{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0275229357798166, "global_step": 330, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 5.555555555555555e-07, "loss": 0.6498, "step": 1 }, { "epoch": 0.02, "learning_rate": 1.111111111111111e-06, "loss": 0.6642, "step": 2 }, { "epoch": 0.03, "learning_rate": 1.6666666666666667e-06, "loss": 0.6479, "step": 3 }, { "epoch": 0.04, "learning_rate": 2.222222222222222e-06, "loss": 0.5813, "step": 4 }, { "epoch": 0.05, "learning_rate": 2.7777777777777783e-06, "loss": 0.6306, "step": 5 }, { "epoch": 0.06, "learning_rate": 3.3333333333333333e-06, "loss": 0.6881, "step": 6 }, { "epoch": 0.06, "learning_rate": 3.88888888888889e-06, "loss": 0.6066, "step": 7 }, { "epoch": 0.07, "learning_rate": 4.444444444444444e-06, "loss": 0.5688, "step": 8 }, { "epoch": 0.08, "learning_rate": 5e-06, "loss": 0.5619, "step": 9 }, { "epoch": 0.09, "learning_rate": 5.555555555555557e-06, "loss": 0.5144, "step": 10 }, { "epoch": 0.1, "learning_rate": 6.111111111111112e-06, "loss": 0.5744, "step": 11 }, { "epoch": 0.11, "learning_rate": 6.666666666666667e-06, "loss": 0.6051, "step": 12 }, { "epoch": 0.12, "learning_rate": 7.222222222222223e-06, "loss": 0.5917, "step": 13 }, { "epoch": 0.13, "learning_rate": 7.77777777777778e-06, "loss": 0.5239, "step": 14 }, { "epoch": 0.14, "learning_rate": 8.333333333333334e-06, "loss": 0.5253, "step": 15 }, { "epoch": 0.15, "learning_rate": 8.888888888888888e-06, "loss": 0.4143, "step": 16 }, { "epoch": 0.16, "learning_rate": 9.444444444444445e-06, "loss": 0.4822, "step": 17 }, { "epoch": 0.17, "learning_rate": 1e-05, "loss": 0.4627, "step": 18 }, { "epoch": 0.17, "learning_rate": 9.999858783596665e-06, "loss": 0.5389, "step": 19 }, { "epoch": 0.18, "learning_rate": 9.999435142363484e-06, "loss": 0.5544, "step": 20 }, { "epoch": 0.19, "learning_rate": 9.998729100230497e-06, "loss": 0.5013, "step": 21 }, { "epoch": 0.2, "learning_rate": 9.997740697079595e-06, "loss": 0.5035, "step": 22 }, { "epoch": 0.21, "learning_rate": 9.99646998874227e-06, "loss": 0.4924, "step": 23 }, { "epoch": 0.22, "learning_rate": 9.994917046996472e-06, "loss": 0.507, "step": 24 }, { "epoch": 0.23, "learning_rate": 9.993081959562539e-06, "loss": 0.4876, "step": 25 }, { "epoch": 0.24, "learning_rate": 9.990964830098246e-06, "loss": 0.555, "step": 26 }, { "epoch": 0.25, "learning_rate": 9.98856577819296e-06, "loss": 0.4692, "step": 27 }, { "epoch": 0.26, "learning_rate": 9.985884939360873e-06, "loss": 0.466, "step": 28 }, { "epoch": 0.27, "learning_rate": 9.98292246503335e-06, "loss": 0.4132, "step": 29 }, { "epoch": 0.28, "learning_rate": 9.979678522550382e-06, "loss": 0.4741, "step": 30 }, { "epoch": 0.28, "learning_rate": 9.976153295151123e-06, "loss": 0.5424, "step": 31 }, { "epoch": 0.29, "learning_rate": 9.972346981963546e-06, "loss": 0.4949, "step": 32 }, { "epoch": 0.3, "learning_rate": 9.968259797993197e-06, "loss": 0.5126, "step": 33 }, { "epoch": 0.31, "learning_rate": 9.963891974111042e-06, "loss": 0.4202, "step": 34 }, { "epoch": 0.32, "learning_rate": 9.959243757040434e-06, "loss": 0.4736, "step": 35 }, { "epoch": 0.33, "learning_rate": 9.95431540934317e-06, "loss": 0.3851, "step": 36 }, { "epoch": 0.34, "learning_rate": 9.949107209404664e-06, "loss": 0.5542, "step": 37 }, { "epoch": 0.35, "learning_rate": 9.943619451418225e-06, "loss": 0.5095, "step": 38 }, { "epoch": 0.36, "learning_rate": 9.937852445368427e-06, "loss": 0.4338, "step": 39 }, { "epoch": 0.37, "learning_rate": 9.931806517013612e-06, "loss": 0.45, "step": 40 }, { "epoch": 0.38, "learning_rate": 9.925482007867485e-06, "loss": 0.4735, "step": 41 }, { "epoch": 0.39, "learning_rate": 9.918879275179819e-06, "loss": 0.4735, "step": 42 }, { "epoch": 0.39, "learning_rate": 9.911998691916275e-06, "loss": 0.5041, "step": 43 }, { "epoch": 0.4, "learning_rate": 9.904840646737346e-06, "loss": 0.4432, "step": 44 }, { "epoch": 0.41, "learning_rate": 9.89740554397639e-06, "loss": 0.4457, "step": 45 }, { "epoch": 0.42, "learning_rate": 9.889693803616793e-06, "loss": 0.4692, "step": 46 }, { "epoch": 0.43, "learning_rate": 9.881705861268252e-06, "loss": 0.4469, "step": 47 }, { "epoch": 0.44, "learning_rate": 9.873442168142158e-06, "loss": 0.4389, "step": 48 }, { "epoch": 0.45, "learning_rate": 9.864903191026125e-06, "loss": 0.4195, "step": 49 }, { "epoch": 0.46, "learning_rate": 9.856089412257605e-06, "loss": 0.4635, "step": 50 }, { "epoch": 0.47, "learning_rate": 9.847001329696653e-06, "loss": 0.4944, "step": 51 }, { "epoch": 0.48, "learning_rate": 9.837639456697802e-06, "loss": 0.4917, "step": 52 }, { "epoch": 0.49, "learning_rate": 9.828004322081067e-06, "loss": 0.4197, "step": 53 }, { "epoch": 0.5, "learning_rate": 9.818096470102067e-06, "loss": 0.4294, "step": 54 }, { "epoch": 0.5, "learning_rate": 9.807916460421294e-06, "loss": 0.4839, "step": 55 }, { "epoch": 0.51, "learning_rate": 9.797464868072489e-06, "loss": 0.4918, "step": 56 }, { "epoch": 0.52, "learning_rate": 9.78674228343016e-06, "loss": 0.5003, "step": 57 }, { "epoch": 0.53, "learning_rate": 9.775749312176249e-06, "loss": 0.4864, "step": 58 }, { "epoch": 0.54, "learning_rate": 9.764486575265893e-06, "loss": 0.4977, "step": 59 }, { "epoch": 0.55, "learning_rate": 9.752954708892379e-06, "loss": 0.4125, "step": 60 }, { "epoch": 0.56, "learning_rate": 9.741154364451179e-06, "loss": 0.4442, "step": 61 }, { "epoch": 0.57, "learning_rate": 9.729086208503174e-06, "loss": 0.4624, "step": 62 }, { "epoch": 0.58, "learning_rate": 9.716750922736998e-06, "loss": 0.3754, "step": 63 }, { "epoch": 0.59, "learning_rate": 9.704149203930522e-06, "loss": 0.4218, "step": 64 }, { "epoch": 0.6, "learning_rate": 9.691281763911513e-06, "loss": 0.4997, "step": 65 }, { "epoch": 0.61, "learning_rate": 9.67814932951741e-06, "loss": 0.3829, "step": 66 }, { "epoch": 0.61, "learning_rate": 9.664752642554272e-06, "loss": 0.3637, "step": 67 }, { "epoch": 0.62, "learning_rate": 9.651092459754879e-06, "loss": 0.5271, "step": 68 }, { "epoch": 0.63, "learning_rate": 9.637169552735985e-06, "loss": 0.4425, "step": 69 }, { "epoch": 0.64, "learning_rate": 9.622984707954732e-06, "loss": 0.4109, "step": 70 }, { "epoch": 0.65, "learning_rate": 9.608538726664224e-06, "loss": 0.3699, "step": 71 }, { "epoch": 0.66, "learning_rate": 9.593832424868271e-06, "loss": 0.4199, "step": 72 }, { "epoch": 0.67, "learning_rate": 9.578866633275289e-06, "loss": 0.4548, "step": 73 }, { "epoch": 0.68, "learning_rate": 9.563642197251382e-06, "loss": 0.4464, "step": 74 }, { "epoch": 0.69, "learning_rate": 9.548159976772593e-06, "loss": 0.4286, "step": 75 }, { "epoch": 0.7, "learning_rate": 9.532420846376316e-06, "loss": 0.397, "step": 76 }, { "epoch": 0.71, "learning_rate": 9.516425695111906e-06, "loss": 0.4522, "step": 77 }, { "epoch": 0.72, "learning_rate": 9.500175426490455e-06, "loss": 0.4763, "step": 78 }, { "epoch": 0.72, "learning_rate": 9.48367095843376e-06, "loss": 0.4386, "step": 79 }, { "epoch": 0.73, "learning_rate": 9.466913223222467e-06, "loss": 0.411, "step": 80 }, { "epoch": 0.74, "learning_rate": 9.449903167443415e-06, "loss": 0.3719, "step": 81 }, { "epoch": 0.75, "learning_rate": 9.432641751936162e-06, "loss": 0.3693, "step": 82 }, { "epoch": 0.76, "learning_rate": 9.415129951738713e-06, "loss": 0.3955, "step": 83 }, { "epoch": 0.77, "learning_rate": 9.397368756032445e-06, "loss": 0.4196, "step": 84 }, { "epoch": 0.78, "learning_rate": 9.379359168086231e-06, "loss": 0.4527, "step": 85 }, { "epoch": 0.79, "learning_rate": 9.361102205199762e-06, "loss": 0.4054, "step": 86 }, { "epoch": 0.8, "learning_rate": 9.34259889864609e-06, "loss": 0.4001, "step": 87 }, { "epoch": 0.81, "learning_rate": 9.32385029361338e-06, "loss": 0.4189, "step": 88 }, { "epoch": 0.82, "learning_rate": 9.304857449145858e-06, "loss": 0.439, "step": 89 }, { "epoch": 0.83, "learning_rate": 9.285621438083997e-06, "loss": 0.4221, "step": 90 }, { "epoch": 0.83, "learning_rate": 9.26614334700392e-06, "loss": 0.4208, "step": 91 }, { "epoch": 0.84, "learning_rate": 9.246424276156008e-06, "loss": 0.4707, "step": 92 }, { "epoch": 0.85, "learning_rate": 9.226465339402768e-06, "loss": 0.3372, "step": 93 }, { "epoch": 0.86, "learning_rate": 9.206267664155906e-06, "loss": 0.3637, "step": 94 }, { "epoch": 0.87, "learning_rate": 9.185832391312644e-06, "loss": 0.4146, "step": 95 }, { "epoch": 0.88, "learning_rate": 9.165160675191272e-06, "loss": 0.393, "step": 96 }, { "epoch": 0.89, "learning_rate": 9.144253683465953e-06, "loss": 0.4712, "step": 97 }, { "epoch": 0.9, "learning_rate": 9.123112597100759e-06, "loss": 0.3904, "step": 98 }, { "epoch": 0.91, "learning_rate": 9.101738610282956e-06, "loss": 0.4192, "step": 99 }, { "epoch": 0.92, "learning_rate": 9.080132930355567e-06, "loss": 0.4039, "step": 100 }, { "epoch": 0.93, "learning_rate": 9.058296777749154e-06, "loss": 0.4516, "step": 101 }, { "epoch": 0.94, "learning_rate": 9.03623138591289e-06, "loss": 0.3348, "step": 102 }, { "epoch": 0.94, "learning_rate": 9.013938001244885e-06, "loss": 0.4256, "step": 103 }, { "epoch": 0.95, "learning_rate": 8.99141788302178e-06, "loss": 0.3774, "step": 104 }, { "epoch": 0.96, "learning_rate": 8.968672303327614e-06, "loss": 0.4471, "step": 105 }, { "epoch": 0.97, "learning_rate": 8.94570254698197e-06, "loss": 0.3771, "step": 106 }, { "epoch": 0.98, "learning_rate": 8.922509911467395e-06, "loss": 0.4137, "step": 107 }, { "epoch": 0.99, "learning_rate": 8.899095706856122e-06, "loss": 0.4269, "step": 108 }, { "epoch": 1.0, "learning_rate": 8.875461255736055e-06, "loss": 0.4102, "step": 109 }, { "epoch": 1.01, "learning_rate": 8.851607893136065e-06, "loss": 0.2619, "step": 110 }, { "epoch": 1.02, "learning_rate": 8.827536966450584e-06, "loss": 0.2864, "step": 111 }, { "epoch": 1.03, "learning_rate": 8.803249835363486e-06, "loss": 0.2712, "step": 112 }, { "epoch": 1.04, "learning_rate": 8.778747871771293e-06, "loss": 0.2186, "step": 113 }, { "epoch": 1.05, "learning_rate": 8.754032459705672e-06, "loss": 0.2585, "step": 114 }, { "epoch": 1.06, "learning_rate": 8.729104995255265e-06, "loss": 0.2711, "step": 115 }, { "epoch": 1.06, "learning_rate": 8.703966886486819e-06, "loss": 0.2801, "step": 116 }, { "epoch": 1.07, "learning_rate": 8.67861955336566e-06, "loss": 0.2586, "step": 117 }, { "epoch": 1.08, "learning_rate": 8.65306442767547e-06, "loss": 0.3134, "step": 118 }, { "epoch": 1.09, "learning_rate": 8.627302952937431e-06, "loss": 0.2067, "step": 119 }, { "epoch": 1.1, "learning_rate": 8.601336584328659e-06, "loss": 0.2637, "step": 120 }, { "epoch": 1.11, "learning_rate": 8.575166788600031e-06, "loss": 0.2445, "step": 121 }, { "epoch": 1.12, "learning_rate": 8.548795043993316e-06, "loss": 0.2456, "step": 122 }, { "epoch": 1.13, "learning_rate": 8.522222840157687e-06, "loss": 0.2963, "step": 123 }, { "epoch": 1.14, "learning_rate": 8.495451678065563e-06, "loss": 0.2781, "step": 124 }, { "epoch": 1.15, "learning_rate": 8.468483069927832e-06, "loss": 0.2572, "step": 125 }, { "epoch": 1.16, "learning_rate": 8.441318539108433e-06, "loss": 0.2473, "step": 126 }, { "epoch": 1.17, "learning_rate": 8.413959620038306e-06, "loss": 0.1962, "step": 127 }, { "epoch": 1.17, "learning_rate": 8.386407858128707e-06, "loss": 0.2644, "step": 128 }, { "epoch": 1.18, "learning_rate": 8.358664809683926e-06, "loss": 0.2451, "step": 129 }, { "epoch": 1.19, "learning_rate": 8.330732041813367e-06, "loss": 0.2801, "step": 130 }, { "epoch": 1.2, "learning_rate": 8.302611132343042e-06, "loss": 0.2425, "step": 131 }, { "epoch": 1.21, "learning_rate": 8.274303669726427e-06, "loss": 0.2113, "step": 132 }, { "epoch": 1.22, "learning_rate": 8.245811252954741e-06, "loss": 0.2918, "step": 133 }, { "epoch": 1.23, "learning_rate": 8.217135491466636e-06, "loss": 0.2356, "step": 134 }, { "epoch": 1.24, "learning_rate": 8.18827800505727e-06, "loss": 0.2809, "step": 135 }, { "epoch": 1.25, "learning_rate": 8.15924042378682e-06, "loss": 0.2559, "step": 136 }, { "epoch": 1.26, "learning_rate": 8.130024387888402e-06, "loss": 0.2446, "step": 137 }, { "epoch": 1.27, "learning_rate": 8.100631547675417e-06, "loss": 0.2314, "step": 138 }, { "epoch": 1.28, "learning_rate": 8.071063563448341e-06, "loss": 0.2729, "step": 139 }, { "epoch": 1.28, "learning_rate": 8.041322105400923e-06, "loss": 0.2017, "step": 140 }, { "epoch": 1.29, "learning_rate": 8.01140885352586e-06, "loss": 0.2195, "step": 141 }, { "epoch": 1.3, "learning_rate": 7.981325497519892e-06, "loss": 0.2934, "step": 142 }, { "epoch": 1.31, "learning_rate": 7.951073736688348e-06, "loss": 0.2598, "step": 143 }, { "epoch": 1.32, "learning_rate": 7.920655279849173e-06, "loss": 0.2776, "step": 144 }, { "epoch": 1.33, "learning_rate": 7.890071845236395e-06, "loss": 0.2622, "step": 145 }, { "epoch": 1.34, "learning_rate": 7.859325160403073e-06, "loss": 0.2621, "step": 146 }, { "epoch": 1.35, "learning_rate": 7.8284169621237e-06, "loss": 0.2499, "step": 147 }, { "epoch": 1.36, "learning_rate": 7.797348996296116e-06, "loss": 0.28, "step": 148 }, { "epoch": 1.37, "learning_rate": 7.766123017842877e-06, "loss": 0.1906, "step": 149 }, { "epoch": 1.38, "learning_rate": 7.734740790612137e-06, "loss": 0.2373, "step": 150 }, { "epoch": 1.39, "learning_rate": 7.703204087277989e-06, "loss": 0.2, "step": 151 }, { "epoch": 1.39, "learning_rate": 7.671514689240366e-06, "loss": 0.2678, "step": 152 }, { "epoch": 1.4, "learning_rate": 7.639674386524395e-06, "loss": 0.2295, "step": 153 }, { "epoch": 1.41, "learning_rate": 7.607684977679284e-06, "loss": 0.2417, "step": 154 }, { "epoch": 1.42, "learning_rate": 7.575548269676741e-06, "loss": 0.2662, "step": 155 }, { "epoch": 1.43, "learning_rate": 7.543266077808893e-06, "loss": 0.2645, "step": 156 }, { "epoch": 1.44, "learning_rate": 7.510840225585749e-06, "loss": 0.2757, "step": 157 }, { "epoch": 1.45, "learning_rate": 7.478272544632204e-06, "loss": 0.2324, "step": 158 }, { "epoch": 1.46, "learning_rate": 7.44556487458456e-06, "loss": 0.2655, "step": 159 }, { "epoch": 1.47, "learning_rate": 7.412719062986632e-06, "loss": 0.2091, "step": 160 }, { "epoch": 1.48, "learning_rate": 7.379736965185369e-06, "loss": 0.2293, "step": 161 }, { "epoch": 1.49, "learning_rate": 7.3466204442260605e-06, "loss": 0.246, "step": 162 }, { "epoch": 1.5, "learning_rate": 7.313371370747104e-06, "loss": 0.2362, "step": 163 }, { "epoch": 1.5, "learning_rate": 7.279991622874319e-06, "loss": 0.2074, "step": 164 }, { "epoch": 1.51, "learning_rate": 7.24648308611489e-06, "loss": 0.2035, "step": 165 }, { "epoch": 1.52, "learning_rate": 7.212847653250828e-06, "loss": 0.3013, "step": 166 }, { "epoch": 1.53, "learning_rate": 7.1790872242320775e-06, "loss": 0.2286, "step": 167 }, { "epoch": 1.54, "learning_rate": 7.145203706069183e-06, "loss": 0.2005, "step": 168 }, { "epoch": 1.55, "learning_rate": 7.1111990127255684e-06, "loss": 0.2351, "step": 169 }, { "epoch": 1.56, "learning_rate": 7.0770750650094335e-06, "loss": 0.2378, "step": 170 }, { "epoch": 1.57, "learning_rate": 7.042833790465241e-06, "loss": 0.2557, "step": 171 }, { "epoch": 1.58, "learning_rate": 7.008477123264849e-06, "loss": 0.1742, "step": 172 }, { "epoch": 1.59, "learning_rate": 6.974007004098243e-06, "loss": 0.2862, "step": 173 }, { "epoch": 1.6, "learning_rate": 6.939425380063924e-06, "loss": 0.2359, "step": 174 }, { "epoch": 1.61, "learning_rate": 6.9047342045589224e-06, "loss": 0.2646, "step": 175 }, { "epoch": 1.61, "learning_rate": 6.869935437168449e-06, "loss": 0.2223, "step": 176 }, { "epoch": 1.62, "learning_rate": 6.835031043555211e-06, "loss": 0.2292, "step": 177 }, { "epoch": 1.63, "learning_rate": 6.800022995348381e-06, "loss": 0.2395, "step": 178 }, { "epoch": 1.64, "learning_rate": 6.76491327003222e-06, "loss": 0.2201, "step": 179 }, { "epoch": 1.65, "learning_rate": 6.729703850834381e-06, "loss": 0.2071, "step": 180 }, { "epoch": 1.66, "learning_rate": 6.694396726613883e-06, "loss": 0.2461, "step": 181 }, { "epoch": 1.67, "learning_rate": 6.65899389174876e-06, "loss": 0.2804, "step": 182 }, { "epoch": 1.68, "learning_rate": 6.6234973460234184e-06, "loss": 0.2859, "step": 183 }, { "epoch": 1.69, "learning_rate": 6.587909094515663e-06, "loss": 0.2358, "step": 184 }, { "epoch": 1.7, "learning_rate": 6.552231147483448e-06, "loss": 0.229, "step": 185 }, { "epoch": 1.71, "learning_rate": 6.5164655202513135e-06, "loss": 0.2673, "step": 186 }, { "epoch": 1.72, "learning_rate": 6.480614233096558e-06, "loss": 0.2169, "step": 187 }, { "epoch": 1.72, "learning_rate": 6.444679311135112e-06, "loss": 0.1908, "step": 188 }, { "epoch": 1.73, "learning_rate": 6.408662784207149e-06, "loss": 0.2167, "step": 189 }, { "epoch": 1.74, "learning_rate": 6.372566686762427e-06, "loss": 0.2533, "step": 190 }, { "epoch": 1.75, "learning_rate": 6.336393057745365e-06, "loss": 0.2361, "step": 191 }, { "epoch": 1.76, "learning_rate": 6.300143940479881e-06, "loss": 0.2396, "step": 192 }, { "epoch": 1.77, "learning_rate": 6.2638213825539595e-06, "loss": 0.2716, "step": 193 }, { "epoch": 1.78, "learning_rate": 6.227427435703997e-06, "loss": 0.2187, "step": 194 }, { "epoch": 1.79, "learning_rate": 6.190964155698903e-06, "loss": 0.2419, "step": 195 }, { "epoch": 1.8, "learning_rate": 6.154433602223979e-06, "loss": 0.2547, "step": 196 }, { "epoch": 1.81, "learning_rate": 6.117837838764579e-06, "loss": 0.2534, "step": 197 }, { "epoch": 1.82, "learning_rate": 6.0811789324895365e-06, "loss": 0.255, "step": 198 }, { "epoch": 1.83, "learning_rate": 6.044458954134411e-06, "loss": 0.2383, "step": 199 }, { "epoch": 1.83, "learning_rate": 6.0076799778845105e-06, "loss": 0.2407, "step": 200 }, { "epoch": 1.84, "learning_rate": 5.970844081257734e-06, "loss": 0.1883, "step": 201 }, { "epoch": 1.85, "learning_rate": 5.933953344987215e-06, "loss": 0.3112, "step": 202 }, { "epoch": 1.86, "learning_rate": 5.897009852903792e-06, "loss": 0.2424, "step": 203 }, { "epoch": 1.87, "learning_rate": 5.860015691818292e-06, "loss": 0.2663, "step": 204 }, { "epoch": 1.88, "learning_rate": 5.82297295140367e-06, "loss": 0.2616, "step": 205 }, { "epoch": 1.89, "learning_rate": 5.78588372407695e-06, "loss": 0.2067, "step": 206 }, { "epoch": 1.9, "learning_rate": 5.748750104881051e-06, "loss": 0.2092, "step": 207 }, { "epoch": 1.91, "learning_rate": 5.711574191366427e-06, "loss": 0.2383, "step": 208 }, { "epoch": 1.92, "learning_rate": 5.674358083472598e-06, "loss": 0.2377, "step": 209 }, { "epoch": 1.93, "learning_rate": 5.637103883409525e-06, "loss": 0.2415, "step": 210 }, { "epoch": 1.94, "learning_rate": 5.599813695538866e-06, "loss": 0.2478, "step": 211 }, { "epoch": 1.94, "learning_rate": 5.562489626255104e-06, "loss": 0.2752, "step": 212 }, { "epoch": 1.95, "learning_rate": 5.52513378386657e-06, "loss": 0.2133, "step": 213 }, { "epoch": 1.96, "learning_rate": 5.487748278476342e-06, "loss": 0.2043, "step": 214 }, { "epoch": 1.97, "learning_rate": 5.450335221863068e-06, "loss": 0.2664, "step": 215 }, { "epoch": 1.98, "learning_rate": 5.412896727361663e-06, "loss": 0.2617, "step": 216 }, { "epoch": 1.99, "learning_rate": 5.375434909743942e-06, "loss": 0.2001, "step": 217 }, { "epoch": 2.0, "learning_rate": 5.337951885099167e-06, "loss": 0.2353, "step": 218 }, { "epoch": 2.01, "learning_rate": 5.300449770714502e-06, "loss": 0.112, "step": 219 }, { "epoch": 2.02, "learning_rate": 5.262930684955439e-06, "loss": 0.1022, "step": 220 }, { "epoch": 2.03, "learning_rate": 5.225396747146112e-06, "loss": 0.1263, "step": 221 }, { "epoch": 2.04, "learning_rate": 5.187850077449604e-06, "loss": 0.1059, "step": 222 }, { "epoch": 2.05, "learning_rate": 5.150292796748174e-06, "loss": 0.1101, "step": 223 }, { "epoch": 2.06, "learning_rate": 5.112727026523461e-06, "loss": 0.1274, "step": 224 }, { "epoch": 2.06, "learning_rate": 5.075154888736653e-06, "loss": 0.1604, "step": 225 }, { "epoch": 2.07, "learning_rate": 5.03757850570861e-06, "loss": 0.1067, "step": 226 }, { "epoch": 2.08, "learning_rate": 5e-06, "loss": 0.1124, "step": 227 }, { "epoch": 2.09, "learning_rate": 4.9624214942913916e-06, "loss": 0.0976, "step": 228 }, { "epoch": 2.1, "learning_rate": 4.924845111263349e-06, "loss": 0.1094, "step": 229 }, { "epoch": 2.11, "learning_rate": 4.88727297347654e-06, "loss": 0.1156, "step": 230 }, { "epoch": 2.12, "learning_rate": 4.8497072032518274e-06, "loss": 0.1201, "step": 231 }, { "epoch": 2.13, "learning_rate": 4.8121499225503974e-06, "loss": 0.1053, "step": 232 }, { "epoch": 2.14, "learning_rate": 4.774603252853889e-06, "loss": 0.1344, "step": 233 }, { "epoch": 2.15, "learning_rate": 4.737069315044562e-06, "loss": 0.1171, "step": 234 }, { "epoch": 2.16, "learning_rate": 4.699550229285499e-06, "loss": 0.0906, "step": 235 }, { "epoch": 2.17, "learning_rate": 4.662048114900837e-06, "loss": 0.1313, "step": 236 }, { "epoch": 2.17, "learning_rate": 4.624565090256059e-06, "loss": 0.1096, "step": 237 }, { "epoch": 2.18, "learning_rate": 4.587103272638339e-06, "loss": 0.1185, "step": 238 }, { "epoch": 2.19, "learning_rate": 4.549664778136933e-06, "loss": 0.1372, "step": 239 }, { "epoch": 2.2, "learning_rate": 4.512251721523659e-06, "loss": 0.1042, "step": 240 }, { "epoch": 2.21, "learning_rate": 4.4748662161334335e-06, "loss": 0.1185, "step": 241 }, { "epoch": 2.22, "learning_rate": 4.437510373744897e-06, "loss": 0.1448, "step": 242 }, { "epoch": 2.23, "learning_rate": 4.400186304461136e-06, "loss": 0.113, "step": 243 }, { "epoch": 2.24, "learning_rate": 4.362896116590475e-06, "loss": 0.1211, "step": 244 }, { "epoch": 2.25, "learning_rate": 4.325641916527405e-06, "loss": 0.132, "step": 245 }, { "epoch": 2.26, "learning_rate": 4.2884258086335755e-06, "loss": 0.1105, "step": 246 }, { "epoch": 2.27, "learning_rate": 4.25124989511895e-06, "loss": 0.0987, "step": 247 }, { "epoch": 2.28, "learning_rate": 4.214116275923051e-06, "loss": 0.1197, "step": 248 }, { "epoch": 2.28, "learning_rate": 4.17702704859633e-06, "loss": 0.1334, "step": 249 }, { "epoch": 2.29, "learning_rate": 4.1399843081817085e-06, "loss": 0.1468, "step": 250 }, { "epoch": 2.3, "learning_rate": 4.1029901470962105e-06, "loss": 0.1161, "step": 251 }, { "epoch": 2.31, "learning_rate": 4.066046655012786e-06, "loss": 0.1106, "step": 252 }, { "epoch": 2.32, "learning_rate": 4.029155918742268e-06, "loss": 0.1234, "step": 253 }, { "epoch": 2.33, "learning_rate": 3.992320022115492e-06, "loss": 0.1261, "step": 254 }, { "epoch": 2.34, "learning_rate": 3.955541045865591e-06, "loss": 0.1011, "step": 255 }, { "epoch": 2.35, "learning_rate": 3.918821067510464e-06, "loss": 0.1058, "step": 256 }, { "epoch": 2.36, "learning_rate": 3.882162161235421e-06, "loss": 0.1002, "step": 257 }, { "epoch": 2.37, "learning_rate": 3.845566397776022e-06, "loss": 0.1023, "step": 258 }, { "epoch": 2.38, "learning_rate": 3.8090358443010993e-06, "loss": 0.1174, "step": 259 }, { "epoch": 2.39, "learning_rate": 3.7725725642960047e-06, "loss": 0.111, "step": 260 }, { "epoch": 2.39, "learning_rate": 3.7361786174460414e-06, "loss": 0.0948, "step": 261 }, { "epoch": 2.4, "learning_rate": 3.6998560595201188e-06, "loss": 0.0972, "step": 262 }, { "epoch": 2.41, "learning_rate": 3.6636069422546363e-06, "loss": 0.0982, "step": 263 }, { "epoch": 2.42, "learning_rate": 3.627433313237576e-06, "loss": 0.1374, "step": 264 }, { "epoch": 2.43, "learning_rate": 3.5913372157928515e-06, "loss": 0.0988, "step": 265 }, { "epoch": 2.44, "learning_rate": 3.555320688864889e-06, "loss": 0.1281, "step": 266 }, { "epoch": 2.45, "learning_rate": 3.519385766903442e-06, "loss": 0.1273, "step": 267 }, { "epoch": 2.46, "learning_rate": 3.483534479748688e-06, "loss": 0.1256, "step": 268 }, { "epoch": 2.47, "learning_rate": 3.447768852516554e-06, "loss": 0.1131, "step": 269 }, { "epoch": 2.48, "learning_rate": 3.4120909054843375e-06, "loss": 0.0974, "step": 270 }, { "epoch": 2.49, "learning_rate": 3.3765026539765832e-06, "loss": 0.1233, "step": 271 }, { "epoch": 2.5, "learning_rate": 3.3410061082512422e-06, "loss": 0.1175, "step": 272 }, { "epoch": 2.5, "learning_rate": 3.3056032733861188e-06, "loss": 0.1178, "step": 273 }, { "epoch": 2.51, "learning_rate": 3.2702961491656197e-06, "loss": 0.11, "step": 274 }, { "epoch": 2.52, "learning_rate": 3.2350867299677802e-06, "loss": 0.1091, "step": 275 }, { "epoch": 2.53, "learning_rate": 3.1999770046516198e-06, "loss": 0.1146, "step": 276 }, { "epoch": 2.54, "learning_rate": 3.164968956444791e-06, "loss": 0.0945, "step": 277 }, { "epoch": 2.55, "learning_rate": 3.130064562831553e-06, "loss": 0.1216, "step": 278 }, { "epoch": 2.56, "learning_rate": 3.0952657954410792e-06, "loss": 0.1273, "step": 279 }, { "epoch": 2.57, "learning_rate": 3.0605746199360755e-06, "loss": 0.1161, "step": 280 }, { "epoch": 2.58, "learning_rate": 3.0259929959017585e-06, "loss": 0.1395, "step": 281 }, { "epoch": 2.59, "learning_rate": 2.991522876735154e-06, "loss": 0.1224, "step": 282 }, { "epoch": 2.6, "learning_rate": 2.95716620953476e-06, "loss": 0.1052, "step": 283 }, { "epoch": 2.61, "learning_rate": 2.9229249349905686e-06, "loss": 0.1383, "step": 284 }, { "epoch": 2.61, "learning_rate": 2.8888009872744332e-06, "loss": 0.0985, "step": 285 }, { "epoch": 2.62, "learning_rate": 2.8547962939308187e-06, "loss": 0.1164, "step": 286 }, { "epoch": 2.63, "learning_rate": 2.8209127757679246e-06, "loss": 0.095, "step": 287 }, { "epoch": 2.64, "learning_rate": 2.787152346749173e-06, "loss": 0.1308, "step": 288 }, { "epoch": 2.65, "learning_rate": 2.7535169138851124e-06, "loss": 0.1336, "step": 289 }, { "epoch": 2.66, "learning_rate": 2.720008377125682e-06, "loss": 0.0952, "step": 290 }, { "epoch": 2.67, "learning_rate": 2.686628629252899e-06, "loss": 0.134, "step": 291 }, { "epoch": 2.68, "learning_rate": 2.6533795557739407e-06, "loss": 0.1008, "step": 292 }, { "epoch": 2.69, "learning_rate": 2.6202630348146323e-06, "loss": 0.0938, "step": 293 }, { "epoch": 2.7, "learning_rate": 2.5872809370133704e-06, "loss": 0.1146, "step": 294 }, { "epoch": 2.71, "learning_rate": 2.5544351254154407e-06, "loss": 0.0909, "step": 295 }, { "epoch": 2.72, "learning_rate": 2.5217274553677975e-06, "loss": 0.1054, "step": 296 }, { "epoch": 2.72, "learning_rate": 2.489159774414252e-06, "loss": 0.0909, "step": 297 }, { "epoch": 2.73, "learning_rate": 2.4567339221911086e-06, "loss": 0.0918, "step": 298 }, { "epoch": 2.74, "learning_rate": 2.424451730323261e-06, "loss": 0.1043, "step": 299 }, { "epoch": 2.75, "learning_rate": 2.3923150223207176e-06, "loss": 0.1139, "step": 300 }, { "epoch": 2.76, "learning_rate": 2.3603256134756066e-06, "loss": 0.0958, "step": 301 }, { "epoch": 2.77, "learning_rate": 2.328485310759635e-06, "loss": 0.1115, "step": 302 }, { "epoch": 2.78, "learning_rate": 2.296795912722014e-06, "loss": 0.092, "step": 303 }, { "epoch": 2.79, "learning_rate": 2.265259209387867e-06, "loss": 0.1011, "step": 304 }, { "epoch": 2.8, "learning_rate": 2.2338769821571225e-06, "loss": 0.106, "step": 305 }, { "epoch": 2.81, "learning_rate": 2.202651003703885e-06, "loss": 0.1053, "step": 306 }, { "epoch": 2.82, "learning_rate": 2.1715830378763025e-06, "loss": 0.1314, "step": 307 }, { "epoch": 2.83, "learning_rate": 2.140674839596931e-06, "loss": 0.1252, "step": 308 }, { "epoch": 2.83, "learning_rate": 2.109928154763606e-06, "loss": 0.1032, "step": 309 }, { "epoch": 2.84, "learning_rate": 2.0793447201508288e-06, "loss": 0.0911, "step": 310 }, { "epoch": 2.85, "learning_rate": 2.0489262633116536e-06, "loss": 0.0865, "step": 311 }, { "epoch": 2.86, "learning_rate": 2.01867450248011e-06, "loss": 0.1087, "step": 312 }, { "epoch": 2.87, "learning_rate": 1.9885911464741413e-06, "loss": 0.1136, "step": 313 }, { "epoch": 2.88, "learning_rate": 1.9586778945990785e-06, "loss": 0.0873, "step": 314 }, { "epoch": 2.89, "learning_rate": 1.928936436551661e-06, "loss": 0.1155, "step": 315 }, { "epoch": 2.9, "learning_rate": 1.8993684523245842e-06, "loss": 0.117, "step": 316 }, { "epoch": 2.91, "learning_rate": 1.8699756121115997e-06, "loss": 0.1083, "step": 317 }, { "epoch": 2.92, "learning_rate": 1.8407595762131814e-06, "loss": 0.1099, "step": 318 }, { "epoch": 2.93, "learning_rate": 1.811721994942731e-06, "loss": 0.1183, "step": 319 }, { "epoch": 2.94, "learning_rate": 1.7828645085333645e-06, "loss": 0.1168, "step": 320 }, { "epoch": 2.94, "learning_rate": 1.7541887470452606e-06, "loss": 0.0985, "step": 321 }, { "epoch": 2.95, "learning_rate": 1.7256963302735752e-06, "loss": 0.1012, "step": 322 }, { "epoch": 2.96, "learning_rate": 1.6973888676569594e-06, "loss": 0.1107, "step": 323 }, { "epoch": 2.97, "learning_rate": 1.6692679581866334e-06, "loss": 0.1301, "step": 324 }, { "epoch": 2.98, "learning_rate": 1.6413351903160763e-06, "loss": 0.1022, "step": 325 }, { "epoch": 2.99, "learning_rate": 1.6135921418712959e-06, "loss": 0.1213, "step": 326 }, { "epoch": 3.0, "learning_rate": 1.5860403799616951e-06, "loss": 0.1093, "step": 327 }, { "epoch": 3.01, "learning_rate": 1.5586814608915673e-06, "loss": 0.0554, "step": 328 }, { "epoch": 3.02, "learning_rate": 1.5315169300721694e-06, "loss": 0.084, "step": 329 }, { "epoch": 3.03, "learning_rate": 1.5045483219344387e-06, "loss": 0.0741, "step": 330 } ], "max_steps": 436, "num_train_epochs": 4, "total_flos": 1.2638375430617825e+18, "trial_name": null, "trial_params": null }