570866425e-08, "loss": 1.1674, "step": 176 }, { "epoch": 0.86, "learning_rate": 9.707346171337894e-09, "loss": 1.2015, "step": 177 }, { "epoch": 0.87, "learning_rate": 9.036800464548155e-09, "loss": 1.2152, "step": 178 }, { "epoch": 0.87, "learning_rate": 8.38915425679304e-09, "loss": 1.1764, "step": 179 }, { "epoch": 0.88, "learning_rate": 7.764570589541875e-09, "loss": 1.1159, "step": 180 }, { "epoch": 0.88, "learning_rate": 7.163206698392743e-09, "loss": 1.1123, "step": 181 }, { "epoch": 0.89, "learning_rate": 6.585213973489334e-09, "loss": 1.1876, "step": 182 }, { "epoch": 0.89, "learning_rate": 6.030737921409168e-09, "loss": 1.2447, "step": 183 }, { "epoch": 0.9, "learning_rate": 5.499918128533154e-09, "loss": 1.245, "step": 184 }, { "epoch": 0.9, "learning_rate": 4.9928882259054674e-09, "loss": 1.1083, "step": 185 }, { "epoch": 0.91, "learning_rate": 4.509775855592613e-09, "loss": 1.1305, "step": 186 }, { "epoch": 0.91, "learning_rate": 4.050702638550274e-09, "loss": 1.1942, "step": 187 }, { "epoch": 0.92, "learning_rate": 3.6157841440057958e-09, "loss": 1.1729, "step": 188 }, { "epoch": 0.92, "learning_rate": 3.205129860364375e-09, "loss": 1.0882, "step": 189 }, { "epoch": 0.93, "learning_rate": 2.8188431676458345e-09, "loss": 1.198, "step": 190 }, { "epoch": 0.93, "learning_rate": 2.457021311459295e-09, "loss": 1.194, "step": 191 }, { "epoch": 0.94, "learning_rate": 2.1197553785221366e-09, "loss": 1.1636, "step": 192 }, { "epoch": 0.94, "learning_rate": 1.8071302737293292e-09, "loss": 1.1798, "step": 193 }, { "epoch": 0.95, "learning_rate": 1.519224698779198e-09, "loss": 1.1553, "step": 194 }, { "epoch": 0.95, "learning_rate": 1.256111132360571e-09, "loss": 1.159, "step": 195 }, { "epoch": 0.96, "learning_rate": 1.0178558119067315e-09, "loss": 1.2488, "step": 196 }, { "epoch": 0.96, "learning_rate": 8.045187169204659e-10, "loss": 1.1661, "step": 197 }, { "epoch": 0.97, "learning_rate": 6.161535538745877e-10, "loss": 1.1698, "step": 198 }, { "epoch": 0.97, "learning_rate": 4.528077426915411e-10, "loss": 1.1729, "step": 199 }, { "epoch": 0.98, "learning_rate": 3.1452240480577264e-10, "loss": 1.1583, "step": 200 }, { "epoch": 0.98, "learning_rate": 2.0133235281156735e-10, "loss": 1.2136, "step": 201 }, { "epoch": 0.99, "learning_rate": 1.132660816992037e-10, "loss": 1.1568, "step": 202 }, { "epoch": 0.99, "learning_rate": 5.0345761681491736e-11, "loss": 1.2338, "step": 203 }, { "epoch": 1.0, "learning_rate": 1.2587232612493171e-11, "loss": 1.2104, "step": 204 }, { "epoch": 1.0, "learning_rate": 0.0, "loss": 1.1763, "step": 205 }, { "epoch": 1.0, "step": 205, "total_flos": 46263470604288.0, "train_loss": 1.2101893407542532, "train_runtime": 1529.7798, "train_samples_per_second": 17.114, "train_steps_per_second": 0.134 } ], "logging_steps": 1.0, "max_steps": 205, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 50000, "total_flos": 46263470604288.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }