{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 2055, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00291970802919708, "grad_norm": 2.4375, "learning_rate": 9.67741935483871e-07, "loss": 1.3209164142608643, "step": 2 }, { "epoch": 0.00583941605839416, "grad_norm": 10.8125, "learning_rate": 2.9032258064516128e-06, "loss": 2.3312554359436035, "step": 4 }, { "epoch": 0.008759124087591242, "grad_norm": 3.921875, "learning_rate": 4.838709677419355e-06, "loss": 1.9411218166351318, "step": 6 }, { "epoch": 0.01167883211678832, "grad_norm": 1.1796875, "learning_rate": 6.774193548387097e-06, "loss": 1.6758718490600586, "step": 8 }, { "epoch": 0.014598540145985401, "grad_norm": 1.8046875, "learning_rate": 8.70967741935484e-06, "loss": 1.7799521684646606, "step": 10 }, { "epoch": 0.017518248175182483, "grad_norm": 3.171875, "learning_rate": 1.0645161290322582e-05, "loss": 1.885763168334961, "step": 12 }, { "epoch": 0.020437956204379562, "grad_norm": 9.0, "learning_rate": 1.2580645161290324e-05, "loss": 1.5260975360870361, "step": 14 }, { "epoch": 0.02335766423357664, "grad_norm": 2.890625, "learning_rate": 1.4516129032258065e-05, "loss": 1.6625151634216309, "step": 16 }, { "epoch": 0.026277372262773723, "grad_norm": 1.828125, "learning_rate": 1.6451612903225807e-05, "loss": 1.7963693141937256, "step": 18 }, { "epoch": 0.029197080291970802, "grad_norm": 2.15625, "learning_rate": 1.838709677419355e-05, "loss": 1.750580072402954, "step": 20 }, { "epoch": 0.032116788321167884, "grad_norm": 1.90625, "learning_rate": 2.032258064516129e-05, "loss": 1.7857003211975098, "step": 22 }, { "epoch": 0.035036496350364967, "grad_norm": 1.6328125, "learning_rate": 2.2258064516129034e-05, "loss": 1.6380910873413086, "step": 24 }, { "epoch": 0.03795620437956204, "grad_norm": 6.375, "learning_rate": 2.4193548387096773e-05, "loss": 1.709096908569336, "step": 26 }, { "epoch": 0.040875912408759124, "grad_norm": 3.734375, "learning_rate": 2.6129032258064516e-05, "loss": 1.9888923168182373, "step": 28 }, { "epoch": 0.043795620437956206, "grad_norm": 1.53125, "learning_rate": 2.806451612903226e-05, "loss": 1.2525625228881836, "step": 30 }, { "epoch": 0.04671532846715328, "grad_norm": 3.15625, "learning_rate": 3e-05, "loss": 1.5124564170837402, "step": 32 }, { "epoch": 0.049635036496350364, "grad_norm": 1.515625, "learning_rate": 3.193548387096774e-05, "loss": 1.342097282409668, "step": 34 }, { "epoch": 0.052554744525547446, "grad_norm": 5.34375, "learning_rate": 3.387096774193549e-05, "loss": 1.1437346935272217, "step": 36 }, { "epoch": 0.05547445255474453, "grad_norm": 2.921875, "learning_rate": 3.580645161290323e-05, "loss": 1.6300029754638672, "step": 38 }, { "epoch": 0.058394160583941604, "grad_norm": 0.9375, "learning_rate": 3.774193548387097e-05, "loss": 1.405299186706543, "step": 40 }, { "epoch": 0.061313868613138686, "grad_norm": 4.40625, "learning_rate": 3.967741935483871e-05, "loss": 1.386954426765442, "step": 42 }, { "epoch": 0.06423357664233577, "grad_norm": 1.5078125, "learning_rate": 4.161290322580646e-05, "loss": 1.3544518947601318, "step": 44 }, { "epoch": 0.06715328467153285, "grad_norm": 3.40625, "learning_rate": 4.3548387096774194e-05, "loss": 1.6254360675811768, "step": 46 }, { "epoch": 0.07007299270072993, "grad_norm": 3.875, "learning_rate": 4.548387096774194e-05, "loss": 1.3457388877868652, "step": 48 }, { "epoch": 0.072992700729927, "grad_norm": 1.0625, "learning_rate": 4.741935483870968e-05, "loss": 1.232921838760376, "step": 50 }, { "epoch": 0.07591240875912408, "grad_norm": 1.640625, "learning_rate": 4.935483870967742e-05, "loss": 1.1997480392456055, "step": 52 }, { "epoch": 0.07883211678832117, "grad_norm": 1.203125, "learning_rate": 5.1290322580645164e-05, "loss": 1.5290566682815552, "step": 54 }, { "epoch": 0.08175182481751825, "grad_norm": 1.234375, "learning_rate": 5.3225806451612906e-05, "loss": 1.553056240081787, "step": 56 }, { "epoch": 0.08467153284671533, "grad_norm": 1.09375, "learning_rate": 5.516129032258064e-05, "loss": 1.3679280281066895, "step": 58 }, { "epoch": 0.08759124087591241, "grad_norm": 1.5234375, "learning_rate": 5.709677419354839e-05, "loss": 1.4744219779968262, "step": 60 }, { "epoch": 0.0905109489051095, "grad_norm": 4.53125, "learning_rate": 5.9032258064516134e-05, "loss": 1.6063017845153809, "step": 62 }, { "epoch": 0.09343065693430656, "grad_norm": 1.828125, "learning_rate": 5.9999968940456305e-05, "loss": 1.4462177753448486, "step": 64 }, { "epoch": 0.09635036496350365, "grad_norm": 5.90625, "learning_rate": 5.999972046456981e-05, "loss": 1.3050040006637573, "step": 66 }, { "epoch": 0.09927007299270073, "grad_norm": 0.96875, "learning_rate": 5.9999223515266435e-05, "loss": 1.4365456104278564, "step": 68 }, { "epoch": 0.10218978102189781, "grad_norm": 1.1953125, "learning_rate": 5.999847809748539e-05, "loss": 1.34994637966156, "step": 70 }, { "epoch": 0.10510948905109489, "grad_norm": 6.28125, "learning_rate": 5.99974842186354e-05, "loss": 0.9953607320785522, "step": 72 }, { "epoch": 0.10802919708029197, "grad_norm": 1.6953125, "learning_rate": 5.999624188859468e-05, "loss": 1.26670241355896, "step": 74 }, { "epoch": 0.11094890510948906, "grad_norm": 1.3671875, "learning_rate": 5.999475111971082e-05, "loss": 1.3074488639831543, "step": 76 }, { "epoch": 0.11386861313868613, "grad_norm": 2.015625, "learning_rate": 5.999301192680064e-05, "loss": 1.4332489967346191, "step": 78 }, { "epoch": 0.11678832116788321, "grad_norm": 1.828125, "learning_rate": 5.999102432715005e-05, "loss": 1.0071167945861816, "step": 80 }, { "epoch": 0.11970802919708029, "grad_norm": 1.5390625, "learning_rate": 5.998878834051391e-05, "loss": 1.5147018432617188, "step": 82 }, { "epoch": 0.12262773722627737, "grad_norm": 1.9921875, "learning_rate": 5.998630398911579e-05, "loss": 1.3494911193847656, "step": 84 }, { "epoch": 0.12554744525547445, "grad_norm": 1.5859375, "learning_rate": 5.998357129764777e-05, "loss": 1.5120820999145508, "step": 86 }, { "epoch": 0.12846715328467154, "grad_norm": 1.25, "learning_rate": 5.998059029327023e-05, "loss": 1.4139865636825562, "step": 88 }, { "epoch": 0.13138686131386862, "grad_norm": 2.3125, "learning_rate": 5.9977361005611484e-05, "loss": 1.223384141921997, "step": 90 }, { "epoch": 0.1343065693430657, "grad_norm": 2.734375, "learning_rate": 5.9973883466767594e-05, "loss": 1.474729299545288, "step": 92 }, { "epoch": 0.13722627737226278, "grad_norm": 0.796875, "learning_rate": 5.997015771130197e-05, "loss": 1.2821106910705566, "step": 94 }, { "epoch": 0.14014598540145987, "grad_norm": 1.46875, "learning_rate": 5.996618377624509e-05, "loss": 1.6019928455352783, "step": 96 }, { "epoch": 0.14306569343065692, "grad_norm": 1.171875, "learning_rate": 5.996196170109409e-05, "loss": 1.3877289295196533, "step": 98 }, { "epoch": 0.145985401459854, "grad_norm": 0.6171875, "learning_rate": 5.995749152781236e-05, "loss": 1.3422930240631104, "step": 100 }, { "epoch": 0.14890510948905109, "grad_norm": 1.5078125, "learning_rate": 5.9952773300829165e-05, "loss": 1.4275691509246826, "step": 102 }, { "epoch": 0.15182481751824817, "grad_norm": 0.91796875, "learning_rate": 5.9947807067039196e-05, "loss": 1.3908946514129639, "step": 104 }, { "epoch": 0.15474452554744525, "grad_norm": 3.265625, "learning_rate": 5.9942592875802085e-05, "loss": 1.590527057647705, "step": 106 }, { "epoch": 0.15766423357664233, "grad_norm": 2.984375, "learning_rate": 5.993713077894193e-05, "loss": 1.6489572525024414, "step": 108 }, { "epoch": 0.16058394160583941, "grad_norm": 2.015625, "learning_rate": 5.993142083074676e-05, "loss": 1.2806832790374756, "step": 110 }, { "epoch": 0.1635036496350365, "grad_norm": 0.8046875, "learning_rate": 5.992546308796806e-05, "loss": 1.4471540451049805, "step": 112 }, { "epoch": 0.16642335766423358, "grad_norm": 1.5859375, "learning_rate": 5.991925760982008e-05, "loss": 1.28468918800354, "step": 114 }, { "epoch": 0.16934306569343066, "grad_norm": 0.84375, "learning_rate": 5.9912804457979376e-05, "loss": 1.4314072132110596, "step": 116 }, { "epoch": 0.17226277372262774, "grad_norm": 2.390625, "learning_rate": 5.990610369658414e-05, "loss": 1.3306201696395874, "step": 118 }, { "epoch": 0.17518248175182483, "grad_norm": 0.9296875, "learning_rate": 5.989915539223357e-05, "loss": 1.5853054523468018, "step": 120 }, { "epoch": 0.1781021897810219, "grad_norm": 0.7890625, "learning_rate": 5.989195961398716e-05, "loss": 1.652395486831665, "step": 122 }, { "epoch": 0.181021897810219, "grad_norm": 0.84765625, "learning_rate": 5.988451643336413e-05, "loss": 1.625920295715332, "step": 124 }, { "epoch": 0.18394160583941604, "grad_norm": 1.21875, "learning_rate": 5.9876825924342584e-05, "loss": 1.5102442502975464, "step": 126 }, { "epoch": 0.18686131386861313, "grad_norm": 0.734375, "learning_rate": 5.9868888163358875e-05, "loss": 1.6674467325210571, "step": 128 }, { "epoch": 0.1897810218978102, "grad_norm": 0.6484375, "learning_rate": 5.986070322930676e-05, "loss": 1.6165754795074463, "step": 130 }, { "epoch": 0.1927007299270073, "grad_norm": 2.65625, "learning_rate": 5.985227120353674e-05, "loss": 1.3306710720062256, "step": 132 }, { "epoch": 0.19562043795620437, "grad_norm": 1.6328125, "learning_rate": 5.9843592169855085e-05, "loss": 1.5615007877349854, "step": 134 }, { "epoch": 0.19854014598540146, "grad_norm": 0.8125, "learning_rate": 5.983466621452313e-05, "loss": 1.549973964691162, "step": 136 }, { "epoch": 0.20145985401459854, "grad_norm": 0.365234375, "learning_rate": 5.982549342625639e-05, "loss": 1.3460943698883057, "step": 138 }, { "epoch": 0.20437956204379562, "grad_norm": 1.7265625, "learning_rate": 5.981607389622362e-05, "loss": 1.4986615180969238, "step": 140 }, { "epoch": 0.2072992700729927, "grad_norm": 0.83984375, "learning_rate": 5.980640771804599e-05, "loss": 1.6788625717163086, "step": 142 }, { "epoch": 0.21021897810218979, "grad_norm": 0.82421875, "learning_rate": 5.979649498779611e-05, "loss": 1.4583231210708618, "step": 144 }, { "epoch": 0.21313868613138687, "grad_norm": 1.875, "learning_rate": 5.978633580399709e-05, "loss": 1.6779685020446777, "step": 146 }, { "epoch": 0.21605839416058395, "grad_norm": 0.7734375, "learning_rate": 5.977593026762152e-05, "loss": 1.3658714294433594, "step": 148 }, { "epoch": 0.21897810218978103, "grad_norm": 0.68359375, "learning_rate": 5.976527848209053e-05, "loss": 1.4529236555099487, "step": 150 }, { "epoch": 0.22189781021897811, "grad_norm": 0.84765625, "learning_rate": 5.975438055327275e-05, "loss": 0.9998324513435364, "step": 152 }, { "epoch": 0.22481751824817517, "grad_norm": 0.6328125, "learning_rate": 5.974323658948319e-05, "loss": 1.1067932844161987, "step": 154 }, { "epoch": 0.22773722627737225, "grad_norm": 3.34375, "learning_rate": 5.973184670148224e-05, "loss": 0.49964094161987305, "step": 156 }, { "epoch": 0.23065693430656933, "grad_norm": 0.796875, "learning_rate": 5.972021100247455e-05, "loss": 1.6203181743621826, "step": 158 }, { "epoch": 0.23357664233576642, "grad_norm": 0.73828125, "learning_rate": 5.9708329608107895e-05, "loss": 1.2763311862945557, "step": 160 }, { "epoch": 0.2364963503649635, "grad_norm": 0.94140625, "learning_rate": 5.9696202636472e-05, "loss": 1.600427508354187, "step": 162 }, { "epoch": 0.23941605839416058, "grad_norm": 0.8671875, "learning_rate": 5.968383020809743e-05, "loss": 1.2396719455718994, "step": 164 }, { "epoch": 0.24233576642335766, "grad_norm": 2.265625, "learning_rate": 5.9671212445954354e-05, "loss": 1.539351224899292, "step": 166 }, { "epoch": 0.24525547445255474, "grad_norm": 0.45703125, "learning_rate": 5.965834947545128e-05, "loss": 1.2752540111541748, "step": 168 }, { "epoch": 0.24817518248175183, "grad_norm": 1.875, "learning_rate": 5.964524142443392e-05, "loss": 1.5187466144561768, "step": 170 }, { "epoch": 0.2510948905109489, "grad_norm": 0.75, "learning_rate": 5.96318884231838e-05, "loss": 1.3301482200622559, "step": 172 }, { "epoch": 0.25401459854014596, "grad_norm": 1.0390625, "learning_rate": 5.961829060441704e-05, "loss": 1.5802184343338013, "step": 174 }, { "epoch": 0.2569343065693431, "grad_norm": 0.64453125, "learning_rate": 5.960444810328303e-05, "loss": 1.2438929080963135, "step": 176 }, { "epoch": 0.25985401459854013, "grad_norm": 0.76953125, "learning_rate": 5.959036105736303e-05, "loss": 1.2162034511566162, "step": 178 }, { "epoch": 0.26277372262773724, "grad_norm": 1.5, "learning_rate": 5.9576029606668856e-05, "loss": 1.469679832458496, "step": 180 }, { "epoch": 0.2656934306569343, "grad_norm": 1.0546875, "learning_rate": 5.9561453893641505e-05, "loss": 1.5327503681182861, "step": 182 }, { "epoch": 0.2686131386861314, "grad_norm": 0.5625, "learning_rate": 5.954663406314968e-05, "loss": 1.210360050201416, "step": 184 }, { "epoch": 0.27153284671532846, "grad_norm": 0.77734375, "learning_rate": 5.953157026248836e-05, "loss": 1.4078571796417236, "step": 186 }, { "epoch": 0.27445255474452557, "grad_norm": 0.9296875, "learning_rate": 5.951626264137741e-05, "loss": 1.2318487167358398, "step": 188 }, { "epoch": 0.2773722627737226, "grad_norm": 0.6640625, "learning_rate": 5.9500711351960006e-05, "loss": 1.3978914022445679, "step": 190 }, { "epoch": 0.28029197080291973, "grad_norm": 2.703125, "learning_rate": 5.9484916548801124e-05, "loss": 1.5884613990783691, "step": 192 }, { "epoch": 0.2832116788321168, "grad_norm": 0.765625, "learning_rate": 5.946887838888612e-05, "loss": 1.6306736469268799, "step": 194 }, { "epoch": 0.28613138686131384, "grad_norm": 0.88671875, "learning_rate": 5.945259703161898e-05, "loss": 1.4341470003128052, "step": 196 }, { "epoch": 0.28905109489051095, "grad_norm": 0.8828125, "learning_rate": 5.943607263882093e-05, "loss": 1.4493606090545654, "step": 198 }, { "epoch": 0.291970802919708, "grad_norm": 1.171875, "learning_rate": 5.9419305374728686e-05, "loss": 1.7190909385681152, "step": 200 }, { "epoch": 0.2948905109489051, "grad_norm": 2.1875, "learning_rate": 5.940229540599288e-05, "loss": 1.4846527576446533, "step": 202 }, { "epoch": 0.29781021897810217, "grad_norm": 0.9453125, "learning_rate": 5.938504290167642e-05, "loss": 1.6837434768676758, "step": 204 }, { "epoch": 0.3007299270072993, "grad_norm": 0.73046875, "learning_rate": 5.936754803325276e-05, "loss": 1.4253846406936646, "step": 206 }, { "epoch": 0.30364963503649633, "grad_norm": 0.84765625, "learning_rate": 5.934981097460425e-05, "loss": 1.7993555068969727, "step": 208 }, { "epoch": 0.30656934306569344, "grad_norm": 0.71484375, "learning_rate": 5.9331831902020325e-05, "loss": 1.6041544675827026, "step": 210 }, { "epoch": 0.3094890510948905, "grad_norm": 0.5546875, "learning_rate": 5.9313610994195884e-05, "loss": 1.1635921001434326, "step": 212 }, { "epoch": 0.3124087591240876, "grad_norm": 0.74609375, "learning_rate": 5.929514843222939e-05, "loss": 1.0920753479003906, "step": 214 }, { "epoch": 0.31532846715328466, "grad_norm": 1.9921875, "learning_rate": 5.9276444399621147e-05, "loss": 1.2639129161834717, "step": 216 }, { "epoch": 0.3182481751824818, "grad_norm": 3.671875, "learning_rate": 5.9257499082271386e-05, "loss": 1.0223140716552734, "step": 218 }, { "epoch": 0.32116788321167883, "grad_norm": 1.1171875, "learning_rate": 5.9238312668478556e-05, "loss": 1.6704933643341064, "step": 220 }, { "epoch": 0.32408759124087594, "grad_norm": 1.7578125, "learning_rate": 5.921888534893733e-05, "loss": 1.6343793869018555, "step": 222 }, { "epoch": 0.327007299270073, "grad_norm": 1.8984375, "learning_rate": 5.919921731673677e-05, "loss": 1.209860920906067, "step": 224 }, { "epoch": 0.32992700729927005, "grad_norm": 1.03125, "learning_rate": 5.917930876735839e-05, "loss": 1.49936044216156, "step": 226 }, { "epoch": 0.33284671532846716, "grad_norm": 10.4375, "learning_rate": 5.9159159898674204e-05, "loss": 1.279283046722412, "step": 228 }, { "epoch": 0.3357664233576642, "grad_norm": 1.4609375, "learning_rate": 5.913877091094479e-05, "loss": 1.3200433254241943, "step": 230 }, { "epoch": 0.3386861313868613, "grad_norm": 3.328125, "learning_rate": 5.911814200681727e-05, "loss": 0.9882813692092896, "step": 232 }, { "epoch": 0.3416058394160584, "grad_norm": 1.09375, "learning_rate": 5.909727339132331e-05, "loss": 1.4972972869873047, "step": 234 }, { "epoch": 0.3445255474452555, "grad_norm": 2.234375, "learning_rate": 5.907616527187707e-05, "loss": 1.658118724822998, "step": 236 }, { "epoch": 0.34744525547445254, "grad_norm": 0.86328125, "learning_rate": 5.905481785827316e-05, "loss": 1.481203556060791, "step": 238 }, { "epoch": 0.35036496350364965, "grad_norm": 1.765625, "learning_rate": 5.9033231362684553e-05, "loss": 1.8912521600723267, "step": 240 }, { "epoch": 0.3532846715328467, "grad_norm": 1.1015625, "learning_rate": 5.901140599966045e-05, "loss": 1.414005994796753, "step": 242 }, { "epoch": 0.3562043795620438, "grad_norm": 1.1328125, "learning_rate": 5.8989341986124194e-05, "loss": 1.3400169610977173, "step": 244 }, { "epoch": 0.35912408759124087, "grad_norm": 0.78125, "learning_rate": 5.896703954137103e-05, "loss": 1.231393814086914, "step": 246 }, { "epoch": 0.362043795620438, "grad_norm": 1.2578125, "learning_rate": 5.894449888706608e-05, "loss": 1.6056867837905884, "step": 248 }, { "epoch": 0.36496350364963503, "grad_norm": 1.3359375, "learning_rate": 5.892172024724195e-05, "loss": 1.2110869884490967, "step": 250 }, { "epoch": 0.3678832116788321, "grad_norm": 3.515625, "learning_rate": 5.889870384829663e-05, "loss": 1.5161144733428955, "step": 252 }, { "epoch": 0.3708029197080292, "grad_norm": 2.015625, "learning_rate": 5.8875449918991225e-05, "loss": 1.719308853149414, "step": 254 }, { "epoch": 0.37372262773722625, "grad_norm": 0.89453125, "learning_rate": 5.885195869044763e-05, "loss": 1.349236249923706, "step": 256 }, { "epoch": 0.37664233576642336, "grad_norm": 1.046875, "learning_rate": 5.8828230396146294e-05, "loss": 1.4929360151290894, "step": 258 }, { "epoch": 0.3795620437956204, "grad_norm": 1.765625, "learning_rate": 5.8804265271923895e-05, "loss": 1.4849441051483154, "step": 260 }, { "epoch": 0.38248175182481753, "grad_norm": 1.15625, "learning_rate": 5.87800635559709e-05, "loss": 1.3683178424835205, "step": 262 }, { "epoch": 0.3854014598540146, "grad_norm": 0.8046875, "learning_rate": 5.875562548882938e-05, "loss": 1.3490722179412842, "step": 264 }, { "epoch": 0.3883211678832117, "grad_norm": 0.78125, "learning_rate": 5.87309513133904e-05, "loss": 1.5665984153747559, "step": 266 }, { "epoch": 0.39124087591240875, "grad_norm": 0.6484375, "learning_rate": 5.870604127489179e-05, "loss": 1.312342643737793, "step": 268 }, { "epoch": 0.39416058394160586, "grad_norm": 2.46875, "learning_rate": 5.868089562091562e-05, "loss": 1.5042738914489746, "step": 270 }, { "epoch": 0.3970802919708029, "grad_norm": 0.74609375, "learning_rate": 5.865551460138574e-05, "loss": 1.2545225620269775, "step": 272 }, { "epoch": 0.4, "grad_norm": 1.6484375, "learning_rate": 5.862989846856532e-05, "loss": 1.7891290187835693, "step": 274 }, { "epoch": 0.4029197080291971, "grad_norm": 0.7890625, "learning_rate": 5.860404747705432e-05, "loss": 1.427787184715271, "step": 276 }, { "epoch": 0.4058394160583942, "grad_norm": 1.1015625, "learning_rate": 5.857796188378699e-05, "loss": 1.2621533870697021, "step": 278 }, { "epoch": 0.40875912408759124, "grad_norm": 0.5390625, "learning_rate": 5.855164194802929e-05, "loss": 1.6562426090240479, "step": 280 }, { "epoch": 0.4116788321167883, "grad_norm": 0.87890625, "learning_rate": 5.85250879313763e-05, "loss": 1.4017126560211182, "step": 282 }, { "epoch": 0.4145985401459854, "grad_norm": 1.0703125, "learning_rate": 5.8498300097749684e-05, "loss": 1.4500172138214111, "step": 284 }, { "epoch": 0.41751824817518246, "grad_norm": 0.609375, "learning_rate": 5.847127871339499e-05, "loss": 1.394424319267273, "step": 286 }, { "epoch": 0.42043795620437957, "grad_norm": 0.88671875, "learning_rate": 5.844402404687906e-05, "loss": 1.4016766548156738, "step": 288 }, { "epoch": 0.4233576642335766, "grad_norm": 0.80078125, "learning_rate": 5.8416536369087306e-05, "loss": 1.4250332117080688, "step": 290 }, { "epoch": 0.42627737226277373, "grad_norm": 0.828125, "learning_rate": 5.8388815953221104e-05, "loss": 1.2282607555389404, "step": 292 }, { "epoch": 0.4291970802919708, "grad_norm": 0.9296875, "learning_rate": 5.836086307479496e-05, "loss": 1.278720736503601, "step": 294 }, { "epoch": 0.4321167883211679, "grad_norm": 1.3671875, "learning_rate": 5.833267801163389e-05, "loss": 1.6063511371612549, "step": 296 }, { "epoch": 0.43503649635036495, "grad_norm": 0.66796875, "learning_rate": 5.8304261043870564e-05, "loss": 1.606541633605957, "step": 298 }, { "epoch": 0.43795620437956206, "grad_norm": 1.8359375, "learning_rate": 5.8275612453942596e-05, "loss": 1.1832222938537598, "step": 300 }, { "epoch": 0.4408759124087591, "grad_norm": 1.5703125, "learning_rate": 5.8246732526589704e-05, "loss": 1.3833293914794922, "step": 302 }, { "epoch": 0.44379562043795623, "grad_norm": 0.83203125, "learning_rate": 5.821762154885085e-05, "loss": 1.4438910484313965, "step": 304 }, { "epoch": 0.4467153284671533, "grad_norm": 1.09375, "learning_rate": 5.818827981006144e-05, "loss": 1.263401985168457, "step": 306 }, { "epoch": 0.44963503649635034, "grad_norm": 3.5625, "learning_rate": 5.815870760185041e-05, "loss": 1.0707682371139526, "step": 308 }, { "epoch": 0.45255474452554745, "grad_norm": 2.28125, "learning_rate": 5.812890521813736e-05, "loss": 1.8427000045776367, "step": 310 }, { "epoch": 0.4554744525547445, "grad_norm": 0.98046875, "learning_rate": 5.8098872955129595e-05, "loss": 1.423462152481079, "step": 312 }, { "epoch": 0.4583941605839416, "grad_norm": 0.69921875, "learning_rate": 5.806861111131923e-05, "loss": 1.3223985433578491, "step": 314 }, { "epoch": 0.46131386861313867, "grad_norm": 0.76953125, "learning_rate": 5.803811998748017e-05, "loss": 1.369985818862915, "step": 316 }, { "epoch": 0.4642335766423358, "grad_norm": 0.91015625, "learning_rate": 5.8007399886665145e-05, "loss": 1.402444839477539, "step": 318 }, { "epoch": 0.46715328467153283, "grad_norm": 0.76171875, "learning_rate": 5.797645111420271e-05, "loss": 1.665392518043518, "step": 320 }, { "epoch": 0.47007299270072994, "grad_norm": 0.84375, "learning_rate": 5.7945273977694205e-05, "loss": 1.399198055267334, "step": 322 }, { "epoch": 0.472992700729927, "grad_norm": 0.72265625, "learning_rate": 5.791386878701066e-05, "loss": 1.3573551177978516, "step": 324 }, { "epoch": 0.4759124087591241, "grad_norm": 1.2265625, "learning_rate": 5.7882235854289785e-05, "loss": 1.3456391096115112, "step": 326 }, { "epoch": 0.47883211678832116, "grad_norm": 1.078125, "learning_rate": 5.78503754939328e-05, "loss": 1.5439651012420654, "step": 328 }, { "epoch": 0.48175182481751827, "grad_norm": 0.9453125, "learning_rate": 5.7818288022601355e-05, "loss": 1.1991386413574219, "step": 330 }, { "epoch": 0.4846715328467153, "grad_norm": 1.0, "learning_rate": 5.7785973759214365e-05, "loss": 1.4791953563690186, "step": 332 }, { "epoch": 0.48759124087591244, "grad_norm": 1.4609375, "learning_rate": 5.7753433024944834e-05, "loss": 1.6375491619110107, "step": 334 }, { "epoch": 0.4905109489051095, "grad_norm": 0.419921875, "learning_rate": 5.772066614321667e-05, "loss": 1.0832901000976562, "step": 336 }, { "epoch": 0.49343065693430654, "grad_norm": 2.546875, "learning_rate": 5.7687673439701484e-05, "loss": 1.662909746170044, "step": 338 }, { "epoch": 0.49635036496350365, "grad_norm": 1.25, "learning_rate": 5.7654455242315315e-05, "loss": 1.498556137084961, "step": 340 }, { "epoch": 0.4992700729927007, "grad_norm": 2.53125, "learning_rate": 5.7621011881215456e-05, "loss": 1.5630645751953125, "step": 342 }, { "epoch": 0.5021897810218978, "grad_norm": 1.21875, "learning_rate": 5.758734368879703e-05, "loss": 1.4207791090011597, "step": 344 }, { "epoch": 0.5051094890510949, "grad_norm": 0.80859375, "learning_rate": 5.755345099968983e-05, "loss": 1.4251861572265625, "step": 346 }, { "epoch": 0.5080291970802919, "grad_norm": 0.6015625, "learning_rate": 5.751933415075492e-05, "loss": 1.4990272521972656, "step": 348 }, { "epoch": 0.5109489051094891, "grad_norm": 0.78125, "learning_rate": 5.74849934810813e-05, "loss": 1.2336238622665405, "step": 350 }, { "epoch": 0.5138686131386861, "grad_norm": 1.0859375, "learning_rate": 5.7450429331982476e-05, "loss": 1.613104224205017, "step": 352 }, { "epoch": 0.5167883211678832, "grad_norm": 0.69140625, "learning_rate": 5.741564204699324e-05, "loss": 1.3290159702301025, "step": 354 }, { "epoch": 0.5197080291970803, "grad_norm": 0.6796875, "learning_rate": 5.738063197186605e-05, "loss": 1.3364830017089844, "step": 356 }, { "epoch": 0.5226277372262774, "grad_norm": 1.109375, "learning_rate": 5.734539945456771e-05, "loss": 1.415923833847046, "step": 358 }, { "epoch": 0.5255474452554745, "grad_norm": 0.65625, "learning_rate": 5.730994484527591e-05, "loss": 1.5084344148635864, "step": 360 }, { "epoch": 0.5284671532846715, "grad_norm": 1.3125, "learning_rate": 5.7274268496375714e-05, "loss": 1.54220449924469, "step": 362 }, { "epoch": 0.5313868613138686, "grad_norm": 0.8203125, "learning_rate": 5.723837076245603e-05, "loss": 1.1953389644622803, "step": 364 }, { "epoch": 0.5343065693430656, "grad_norm": 1.7578125, "learning_rate": 5.7202252000306173e-05, "loss": 1.7339110374450684, "step": 366 }, { "epoch": 0.5372262773722628, "grad_norm": 0.77734375, "learning_rate": 5.716591256891225e-05, "loss": 1.3485782146453857, "step": 368 }, { "epoch": 0.5401459854014599, "grad_norm": 0.7109375, "learning_rate": 5.712935282945359e-05, "loss": 1.2649073600769043, "step": 370 }, { "epoch": 0.5430656934306569, "grad_norm": 1.7421875, "learning_rate": 5.709257314529922e-05, "loss": 1.0904498100280762, "step": 372 }, { "epoch": 0.545985401459854, "grad_norm": 0.8046875, "learning_rate": 5.705557388200414e-05, "loss": 1.5211831331253052, "step": 374 }, { "epoch": 0.5489051094890511, "grad_norm": 2.5, "learning_rate": 5.7018355407305835e-05, "loss": 1.6481215953826904, "step": 376 }, { "epoch": 0.5518248175182482, "grad_norm": 3.21875, "learning_rate": 5.6980918091120475e-05, "loss": 1.6256749629974365, "step": 378 }, { "epoch": 0.5547445255474452, "grad_norm": 0.859375, "learning_rate": 5.694326230553934e-05, "loss": 1.287406086921692, "step": 380 }, { "epoch": 0.5576642335766423, "grad_norm": 0.76171875, "learning_rate": 5.690538842482509e-05, "loss": 1.3722602128982544, "step": 382 }, { "epoch": 0.5605839416058395, "grad_norm": 0.69921875, "learning_rate": 5.686729682540806e-05, "loss": 1.3319132328033447, "step": 384 }, { "epoch": 0.5635036496350365, "grad_norm": 0.61328125, "learning_rate": 5.6828987885882424e-05, "loss": 1.3779847621917725, "step": 386 }, { "epoch": 0.5664233576642336, "grad_norm": 0.89453125, "learning_rate": 5.679046198700261e-05, "loss": 1.652022361755371, "step": 388 }, { "epoch": 0.5693430656934306, "grad_norm": 0.57421875, "learning_rate": 5.675171951167935e-05, "loss": 1.425034999847412, "step": 390 }, { "epoch": 0.5722627737226277, "grad_norm": 0.64453125, "learning_rate": 5.6712760844975945e-05, "loss": 1.275322437286377, "step": 392 }, { "epoch": 0.5751824817518248, "grad_norm": 1.15625, "learning_rate": 5.6673586374104475e-05, "loss": 1.4515531063079834, "step": 394 }, { "epoch": 0.5781021897810219, "grad_norm": 1.703125, "learning_rate": 5.663419648842187e-05, "loss": 1.622361421585083, "step": 396 }, { "epoch": 0.581021897810219, "grad_norm": 0.7421875, "learning_rate": 5.6594591579426065e-05, "loss": 1.6139750480651855, "step": 398 }, { "epoch": 0.583941605839416, "grad_norm": 1.015625, "learning_rate": 5.655477204075218e-05, "loss": 1.1461379528045654, "step": 400 }, { "epoch": 0.5868613138686132, "grad_norm": 0.69140625, "learning_rate": 5.651473826816849e-05, "loss": 1.5190107822418213, "step": 402 }, { "epoch": 0.5897810218978102, "grad_norm": 1.0859375, "learning_rate": 5.6474490659572564e-05, "loss": 1.3880577087402344, "step": 404 }, { "epoch": 0.5927007299270073, "grad_norm": 0.5703125, "learning_rate": 5.643402961498734e-05, "loss": 1.2586677074432373, "step": 406 }, { "epoch": 0.5956204379562043, "grad_norm": 1.2421875, "learning_rate": 5.6393355536557044e-05, "loss": 1.5116124153137207, "step": 408 }, { "epoch": 0.5985401459854015, "grad_norm": 0.8515625, "learning_rate": 5.635246882854331e-05, "loss": 1.4478124380111694, "step": 410 }, { "epoch": 0.6014598540145986, "grad_norm": 0.68359375, "learning_rate": 5.631136989732107e-05, "loss": 1.2513399124145508, "step": 412 }, { "epoch": 0.6043795620437956, "grad_norm": 1.109375, "learning_rate": 5.6270059151374555e-05, "loss": 1.342003345489502, "step": 414 }, { "epoch": 0.6072992700729927, "grad_norm": 0.6015625, "learning_rate": 5.622853700129328e-05, "loss": 1.3598436117172241, "step": 416 }, { "epoch": 0.6102189781021898, "grad_norm": 5.09375, "learning_rate": 5.6186803859767854e-05, "loss": 1.286314845085144, "step": 418 }, { "epoch": 0.6131386861313869, "grad_norm": 2.328125, "learning_rate": 5.614486014158596e-05, "loss": 1.7422081232070923, "step": 420 }, { "epoch": 0.6160583941605839, "grad_norm": 1.015625, "learning_rate": 5.610270626362824e-05, "loss": 1.3796322345733643, "step": 422 }, { "epoch": 0.618978102189781, "grad_norm": 1.125, "learning_rate": 5.6060342644864076e-05, "loss": 1.5898281335830688, "step": 424 }, { "epoch": 0.621897810218978, "grad_norm": 0.703125, "learning_rate": 5.6017769706347517e-05, "loss": 1.2560129165649414, "step": 426 }, { "epoch": 0.6248175182481752, "grad_norm": 0.89453125, "learning_rate": 5.597498787121305e-05, "loss": 1.4159992933273315, "step": 428 }, { "epoch": 0.6277372262773723, "grad_norm": 0.9296875, "learning_rate": 5.593199756467134e-05, "loss": 1.3794710636138916, "step": 430 }, { "epoch": 0.6306569343065693, "grad_norm": 2.265625, "learning_rate": 5.588879921400514e-05, "loss": 1.5122029781341553, "step": 432 }, { "epoch": 0.6335766423357664, "grad_norm": 0.87890625, "learning_rate": 5.584539324856488e-05, "loss": 1.33951997756958, "step": 434 }, { "epoch": 0.6364963503649635, "grad_norm": 0.62890625, "learning_rate": 5.580178009976456e-05, "loss": 1.353576421737671, "step": 436 }, { "epoch": 0.6394160583941606, "grad_norm": 2.5, "learning_rate": 5.575796020107735e-05, "loss": 1.1946585178375244, "step": 438 }, { "epoch": 0.6423357664233577, "grad_norm": 0.77734375, "learning_rate": 5.571393398803129e-05, "loss": 1.3554836511611938, "step": 440 }, { "epoch": 0.6452554744525547, "grad_norm": 0.8671875, "learning_rate": 5.566970189820504e-05, "loss": 1.3375674486160278, "step": 442 }, { "epoch": 0.6481751824817519, "grad_norm": 0.55859375, "learning_rate": 5.562526437122347e-05, "loss": 1.5423380136489868, "step": 444 }, { "epoch": 0.6510948905109489, "grad_norm": 0.7109375, "learning_rate": 5.558062184875324e-05, "loss": 1.326158046722412, "step": 446 }, { "epoch": 0.654014598540146, "grad_norm": 2.390625, "learning_rate": 5.553577477449855e-05, "loss": 1.5753459930419922, "step": 448 }, { "epoch": 0.656934306569343, "grad_norm": 0.62109375, "learning_rate": 5.5490723594196604e-05, "loss": 1.6284210681915283, "step": 450 }, { "epoch": 0.6598540145985401, "grad_norm": 0.63671875, "learning_rate": 5.544546875561323e-05, "loss": 1.5892188549041748, "step": 452 }, { "epoch": 0.6627737226277373, "grad_norm": 0.373046875, "learning_rate": 5.5400010708538445e-05, "loss": 1.3458421230316162, "step": 454 }, { "epoch": 0.6656934306569343, "grad_norm": 0.66796875, "learning_rate": 5.535434990478194e-05, "loss": 1.395503044128418, "step": 456 }, { "epoch": 0.6686131386861314, "grad_norm": 3.078125, "learning_rate": 5.530848679816864e-05, "loss": 1.215213656425476, "step": 458 }, { "epoch": 0.6715328467153284, "grad_norm": 0.609375, "learning_rate": 5.526242184453416e-05, "loss": 1.3465626239776611, "step": 460 }, { "epoch": 0.6744525547445256, "grad_norm": 0.6171875, "learning_rate": 5.521615550172024e-05, "loss": 1.3042705059051514, "step": 462 }, { "epoch": 0.6773722627737226, "grad_norm": 0.7890625, "learning_rate": 5.516968822957031e-05, "loss": 1.0415093898773193, "step": 464 }, { "epoch": 0.6802919708029197, "grad_norm": 1.0, "learning_rate": 5.5123020489924794e-05, "loss": 1.292935848236084, "step": 466 }, { "epoch": 0.6832116788321168, "grad_norm": 0.60546875, "learning_rate": 5.507615274661659e-05, "loss": 1.3761882781982422, "step": 468 }, { "epoch": 0.6861313868613139, "grad_norm": 3.3125, "learning_rate": 5.502908546546645e-05, "loss": 1.4313558340072632, "step": 470 }, { "epoch": 0.689051094890511, "grad_norm": 0.9765625, "learning_rate": 5.498181911427831e-05, "loss": 1.359558343887329, "step": 472 }, { "epoch": 0.691970802919708, "grad_norm": 0.9921875, "learning_rate": 5.493435416283471e-05, "loss": 1.3049167394638062, "step": 474 }, { "epoch": 0.6948905109489051, "grad_norm": 0.79296875, "learning_rate": 5.488669108289207e-05, "loss": 1.2260148525238037, "step": 476 }, { "epoch": 0.6978102189781021, "grad_norm": 0.7890625, "learning_rate": 5.4838830348176e-05, "loss": 1.3627452850341797, "step": 478 }, { "epoch": 0.7007299270072993, "grad_norm": 4.21875, "learning_rate": 5.479077243437665e-05, "loss": 1.3585658073425293, "step": 480 }, { "epoch": 0.7036496350364964, "grad_norm": 1.1484375, "learning_rate": 5.474251781914393e-05, "loss": 1.4558300971984863, "step": 482 }, { "epoch": 0.7065693430656934, "grad_norm": 1.0234375, "learning_rate": 5.469406698208276e-05, "loss": 1.4577538967132568, "step": 484 }, { "epoch": 0.7094890510948905, "grad_norm": 0.6484375, "learning_rate": 5.4645420404748316e-05, "loss": 1.3028912544250488, "step": 486 }, { "epoch": 0.7124087591240876, "grad_norm": 0.515625, "learning_rate": 5.459657857064128e-05, "loss": 1.2662179470062256, "step": 488 }, { "epoch": 0.7153284671532847, "grad_norm": 0.578125, "learning_rate": 5.454754196520294e-05, "loss": 1.5223736763000488, "step": 490 }, { "epoch": 0.7182481751824817, "grad_norm": 0.91015625, "learning_rate": 5.449831107581049e-05, "loss": 1.2283718585968018, "step": 492 }, { "epoch": 0.7211678832116788, "grad_norm": 0.64453125, "learning_rate": 5.444888639177207e-05, "loss": 1.3435661792755127, "step": 494 }, { "epoch": 0.724087591240876, "grad_norm": 1.5625, "learning_rate": 5.439926840432201e-05, "loss": 1.3870526552200317, "step": 496 }, { "epoch": 0.727007299270073, "grad_norm": 0.640625, "learning_rate": 5.434945760661583e-05, "loss": 1.2267348766326904, "step": 498 }, { "epoch": 0.7299270072992701, "grad_norm": 0.7265625, "learning_rate": 5.429945449372547e-05, "loss": 1.3859362602233887, "step": 500 }, { "epoch": 0.7328467153284671, "grad_norm": 0.65234375, "learning_rate": 5.42492595626342e-05, "loss": 1.4018727540969849, "step": 502 }, { "epoch": 0.7357664233576642, "grad_norm": 0.66796875, "learning_rate": 5.4198873312231875e-05, "loss": 1.6720361709594727, "step": 504 }, { "epoch": 0.7386861313868613, "grad_norm": 0.5, "learning_rate": 5.414829624330986e-05, "loss": 1.219796895980835, "step": 506 }, { "epoch": 0.7416058394160584, "grad_norm": 0.5546875, "learning_rate": 5.4097528858556036e-05, "loss": 1.0394165515899658, "step": 508 }, { "epoch": 0.7445255474452555, "grad_norm": 0.625, "learning_rate": 5.404657166254986e-05, "loss": 0.7834327220916748, "step": 510 }, { "epoch": 0.7474452554744525, "grad_norm": 0.77734375, "learning_rate": 5.399542516175736e-05, "loss": 1.3642323017120361, "step": 512 }, { "epoch": 0.7503649635036497, "grad_norm": 0.6015625, "learning_rate": 5.394408986452606e-05, "loss": 1.352614164352417, "step": 514 }, { "epoch": 0.7532846715328467, "grad_norm": 0.57421875, "learning_rate": 5.3892566281079915e-05, "loss": 1.3700366020202637, "step": 516 }, { "epoch": 0.7562043795620438, "grad_norm": 1.375, "learning_rate": 5.384085492351428e-05, "loss": 1.157553791999817, "step": 518 }, { "epoch": 0.7591240875912408, "grad_norm": 6.46875, "learning_rate": 5.378895630579083e-05, "loss": 0.9500113725662231, "step": 520 }, { "epoch": 0.762043795620438, "grad_norm": 0.6953125, "learning_rate": 5.3736870943732385e-05, "loss": 1.3946928977966309, "step": 522 }, { "epoch": 0.7649635036496351, "grad_norm": 1.3515625, "learning_rate": 5.3684599355017876e-05, "loss": 1.552666425704956, "step": 524 }, { "epoch": 0.7678832116788321, "grad_norm": 2.625, "learning_rate": 5.363214205917708e-05, "loss": 1.5947957038879395, "step": 526 }, { "epoch": 0.7708029197080292, "grad_norm": 1.03125, "learning_rate": 5.3579499577585615e-05, "loss": 1.2542707920074463, "step": 528 }, { "epoch": 0.7737226277372263, "grad_norm": 1.0703125, "learning_rate": 5.352667243345962e-05, "loss": 1.2413777112960815, "step": 530 }, { "epoch": 0.7766423357664234, "grad_norm": 2.375, "learning_rate": 5.3473661151850554e-05, "loss": 1.6848254203796387, "step": 532 }, { "epoch": 0.7795620437956204, "grad_norm": 2.171875, "learning_rate": 5.342046625964013e-05, "loss": 1.7417545318603516, "step": 534 }, { "epoch": 0.7824817518248175, "grad_norm": 0.82421875, "learning_rate": 5.33670882855349e-05, "loss": 1.4168747663497925, "step": 536 }, { "epoch": 0.7854014598540145, "grad_norm": 1.15625, "learning_rate": 5.331352776006112e-05, "loss": 1.4896337985992432, "step": 538 }, { "epoch": 0.7883211678832117, "grad_norm": 0.671875, "learning_rate": 5.325978521555939e-05, "loss": 1.3501195907592773, "step": 540 }, { "epoch": 0.7912408759124088, "grad_norm": 1.7890625, "learning_rate": 5.3205861186179446e-05, "loss": 1.5467338562011719, "step": 542 }, { "epoch": 0.7941605839416058, "grad_norm": 1.2890625, "learning_rate": 5.31517562078748e-05, "loss": 1.7124128341674805, "step": 544 }, { "epoch": 0.7970802919708029, "grad_norm": 0.62890625, "learning_rate": 5.3097470818397395e-05, "loss": 1.580247163772583, "step": 546 }, { "epoch": 0.8, "grad_norm": 1.328125, "learning_rate": 5.304300555729233e-05, "loss": 1.5659313201904297, "step": 548 }, { "epoch": 0.8029197080291971, "grad_norm": 1.0, "learning_rate": 5.298836096589243e-05, "loss": 1.5881831645965576, "step": 550 }, { "epoch": 0.8058394160583942, "grad_norm": 0.70703125, "learning_rate": 5.2933537587312905e-05, "loss": 1.5077919960021973, "step": 552 }, { "epoch": 0.8087591240875912, "grad_norm": 1.53125, "learning_rate": 5.2878535966445946e-05, "loss": 1.55655837059021, "step": 554 }, { "epoch": 0.8116788321167884, "grad_norm": 0.61328125, "learning_rate": 5.282335664995526e-05, "loss": 1.2932486534118652, "step": 556 }, { "epoch": 0.8145985401459854, "grad_norm": 0.71484375, "learning_rate": 5.2768000186270745e-05, "loss": 1.3656586408615112, "step": 558 }, { "epoch": 0.8175182481751825, "grad_norm": 1.3671875, "learning_rate": 5.2712467125582924e-05, "loss": 1.5890371799468994, "step": 560 }, { "epoch": 0.8204379562043795, "grad_norm": 0.65625, "learning_rate": 5.2656758019837565e-05, "loss": 1.6321032047271729, "step": 562 }, { "epoch": 0.8233576642335766, "grad_norm": 0.85546875, "learning_rate": 5.2600873422730094e-05, "loss": 1.3365864753723145, "step": 564 }, { "epoch": 0.8262773722627738, "grad_norm": 0.55078125, "learning_rate": 5.254481388970022e-05, "loss": 1.2258787155151367, "step": 566 }, { "epoch": 0.8291970802919708, "grad_norm": 1.4609375, "learning_rate": 5.248857997792632e-05, "loss": 1.5853493213653564, "step": 568 }, { "epoch": 0.8321167883211679, "grad_norm": 0.73828125, "learning_rate": 5.2432172246319925e-05, "loss": 1.3046971559524536, "step": 570 }, { "epoch": 0.8350364963503649, "grad_norm": 0.74609375, "learning_rate": 5.2375591255520176e-05, "loss": 1.4615299701690674, "step": 572 }, { "epoch": 0.8379562043795621, "grad_norm": 0.5390625, "learning_rate": 5.231883756788825e-05, "loss": 1.32985520362854, "step": 574 }, { "epoch": 0.8408759124087591, "grad_norm": 0.765625, "learning_rate": 5.226191174750177e-05, "loss": 1.5649585723876953, "step": 576 }, { "epoch": 0.8437956204379562, "grad_norm": 1.3125, "learning_rate": 5.220481436014916e-05, "loss": 1.497056007385254, "step": 578 }, { "epoch": 0.8467153284671532, "grad_norm": 0.7734375, "learning_rate": 5.2147545973324114e-05, "loss": 1.3932712078094482, "step": 580 }, { "epoch": 0.8496350364963504, "grad_norm": 0.640625, "learning_rate": 5.209010715621985e-05, "loss": 1.2655208110809326, "step": 582 }, { "epoch": 0.8525547445255475, "grad_norm": 0.75390625, "learning_rate": 5.2032498479723515e-05, "loss": 1.1907432079315186, "step": 584 }, { "epoch": 0.8554744525547445, "grad_norm": 1.1171875, "learning_rate": 5.19747205164105e-05, "loss": 1.4104052782058716, "step": 586 }, { "epoch": 0.8583941605839416, "grad_norm": 0.5, "learning_rate": 5.1916773840538756e-05, "loss": 1.3563584089279175, "step": 588 }, { "epoch": 0.8613138686131386, "grad_norm": 0.96875, "learning_rate": 5.185865902804304e-05, "loss": 1.255893349647522, "step": 590 }, { "epoch": 0.8642335766423358, "grad_norm": 0.62109375, "learning_rate": 5.180037665652928e-05, "loss": 1.231784701347351, "step": 592 }, { "epoch": 0.8671532846715329, "grad_norm": 0.640625, "learning_rate": 5.1741927305268744e-05, "loss": 1.2044652700424194, "step": 594 }, { "epoch": 0.8700729927007299, "grad_norm": 0.515625, "learning_rate": 5.1683311555192316e-05, "loss": 1.2496631145477295, "step": 596 }, { "epoch": 0.872992700729927, "grad_norm": 0.53515625, "learning_rate": 5.162452998888474e-05, "loss": 1.2133772373199463, "step": 598 }, { "epoch": 0.8759124087591241, "grad_norm": 0.67578125, "learning_rate": 5.1565583190578835e-05, "loss": 1.2104904651641846, "step": 600 }, { "epoch": 0.8788321167883212, "grad_norm": 1.953125, "learning_rate": 5.150647174614963e-05, "loss": 1.2359637022018433, "step": 602 }, { "epoch": 0.8817518248175182, "grad_norm": 0.67578125, "learning_rate": 5.144719624310861e-05, "loss": 1.3523967266082764, "step": 604 }, { "epoch": 0.8846715328467153, "grad_norm": 0.859375, "learning_rate": 5.138775727059781e-05, "loss": 1.2998895645141602, "step": 606 }, { "epoch": 0.8875912408759125, "grad_norm": 2.78125, "learning_rate": 5.132815541938404e-05, "loss": 1.2893991470336914, "step": 608 }, { "epoch": 0.8905109489051095, "grad_norm": 1.21875, "learning_rate": 5.126839128185294e-05, "loss": 1.493467092514038, "step": 610 }, { "epoch": 0.8934306569343066, "grad_norm": 0.49609375, "learning_rate": 5.120846545200312e-05, "loss": 1.2961106300354004, "step": 612 }, { "epoch": 0.8963503649635036, "grad_norm": 2.40625, "learning_rate": 5.11483785254403e-05, "loss": 1.7353638410568237, "step": 614 }, { "epoch": 0.8992700729927007, "grad_norm": 1.125, "learning_rate": 5.108813109937129e-05, "loss": 1.246555209159851, "step": 616 }, { "epoch": 0.9021897810218978, "grad_norm": 2.140625, "learning_rate": 5.102772377259815e-05, "loss": 0.9241611957550049, "step": 618 }, { "epoch": 0.9051094890510949, "grad_norm": 0.96875, "learning_rate": 5.096715714551218e-05, "loss": 1.3003945350646973, "step": 620 }, { "epoch": 0.908029197080292, "grad_norm": 1.40625, "learning_rate": 5.0906431820087985e-05, "loss": 1.478858232498169, "step": 622 }, { "epoch": 0.910948905109489, "grad_norm": 0.78515625, "learning_rate": 5.0845548399877476e-05, "loss": 1.355273723602295, "step": 624 }, { "epoch": 0.9138686131386862, "grad_norm": 0.6484375, "learning_rate": 5.078450749000388e-05, "loss": 1.3407752513885498, "step": 626 }, { "epoch": 0.9167883211678832, "grad_norm": 2.4375, "learning_rate": 5.072330969715571e-05, "loss": 1.6535050868988037, "step": 628 }, { "epoch": 0.9197080291970803, "grad_norm": 0.88671875, "learning_rate": 5.066195562958076e-05, "loss": 1.3024041652679443, "step": 630 }, { "epoch": 0.9226277372262773, "grad_norm": 0.7421875, "learning_rate": 5.0600445897080045e-05, "loss": 1.3594764471054077, "step": 632 }, { "epoch": 0.9255474452554745, "grad_norm": 0.609375, "learning_rate": 5.053878111100175e-05, "loss": 1.189161777496338, "step": 634 }, { "epoch": 0.9284671532846716, "grad_norm": 0.462890625, "learning_rate": 5.04769618842351e-05, "loss": 0.995853841304779, "step": 636 }, { "epoch": 0.9313868613138686, "grad_norm": 0.796875, "learning_rate": 5.041498883120437e-05, "loss": 1.3354415893554688, "step": 638 }, { "epoch": 0.9343065693430657, "grad_norm": 0.58984375, "learning_rate": 5.035286256786269e-05, "loss": 1.3200321197509766, "step": 640 }, { "epoch": 0.9372262773722628, "grad_norm": 0.98828125, "learning_rate": 5.029058371168596e-05, "loss": 1.3042809963226318, "step": 642 }, { "epoch": 0.9401459854014599, "grad_norm": 0.5546875, "learning_rate": 5.022815288166671e-05, "loss": 1.3093239068984985, "step": 644 }, { "epoch": 0.9430656934306569, "grad_norm": 0.7109375, "learning_rate": 5.0165570698307945e-05, "loss": 1.2666034698486328, "step": 646 }, { "epoch": 0.945985401459854, "grad_norm": 3.671875, "learning_rate": 5.010283778361697e-05, "loss": 0.7473506927490234, "step": 648 }, { "epoch": 0.948905109489051, "grad_norm": 1.09375, "learning_rate": 5.003995476109922e-05, "loss": 1.4341310262680054, "step": 650 }, { "epoch": 0.9518248175182482, "grad_norm": 0.75, "learning_rate": 4.9976922255752084e-05, "loss": 1.3149964809417725, "step": 652 }, { "epoch": 0.9547445255474453, "grad_norm": 2.25, "learning_rate": 4.9913740894058637e-05, "loss": 1.3723666667938232, "step": 654 }, { "epoch": 0.9576642335766423, "grad_norm": 1.3984375, "learning_rate": 4.9850411303981446e-05, "loss": 1.3720585107803345, "step": 656 }, { "epoch": 0.9605839416058394, "grad_norm": 0.6875, "learning_rate": 4.978693411495634e-05, "loss": 1.2820851802825928, "step": 658 }, { "epoch": 0.9635036496350365, "grad_norm": 0.90625, "learning_rate": 4.9723309957886124e-05, "loss": 1.4141497611999512, "step": 660 }, { "epoch": 0.9664233576642336, "grad_norm": 0.57421875, "learning_rate": 4.965953946513437e-05, "loss": 1.1847033500671387, "step": 662 }, { "epoch": 0.9693430656934306, "grad_norm": 0.66015625, "learning_rate": 4.9595623270519006e-05, "loss": 1.2596821784973145, "step": 664 }, { "epoch": 0.9722627737226277, "grad_norm": 2.421875, "learning_rate": 4.9531562009306176e-05, "loss": 1.5929439067840576, "step": 666 }, { "epoch": 0.9751824817518249, "grad_norm": 3.5, "learning_rate": 4.946735631820382e-05, "loss": 1.7425875663757324, "step": 668 }, { "epoch": 0.9781021897810219, "grad_norm": 2.390625, "learning_rate": 4.940300683535535e-05, "loss": 1.5199015140533447, "step": 670 }, { "epoch": 0.981021897810219, "grad_norm": 0.55078125, "learning_rate": 4.933851420033338e-05, "loss": 1.261643409729004, "step": 672 }, { "epoch": 0.983941605839416, "grad_norm": 1.9453125, "learning_rate": 4.927387905413328e-05, "loss": 1.4657907485961914, "step": 674 }, { "epoch": 0.9868613138686131, "grad_norm": 1.921875, "learning_rate": 4.9209102039166846e-05, "loss": 0.47202789783477783, "step": 676 }, { "epoch": 0.9897810218978103, "grad_norm": 0.98828125, "learning_rate": 4.914418379925595e-05, "loss": 1.2479331493377686, "step": 678 }, { "epoch": 0.9927007299270073, "grad_norm": 1.2109375, "learning_rate": 4.907912497962606e-05, "loss": 1.3170100450515747, "step": 680 }, { "epoch": 0.9956204379562044, "grad_norm": 1.0, "learning_rate": 4.901392622689992e-05, "loss": 1.3046462535858154, "step": 682 }, { "epoch": 0.9985401459854014, "grad_norm": 2.203125, "learning_rate": 4.894858818909101e-05, "loss": 1.5552364587783813, "step": 684 }, { "epoch": 1.0014598540145985, "grad_norm": 1.0625, "learning_rate": 4.888311151559726e-05, "loss": 1.3213224411010742, "step": 686 }, { "epoch": 1.0043795620437956, "grad_norm": 0.9375, "learning_rate": 4.8817496857194416e-05, "loss": 1.2282832860946655, "step": 688 }, { "epoch": 1.0072992700729928, "grad_norm": 2.53125, "learning_rate": 4.875174486602973e-05, "loss": 1.4474010467529297, "step": 690 }, { "epoch": 1.0102189781021897, "grad_norm": 2.578125, "learning_rate": 4.868585619561538e-05, "loss": 1.3594729900360107, "step": 692 }, { "epoch": 1.013138686131387, "grad_norm": 53.5, "learning_rate": 4.8619831500822e-05, "loss": 1.2219158411026, "step": 694 }, { "epoch": 1.0160583941605839, "grad_norm": 0.640625, "learning_rate": 4.855367143787222e-05, "loss": 1.1806716918945312, "step": 696 }, { "epoch": 1.018978102189781, "grad_norm": 0.76171875, "learning_rate": 4.848737666433404e-05, "loss": 1.3240623474121094, "step": 698 }, { "epoch": 1.0218978102189782, "grad_norm": 1.046875, "learning_rate": 4.842094783911438e-05, "loss": 1.0889549255371094, "step": 700 }, { "epoch": 1.0248175182481751, "grad_norm": 1.1640625, "learning_rate": 4.8354385622452534e-05, "loss": 1.1705904006958008, "step": 702 }, { "epoch": 1.0277372262773723, "grad_norm": 1.3515625, "learning_rate": 4.8287690675913536e-05, "loss": 1.2288353443145752, "step": 704 }, { "epoch": 1.0306569343065692, "grad_norm": 0.6328125, "learning_rate": 4.822086366238166e-05, "loss": 1.258807897567749, "step": 706 }, { "epoch": 1.0335766423357664, "grad_norm": 0.86328125, "learning_rate": 4.815390524605377e-05, "loss": 1.1945178508758545, "step": 708 }, { "epoch": 1.0364963503649636, "grad_norm": 1.6171875, "learning_rate": 4.808681609243279e-05, "loss": 1.3580873012542725, "step": 710 }, { "epoch": 1.0394160583941605, "grad_norm": 4.65625, "learning_rate": 4.8019596868321005e-05, "loss": 1.1779024600982666, "step": 712 }, { "epoch": 1.0423357664233577, "grad_norm": 2.046875, "learning_rate": 4.79522482418135e-05, "loss": 0.9665200710296631, "step": 714 }, { "epoch": 1.0452554744525548, "grad_norm": 1.0625, "learning_rate": 4.788477088229151e-05, "loss": 1.090451955795288, "step": 716 }, { "epoch": 1.0481751824817518, "grad_norm": 2.40625, "learning_rate": 4.7817165460415734e-05, "loss": 0.9157304167747498, "step": 718 }, { "epoch": 1.051094890510949, "grad_norm": 1.296875, "learning_rate": 4.7749432648119675e-05, "loss": 1.003751516342163, "step": 720 }, { "epoch": 1.054014598540146, "grad_norm": 4.65625, "learning_rate": 4.7681573118603024e-05, "loss": 0.7297412157058716, "step": 722 }, { "epoch": 1.056934306569343, "grad_norm": 1.46875, "learning_rate": 4.761358754632483e-05, "loss": 1.176282525062561, "step": 724 }, { "epoch": 1.0598540145985402, "grad_norm": 0.60546875, "learning_rate": 4.7545476606997025e-05, "loss": 0.9568055868148804, "step": 726 }, { "epoch": 1.0627737226277372, "grad_norm": 0.8515625, "learning_rate": 4.747724097757743e-05, "loss": 0.849090576171875, "step": 728 }, { "epoch": 1.0656934306569343, "grad_norm": 3.0625, "learning_rate": 4.740888133626324e-05, "loss": 1.0601378679275513, "step": 730 }, { "epoch": 1.0686131386861315, "grad_norm": 0.8125, "learning_rate": 4.7340398362484226e-05, "loss": 0.9943457245826721, "step": 732 }, { "epoch": 1.0715328467153284, "grad_norm": 2.40625, "learning_rate": 4.727179273689591e-05, "loss": 0.6839646100997925, "step": 734 }, { "epoch": 1.0744525547445256, "grad_norm": 1.0703125, "learning_rate": 4.72030651413729e-05, "loss": 1.0361661911010742, "step": 736 }, { "epoch": 1.0773722627737226, "grad_norm": 2.0625, "learning_rate": 4.713421625900203e-05, "loss": 1.017957329750061, "step": 738 }, { "epoch": 1.0802919708029197, "grad_norm": 1.0625, "learning_rate": 4.706524677407567e-05, "loss": 0.6705314517021179, "step": 740 }, { "epoch": 1.0832116788321169, "grad_norm": 0.6640625, "learning_rate": 4.69961573720848e-05, "loss": 0.906855583190918, "step": 742 }, { "epoch": 1.0861313868613138, "grad_norm": 1.9921875, "learning_rate": 4.69269487397123e-05, "loss": 0.8631948232650757, "step": 744 }, { "epoch": 1.089051094890511, "grad_norm": 0.85546875, "learning_rate": 4.6857621564826066e-05, "loss": 0.9076054096221924, "step": 746 }, { "epoch": 1.091970802919708, "grad_norm": 0.98046875, "learning_rate": 4.678817653647223e-05, "loss": 1.0434634685516357, "step": 748 }, { "epoch": 1.094890510948905, "grad_norm": 0.6796875, "learning_rate": 4.671861434486821e-05, "loss": 0.9549797773361206, "step": 750 }, { "epoch": 1.0978102189781023, "grad_norm": 0.578125, "learning_rate": 4.6648935681395996e-05, "loss": 0.9440876245498657, "step": 752 }, { "epoch": 1.1007299270072992, "grad_norm": 0.62890625, "learning_rate": 4.657914123859512e-05, "loss": 0.8400392532348633, "step": 754 }, { "epoch": 1.1036496350364964, "grad_norm": 1.484375, "learning_rate": 4.65092317101559e-05, "loss": 0.8279464244842529, "step": 756 }, { "epoch": 1.1065693430656935, "grad_norm": 1.0234375, "learning_rate": 4.643920779091249e-05, "loss": 0.6409823894500732, "step": 758 }, { "epoch": 1.1094890510948905, "grad_norm": 0.984375, "learning_rate": 4.6369070176835943e-05, "loss": 0.7931218147277832, "step": 760 }, { "epoch": 1.1124087591240877, "grad_norm": 0.84375, "learning_rate": 4.629881956502739e-05, "loss": 0.9238249063491821, "step": 762 }, { "epoch": 1.1153284671532846, "grad_norm": 0.6328125, "learning_rate": 4.6228456653711005e-05, "loss": 0.9794689416885376, "step": 764 }, { "epoch": 1.1182481751824818, "grad_norm": 1.65625, "learning_rate": 4.615798214222713e-05, "loss": 0.7727400660514832, "step": 766 }, { "epoch": 1.121167883211679, "grad_norm": 0.73828125, "learning_rate": 4.6087396731025305e-05, "loss": 0.7949696779251099, "step": 768 }, { "epoch": 1.1240875912408759, "grad_norm": 0.6875, "learning_rate": 4.6016701121657315e-05, "loss": 0.9301152229309082, "step": 770 }, { "epoch": 1.127007299270073, "grad_norm": 0.56640625, "learning_rate": 4.5945896016770174e-05, "loss": 0.8954685926437378, "step": 772 }, { "epoch": 1.12992700729927, "grad_norm": 0.65234375, "learning_rate": 4.5874982120099244e-05, "loss": 0.790676474571228, "step": 774 }, { "epoch": 1.1328467153284671, "grad_norm": 2.3125, "learning_rate": 4.5803960136461095e-05, "loss": 0.8866307735443115, "step": 776 }, { "epoch": 1.1357664233576643, "grad_norm": 0.5546875, "learning_rate": 4.573283077174664e-05, "loss": 1.0086119174957275, "step": 778 }, { "epoch": 1.1386861313868613, "grad_norm": 0.625, "learning_rate": 4.566159473291403e-05, "loss": 0.9111287593841553, "step": 780 }, { "epoch": 1.1416058394160584, "grad_norm": 0.60546875, "learning_rate": 4.5590252727981665e-05, "loss": 0.7828556299209595, "step": 782 }, { "epoch": 1.1445255474452556, "grad_norm": 1.28125, "learning_rate": 4.551880546602113e-05, "loss": 0.8504880666732788, "step": 784 }, { "epoch": 1.1474452554744525, "grad_norm": 0.875, "learning_rate": 4.54472536571502e-05, "loss": 0.6608180999755859, "step": 786 }, { "epoch": 1.1503649635036497, "grad_norm": 0.50390625, "learning_rate": 4.537559801252571e-05, "loss": 0.8004018068313599, "step": 788 }, { "epoch": 1.1532846715328466, "grad_norm": 0.49609375, "learning_rate": 4.530383924433653e-05, "loss": 0.8180815577507019, "step": 790 }, { "epoch": 1.1562043795620438, "grad_norm": 0.376953125, "learning_rate": 4.523197806579649e-05, "loss": 0.8199501037597656, "step": 792 }, { "epoch": 1.159124087591241, "grad_norm": 0.60546875, "learning_rate": 4.51600151911373e-05, "loss": 0.8051800727844238, "step": 794 }, { "epoch": 1.162043795620438, "grad_norm": 0.5625, "learning_rate": 4.50879513356014e-05, "loss": 0.838517427444458, "step": 796 }, { "epoch": 1.164963503649635, "grad_norm": 0.84765625, "learning_rate": 4.501578721543488e-05, "loss": 0.7846630215644836, "step": 798 }, { "epoch": 1.167883211678832, "grad_norm": 0.515625, "learning_rate": 4.4943523547880396e-05, "loss": 0.8626841902732849, "step": 800 }, { "epoch": 1.1708029197080292, "grad_norm": 0.578125, "learning_rate": 4.487116105117e-05, "loss": 0.834052324295044, "step": 802 }, { "epoch": 1.1737226277372264, "grad_norm": 1.1171875, "learning_rate": 4.479870044451797e-05, "loss": 0.7536060810089111, "step": 804 }, { "epoch": 1.1766423357664233, "grad_norm": 2.0, "learning_rate": 4.472614244811379e-05, "loss": 0.5739097595214844, "step": 806 }, { "epoch": 1.1795620437956205, "grad_norm": 1.1875, "learning_rate": 4.4653487783114824e-05, "loss": 0.8073641657829285, "step": 808 }, { "epoch": 1.1824817518248176, "grad_norm": 1.0859375, "learning_rate": 4.458073717163927e-05, "loss": 0.7964679002761841, "step": 810 }, { "epoch": 1.1854014598540146, "grad_norm": 1.1953125, "learning_rate": 4.4507891336758944e-05, "loss": 0.727637529373169, "step": 812 }, { "epoch": 1.1883211678832117, "grad_norm": 1.234375, "learning_rate": 4.443495100249206e-05, "loss": 0.759423017501831, "step": 814 }, { "epoch": 1.1912408759124087, "grad_norm": 0.59765625, "learning_rate": 4.43619168937961e-05, "loss": 0.9602820873260498, "step": 816 }, { "epoch": 1.1941605839416058, "grad_norm": 0.44140625, "learning_rate": 4.4288789736560544e-05, "loss": 0.5548771619796753, "step": 818 }, { "epoch": 1.197080291970803, "grad_norm": 1.0625, "learning_rate": 4.421557025759974e-05, "loss": 0.8706461191177368, "step": 820 }, { "epoch": 1.2, "grad_norm": 0.59765625, "learning_rate": 4.414225918464556e-05, "loss": 0.7758599519729614, "step": 822 }, { "epoch": 1.2029197080291971, "grad_norm": 0.44140625, "learning_rate": 4.406885724634027e-05, "loss": 0.8401455879211426, "step": 824 }, { "epoch": 1.205839416058394, "grad_norm": 1.5546875, "learning_rate": 4.399536517222924e-05, "loss": 0.4931638836860657, "step": 826 }, { "epoch": 1.2087591240875912, "grad_norm": 0.60546875, "learning_rate": 4.3921783692753726e-05, "loss": 0.8398951888084412, "step": 828 }, { "epoch": 1.2116788321167884, "grad_norm": 1.21875, "learning_rate": 4.384811353924356e-05, "loss": 0.5683646202087402, "step": 830 }, { "epoch": 1.2145985401459853, "grad_norm": 0.5078125, "learning_rate": 4.3774355443909906e-05, "loss": 0.7751730680465698, "step": 832 }, { "epoch": 1.2175182481751825, "grad_norm": 0.4609375, "learning_rate": 4.3700510139838025e-05, "loss": 0.7935394048690796, "step": 834 }, { "epoch": 1.2204379562043797, "grad_norm": 1.265625, "learning_rate": 4.36265783609799e-05, "loss": 0.45605871081352234, "step": 836 }, { "epoch": 1.2233576642335766, "grad_norm": 1.4921875, "learning_rate": 4.355256084214703e-05, "loss": 0.39349085092544556, "step": 838 }, { "epoch": 1.2262773722627738, "grad_norm": 0.72265625, "learning_rate": 4.3478458319003076e-05, "loss": 0.5289195775985718, "step": 840 }, { "epoch": 1.2291970802919707, "grad_norm": 1.3046875, "learning_rate": 4.340427152805655e-05, "loss": 0.3823907971382141, "step": 842 }, { "epoch": 1.2321167883211679, "grad_norm": 0.7734375, "learning_rate": 4.333000120665351e-05, "loss": 0.7731255292892456, "step": 844 }, { "epoch": 1.235036496350365, "grad_norm": 0.98046875, "learning_rate": 4.325564809297025e-05, "loss": 0.7135635614395142, "step": 846 }, { "epoch": 1.237956204379562, "grad_norm": 0.625, "learning_rate": 4.31812129260059e-05, "loss": 0.7766275405883789, "step": 848 }, { "epoch": 1.2408759124087592, "grad_norm": 0.4140625, "learning_rate": 4.310669644557515e-05, "loss": 0.7377574443817139, "step": 850 }, { "epoch": 1.243795620437956, "grad_norm": 0.50390625, "learning_rate": 4.303209939230086e-05, "loss": 0.6263979077339172, "step": 852 }, { "epoch": 1.2467153284671533, "grad_norm": 0.5390625, "learning_rate": 4.295742250760674e-05, "loss": 0.8428114652633667, "step": 854 }, { "epoch": 1.2496350364963504, "grad_norm": 0.609375, "learning_rate": 4.288266653370986e-05, "loss": 0.6811407804489136, "step": 856 }, { "epoch": 1.2525547445255474, "grad_norm": 1.265625, "learning_rate": 4.280783221361346e-05, "loss": 0.6165977120399475, "step": 858 }, { "epoch": 1.2554744525547445, "grad_norm": 1.03125, "learning_rate": 4.2732920291099407e-05, "loss": 0.4194489121437073, "step": 860 }, { "epoch": 1.2583941605839417, "grad_norm": 0.67578125, "learning_rate": 4.265793151072089e-05, "loss": 0.7688656449317932, "step": 862 }, { "epoch": 1.2613138686131387, "grad_norm": 0.498046875, "learning_rate": 4.258286661779498e-05, "loss": 0.7792798280715942, "step": 864 }, { "epoch": 1.2642335766423358, "grad_norm": 1.0703125, "learning_rate": 4.2507726358395236e-05, "loss": 0.529695451259613, "step": 866 }, { "epoch": 1.2671532846715328, "grad_norm": 0.53125, "learning_rate": 4.243251147934428e-05, "loss": 0.7187914848327637, "step": 868 }, { "epoch": 1.27007299270073, "grad_norm": 0.6015625, "learning_rate": 4.2357222728206404e-05, "loss": 0.8103936314582825, "step": 870 }, { "epoch": 1.2729927007299269, "grad_norm": 0.4375, "learning_rate": 4.22818608532801e-05, "loss": 0.7720659971237183, "step": 872 }, { "epoch": 1.275912408759124, "grad_norm": 0.494140625, "learning_rate": 4.2206426603590644e-05, "loss": 0.7679579257965088, "step": 874 }, { "epoch": 1.2788321167883212, "grad_norm": 0.455078125, "learning_rate": 4.213092072888264e-05, "loss": 0.7206258177757263, "step": 876 }, { "epoch": 1.2817518248175181, "grad_norm": 1.984375, "learning_rate": 4.205534397961262e-05, "loss": 0.5508802533149719, "step": 878 }, { "epoch": 1.2846715328467153, "grad_norm": 0.703125, "learning_rate": 4.197969710694146e-05, "loss": 0.717147707939148, "step": 880 }, { "epoch": 1.2875912408759125, "grad_norm": 0.53125, "learning_rate": 4.190398086272708e-05, "loss": 0.7146405577659607, "step": 882 }, { "epoch": 1.2905109489051094, "grad_norm": 1.15625, "learning_rate": 4.182819599951685e-05, "loss": 0.8259648084640503, "step": 884 }, { "epoch": 1.2934306569343066, "grad_norm": 0.55078125, "learning_rate": 4.1752343270540136e-05, "loss": 0.6179315447807312, "step": 886 }, { "epoch": 1.2963503649635038, "grad_norm": 0.8203125, "learning_rate": 4.167642342970084e-05, "loss": 0.7542316913604736, "step": 888 }, { "epoch": 1.2992700729927007, "grad_norm": 0.5078125, "learning_rate": 4.160043723156989e-05, "loss": 0.7624303102493286, "step": 890 }, { "epoch": 1.3021897810218979, "grad_norm": 1.5, "learning_rate": 4.152438543137773e-05, "loss": 0.9253525733947754, "step": 892 }, { "epoch": 1.305109489051095, "grad_norm": 1.1796875, "learning_rate": 4.144826878500687e-05, "loss": 0.6559224128723145, "step": 894 }, { "epoch": 1.308029197080292, "grad_norm": 0.40625, "learning_rate": 4.137208804898428e-05, "loss": 0.8019828796386719, "step": 896 }, { "epoch": 1.310948905109489, "grad_norm": 0.6015625, "learning_rate": 4.1295843980473924e-05, "loss": 0.9176638126373291, "step": 898 }, { "epoch": 1.313868613138686, "grad_norm": 0.453125, "learning_rate": 4.121953733726925e-05, "loss": 0.8651784658432007, "step": 900 }, { "epoch": 1.3167883211678832, "grad_norm": 0.59375, "learning_rate": 4.114316887778564e-05, "loss": 0.8026199340820312, "step": 902 }, { "epoch": 1.3197080291970802, "grad_norm": 1.546875, "learning_rate": 4.106673936105287e-05, "loss": 0.5445064306259155, "step": 904 }, { "epoch": 1.3226277372262774, "grad_norm": 1.140625, "learning_rate": 4.0990249546707556e-05, "loss": 0.5602635145187378, "step": 906 }, { "epoch": 1.3255474452554745, "grad_norm": 3.15625, "learning_rate": 4.091370019498561e-05, "loss": 0.8074881434440613, "step": 908 }, { "epoch": 1.3284671532846715, "grad_norm": 0.48046875, "learning_rate": 4.0837092066714694e-05, "loss": 0.8778038024902344, "step": 910 }, { "epoch": 1.3313868613138686, "grad_norm": 1.296875, "learning_rate": 4.076042592330668e-05, "loss": 0.8802911639213562, "step": 912 }, { "epoch": 1.3343065693430658, "grad_norm": 0.65625, "learning_rate": 4.068370252675002e-05, "loss": 1.0339293479919434, "step": 914 }, { "epoch": 1.3372262773722627, "grad_norm": 3.171875, "learning_rate": 4.0606922639602215e-05, "loss": 0.7693579196929932, "step": 916 }, { "epoch": 1.34014598540146, "grad_norm": 0.66015625, "learning_rate": 4.0530087024982245e-05, "loss": 0.545236349105835, "step": 918 }, { "epoch": 1.343065693430657, "grad_norm": 1.0, "learning_rate": 4.045319644656295e-05, "loss": 0.7232102155685425, "step": 920 }, { "epoch": 1.345985401459854, "grad_norm": 0.83984375, "learning_rate": 4.037625166856347e-05, "loss": 0.5631003975868225, "step": 922 }, { "epoch": 1.348905109489051, "grad_norm": 0.84765625, "learning_rate": 4.0299253455741626e-05, "loss": 0.75032639503479, "step": 924 }, { "epoch": 1.3518248175182481, "grad_norm": 0.546875, "learning_rate": 4.022220257338635e-05, "loss": 0.7473883628845215, "step": 926 }, { "epoch": 1.3547445255474453, "grad_norm": 0.66015625, "learning_rate": 4.014509978731002e-05, "loss": 0.6389172077178955, "step": 928 }, { "epoch": 1.3576642335766422, "grad_norm": 0.52734375, "learning_rate": 4.006794586384093e-05, "loss": 0.8051577210426331, "step": 930 }, { "epoch": 1.3605839416058394, "grad_norm": 1.640625, "learning_rate": 3.999074156981561e-05, "loss": 0.7719905376434326, "step": 932 }, { "epoch": 1.3635036496350366, "grad_norm": 1.3125, "learning_rate": 3.9913487672571246e-05, "loss": 0.7920917272567749, "step": 934 }, { "epoch": 1.3664233576642335, "grad_norm": 0.380859375, "learning_rate": 3.983618493993799e-05, "loss": 0.6346763372421265, "step": 936 }, { "epoch": 1.3693430656934307, "grad_norm": 0.90625, "learning_rate": 3.975883414023141e-05, "loss": 0.7190485000610352, "step": 938 }, { "epoch": 1.3722627737226278, "grad_norm": 0.474609375, "learning_rate": 3.96814360422448e-05, "loss": 0.6966416835784912, "step": 940 }, { "epoch": 1.3751824817518248, "grad_norm": 0.427734375, "learning_rate": 3.9603991415241556e-05, "loss": 0.8331156373023987, "step": 942 }, { "epoch": 1.378102189781022, "grad_norm": 0.44921875, "learning_rate": 3.9526501028947525e-05, "loss": 0.7017558813095093, "step": 944 }, { "epoch": 1.3810218978102191, "grad_norm": 0.55859375, "learning_rate": 3.9448965653543374e-05, "loss": 0.7089091539382935, "step": 946 }, { "epoch": 1.383941605839416, "grad_norm": 0.5859375, "learning_rate": 3.937138605965689e-05, "loss": 0.8313618898391724, "step": 948 }, { "epoch": 1.3868613138686132, "grad_norm": 1.265625, "learning_rate": 3.929376301835538e-05, "loss": 0.5355940461158752, "step": 950 }, { "epoch": 1.3897810218978102, "grad_norm": 0.546875, "learning_rate": 3.9216097301137943e-05, "loss": 0.8322726488113403, "step": 952 }, { "epoch": 1.3927007299270073, "grad_norm": 0.5234375, "learning_rate": 3.9138389679927876e-05, "loss": 0.7144240736961365, "step": 954 }, { "epoch": 1.3956204379562043, "grad_norm": 0.71484375, "learning_rate": 3.906064092706493e-05, "loss": 0.877200722694397, "step": 956 }, { "epoch": 1.3985401459854014, "grad_norm": 1.4453125, "learning_rate": 3.898285181529768e-05, "loss": 0.5701913833618164, "step": 958 }, { "epoch": 1.4014598540145986, "grad_norm": 0.58203125, "learning_rate": 3.890502311777583e-05, "loss": 0.6877053380012512, "step": 960 }, { "epoch": 1.4043795620437955, "grad_norm": 1.234375, "learning_rate": 3.882715560804251e-05, "loss": 0.7841130495071411, "step": 962 }, { "epoch": 1.4072992700729927, "grad_norm": 0.91796875, "learning_rate": 3.8749250060026645e-05, "loss": 0.6575303077697754, "step": 964 }, { "epoch": 1.4102189781021899, "grad_norm": 0.482421875, "learning_rate": 3.8671307248035183e-05, "loss": 0.7847940921783447, "step": 966 }, { "epoch": 1.4131386861313868, "grad_norm": 0.56640625, "learning_rate": 3.8593327946745465e-05, "loss": 0.6031632423400879, "step": 968 }, { "epoch": 1.416058394160584, "grad_norm": 0.33984375, "learning_rate": 3.85153129311975e-05, "loss": 0.7221322655677795, "step": 970 }, { "epoch": 1.4189781021897812, "grad_norm": 0.494140625, "learning_rate": 3.843726297678625e-05, "loss": 0.6405712366104126, "step": 972 }, { "epoch": 1.421897810218978, "grad_norm": 0.55859375, "learning_rate": 3.835917885925394e-05, "loss": 0.8854905962944031, "step": 974 }, { "epoch": 1.4248175182481753, "grad_norm": 0.33984375, "learning_rate": 3.828106135468236e-05, "loss": 0.8395616412162781, "step": 976 }, { "epoch": 1.4277372262773722, "grad_norm": 0.953125, "learning_rate": 3.820291123948512e-05, "loss": 0.7625291347503662, "step": 978 }, { "epoch": 1.4306569343065694, "grad_norm": 0.5390625, "learning_rate": 3.812472929039994e-05, "loss": 0.6698781251907349, "step": 980 }, { "epoch": 1.4335766423357663, "grad_norm": 1.71875, "learning_rate": 3.804651628448095e-05, "loss": 0.46671491861343384, "step": 982 }, { "epoch": 1.4364963503649635, "grad_norm": 3.59375, "learning_rate": 3.796827299909097e-05, "loss": 0.7954891920089722, "step": 984 }, { "epoch": 1.4394160583941606, "grad_norm": 0.5390625, "learning_rate": 3.789000021189374e-05, "loss": 0.7973778247833252, "step": 986 }, { "epoch": 1.4423357664233576, "grad_norm": 0.6171875, "learning_rate": 3.7811698700846245e-05, "loss": 0.77935791015625, "step": 988 }, { "epoch": 1.4452554744525548, "grad_norm": 0.2890625, "learning_rate": 3.773336924419094e-05, "loss": 0.4812260866165161, "step": 990 }, { "epoch": 1.448175182481752, "grad_norm": 0.58984375, "learning_rate": 3.765501262044804e-05, "loss": 0.6503823399543762, "step": 992 }, { "epoch": 1.4510948905109489, "grad_norm": 1.1875, "learning_rate": 3.757662960840776e-05, "loss": 0.7457800507545471, "step": 994 }, { "epoch": 1.454014598540146, "grad_norm": 0.78515625, "learning_rate": 3.749822098712264e-05, "loss": 0.729491114616394, "step": 996 }, { "epoch": 1.4569343065693432, "grad_norm": 1.0, "learning_rate": 3.741978753589968e-05, "loss": 0.7310203313827515, "step": 998 }, { "epoch": 1.4598540145985401, "grad_norm": 0.78515625, "learning_rate": 3.7341330034292735e-05, "loss": 0.8127636909484863, "step": 1000 }, { "epoch": 1.4627737226277373, "grad_norm": 0.96875, "learning_rate": 3.7262849262094626e-05, "loss": 0.744979202747345, "step": 1002 }, { "epoch": 1.4656934306569342, "grad_norm": 2.8125, "learning_rate": 3.718434599932953e-05, "loss": 0.6435647010803223, "step": 1004 }, { "epoch": 1.4686131386861314, "grad_norm": 0.53125, "learning_rate": 3.710582102624511e-05, "loss": 0.7206861972808838, "step": 1006 }, { "epoch": 1.4715328467153284, "grad_norm": 0.494140625, "learning_rate": 3.702727512330484e-05, "loss": 0.6581206321716309, "step": 1008 }, { "epoch": 1.4744525547445255, "grad_norm": 0.6015625, "learning_rate": 3.69487090711802e-05, "loss": 0.6592334508895874, "step": 1010 }, { "epoch": 1.4773722627737227, "grad_norm": 1.5625, "learning_rate": 3.687012365074293e-05, "loss": 0.8629540205001831, "step": 1012 }, { "epoch": 1.4802919708029196, "grad_norm": 0.80078125, "learning_rate": 3.67915196430573e-05, "loss": 0.786446213722229, "step": 1014 }, { "epoch": 1.4832116788321168, "grad_norm": 1.640625, "learning_rate": 3.671289782937228e-05, "loss": 0.7205163240432739, "step": 1016 }, { "epoch": 1.486131386861314, "grad_norm": 0.78515625, "learning_rate": 3.663425899111385e-05, "loss": 0.6132459044456482, "step": 1018 }, { "epoch": 1.489051094890511, "grad_norm": 0.478515625, "learning_rate": 3.6555603909877175e-05, "loss": 0.7237584590911865, "step": 1020 }, { "epoch": 1.491970802919708, "grad_norm": 0.5859375, "learning_rate": 3.647693336741887e-05, "loss": 0.7926473617553711, "step": 1022 }, { "epoch": 1.4948905109489052, "grad_norm": 0.6328125, "learning_rate": 3.639824814564921e-05, "loss": 0.8149737119674683, "step": 1024 }, { "epoch": 1.4978102189781022, "grad_norm": 0.65234375, "learning_rate": 3.63195490266244e-05, "loss": 0.590778112411499, "step": 1026 }, { "epoch": 1.5007299270072991, "grad_norm": 0.640625, "learning_rate": 3.6240836792538715e-05, "loss": 0.560297966003418, "step": 1028 }, { "epoch": 1.5036496350364965, "grad_norm": 0.5625, "learning_rate": 3.616211222571683e-05, "loss": 0.73332679271698, "step": 1030 }, { "epoch": 1.5065693430656935, "grad_norm": 1.2421875, "learning_rate": 3.608337610860598e-05, "loss": 0.5995861291885376, "step": 1032 }, { "epoch": 1.5094890510948904, "grad_norm": 0.6015625, "learning_rate": 3.600462922376819e-05, "loss": 0.6911063194274902, "step": 1034 }, { "epoch": 1.5124087591240876, "grad_norm": 0.984375, "learning_rate": 3.592587235387251e-05, "loss": 0.7384026050567627, "step": 1036 }, { "epoch": 1.5153284671532847, "grad_norm": 1.5390625, "learning_rate": 3.5847106281687245e-05, "loss": 0.44528472423553467, "step": 1038 }, { "epoch": 1.5182481751824817, "grad_norm": 0.46875, "learning_rate": 3.576833179007214e-05, "loss": 0.7749471664428711, "step": 1040 }, { "epoch": 1.5211678832116788, "grad_norm": 0.7109375, "learning_rate": 3.5689549661970667e-05, "loss": 0.8797944188117981, "step": 1042 }, { "epoch": 1.524087591240876, "grad_norm": 4.71875, "learning_rate": 3.561076068040211e-05, "loss": 0.6449779272079468, "step": 1044 }, { "epoch": 1.527007299270073, "grad_norm": 1.1796875, "learning_rate": 3.553196562845396e-05, "loss": 0.7924333810806274, "step": 1046 }, { "epoch": 1.5299270072992701, "grad_norm": 0.443359375, "learning_rate": 3.545316528927401e-05, "loss": 0.727583646774292, "step": 1048 }, { "epoch": 1.5328467153284673, "grad_norm": 1.25, "learning_rate": 3.537436044606257e-05, "loss": 0.6008315682411194, "step": 1050 }, { "epoch": 1.5357664233576642, "grad_norm": 0.46875, "learning_rate": 3.529555188206476e-05, "loss": 0.5438330769538879, "step": 1052 }, { "epoch": 1.5386861313868612, "grad_norm": 1.453125, "learning_rate": 3.521674038056267e-05, "loss": 0.8460222482681274, "step": 1054 }, { "epoch": 1.5416058394160586, "grad_norm": 0.53515625, "learning_rate": 3.5137926724867566e-05, "loss": 0.7475937604904175, "step": 1056 }, { "epoch": 1.5445255474452555, "grad_norm": 1.125, "learning_rate": 3.505911169831216e-05, "loss": 0.31257715821266174, "step": 1058 }, { "epoch": 1.5474452554744524, "grad_norm": 1.8671875, "learning_rate": 3.498029608424276e-05, "loss": 0.5920336246490479, "step": 1060 }, { "epoch": 1.5503649635036496, "grad_norm": 0.5546875, "learning_rate": 3.490148066601151e-05, "loss": 0.8748451471328735, "step": 1062 }, { "epoch": 1.5532846715328468, "grad_norm": 0.486328125, "learning_rate": 3.482266622696863e-05, "loss": 0.7054937481880188, "step": 1064 }, { "epoch": 1.5562043795620437, "grad_norm": 0.75, "learning_rate": 3.474385355045462e-05, "loss": 0.839794397354126, "step": 1066 }, { "epoch": 1.5591240875912409, "grad_norm": 1.3515625, "learning_rate": 3.46650434197924e-05, "loss": 0.6114320755004883, "step": 1068 }, { "epoch": 1.562043795620438, "grad_norm": 0.53515625, "learning_rate": 3.4586236618279654e-05, "loss": 0.6584864854812622, "step": 1070 }, { "epoch": 1.564963503649635, "grad_norm": 1.046875, "learning_rate": 3.4507433929180924e-05, "loss": 0.6659724712371826, "step": 1072 }, { "epoch": 1.5678832116788322, "grad_norm": 0.75390625, "learning_rate": 3.4428636135719936e-05, "loss": 0.6218416690826416, "step": 1074 }, { "epoch": 1.5708029197080293, "grad_norm": 1.1328125, "learning_rate": 3.4349844021071694e-05, "loss": 0.41415178775787354, "step": 1076 }, { "epoch": 1.5737226277372263, "grad_norm": 0.5234375, "learning_rate": 3.427105836835482e-05, "loss": 0.7918195724487305, "step": 1078 }, { "epoch": 1.5766423357664232, "grad_norm": 0.62109375, "learning_rate": 3.4192279960623674e-05, "loss": 0.5616426467895508, "step": 1080 }, { "epoch": 1.5795620437956206, "grad_norm": 1.390625, "learning_rate": 3.411350958086061e-05, "loss": 0.48918431997299194, "step": 1082 }, { "epoch": 1.5824817518248175, "grad_norm": 0.447265625, "learning_rate": 3.403474801196821e-05, "loss": 0.8120974898338318, "step": 1084 }, { "epoch": 1.5854014598540145, "grad_norm": 0.84375, "learning_rate": 3.3955996036761454e-05, "loss": 0.6056236624717712, "step": 1086 }, { "epoch": 1.5883211678832116, "grad_norm": 0.578125, "learning_rate": 3.3877254437960014e-05, "loss": 0.7579033374786377, "step": 1088 }, { "epoch": 1.5912408759124088, "grad_norm": 0.7890625, "learning_rate": 3.3798523998180374e-05, "loss": 0.6755779981613159, "step": 1090 }, { "epoch": 1.5941605839416058, "grad_norm": 0.328125, "learning_rate": 3.371980549992817e-05, "loss": 0.6938905119895935, "step": 1092 }, { "epoch": 1.597080291970803, "grad_norm": 0.412109375, "learning_rate": 3.36410997255903e-05, "loss": 0.5442206859588623, "step": 1094 }, { "epoch": 1.6, "grad_norm": 0.443359375, "learning_rate": 3.3562407457427236e-05, "loss": 0.6741411685943604, "step": 1096 }, { "epoch": 1.602919708029197, "grad_norm": 0.435546875, "learning_rate": 3.3483729477565163e-05, "loss": 0.8024434447288513, "step": 1098 }, { "epoch": 1.6058394160583942, "grad_norm": 0.55078125, "learning_rate": 3.340506656798831e-05, "loss": 0.887005090713501, "step": 1100 }, { "epoch": 1.6087591240875914, "grad_norm": 1.4296875, "learning_rate": 3.332641951053108e-05, "loss": 0.5712992548942566, "step": 1102 }, { "epoch": 1.6116788321167883, "grad_norm": 0.380859375, "learning_rate": 3.3247789086870364e-05, "loss": 0.27882304787635803, "step": 1104 }, { "epoch": 1.6145985401459853, "grad_norm": 0.416015625, "learning_rate": 3.3169176078517666e-05, "loss": 0.7044307589530945, "step": 1106 }, { "epoch": 1.6175182481751826, "grad_norm": 1.171875, "learning_rate": 3.3090581266811454e-05, "loss": 0.4385440945625305, "step": 1108 }, { "epoch": 1.6204379562043796, "grad_norm": 3.828125, "learning_rate": 3.301200543290932e-05, "loss": 0.5933332443237305, "step": 1110 }, { "epoch": 1.6233576642335765, "grad_norm": 0.48828125, "learning_rate": 3.293344935778023e-05, "loss": 0.7244850397109985, "step": 1112 }, { "epoch": 1.6262773722627737, "grad_norm": 0.48046875, "learning_rate": 3.2854913822196774e-05, "loss": 0.725429117679596, "step": 1114 }, { "epoch": 1.6291970802919709, "grad_norm": 0.83203125, "learning_rate": 3.277639960672739e-05, "loss": 0.6351516246795654, "step": 1116 }, { "epoch": 1.6321167883211678, "grad_norm": 0.50390625, "learning_rate": 3.269790749172861e-05, "loss": 0.5855380296707153, "step": 1118 }, { "epoch": 1.635036496350365, "grad_norm": 0.33203125, "learning_rate": 3.261943825733734e-05, "loss": 0.5529741644859314, "step": 1120 }, { "epoch": 1.6379562043795621, "grad_norm": 0.62890625, "learning_rate": 3.254099268346303e-05, "loss": 0.7271662354469299, "step": 1122 }, { "epoch": 1.640875912408759, "grad_norm": 0.4140625, "learning_rate": 3.2462571549780006e-05, "loss": 0.3429512083530426, "step": 1124 }, { "epoch": 1.6437956204379562, "grad_norm": 0.5703125, "learning_rate": 3.238417563571965e-05, "loss": 0.8319324254989624, "step": 1126 }, { "epoch": 1.6467153284671534, "grad_norm": 1.2890625, "learning_rate": 3.230580572046272e-05, "loss": 0.5692154169082642, "step": 1128 }, { "epoch": 1.6496350364963503, "grad_norm": 0.55859375, "learning_rate": 3.222746258293153e-05, "loss": 0.789124608039856, "step": 1130 }, { "epoch": 1.6525547445255473, "grad_norm": 0.7578125, "learning_rate": 3.2149147001782275e-05, "loss": 0.5931746959686279, "step": 1132 }, { "epoch": 1.6554744525547447, "grad_norm": 1.1015625, "learning_rate": 3.207085975539727e-05, "loss": 0.5728804469108582, "step": 1134 }, { "epoch": 1.6583941605839416, "grad_norm": 1.046875, "learning_rate": 3.199260162187719e-05, "loss": 0.5378882884979248, "step": 1136 }, { "epoch": 1.6613138686131386, "grad_norm": 0.8046875, "learning_rate": 3.191437337903336e-05, "loss": 0.7326488494873047, "step": 1138 }, { "epoch": 1.6642335766423357, "grad_norm": 0.474609375, "learning_rate": 3.183617580438004e-05, "loss": 0.5117689371109009, "step": 1140 }, { "epoch": 1.667153284671533, "grad_norm": 2.484375, "learning_rate": 3.1758009675126656e-05, "loss": 0.7147212028503418, "step": 1142 }, { "epoch": 1.6700729927007298, "grad_norm": 0.65625, "learning_rate": 3.16798757681701e-05, "loss": 0.4904450476169586, "step": 1144 }, { "epoch": 1.672992700729927, "grad_norm": 0.5390625, "learning_rate": 3.160177486008702e-05, "loss": 0.6540476083755493, "step": 1146 }, { "epoch": 1.6759124087591242, "grad_norm": 4.15625, "learning_rate": 3.152370772712605e-05, "loss": 0.43966585397720337, "step": 1148 }, { "epoch": 1.6788321167883211, "grad_norm": 0.5, "learning_rate": 3.144567514520019e-05, "loss": 0.6375809907913208, "step": 1150 }, { "epoch": 1.6817518248175183, "grad_norm": 0.5546875, "learning_rate": 3.136767788987898e-05, "loss": 0.5703507661819458, "step": 1152 }, { "epoch": 1.6846715328467154, "grad_norm": 0.5625, "learning_rate": 3.128971673638088e-05, "loss": 0.8321278691291809, "step": 1154 }, { "epoch": 1.6875912408759124, "grad_norm": 0.68359375, "learning_rate": 3.121179245956552e-05, "loss": 0.6486014127731323, "step": 1156 }, { "epoch": 1.6905109489051093, "grad_norm": 2.125, "learning_rate": 3.1133905833926004e-05, "loss": 0.8116127252578735, "step": 1158 }, { "epoch": 1.6934306569343067, "grad_norm": 0.546875, "learning_rate": 3.105605763358124e-05, "loss": 0.9435343742370605, "step": 1160 }, { "epoch": 1.6963503649635037, "grad_norm": 0.71484375, "learning_rate": 3.097824863226821e-05, "loss": 0.7895150184631348, "step": 1162 }, { "epoch": 1.6992700729927006, "grad_norm": 1.2265625, "learning_rate": 3.090047960333428e-05, "loss": 0.6546390056610107, "step": 1164 }, { "epoch": 1.7021897810218978, "grad_norm": 0.89453125, "learning_rate": 3.082275131972958e-05, "loss": 0.4520113468170166, "step": 1166 }, { "epoch": 1.705109489051095, "grad_norm": 0.625, "learning_rate": 3.07450645539992e-05, "loss": 0.6498976945877075, "step": 1168 }, { "epoch": 1.7080291970802919, "grad_norm": 0.5, "learning_rate": 3.066742007827566e-05, "loss": 0.625820517539978, "step": 1170 }, { "epoch": 1.710948905109489, "grad_norm": 0.431640625, "learning_rate": 3.058981866427107e-05, "loss": 0.6509993076324463, "step": 1172 }, { "epoch": 1.7138686131386862, "grad_norm": 3.6875, "learning_rate": 3.0512261083269633e-05, "loss": 0.6243072152137756, "step": 1174 }, { "epoch": 1.7167883211678832, "grad_norm": 0.5078125, "learning_rate": 3.043474810611983e-05, "loss": 0.5908772349357605, "step": 1176 }, { "epoch": 1.7197080291970803, "grad_norm": 0.59765625, "learning_rate": 3.035728050322687e-05, "loss": 0.7354363203048706, "step": 1178 }, { "epoch": 1.7226277372262775, "grad_norm": 0.6015625, "learning_rate": 3.027985904454493e-05, "loss": 0.5529353618621826, "step": 1180 }, { "epoch": 1.7255474452554744, "grad_norm": 0.87109375, "learning_rate": 3.0202484499569614e-05, "loss": 0.6982197761535645, "step": 1182 }, { "epoch": 1.7284671532846714, "grad_norm": 0.6875, "learning_rate": 3.012515763733021e-05, "loss": 0.7173529863357544, "step": 1184 }, { "epoch": 1.7313868613138688, "grad_norm": 0.88671875, "learning_rate": 3.0047879226382115e-05, "loss": 0.5552504062652588, "step": 1186 }, { "epoch": 1.7343065693430657, "grad_norm": 1.359375, "learning_rate": 2.9970650034799123e-05, "loss": 0.7216619253158569, "step": 1188 }, { "epoch": 1.7372262773722627, "grad_norm": 1.0546875, "learning_rate": 2.9893470830165898e-05, "loss": 0.679259717464447, "step": 1190 }, { "epoch": 1.7401459854014598, "grad_norm": 1.609375, "learning_rate": 2.9816342379570198e-05, "loss": 0.8096989393234253, "step": 1192 }, { "epoch": 1.743065693430657, "grad_norm": 0.5703125, "learning_rate": 2.9739265449595435e-05, "loss": 0.9562541246414185, "step": 1194 }, { "epoch": 1.745985401459854, "grad_norm": 1.4765625, "learning_rate": 2.9662240806312853e-05, "loss": 0.23433877527713776, "step": 1196 }, { "epoch": 1.748905109489051, "grad_norm": 0.31640625, "learning_rate": 2.9585269215274096e-05, "loss": 0.6434130668640137, "step": 1198 }, { "epoch": 1.7518248175182483, "grad_norm": 0.98828125, "learning_rate": 2.9508351441503465e-05, "loss": 0.4967500567436218, "step": 1200 }, { "epoch": 1.7547445255474452, "grad_norm": 0.6875, "learning_rate": 2.9431488249490415e-05, "loss": 0.6541959047317505, "step": 1202 }, { "epoch": 1.7576642335766424, "grad_norm": 0.59375, "learning_rate": 2.9354680403181856e-05, "loss": 0.39902958273887634, "step": 1204 }, { "epoch": 1.7605839416058395, "grad_norm": 0.671875, "learning_rate": 2.9277928665974674e-05, "loss": 0.45974647998809814, "step": 1206 }, { "epoch": 1.7635036496350365, "grad_norm": 0.5546875, "learning_rate": 2.9201233800708025e-05, "loss": 0.6035457849502563, "step": 1208 }, { "epoch": 1.7664233576642334, "grad_norm": 0.58203125, "learning_rate": 2.9124596569655876e-05, "loss": 0.6182115077972412, "step": 1210 }, { "epoch": 1.7693430656934308, "grad_norm": 0.60546875, "learning_rate": 2.9048017734519298e-05, "loss": 0.7187447547912598, "step": 1212 }, { "epoch": 1.7722627737226277, "grad_norm": 0.412109375, "learning_rate": 2.8971498056419018e-05, "loss": 0.928739070892334, "step": 1214 }, { "epoch": 1.7751824817518247, "grad_norm": 1.6484375, "learning_rate": 2.8895038295887756e-05, "loss": 0.551633894443512, "step": 1216 }, { "epoch": 1.7781021897810219, "grad_norm": 0.96875, "learning_rate": 2.8818639212862733e-05, "loss": 0.49139732122421265, "step": 1218 }, { "epoch": 1.781021897810219, "grad_norm": 0.62109375, "learning_rate": 2.8742301566678107e-05, "loss": 0.5061070919036865, "step": 1220 }, { "epoch": 1.783941605839416, "grad_norm": 0.765625, "learning_rate": 2.8666026116057362e-05, "loss": 0.5571912527084351, "step": 1222 }, { "epoch": 1.7868613138686131, "grad_norm": 0.62890625, "learning_rate": 2.8589813619105876e-05, "loss": 0.7460435032844543, "step": 1224 }, { "epoch": 1.7897810218978103, "grad_norm": 0.67578125, "learning_rate": 2.851366483330327e-05, "loss": 0.7260211706161499, "step": 1226 }, { "epoch": 1.7927007299270072, "grad_norm": 1.5625, "learning_rate": 2.8437580515496004e-05, "loss": 0.523064374923706, "step": 1228 }, { "epoch": 1.7956204379562044, "grad_norm": 1.703125, "learning_rate": 2.8361561421889713e-05, "loss": 0.5164039134979248, "step": 1230 }, { "epoch": 1.7985401459854016, "grad_norm": 1.4609375, "learning_rate": 2.828560830804183e-05, "loss": 0.6638224124908447, "step": 1232 }, { "epoch": 1.8014598540145985, "grad_norm": 0.515625, "learning_rate": 2.8209721928853964e-05, "loss": 0.717475950717926, "step": 1234 }, { "epoch": 1.8043795620437955, "grad_norm": 1.171875, "learning_rate": 2.8133903038564492e-05, "loss": 0.5235683917999268, "step": 1236 }, { "epoch": 1.8072992700729928, "grad_norm": 0.404296875, "learning_rate": 2.8058152390740954e-05, "loss": 0.5262608528137207, "step": 1238 }, { "epoch": 1.8102189781021898, "grad_norm": 0.69921875, "learning_rate": 2.7982470738272697e-05, "loss": 0.7407115697860718, "step": 1240 }, { "epoch": 1.8131386861313867, "grad_norm": 0.87109375, "learning_rate": 2.7906858833363256e-05, "loss": 0.6663039922714233, "step": 1242 }, { "epoch": 1.816058394160584, "grad_norm": 0.474609375, "learning_rate": 2.783131742752298e-05, "loss": 0.6756942868232727, "step": 1244 }, { "epoch": 1.818978102189781, "grad_norm": 2.0625, "learning_rate": 2.775584727156151e-05, "loss": 0.39022934436798096, "step": 1246 }, { "epoch": 1.821897810218978, "grad_norm": 0.5859375, "learning_rate": 2.768044911558034e-05, "loss": 0.7429840564727783, "step": 1248 }, { "epoch": 1.8248175182481752, "grad_norm": 0.5, "learning_rate": 2.7605123708965334e-05, "loss": 0.7059404850006104, "step": 1250 }, { "epoch": 1.8277372262773723, "grad_norm": 0.609375, "learning_rate": 2.752987180037932e-05, "loss": 0.7387444972991943, "step": 1252 }, { "epoch": 1.8306569343065693, "grad_norm": 0.76953125, "learning_rate": 2.7454694137754615e-05, "loss": 0.5452640652656555, "step": 1254 }, { "epoch": 1.8335766423357664, "grad_norm": 0.98046875, "learning_rate": 2.73795914682856e-05, "loss": 0.6368634700775146, "step": 1256 }, { "epoch": 1.8364963503649636, "grad_norm": 0.69140625, "learning_rate": 2.7304564538421298e-05, "loss": 0.528918981552124, "step": 1258 }, { "epoch": 1.8394160583941606, "grad_norm": 2.328125, "learning_rate": 2.7229614093857955e-05, "loss": 0.4514443278312683, "step": 1260 }, { "epoch": 1.8423357664233575, "grad_norm": 1.125, "learning_rate": 2.7154740879531608e-05, "loss": 0.5287116765975952, "step": 1262 }, { "epoch": 1.845255474452555, "grad_norm": 1.3671875, "learning_rate": 2.7079945639610734e-05, "loss": 0.7141860723495483, "step": 1264 }, { "epoch": 1.8481751824817518, "grad_norm": 0.498046875, "learning_rate": 2.700522911748878e-05, "loss": 0.7375243306159973, "step": 1266 }, { "epoch": 1.8510948905109488, "grad_norm": 0.90625, "learning_rate": 2.693059205577685e-05, "loss": 0.5961591601371765, "step": 1268 }, { "epoch": 1.854014598540146, "grad_norm": 0.57421875, "learning_rate": 2.6856035196296247e-05, "loss": 0.8086559772491455, "step": 1270 }, { "epoch": 1.856934306569343, "grad_norm": 0.4921875, "learning_rate": 2.6781559280071186e-05, "loss": 0.5709831714630127, "step": 1272 }, { "epoch": 1.85985401459854, "grad_norm": 0.69921875, "learning_rate": 2.6707165047321337e-05, "loss": 0.6018452644348145, "step": 1274 }, { "epoch": 1.8627737226277372, "grad_norm": 1.6015625, "learning_rate": 2.6632853237454555e-05, "loss": 0.9354125261306763, "step": 1276 }, { "epoch": 1.8656934306569344, "grad_norm": 0.90625, "learning_rate": 2.6558624589059454e-05, "loss": 0.7831665873527527, "step": 1278 }, { "epoch": 1.8686131386861313, "grad_norm": 0.380859375, "learning_rate": 2.6484479839898127e-05, "loss": 0.7829075455665588, "step": 1280 }, { "epoch": 1.8715328467153285, "grad_norm": 0.6640625, "learning_rate": 2.6410419726898782e-05, "loss": 0.6603153944015503, "step": 1282 }, { "epoch": 1.8744525547445257, "grad_norm": 0.68359375, "learning_rate": 2.633644498614842e-05, "loss": 0.4937764108181, "step": 1284 }, { "epoch": 1.8773722627737226, "grad_norm": 0.59375, "learning_rate": 2.6262556352885528e-05, "loss": 0.7518973350524902, "step": 1286 }, { "epoch": 1.8802919708029195, "grad_norm": 0.56640625, "learning_rate": 2.618875456149276e-05, "loss": 0.8098867535591125, "step": 1288 }, { "epoch": 1.883211678832117, "grad_norm": 0.7890625, "learning_rate": 2.6115040345489654e-05, "loss": 0.8065282702445984, "step": 1290 }, { "epoch": 1.8861313868613139, "grad_norm": 1.9296875, "learning_rate": 2.604141443752532e-05, "loss": 0.45944589376449585, "step": 1292 }, { "epoch": 1.8890510948905108, "grad_norm": 0.6328125, "learning_rate": 2.5967877569371193e-05, "loss": 0.41555699706077576, "step": 1294 }, { "epoch": 1.891970802919708, "grad_norm": 0.5859375, "learning_rate": 2.5894430471913706e-05, "loss": 0.5259090662002563, "step": 1296 }, { "epoch": 1.8948905109489051, "grad_norm": 1.7109375, "learning_rate": 2.5821073875147074e-05, "loss": 0.5657607316970825, "step": 1298 }, { "epoch": 1.897810218978102, "grad_norm": 0.9921875, "learning_rate": 2.5747808508166022e-05, "loss": 0.5103273391723633, "step": 1300 }, { "epoch": 1.9007299270072993, "grad_norm": 0.5078125, "learning_rate": 2.567463509915854e-05, "loss": 0.7555137872695923, "step": 1302 }, { "epoch": 1.9036496350364964, "grad_norm": 0.3125, "learning_rate": 2.5601554375398626e-05, "loss": 0.22839674353599548, "step": 1304 }, { "epoch": 1.9065693430656934, "grad_norm": 1.3984375, "learning_rate": 2.55285670632391e-05, "loss": 0.5510879158973694, "step": 1306 }, { "epoch": 1.9094890510948905, "grad_norm": 2.75, "learning_rate": 2.545567388810434e-05, "loss": 0.9097703099250793, "step": 1308 }, { "epoch": 1.9124087591240877, "grad_norm": 0.671875, "learning_rate": 2.5382875574483104e-05, "loss": 0.5909059047698975, "step": 1310 }, { "epoch": 1.9153284671532846, "grad_norm": 1.3046875, "learning_rate": 2.53101728459213e-05, "loss": 0.6244196891784668, "step": 1312 }, { "epoch": 1.9182481751824818, "grad_norm": 0.51953125, "learning_rate": 2.523756642501483e-05, "loss": 0.5189331769943237, "step": 1314 }, { "epoch": 1.921167883211679, "grad_norm": 0.875, "learning_rate": 2.5165057033402388e-05, "loss": 0.6180688738822937, "step": 1316 }, { "epoch": 1.924087591240876, "grad_norm": 0.345703125, "learning_rate": 2.5092645391758282e-05, "loss": 0.7852336764335632, "step": 1318 }, { "epoch": 1.9270072992700729, "grad_norm": 0.44921875, "learning_rate": 2.5020332219785266e-05, "loss": 0.6935960054397583, "step": 1320 }, { "epoch": 1.92992700729927, "grad_norm": 0.5390625, "learning_rate": 2.4948118236207428e-05, "loss": 0.7328112721443176, "step": 1322 }, { "epoch": 1.9328467153284672, "grad_norm": 0.25, "learning_rate": 2.4876004158763e-05, "loss": 0.4806957244873047, "step": 1324 }, { "epoch": 1.9357664233576641, "grad_norm": 0.59375, "learning_rate": 2.4803990704197254e-05, "loss": 0.6416811943054199, "step": 1326 }, { "epoch": 1.9386861313868613, "grad_norm": 0.59375, "learning_rate": 2.4732078588255335e-05, "loss": 0.5719661712646484, "step": 1328 }, { "epoch": 1.9416058394160585, "grad_norm": 0.52734375, "learning_rate": 2.4660268525675224e-05, "loss": 0.7179700136184692, "step": 1330 }, { "epoch": 1.9445255474452554, "grad_norm": 0.6015625, "learning_rate": 2.458856123018055e-05, "loss": 0.698421835899353, "step": 1332 }, { "epoch": 1.9474452554744526, "grad_norm": 1.640625, "learning_rate": 2.4516957414473565e-05, "loss": 0.2941528260707855, "step": 1334 }, { "epoch": 1.9503649635036497, "grad_norm": 0.72265625, "learning_rate": 2.4445457790228005e-05, "loss": 0.7893608808517456, "step": 1336 }, { "epoch": 1.9532846715328467, "grad_norm": 0.69921875, "learning_rate": 2.4374063068082062e-05, "loss": 0.749394953250885, "step": 1338 }, { "epoch": 1.9562043795620438, "grad_norm": 0.396484375, "learning_rate": 2.430277395763128e-05, "loss": 0.588677704334259, "step": 1340 }, { "epoch": 1.959124087591241, "grad_norm": 0.6171875, "learning_rate": 2.423159116742155e-05, "loss": 0.7306614518165588, "step": 1342 }, { "epoch": 1.962043795620438, "grad_norm": 2.1875, "learning_rate": 2.416051540494202e-05, "loss": 0.43609586358070374, "step": 1344 }, { "epoch": 1.964963503649635, "grad_norm": 0.74609375, "learning_rate": 2.40895473766181e-05, "loss": 0.6756691336631775, "step": 1346 }, { "epoch": 1.967883211678832, "grad_norm": 0.49609375, "learning_rate": 2.4018687787804397e-05, "loss": 0.7406471967697144, "step": 1348 }, { "epoch": 1.9708029197080292, "grad_norm": 0.76953125, "learning_rate": 2.394793734277777e-05, "loss": 0.6689704656600952, "step": 1350 }, { "epoch": 1.9737226277372262, "grad_norm": 0.4921875, "learning_rate": 2.387729674473027e-05, "loss": 0.49914392828941345, "step": 1352 }, { "epoch": 1.9766423357664233, "grad_norm": 0.5625, "learning_rate": 2.380676669576218e-05, "loss": 0.5775256156921387, "step": 1354 }, { "epoch": 1.9795620437956205, "grad_norm": 0.65234375, "learning_rate": 2.373634789687502e-05, "loss": 0.8281344175338745, "step": 1356 }, { "epoch": 1.9824817518248175, "grad_norm": 0.384765625, "learning_rate": 2.3666041047964605e-05, "loss": 0.9128406047821045, "step": 1358 }, { "epoch": 1.9854014598540146, "grad_norm": 1.65625, "learning_rate": 2.359584684781407e-05, "loss": 0.4032854437828064, "step": 1360 }, { "epoch": 1.9883211678832118, "grad_norm": 1.203125, "learning_rate": 2.3525765994086912e-05, "loss": 0.507628858089447, "step": 1362 }, { "epoch": 1.9912408759124087, "grad_norm": 0.470703125, "learning_rate": 2.3455799183320102e-05, "loss": 0.6595430374145508, "step": 1364 }, { "epoch": 1.994160583941606, "grad_norm": 0.55859375, "learning_rate": 2.3385947110917122e-05, "loss": 0.7209213972091675, "step": 1366 }, { "epoch": 1.997080291970803, "grad_norm": 0.9296875, "learning_rate": 2.331621047114105e-05, "loss": 0.6316984295845032, "step": 1368 }, { "epoch": 2.0, "grad_norm": 1.265625, "learning_rate": 2.3246589957107666e-05, "loss": 0.5848791599273682, "step": 1370 }, { "epoch": 2.002919708029197, "grad_norm": 0.451171875, "learning_rate": 2.3177086260778612e-05, "loss": 1.0494563579559326, "step": 1372 }, { "epoch": 2.0058394160583943, "grad_norm": 1.140625, "learning_rate": 2.310770007295445e-05, "loss": 0.41478756070137024, "step": 1374 }, { "epoch": 2.0087591240875913, "grad_norm": 0.62109375, "learning_rate": 2.3038432083267808e-05, "loss": 0.547417163848877, "step": 1376 }, { "epoch": 2.011678832116788, "grad_norm": 0.56640625, "learning_rate": 2.2969282980176526e-05, "loss": 0.6029900312423706, "step": 1378 }, { "epoch": 2.0145985401459856, "grad_norm": 0.392578125, "learning_rate": 2.2900253450956885e-05, "loss": 0.8512624502182007, "step": 1380 }, { "epoch": 2.0175182481751825, "grad_norm": 0.498046875, "learning_rate": 2.2831344181696654e-05, "loss": 0.5808537006378174, "step": 1382 }, { "epoch": 2.0204379562043795, "grad_norm": 0.65625, "learning_rate": 2.2762555857288363e-05, "loss": 0.7059779167175293, "step": 1384 }, { "epoch": 2.0233576642335764, "grad_norm": 1.5859375, "learning_rate": 2.269388916142244e-05, "loss": 0.698493480682373, "step": 1386 }, { "epoch": 2.026277372262774, "grad_norm": 0.57421875, "learning_rate": 2.2625344776580486e-05, "loss": 0.6546064019203186, "step": 1388 }, { "epoch": 2.0291970802919708, "grad_norm": 0.87890625, "learning_rate": 2.2556923384028394e-05, "loss": 0.6731595993041992, "step": 1390 }, { "epoch": 2.0321167883211677, "grad_norm": 0.703125, "learning_rate": 2.248862566380967e-05, "loss": 0.566330075263977, "step": 1392 }, { "epoch": 2.035036496350365, "grad_norm": 0.455078125, "learning_rate": 2.2420452294738604e-05, "loss": 0.5171869993209839, "step": 1394 }, { "epoch": 2.037956204379562, "grad_norm": 1.046875, "learning_rate": 2.2352403954393592e-05, "loss": 0.3384949564933777, "step": 1396 }, { "epoch": 2.040875912408759, "grad_norm": 1.078125, "learning_rate": 2.2284481319110333e-05, "loss": 0.41768980026245117, "step": 1398 }, { "epoch": 2.0437956204379564, "grad_norm": 0.5859375, "learning_rate": 2.2216685063975147e-05, "loss": 0.6205852627754211, "step": 1400 }, { "epoch": 2.0467153284671533, "grad_norm": 0.59765625, "learning_rate": 2.214901586281824e-05, "loss": 0.42609018087387085, "step": 1402 }, { "epoch": 2.0496350364963503, "grad_norm": 0.69921875, "learning_rate": 2.2081474388207062e-05, "loss": 0.5273014307022095, "step": 1404 }, { "epoch": 2.0525547445255476, "grad_norm": 0.95703125, "learning_rate": 2.2014061311439547e-05, "loss": 0.2152610421180725, "step": 1406 }, { "epoch": 2.0554744525547446, "grad_norm": 0.546875, "learning_rate": 2.1946777302537487e-05, "loss": 0.42848068475723267, "step": 1408 }, { "epoch": 2.0583941605839415, "grad_norm": 1.46875, "learning_rate": 2.1879623030239855e-05, "loss": 0.5728009343147278, "step": 1410 }, { "epoch": 2.0613138686131385, "grad_norm": 0.85546875, "learning_rate": 2.1812599161996192e-05, "loss": 0.2542463541030884, "step": 1412 }, { "epoch": 2.064233576642336, "grad_norm": 0.7109375, "learning_rate": 2.1745706363959914e-05, "loss": 0.3851102590560913, "step": 1414 }, { "epoch": 2.067153284671533, "grad_norm": 0.85546875, "learning_rate": 2.1678945300981753e-05, "loss": 0.23180317878723145, "step": 1416 }, { "epoch": 2.0700729927007298, "grad_norm": 0.95703125, "learning_rate": 2.1612316636603082e-05, "loss": 0.31533747911453247, "step": 1418 }, { "epoch": 2.072992700729927, "grad_norm": 0.486328125, "learning_rate": 2.1545821033049407e-05, "loss": 0.22496944665908813, "step": 1420 }, { "epoch": 2.075912408759124, "grad_norm": 0.455078125, "learning_rate": 2.14794591512237e-05, "loss": 0.8437415361404419, "step": 1422 }, { "epoch": 2.078832116788321, "grad_norm": 0.671875, "learning_rate": 2.1413231650699877e-05, "loss": 0.28613120317459106, "step": 1424 }, { "epoch": 2.0817518248175184, "grad_norm": 0.51953125, "learning_rate": 2.1347139189716225e-05, "loss": 0.2427263855934143, "step": 1426 }, { "epoch": 2.0846715328467154, "grad_norm": 0.53125, "learning_rate": 2.1281182425168885e-05, "loss": 0.35121026635169983, "step": 1428 }, { "epoch": 2.0875912408759123, "grad_norm": 0.6328125, "learning_rate": 2.1215362012605297e-05, "loss": 0.2667168974876404, "step": 1430 }, { "epoch": 2.0905109489051097, "grad_norm": 1.1953125, "learning_rate": 2.114967860621766e-05, "loss": 0.30120381712913513, "step": 1432 }, { "epoch": 2.0934306569343066, "grad_norm": 0.55859375, "learning_rate": 2.1084132858836522e-05, "loss": 0.27142685651779175, "step": 1434 }, { "epoch": 2.0963503649635036, "grad_norm": 0.4375, "learning_rate": 2.1018725421924195e-05, "loss": 0.34424224495887756, "step": 1436 }, { "epoch": 2.0992700729927005, "grad_norm": 0.6171875, "learning_rate": 2.095345694556831e-05, "loss": 0.3176770508289337, "step": 1438 }, { "epoch": 2.102189781021898, "grad_norm": 0.51953125, "learning_rate": 2.0888328078475364e-05, "loss": 0.2459079474210739, "step": 1440 }, { "epoch": 2.105109489051095, "grad_norm": 0.94921875, "learning_rate": 2.0823339467964312e-05, "loss": 0.4240154027938843, "step": 1442 }, { "epoch": 2.108029197080292, "grad_norm": 0.5625, "learning_rate": 2.0758491759960037e-05, "loss": 0.508914053440094, "step": 1444 }, { "epoch": 2.110948905109489, "grad_norm": 0.474609375, "learning_rate": 2.0693785598987023e-05, "loss": 0.4890514314174652, "step": 1446 }, { "epoch": 2.113868613138686, "grad_norm": 0.5078125, "learning_rate": 2.0629221628162876e-05, "loss": 0.28382956981658936, "step": 1448 }, { "epoch": 2.116788321167883, "grad_norm": 0.38671875, "learning_rate": 2.0564800489192013e-05, "loss": 0.6701063513755798, "step": 1450 }, { "epoch": 2.1197080291970805, "grad_norm": 0.484375, "learning_rate": 2.0500522822359208e-05, "loss": 0.2041202187538147, "step": 1452 }, { "epoch": 2.1226277372262774, "grad_norm": 0.54296875, "learning_rate": 2.0436389266523257e-05, "loss": 0.3254830241203308, "step": 1454 }, { "epoch": 2.1255474452554743, "grad_norm": 0.62109375, "learning_rate": 2.0372400459110633e-05, "loss": 0.2847502827644348, "step": 1456 }, { "epoch": 2.1284671532846717, "grad_norm": 0.46484375, "learning_rate": 2.0308557036109167e-05, "loss": 0.250871479511261, "step": 1458 }, { "epoch": 2.1313868613138687, "grad_norm": 0.498046875, "learning_rate": 2.024485963206169e-05, "loss": 0.37414756417274475, "step": 1460 }, { "epoch": 2.1343065693430656, "grad_norm": 0.5234375, "learning_rate": 2.018130888005974e-05, "loss": 0.5036002397537231, "step": 1462 }, { "epoch": 2.137226277372263, "grad_norm": 0.4609375, "learning_rate": 2.0117905411737267e-05, "loss": 0.510346531867981, "step": 1464 }, { "epoch": 2.14014598540146, "grad_norm": 3.953125, "learning_rate": 2.005464985726441e-05, "loss": 0.1956787109375, "step": 1466 }, { "epoch": 2.143065693430657, "grad_norm": 0.4921875, "learning_rate": 1.999154284534111e-05, "loss": 0.24436712265014648, "step": 1468 }, { "epoch": 2.145985401459854, "grad_norm": 0.2578125, "learning_rate": 1.9928585003191e-05, "loss": 0.20835143327713013, "step": 1470 }, { "epoch": 2.1489051094890512, "grad_norm": 0.4609375, "learning_rate": 1.9865776956555066e-05, "loss": 0.2522450089454651, "step": 1472 }, { "epoch": 2.151824817518248, "grad_norm": 0.427734375, "learning_rate": 1.9803119329685513e-05, "loss": 0.23127983510494232, "step": 1474 }, { "epoch": 2.154744525547445, "grad_norm": 0.94921875, "learning_rate": 1.974061274533946e-05, "loss": 0.19277967512607574, "step": 1476 }, { "epoch": 2.1576642335766425, "grad_norm": 1.1328125, "learning_rate": 1.9678257824772857e-05, "loss": 0.4629068374633789, "step": 1478 }, { "epoch": 2.1605839416058394, "grad_norm": 0.41796875, "learning_rate": 1.961605518773422e-05, "loss": 0.49624115228652954, "step": 1480 }, { "epoch": 2.1635036496350364, "grad_norm": 0.453125, "learning_rate": 1.9554005452458563e-05, "loss": 0.20151491463184357, "step": 1482 }, { "epoch": 2.1664233576642338, "grad_norm": 0.8125, "learning_rate": 1.949210923566114e-05, "loss": 0.491416871547699, "step": 1484 }, { "epoch": 2.1693430656934307, "grad_norm": 0.46875, "learning_rate": 1.9430367152531432e-05, "loss": 0.21659183502197266, "step": 1486 }, { "epoch": 2.1722627737226277, "grad_norm": 1.484375, "learning_rate": 1.9368779816726938e-05, "loss": 0.41106492280960083, "step": 1488 }, { "epoch": 2.1751824817518246, "grad_norm": 0.875, "learning_rate": 1.9307347840367156e-05, "loss": 0.1200709193944931, "step": 1490 }, { "epoch": 2.178102189781022, "grad_norm": 0.80859375, "learning_rate": 1.9246071834027406e-05, "loss": 0.16358846426010132, "step": 1492 }, { "epoch": 2.181021897810219, "grad_norm": 0.5234375, "learning_rate": 1.9184952406732864e-05, "loss": 0.1703709363937378, "step": 1494 }, { "epoch": 2.183941605839416, "grad_norm": 0.64453125, "learning_rate": 1.9123990165952438e-05, "loss": 0.17072588205337524, "step": 1496 }, { "epoch": 2.1868613138686133, "grad_norm": 0.44921875, "learning_rate": 1.906318571759274e-05, "loss": 0.16185638308525085, "step": 1498 }, { "epoch": 2.18978102189781, "grad_norm": 0.609375, "learning_rate": 1.900253966599211e-05, "loss": 0.4327083230018616, "step": 1500 }, { "epoch": 2.192700729927007, "grad_norm": 1.1328125, "learning_rate": 1.8942052613914533e-05, "loss": 0.5814958810806274, "step": 1502 }, { "epoch": 2.1956204379562045, "grad_norm": 0.86328125, "learning_rate": 1.8881725162543743e-05, "loss": 0.13417096436023712, "step": 1504 }, { "epoch": 2.1985401459854015, "grad_norm": 0.42578125, "learning_rate": 1.8821557911477123e-05, "loss": 0.2804584503173828, "step": 1506 }, { "epoch": 2.2014598540145984, "grad_norm": 0.35546875, "learning_rate": 1.8761551458719887e-05, "loss": 0.4399762749671936, "step": 1508 }, { "epoch": 2.204379562043796, "grad_norm": 0.578125, "learning_rate": 1.8701706400679023e-05, "loss": 0.1459416151046753, "step": 1510 }, { "epoch": 2.2072992700729928, "grad_norm": 0.482421875, "learning_rate": 1.8642023332157446e-05, "loss": 0.19403180480003357, "step": 1512 }, { "epoch": 2.2102189781021897, "grad_norm": 0.439453125, "learning_rate": 1.8582502846347988e-05, "loss": 0.22103887796401978, "step": 1514 }, { "epoch": 2.213138686131387, "grad_norm": 0.9921875, "learning_rate": 1.8523145534827637e-05, "loss": 0.12535154819488525, "step": 1516 }, { "epoch": 2.216058394160584, "grad_norm": 0.458984375, "learning_rate": 1.8463951987551536e-05, "loss": 0.20944124460220337, "step": 1518 }, { "epoch": 2.218978102189781, "grad_norm": 0.44921875, "learning_rate": 1.8404922792847193e-05, "loss": 0.23369359970092773, "step": 1520 }, { "epoch": 2.221897810218978, "grad_norm": 0.451171875, "learning_rate": 1.8346058537408573e-05, "loss": 0.04991687834262848, "step": 1522 }, { "epoch": 2.2248175182481753, "grad_norm": 0.9609375, "learning_rate": 1.8287359806290346e-05, "loss": 0.39661896228790283, "step": 1524 }, { "epoch": 2.2277372262773723, "grad_norm": 0.32421875, "learning_rate": 1.8228827182902e-05, "loss": 0.027169307693839073, "step": 1526 }, { "epoch": 2.230656934306569, "grad_norm": 0.546875, "learning_rate": 1.8170461249002077e-05, "loss": 0.15711988508701324, "step": 1528 }, { "epoch": 2.2335766423357666, "grad_norm": 0.8203125, "learning_rate": 1.811226258469238e-05, "loss": 0.46470558643341064, "step": 1530 }, { "epoch": 2.2364963503649635, "grad_norm": 0.49609375, "learning_rate": 1.8054231768412225e-05, "loss": 0.14680901169776917, "step": 1532 }, { "epoch": 2.2394160583941605, "grad_norm": 0.376953125, "learning_rate": 1.7996369376932668e-05, "loss": 0.44296467304229736, "step": 1534 }, { "epoch": 2.242335766423358, "grad_norm": 0.8515625, "learning_rate": 1.7938675985350778e-05, "loss": 0.11045099794864655, "step": 1536 }, { "epoch": 2.245255474452555, "grad_norm": 0.484375, "learning_rate": 1.788115216708393e-05, "loss": 0.46650534868240356, "step": 1538 }, { "epoch": 2.2481751824817517, "grad_norm": 0.85546875, "learning_rate": 1.7823798493864113e-05, "loss": 0.16499537229537964, "step": 1540 }, { "epoch": 2.2510948905109487, "grad_norm": 0.41015625, "learning_rate": 1.776661553573222e-05, "loss": 0.4719018340110779, "step": 1542 }, { "epoch": 2.254014598540146, "grad_norm": 0.62890625, "learning_rate": 1.7709603861032415e-05, "loss": 0.10700561106204987, "step": 1544 }, { "epoch": 2.256934306569343, "grad_norm": 0.447265625, "learning_rate": 1.7652764036406445e-05, "loss": 0.14236462116241455, "step": 1546 }, { "epoch": 2.25985401459854, "grad_norm": 0.96875, "learning_rate": 1.7596096626788055e-05, "loss": 0.4474194049835205, "step": 1548 }, { "epoch": 2.2627737226277373, "grad_norm": 0.70703125, "learning_rate": 1.7539602195397337e-05, "loss": 0.13715533912181854, "step": 1550 }, { "epoch": 2.2656934306569343, "grad_norm": 0.470703125, "learning_rate": 1.7483281303735143e-05, "loss": 0.12302975356578827, "step": 1552 }, { "epoch": 2.2686131386861312, "grad_norm": 0.44140625, "learning_rate": 1.7427134511577503e-05, "loss": 0.3969992995262146, "step": 1554 }, { "epoch": 2.2715328467153286, "grad_norm": 0.47265625, "learning_rate": 1.737116237697008e-05, "loss": 0.18677493929862976, "step": 1556 }, { "epoch": 2.2744525547445256, "grad_norm": 0.58984375, "learning_rate": 1.7315365456222585e-05, "loss": 0.4171912670135498, "step": 1558 }, { "epoch": 2.2773722627737225, "grad_norm": 0.50390625, "learning_rate": 1.7259744303903292e-05, "loss": 0.2462674379348755, "step": 1560 }, { "epoch": 2.28029197080292, "grad_norm": 0.57421875, "learning_rate": 1.7204299472833474e-05, "loss": 0.13167493045330048, "step": 1562 }, { "epoch": 2.283211678832117, "grad_norm": 0.56640625, "learning_rate": 1.7149031514081978e-05, "loss": 0.10060621798038483, "step": 1564 }, { "epoch": 2.286131386861314, "grad_norm": 0.48828125, "learning_rate": 1.7093940976959682e-05, "loss": 0.20011527836322784, "step": 1566 }, { "epoch": 2.289051094890511, "grad_norm": 0.455078125, "learning_rate": 1.7039028409014057e-05, "loss": 0.2114190012216568, "step": 1568 }, { "epoch": 2.291970802919708, "grad_norm": 0.412109375, "learning_rate": 1.6984294356023756e-05, "loss": 0.11579601466655731, "step": 1570 }, { "epoch": 2.294890510948905, "grad_norm": 0.484375, "learning_rate": 1.6929739361993152e-05, "loss": 0.14887449145317078, "step": 1572 }, { "epoch": 2.297810218978102, "grad_norm": 0.443359375, "learning_rate": 1.6875363969146922e-05, "loss": 0.14829905331134796, "step": 1574 }, { "epoch": 2.3007299270072994, "grad_norm": 0.48046875, "learning_rate": 1.6821168717924696e-05, "loss": 0.1990872621536255, "step": 1576 }, { "epoch": 2.3036496350364963, "grad_norm": 0.408203125, "learning_rate": 1.6767154146975678e-05, "loss": 0.17175400257110596, "step": 1578 }, { "epoch": 2.3065693430656933, "grad_norm": 0.419921875, "learning_rate": 1.6713320793153272e-05, "loss": 0.11114311963319778, "step": 1580 }, { "epoch": 2.3094890510948907, "grad_norm": 0.37109375, "learning_rate": 1.665966919150975e-05, "loss": 0.7707895040512085, "step": 1582 }, { "epoch": 2.3124087591240876, "grad_norm": 1.21875, "learning_rate": 1.6606199875290936e-05, "loss": 0.6511590480804443, "step": 1584 }, { "epoch": 2.3153284671532846, "grad_norm": 0.53515625, "learning_rate": 1.655291337593094e-05, "loss": 0.48480451107025146, "step": 1586 }, { "epoch": 2.318248175182482, "grad_norm": 1.2578125, "learning_rate": 1.6499810223046812e-05, "loss": 0.33573096990585327, "step": 1588 }, { "epoch": 2.321167883211679, "grad_norm": 0.609375, "learning_rate": 1.6446890944433333e-05, "loss": 0.2068139910697937, "step": 1590 }, { "epoch": 2.324087591240876, "grad_norm": 0.73046875, "learning_rate": 1.6394156066057737e-05, "loss": 0.1301928609609604, "step": 1592 }, { "epoch": 2.3270072992700728, "grad_norm": 0.53515625, "learning_rate": 1.6341606112054513e-05, "loss": 0.8669120073318481, "step": 1594 }, { "epoch": 2.32992700729927, "grad_norm": 0.546875, "learning_rate": 1.6289241604720157e-05, "loss": 0.18708989024162292, "step": 1596 }, { "epoch": 2.332846715328467, "grad_norm": 0.283203125, "learning_rate": 1.6237063064508013e-05, "loss": 0.5843273401260376, "step": 1598 }, { "epoch": 2.335766423357664, "grad_norm": 0.6640625, "learning_rate": 1.6185071010023086e-05, "loss": 0.4750453233718872, "step": 1600 }, { "epoch": 2.3386861313868614, "grad_norm": 0.435546875, "learning_rate": 1.613326595801689e-05, "loss": 0.09889721870422363, "step": 1602 }, { "epoch": 2.3416058394160584, "grad_norm": 0.59765625, "learning_rate": 1.6081648423382324e-05, "loss": 0.1712934672832489, "step": 1604 }, { "epoch": 2.3445255474452553, "grad_norm": 1.28125, "learning_rate": 1.6030218919148516e-05, "loss": 0.16689908504486084, "step": 1606 }, { "epoch": 2.3474452554744527, "grad_norm": 0.43359375, "learning_rate": 1.597897795647577e-05, "loss": 0.19200465083122253, "step": 1608 }, { "epoch": 2.3503649635036497, "grad_norm": 0.72265625, "learning_rate": 1.592792604465048e-05, "loss": 0.10364709794521332, "step": 1610 }, { "epoch": 2.3532846715328466, "grad_norm": 0.984375, "learning_rate": 1.587706369108003e-05, "loss": 0.31456390023231506, "step": 1612 }, { "epoch": 2.356204379562044, "grad_norm": 0.3984375, "learning_rate": 1.5826391401287778e-05, "loss": 0.1509348750114441, "step": 1614 }, { "epoch": 2.359124087591241, "grad_norm": 0.85546875, "learning_rate": 1.5775909678908028e-05, "loss": 0.5223057270050049, "step": 1616 }, { "epoch": 2.362043795620438, "grad_norm": 1.2265625, "learning_rate": 1.572561902568105e-05, "loss": 0.4159151017665863, "step": 1618 }, { "epoch": 2.3649635036496353, "grad_norm": 0.498046875, "learning_rate": 1.5675519941448042e-05, "loss": 0.4469154179096222, "step": 1620 }, { "epoch": 2.367883211678832, "grad_norm": 0.61328125, "learning_rate": 1.56256129241462e-05, "loss": 0.12339417636394501, "step": 1622 }, { "epoch": 2.370802919708029, "grad_norm": 0.69140625, "learning_rate": 1.5575898469803747e-05, "loss": 0.0839366763830185, "step": 1624 }, { "epoch": 2.373722627737226, "grad_norm": 2.3125, "learning_rate": 1.552637707253503e-05, "loss": 0.5269365310668945, "step": 1626 }, { "epoch": 2.3766423357664235, "grad_norm": 0.54296875, "learning_rate": 1.5477049224535585e-05, "loss": 0.12736204266548157, "step": 1628 }, { "epoch": 2.3795620437956204, "grad_norm": 1.203125, "learning_rate": 1.5427915416077247e-05, "loss": 0.46538758277893066, "step": 1630 }, { "epoch": 2.3824817518248174, "grad_norm": 0.57421875, "learning_rate": 1.5378976135503283e-05, "loss": 0.4754159450531006, "step": 1632 }, { "epoch": 2.3854014598540147, "grad_norm": 0.365234375, "learning_rate": 1.533023186922356e-05, "loss": 0.23388564586639404, "step": 1634 }, { "epoch": 2.3883211678832117, "grad_norm": 0.6484375, "learning_rate": 1.5281683101709646e-05, "loss": 0.13159261643886566, "step": 1636 }, { "epoch": 2.3912408759124086, "grad_norm": 0.423828125, "learning_rate": 1.5233330315490072e-05, "loss": 0.4673286974430084, "step": 1638 }, { "epoch": 2.394160583941606, "grad_norm": 1.40625, "learning_rate": 1.5185173991145504e-05, "loss": 0.14024242758750916, "step": 1640 }, { "epoch": 2.397080291970803, "grad_norm": 0.44921875, "learning_rate": 1.5137214607303939e-05, "loss": 0.4857579171657562, "step": 1642 }, { "epoch": 2.4, "grad_norm": 0.5078125, "learning_rate": 1.5089452640635994e-05, "loss": 0.06180737540125847, "step": 1644 }, { "epoch": 2.402919708029197, "grad_norm": 0.33984375, "learning_rate": 1.5041888565850121e-05, "loss": 0.1659780740737915, "step": 1646 }, { "epoch": 2.4058394160583942, "grad_norm": 0.51953125, "learning_rate": 1.499452285568796e-05, "loss": 0.5283091068267822, "step": 1648 }, { "epoch": 2.408759124087591, "grad_norm": 0.484375, "learning_rate": 1.4947355980919554e-05, "loss": 0.1580996811389923, "step": 1650 }, { "epoch": 2.411678832116788, "grad_norm": 0.6171875, "learning_rate": 1.4900388410338727e-05, "loss": 0.18507535755634308, "step": 1652 }, { "epoch": 2.4145985401459855, "grad_norm": 0.4453125, "learning_rate": 1.4853620610758403e-05, "loss": 0.13384950160980225, "step": 1654 }, { "epoch": 2.4175182481751825, "grad_norm": 0.380859375, "learning_rate": 1.4807053047005992e-05, "loss": 0.15351298451423645, "step": 1656 }, { "epoch": 2.4204379562043794, "grad_norm": 0.396484375, "learning_rate": 1.4760686181918726e-05, "loss": 0.43027734756469727, "step": 1658 }, { "epoch": 2.423357664233577, "grad_norm": 0.451171875, "learning_rate": 1.4714520476339096e-05, "loss": 0.21505692601203918, "step": 1660 }, { "epoch": 2.4262773722627737, "grad_norm": 0.48046875, "learning_rate": 1.4668556389110251e-05, "loss": 0.4807857871055603, "step": 1662 }, { "epoch": 2.4291970802919707, "grad_norm": 0.455078125, "learning_rate": 1.4622794377071474e-05, "loss": 0.3995106816291809, "step": 1664 }, { "epoch": 2.432116788321168, "grad_norm": 0.482421875, "learning_rate": 1.4577234895053557e-05, "loss": 0.0785241350531578, "step": 1666 }, { "epoch": 2.435036496350365, "grad_norm": 0.478515625, "learning_rate": 1.4531878395874396e-05, "loss": 0.13907024264335632, "step": 1668 }, { "epoch": 2.437956204379562, "grad_norm": 0.51171875, "learning_rate": 1.4486725330334389e-05, "loss": 0.5620059967041016, "step": 1670 }, { "epoch": 2.4408759124087593, "grad_norm": 0.48828125, "learning_rate": 1.444177614721203e-05, "loss": 0.21153351664543152, "step": 1672 }, { "epoch": 2.4437956204379563, "grad_norm": 0.400390625, "learning_rate": 1.4397031293259377e-05, "loss": 0.15002240240573883, "step": 1674 }, { "epoch": 2.4467153284671532, "grad_norm": 0.423828125, "learning_rate": 1.435249121319769e-05, "loss": 0.06266305595636368, "step": 1676 }, { "epoch": 2.44963503649635, "grad_norm": 0.50390625, "learning_rate": 1.4308156349712946e-05, "loss": 0.6670635938644409, "step": 1678 }, { "epoch": 2.4525547445255476, "grad_norm": 0.8828125, "learning_rate": 1.4264027143451482e-05, "loss": 0.0652037113904953, "step": 1680 }, { "epoch": 2.4554744525547445, "grad_norm": 0.515625, "learning_rate": 1.4220104033015571e-05, "loss": 0.21574798226356506, "step": 1682 }, { "epoch": 2.4583941605839414, "grad_norm": 1.671875, "learning_rate": 1.4176387454959128e-05, "loss": 0.42671555280685425, "step": 1684 }, { "epoch": 2.461313868613139, "grad_norm": 0.427734375, "learning_rate": 1.4132877843783294e-05, "loss": 0.19699843227863312, "step": 1686 }, { "epoch": 2.4642335766423358, "grad_norm": 0.435546875, "learning_rate": 1.4089575631932194e-05, "loss": 0.20942422747612, "step": 1688 }, { "epoch": 2.4671532846715327, "grad_norm": 0.453125, "learning_rate": 1.4046481249788559e-05, "loss": 0.11901196837425232, "step": 1690 }, { "epoch": 2.47007299270073, "grad_norm": 0.703125, "learning_rate": 1.4003595125669534e-05, "loss": 0.20953388512134552, "step": 1692 }, { "epoch": 2.472992700729927, "grad_norm": 0.423828125, "learning_rate": 1.3960917685822347e-05, "loss": 0.18220821022987366, "step": 1694 }, { "epoch": 2.475912408759124, "grad_norm": 0.41796875, "learning_rate": 1.3918449354420113e-05, "loss": 0.3743027448654175, "step": 1696 }, { "epoch": 2.478832116788321, "grad_norm": 0.56640625, "learning_rate": 1.3876190553557605e-05, "loss": 0.13356122374534607, "step": 1698 }, { "epoch": 2.4817518248175183, "grad_norm": 0.4453125, "learning_rate": 1.3834141703247076e-05, "loss": 0.433431476354599, "step": 1700 }, { "epoch": 2.4846715328467153, "grad_norm": 0.3984375, "learning_rate": 1.3792303221414052e-05, "loss": 0.18251864612102509, "step": 1702 }, { "epoch": 2.487591240875912, "grad_norm": 0.83984375, "learning_rate": 1.3750675523893204e-05, "loss": 0.10885253548622131, "step": 1704 }, { "epoch": 2.4905109489051096, "grad_norm": 0.326171875, "learning_rate": 1.3709259024424204e-05, "loss": 0.7609469294548035, "step": 1706 }, { "epoch": 2.4934306569343065, "grad_norm": 0.96484375, "learning_rate": 1.3668054134647623e-05, "loss": 0.14499390125274658, "step": 1708 }, { "epoch": 2.4963503649635035, "grad_norm": 0.6640625, "learning_rate": 1.362706126410085e-05, "loss": 0.1034422516822815, "step": 1710 }, { "epoch": 2.499270072992701, "grad_norm": 0.51953125, "learning_rate": 1.358628082021395e-05, "loss": 0.13897135853767395, "step": 1712 }, { "epoch": 2.502189781021898, "grad_norm": 0.46875, "learning_rate": 1.3545713208305729e-05, "loss": 0.1568652242422104, "step": 1714 }, { "epoch": 2.5051094890510948, "grad_norm": 0.412109375, "learning_rate": 1.3505358831579612e-05, "loss": 0.15259164571762085, "step": 1716 }, { "epoch": 2.508029197080292, "grad_norm": 0.423828125, "learning_rate": 1.3465218091119688e-05, "loss": 0.12042444944381714, "step": 1718 }, { "epoch": 2.510948905109489, "grad_norm": 0.6953125, "learning_rate": 1.342529138588668e-05, "loss": 0.4051997661590576, "step": 1720 }, { "epoch": 2.513868613138686, "grad_norm": 0.90625, "learning_rate": 1.3385579112714033e-05, "loss": 0.0928339958190918, "step": 1722 }, { "epoch": 2.5167883211678834, "grad_norm": 0.58203125, "learning_rate": 1.3346081666303925e-05, "loss": 0.3658754527568817, "step": 1724 }, { "epoch": 2.5197080291970804, "grad_norm": 0.515625, "learning_rate": 1.3306799439223361e-05, "loss": 0.3925940990447998, "step": 1726 }, { "epoch": 2.5226277372262773, "grad_norm": 0.5390625, "learning_rate": 1.3267732821900273e-05, "loss": 0.41492193937301636, "step": 1728 }, { "epoch": 2.5255474452554747, "grad_norm": 0.4375, "learning_rate": 1.3228882202619645e-05, "loss": 0.1794070452451706, "step": 1730 }, { "epoch": 2.5284671532846716, "grad_norm": 0.49609375, "learning_rate": 1.3190247967519638e-05, "loss": 0.164272278547287, "step": 1732 }, { "epoch": 2.5313868613138686, "grad_norm": 0.82421875, "learning_rate": 1.3151830500587755e-05, "loss": 0.4470779299736023, "step": 1734 }, { "epoch": 2.5343065693430655, "grad_norm": 0.7578125, "learning_rate": 1.3113630183657041e-05, "loss": 0.056883249431848526, "step": 1736 }, { "epoch": 2.537226277372263, "grad_norm": 0.546875, "learning_rate": 1.3075647396402274e-05, "loss": 0.1761680245399475, "step": 1738 }, { "epoch": 2.54014598540146, "grad_norm": 0.478515625, "learning_rate": 1.3037882516336194e-05, "loss": 0.5256772041320801, "step": 1740 }, { "epoch": 2.543065693430657, "grad_norm": 0.1953125, "learning_rate": 1.3000335918805743e-05, "loss": 0.12634223699569702, "step": 1742 }, { "epoch": 2.5459854014598537, "grad_norm": 0.60546875, "learning_rate": 1.2963007976988347e-05, "loss": 0.14479045569896698, "step": 1744 }, { "epoch": 2.548905109489051, "grad_norm": 1.046875, "learning_rate": 1.2925899061888217e-05, "loss": 0.4536617398262024, "step": 1746 }, { "epoch": 2.551824817518248, "grad_norm": 1.03125, "learning_rate": 1.2889009542332625e-05, "loss": 0.10153871029615402, "step": 1748 }, { "epoch": 2.554744525547445, "grad_norm": 0.3828125, "learning_rate": 1.285233978496827e-05, "loss": 0.5315350294113159, "step": 1750 }, { "epoch": 2.5576642335766424, "grad_norm": 0.515625, "learning_rate": 1.281589015425762e-05, "loss": 0.20661664009094238, "step": 1752 }, { "epoch": 2.5605839416058394, "grad_norm": 0.484375, "learning_rate": 1.2779661012475309e-05, "loss": 0.18551045656204224, "step": 1754 }, { "epoch": 2.5635036496350363, "grad_norm": 0.62890625, "learning_rate": 1.2743652719704498e-05, "loss": 0.2083350121974945, "step": 1756 }, { "epoch": 2.5664233576642337, "grad_norm": 0.98046875, "learning_rate": 1.2707865633833338e-05, "loss": 0.14626437425613403, "step": 1758 }, { "epoch": 2.5693430656934306, "grad_norm": 0.396484375, "learning_rate": 1.2672300110551376e-05, "loss": 0.16228264570236206, "step": 1760 }, { "epoch": 2.5722627737226276, "grad_norm": 0.578125, "learning_rate": 1.2636956503346062e-05, "loss": 0.20038333535194397, "step": 1762 }, { "epoch": 2.575182481751825, "grad_norm": 0.50390625, "learning_rate": 1.2601835163499188e-05, "loss": 0.36375898122787476, "step": 1764 }, { "epoch": 2.578102189781022, "grad_norm": 0.59375, "learning_rate": 1.2566936440083432e-05, "loss": 0.10441319644451141, "step": 1766 }, { "epoch": 2.581021897810219, "grad_norm": 0.515625, "learning_rate": 1.253226067995887e-05, "loss": 0.15817835927009583, "step": 1768 }, { "epoch": 2.5839416058394162, "grad_norm": 0.515625, "learning_rate": 1.2497808227769554e-05, "loss": 0.7072803974151611, "step": 1770 }, { "epoch": 2.586861313868613, "grad_norm": 0.44140625, "learning_rate": 1.2463579425940038e-05, "loss": 0.12276852875947952, "step": 1772 }, { "epoch": 2.58978102189781, "grad_norm": 1.4765625, "learning_rate": 1.242957461467203e-05, "loss": 0.22830504179000854, "step": 1774 }, { "epoch": 2.5927007299270075, "grad_norm": 0.380859375, "learning_rate": 1.2395794131940965e-05, "loss": 0.42617326974868774, "step": 1776 }, { "epoch": 2.5956204379562045, "grad_norm": 0.404296875, "learning_rate": 1.2362238313492683e-05, "loss": 0.0888446643948555, "step": 1778 }, { "epoch": 2.5985401459854014, "grad_norm": 0.625, "learning_rate": 1.232890749284007e-05, "loss": 0.11126016080379486, "step": 1780 }, { "epoch": 2.601459854014599, "grad_norm": 0.46484375, "learning_rate": 1.2295802001259735e-05, "loss": 0.5219573974609375, "step": 1782 }, { "epoch": 2.6043795620437957, "grad_norm": 1.9765625, "learning_rate": 1.2262922167788754e-05, "loss": 0.5340749621391296, "step": 1784 }, { "epoch": 2.6072992700729927, "grad_norm": 0.421875, "learning_rate": 1.223026831922136e-05, "loss": 0.19568416476249695, "step": 1786 }, { "epoch": 2.61021897810219, "grad_norm": 0.05859375, "learning_rate": 1.2197840780105713e-05, "loss": 0.04643552750349045, "step": 1788 }, { "epoch": 2.613138686131387, "grad_norm": 0.80859375, "learning_rate": 1.216563987274067e-05, "loss": 0.10973326861858368, "step": 1790 }, { "epoch": 2.616058394160584, "grad_norm": 0.296875, "learning_rate": 1.21336659171726e-05, "loss": 0.13602161407470703, "step": 1792 }, { "epoch": 2.618978102189781, "grad_norm": 0.35546875, "learning_rate": 1.2101919231192166e-05, "loss": 0.034216154366731644, "step": 1794 }, { "epoch": 2.621897810218978, "grad_norm": 0.486328125, "learning_rate": 1.2070400130331205e-05, "loss": 0.5200577974319458, "step": 1796 }, { "epoch": 2.624817518248175, "grad_norm": 0.890625, "learning_rate": 1.2039108927859557e-05, "loss": 0.19371935725212097, "step": 1798 }, { "epoch": 2.627737226277372, "grad_norm": 0.380859375, "learning_rate": 1.2008045934781995e-05, "loss": 0.21105724573135376, "step": 1800 }, { "epoch": 2.630656934306569, "grad_norm": 1.1015625, "learning_rate": 1.1977211459835078e-05, "loss": 0.14942879974842072, "step": 1802 }, { "epoch": 2.6335766423357665, "grad_norm": 0.419921875, "learning_rate": 1.1946605809484135e-05, "loss": 0.20162081718444824, "step": 1804 }, { "epoch": 2.6364963503649634, "grad_norm": 0.46875, "learning_rate": 1.191622928792018e-05, "loss": 0.16653478145599365, "step": 1806 }, { "epoch": 2.6394160583941604, "grad_norm": 0.98046875, "learning_rate": 1.1886082197056922e-05, "loss": 0.14720329642295837, "step": 1808 }, { "epoch": 2.6423357664233578, "grad_norm": 0.435546875, "learning_rate": 1.1856164836527736e-05, "loss": 0.4846619963645935, "step": 1810 }, { "epoch": 2.6452554744525547, "grad_norm": 0.455078125, "learning_rate": 1.1826477503682697e-05, "loss": 0.18337565660476685, "step": 1812 }, { "epoch": 2.6481751824817517, "grad_norm": 1.03125, "learning_rate": 1.179702049358562e-05, "loss": 0.3766210377216339, "step": 1814 }, { "epoch": 2.651094890510949, "grad_norm": 0.76171875, "learning_rate": 1.1767794099011148e-05, "loss": 0.33012354373931885, "step": 1816 }, { "epoch": 2.654014598540146, "grad_norm": 0.8515625, "learning_rate": 1.1738798610441806e-05, "loss": 0.13045457005500793, "step": 1818 }, { "epoch": 2.656934306569343, "grad_norm": 0.34375, "learning_rate": 1.1710034316065145e-05, "loss": 0.09453101456165314, "step": 1820 }, { "epoch": 2.6598540145985403, "grad_norm": 0.443359375, "learning_rate": 1.1681501501770855e-05, "loss": 0.16232267022132874, "step": 1822 }, { "epoch": 2.6627737226277373, "grad_norm": 0.1533203125, "learning_rate": 1.1653200451147953e-05, "loss": 0.1418476402759552, "step": 1824 }, { "epoch": 2.665693430656934, "grad_norm": 0.62890625, "learning_rate": 1.1625131445481925e-05, "loss": 0.188279390335083, "step": 1826 }, { "epoch": 2.6686131386861316, "grad_norm": 1.9140625, "learning_rate": 1.1597294763751962e-05, "loss": 0.14047500491142273, "step": 1828 }, { "epoch": 2.6715328467153285, "grad_norm": 0.443359375, "learning_rate": 1.1569690682628173e-05, "loss": 0.2133643925189972, "step": 1830 }, { "epoch": 2.6744525547445255, "grad_norm": 0.431640625, "learning_rate": 1.1542319476468853e-05, "loss": 0.18978801369667053, "step": 1832 }, { "epoch": 2.677372262773723, "grad_norm": 0.5859375, "learning_rate": 1.1515181417317706e-05, "loss": 0.095926433801651, "step": 1834 }, { "epoch": 2.68029197080292, "grad_norm": 0.65625, "learning_rate": 1.1488276774901216e-05, "loss": 0.1508825123310089, "step": 1836 }, { "epoch": 2.6832116788321168, "grad_norm": 0.39453125, "learning_rate": 1.1461605816625904e-05, "loss": 0.14228323101997375, "step": 1838 }, { "epoch": 2.686131386861314, "grad_norm": 0.53125, "learning_rate": 1.1435168807575707e-05, "loss": 0.42970946431159973, "step": 1840 }, { "epoch": 2.689051094890511, "grad_norm": 0.3984375, "learning_rate": 1.140896601050931e-05, "loss": 0.3935721814632416, "step": 1842 }, { "epoch": 2.691970802919708, "grad_norm": 0.66796875, "learning_rate": 1.1382997685857571e-05, "loss": 0.8472050428390503, "step": 1844 }, { "epoch": 2.694890510948905, "grad_norm": 0.400390625, "learning_rate": 1.135726409172091e-05, "loss": 0.4143994450569153, "step": 1846 }, { "epoch": 2.697810218978102, "grad_norm": 0.416015625, "learning_rate": 1.1331765483866754e-05, "loss": 0.44374924898147583, "step": 1848 }, { "epoch": 2.7007299270072993, "grad_norm": 1.4140625, "learning_rate": 1.130650211572698e-05, "loss": 0.16854047775268555, "step": 1850 }, { "epoch": 2.7036496350364962, "grad_norm": 0.53515625, "learning_rate": 1.1281474238395413e-05, "loss": 0.12173114717006683, "step": 1852 }, { "epoch": 2.706569343065693, "grad_norm": 0.37109375, "learning_rate": 1.1256682100625323e-05, "loss": 0.12239767611026764, "step": 1854 }, { "epoch": 2.7094890510948906, "grad_norm": 0.4375, "learning_rate": 1.1232125948826955e-05, "loss": 0.2048071324825287, "step": 1856 }, { "epoch": 2.7124087591240875, "grad_norm": 0.380859375, "learning_rate": 1.120780602706507e-05, "loss": 0.45423823595046997, "step": 1858 }, { "epoch": 2.7153284671532845, "grad_norm": 0.462890625, "learning_rate": 1.1183722577056533e-05, "loss": 0.10558777302503586, "step": 1860 }, { "epoch": 2.718248175182482, "grad_norm": 1.375, "learning_rate": 1.1159875838167915e-05, "loss": 0.41063955426216125, "step": 1862 }, { "epoch": 2.721167883211679, "grad_norm": 0.42578125, "learning_rate": 1.1136266047413083e-05, "loss": 0.15179383754730225, "step": 1864 }, { "epoch": 2.7240875912408757, "grad_norm": 0.369140625, "learning_rate": 1.111289343945088e-05, "loss": 0.11223858594894409, "step": 1866 }, { "epoch": 2.727007299270073, "grad_norm": 0.451171875, "learning_rate": 1.1089758246582772e-05, "loss": 0.44403815269470215, "step": 1868 }, { "epoch": 2.72992700729927, "grad_norm": 0.50390625, "learning_rate": 1.1066860698750556e-05, "loss": 0.1581033617258072, "step": 1870 }, { "epoch": 2.732846715328467, "grad_norm": 1.265625, "learning_rate": 1.104420102353404e-05, "loss": 0.36841726303100586, "step": 1872 }, { "epoch": 2.7357664233576644, "grad_norm": 0.6796875, "learning_rate": 1.1021779446148827e-05, "loss": 0.1277724653482437, "step": 1874 }, { "epoch": 2.7386861313868613, "grad_norm": 1.578125, "learning_rate": 1.0999596189444048e-05, "loss": 0.4650922417640686, "step": 1876 }, { "epoch": 2.7416058394160583, "grad_norm": 0.796875, "learning_rate": 1.0977651473900154e-05, "loss": 0.7138586640357971, "step": 1878 }, { "epoch": 2.7445255474452557, "grad_norm": 0.1162109375, "learning_rate": 1.0955945517626712e-05, "loss": 0.4540976881980896, "step": 1880 }, { "epoch": 2.7474452554744526, "grad_norm": 0.66015625, "learning_rate": 1.0934478536360268e-05, "loss": 0.3492112159729004, "step": 1882 }, { "epoch": 2.7503649635036496, "grad_norm": 0.52734375, "learning_rate": 1.0913250743462172e-05, "loss": 0.17992794513702393, "step": 1884 }, { "epoch": 2.753284671532847, "grad_norm": 0.498046875, "learning_rate": 1.0892262349916478e-05, "loss": 0.11098943650722504, "step": 1886 }, { "epoch": 2.756204379562044, "grad_norm": 0.89453125, "learning_rate": 1.0871513564327818e-05, "loss": 0.07393358647823334, "step": 1888 }, { "epoch": 2.759124087591241, "grad_norm": 0.296875, "learning_rate": 1.0851004592919381e-05, "loss": 0.13731540739536285, "step": 1890 }, { "epoch": 2.7620437956204382, "grad_norm": 0.439453125, "learning_rate": 1.0830735639530804e-05, "loss": 0.20814824104309082, "step": 1892 }, { "epoch": 2.764963503649635, "grad_norm": 0.5234375, "learning_rate": 1.0810706905616185e-05, "loss": 0.1031702533364296, "step": 1894 }, { "epoch": 2.767883211678832, "grad_norm": 3.25, "learning_rate": 1.0790918590242073e-05, "loss": 0.1332436501979828, "step": 1896 }, { "epoch": 2.770802919708029, "grad_norm": 0.55859375, "learning_rate": 1.0771370890085475e-05, "loss": 0.43151581287384033, "step": 1898 }, { "epoch": 2.7737226277372264, "grad_norm": 2.53125, "learning_rate": 1.0752063999431919e-05, "loss": 0.5219103097915649, "step": 1900 }, { "epoch": 2.7766423357664234, "grad_norm": 0.26171875, "learning_rate": 1.0732998110173518e-05, "loss": 0.04137423634529114, "step": 1902 }, { "epoch": 2.7795620437956203, "grad_norm": 0.87890625, "learning_rate": 1.071417341180705e-05, "loss": 0.1267162263393402, "step": 1904 }, { "epoch": 2.7824817518248173, "grad_norm": 0.60546875, "learning_rate": 1.0695590091432096e-05, "loss": 0.2184256613254547, "step": 1906 }, { "epoch": 2.7854014598540147, "grad_norm": 0.61328125, "learning_rate": 1.0677248333749168e-05, "loss": 0.1326974630355835, "step": 1908 }, { "epoch": 2.7883211678832116, "grad_norm": 0.73046875, "learning_rate": 1.065914832105787e-05, "loss": 0.25122809410095215, "step": 1910 }, { "epoch": 2.7912408759124085, "grad_norm": 0.6484375, "learning_rate": 1.0641290233255086e-05, "loss": 0.14174973964691162, "step": 1912 }, { "epoch": 2.794160583941606, "grad_norm": 0.52734375, "learning_rate": 1.0623674247833218e-05, "loss": 0.05495617538690567, "step": 1914 }, { "epoch": 2.797080291970803, "grad_norm": 0.65234375, "learning_rate": 1.060630053987838e-05, "loss": 0.15820512175559998, "step": 1916 }, { "epoch": 2.8, "grad_norm": 0.90625, "learning_rate": 1.0589169282068684e-05, "loss": 0.15590894222259521, "step": 1918 }, { "epoch": 2.802919708029197, "grad_norm": 1.2734375, "learning_rate": 1.0572280644672525e-05, "loss": 0.1760368049144745, "step": 1920 }, { "epoch": 2.805839416058394, "grad_norm": 0.68359375, "learning_rate": 1.055563479554688e-05, "loss": 0.11797623336315155, "step": 1922 }, { "epoch": 2.808759124087591, "grad_norm": 1.2109375, "learning_rate": 1.053923190013563e-05, "loss": 0.10800056159496307, "step": 1924 }, { "epoch": 2.8116788321167885, "grad_norm": 0.5234375, "learning_rate": 1.0523072121467941e-05, "loss": 0.5103263854980469, "step": 1926 }, { "epoch": 2.8145985401459854, "grad_norm": 0.59375, "learning_rate": 1.0507155620156628e-05, "loss": 0.233617901802063, "step": 1928 }, { "epoch": 2.8175182481751824, "grad_norm": 0.373046875, "learning_rate": 1.0491482554396559e-05, "loss": 0.12097039818763733, "step": 1930 }, { "epoch": 2.8204379562043798, "grad_norm": 0.50390625, "learning_rate": 1.047605307996308e-05, "loss": 0.16398653388023376, "step": 1932 }, { "epoch": 2.8233576642335767, "grad_norm": 0.55859375, "learning_rate": 1.0460867350210475e-05, "loss": 0.29245248436927795, "step": 1934 }, { "epoch": 2.8262773722627736, "grad_norm": 0.5703125, "learning_rate": 1.0445925516070444e-05, "loss": 0.4952959716320038, "step": 1936 }, { "epoch": 2.829197080291971, "grad_norm": 0.474609375, "learning_rate": 1.0431227726050582e-05, "loss": 0.10879136621952057, "step": 1938 }, { "epoch": 2.832116788321168, "grad_norm": 0.53515625, "learning_rate": 1.0416774126232934e-05, "loss": 0.20576564967632294, "step": 1940 }, { "epoch": 2.835036496350365, "grad_norm": 0.376953125, "learning_rate": 1.0402564860272515e-05, "loss": 0.16286109387874603, "step": 1942 }, { "epoch": 2.8379562043795623, "grad_norm": 0.388671875, "learning_rate": 1.0388600069395904e-05, "loss": 0.1414416879415512, "step": 1944 }, { "epoch": 2.8408759124087593, "grad_norm": 0.6875, "learning_rate": 1.0374879892399819e-05, "loss": 0.05496169626712799, "step": 1946 }, { "epoch": 2.843795620437956, "grad_norm": 0.671875, "learning_rate": 1.036140446564976e-05, "loss": 0.17887184023857117, "step": 1948 }, { "epoch": 2.846715328467153, "grad_norm": 0.390625, "learning_rate": 1.0348173923078633e-05, "loss": 0.39219170808792114, "step": 1950 }, { "epoch": 2.8496350364963505, "grad_norm": 0.482421875, "learning_rate": 1.0335188396185437e-05, "loss": 0.18130755424499512, "step": 1952 }, { "epoch": 2.8525547445255475, "grad_norm": 1.09375, "learning_rate": 1.0322448014033946e-05, "loss": 0.4997798204421997, "step": 1954 }, { "epoch": 2.8554744525547444, "grad_norm": 0.58984375, "learning_rate": 1.030995290325142e-05, "loss": 0.15554431080818176, "step": 1956 }, { "epoch": 2.8583941605839414, "grad_norm": 0.494140625, "learning_rate": 1.0297703188027368e-05, "loss": 0.23963004350662231, "step": 1958 }, { "epoch": 2.8613138686131387, "grad_norm": 0.259765625, "learning_rate": 1.0285698990112294e-05, "loss": 0.4507202208042145, "step": 1960 }, { "epoch": 2.8642335766423357, "grad_norm": 0.54296875, "learning_rate": 1.0273940428816499e-05, "loss": 0.563029944896698, "step": 1962 }, { "epoch": 2.8671532846715326, "grad_norm": 0.453125, "learning_rate": 1.026242762100888e-05, "loss": 0.48327958583831787, "step": 1964 }, { "epoch": 2.87007299270073, "grad_norm": 0.453125, "learning_rate": 1.0251160681115785e-05, "loss": 0.5378194451332092, "step": 1966 }, { "epoch": 2.872992700729927, "grad_norm": 0.25390625, "learning_rate": 1.024013972111987e-05, "loss": 0.18405146896839142, "step": 1968 }, { "epoch": 2.875912408759124, "grad_norm": 1.28125, "learning_rate": 1.0229364850558985e-05, "loss": 0.4143378734588623, "step": 1970 }, { "epoch": 2.8788321167883213, "grad_norm": 0.66796875, "learning_rate": 1.021883617652508e-05, "loss": 0.4933127760887146, "step": 1972 }, { "epoch": 2.8817518248175182, "grad_norm": 0.55078125, "learning_rate": 1.0208553803663148e-05, "loss": 0.4699753522872925, "step": 1974 }, { "epoch": 2.884671532846715, "grad_norm": 0.5625, "learning_rate": 1.019851783417019e-05, "loss": 0.21085891127586365, "step": 1976 }, { "epoch": 2.8875912408759126, "grad_norm": 0.14453125, "learning_rate": 1.0188728367794187e-05, "loss": 0.027944084256887436, "step": 1978 }, { "epoch": 2.8905109489051095, "grad_norm": 0.310546875, "learning_rate": 1.0179185501833107e-05, "loss": 0.1497136652469635, "step": 1980 }, { "epoch": 2.8934306569343065, "grad_norm": 0.53515625, "learning_rate": 1.0169889331133957e-05, "loss": 0.2530011534690857, "step": 1982 }, { "epoch": 2.896350364963504, "grad_norm": 0.8671875, "learning_rate": 1.0160839948091817e-05, "loss": 0.07604291290044785, "step": 1984 }, { "epoch": 2.899270072992701, "grad_norm": 0.53515625, "learning_rate": 1.015203744264894e-05, "loss": 0.38429468870162964, "step": 1986 }, { "epoch": 2.9021897810218977, "grad_norm": 0.0230712890625, "learning_rate": 1.014348190229385e-05, "loss": 0.12912079691886902, "step": 1988 }, { "epoch": 2.905109489051095, "grad_norm": 0.546875, "learning_rate": 1.013517341206047e-05, "loss": 0.16505561769008636, "step": 1990 }, { "epoch": 2.908029197080292, "grad_norm": 0.60546875, "learning_rate": 1.012711205452728e-05, "loss": 0.1626954823732376, "step": 1992 }, { "epoch": 2.910948905109489, "grad_norm": 1.09375, "learning_rate": 1.0119297909816503e-05, "loss": 0.38482239842414856, "step": 1994 }, { "epoch": 2.9138686131386864, "grad_norm": 0.9375, "learning_rate": 1.0111731055593291e-05, "loss": 0.2643647789955139, "step": 1996 }, { "epoch": 2.9167883211678833, "grad_norm": 0.59765625, "learning_rate": 1.0104411567064973e-05, "loss": 0.05413287878036499, "step": 1998 }, { "epoch": 2.9197080291970803, "grad_norm": 0.54296875, "learning_rate": 1.0097339516980295e-05, "loss": 0.21008609235286713, "step": 2000 }, { "epoch": 2.9226277372262772, "grad_norm": 0.52734375, "learning_rate": 1.0090514975628703e-05, "loss": 0.20540976524353027, "step": 2002 }, { "epoch": 2.9255474452554746, "grad_norm": 0.87890625, "learning_rate": 1.0083938010839634e-05, "loss": 0.5189849734306335, "step": 2004 }, { "epoch": 2.9284671532846716, "grad_norm": 0.61328125, "learning_rate": 1.007760868798186e-05, "loss": 0.7012931108474731, "step": 2006 }, { "epoch": 2.9313868613138685, "grad_norm": 0.515625, "learning_rate": 1.0071527069962825e-05, "loss": 0.24337317049503326, "step": 2008 }, { "epoch": 2.9343065693430654, "grad_norm": 0.443359375, "learning_rate": 1.0065693217228018e-05, "loss": 0.15759816765785217, "step": 2010 }, { "epoch": 2.937226277372263, "grad_norm": 0.455078125, "learning_rate": 1.0060107187760375e-05, "loss": 0.1873372197151184, "step": 2012 }, { "epoch": 2.9401459854014598, "grad_norm": 0.49609375, "learning_rate": 1.0054769037079724e-05, "loss": 0.4337967038154602, "step": 2014 }, { "epoch": 2.9430656934306567, "grad_norm": 0.470703125, "learning_rate": 1.0049678818242187e-05, "loss": 0.19185805320739746, "step": 2016 }, { "epoch": 2.945985401459854, "grad_norm": 0.0986328125, "learning_rate": 1.0044836581839697e-05, "loss": 0.3550591468811035, "step": 2018 }, { "epoch": 2.948905109489051, "grad_norm": 0.5234375, "learning_rate": 1.0040242375999476e-05, "loss": 0.4003986120223999, "step": 2020 }, { "epoch": 2.951824817518248, "grad_norm": 0.5390625, "learning_rate": 1.0035896246383552e-05, "loss": 0.21828565001487732, "step": 2022 }, { "epoch": 2.9547445255474454, "grad_norm": 1.1015625, "learning_rate": 1.0031798236188314e-05, "loss": 0.39733898639678955, "step": 2024 }, { "epoch": 2.9576642335766423, "grad_norm": 2.25, "learning_rate": 1.0027948386144085e-05, "loss": 0.3748105764389038, "step": 2026 }, { "epoch": 2.9605839416058393, "grad_norm": 0.5625, "learning_rate": 1.0024346734514706e-05, "loss": 0.42706161737442017, "step": 2028 }, { "epoch": 2.9635036496350367, "grad_norm": 0.53125, "learning_rate": 1.0020993317097167e-05, "loss": 0.13745678961277008, "step": 2030 }, { "epoch": 2.9664233576642336, "grad_norm": 0.54296875, "learning_rate": 1.0017888167221238e-05, "loss": 0.5219398736953735, "step": 2032 }, { "epoch": 2.9693430656934305, "grad_norm": 0.49609375, "learning_rate": 1.0015031315749157e-05, "loss": 0.21247032284736633, "step": 2034 }, { "epoch": 2.972262773722628, "grad_norm": 0.5546875, "learning_rate": 1.0012422791075309e-05, "loss": 0.13911724090576172, "step": 2036 }, { "epoch": 2.975182481751825, "grad_norm": 1.2890625, "learning_rate": 1.0010062619125943e-05, "loss": 0.1428239345550537, "step": 2038 }, { "epoch": 2.978102189781022, "grad_norm": 0.99609375, "learning_rate": 1.0007950823358923e-05, "loss": 0.20074939727783203, "step": 2040 }, { "epoch": 2.981021897810219, "grad_norm": 1.03125, "learning_rate": 1.0006087424763482e-05, "loss": 0.5967737436294556, "step": 2042 }, { "epoch": 2.983941605839416, "grad_norm": 0.875, "learning_rate": 1.0004472441860033e-05, "loss": 0.43237411975860596, "step": 2044 }, { "epoch": 2.986861313868613, "grad_norm": 0.267578125, "learning_rate": 1.0003105890699969e-05, "loss": 0.01062680408358574, "step": 2046 }, { "epoch": 2.9897810218978105, "grad_norm": 0.55859375, "learning_rate": 1.0001987784865506e-05, "loss": 0.538578450679779, "step": 2048 }, { "epoch": 2.9927007299270074, "grad_norm": 0.5703125, "learning_rate": 1.0001118135469548e-05, "loss": 0.28526389598846436, "step": 2050 }, { "epoch": 2.9956204379562044, "grad_norm": 6.21875, "learning_rate": 1.0000496951155586e-05, "loss": 0.3971728980541229, "step": 2052 }, { "epoch": 2.9985401459854013, "grad_norm": 0.6484375, "learning_rate": 1.0000124238097599e-05, "loss": 0.07702690362930298, "step": 2054 }, { "epoch": 3.0, "step": 2055, "total_flos": 2.976816812190073e+18, "train_loss": 0.8063934714925405, "train_runtime": 19281.3988, "train_samples_per_second": 1.705, "train_steps_per_second": 0.107 } ], "logging_steps": 2, "max_steps": 2055, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 9999999, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2.976816812190073e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }