diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5170 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "global_step": 1718,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 2e-05,
+      "loss": 1.0508,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9999932966068875e-05,
+      "loss": 0.9636,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9999731865174213e-05,
+      "loss": 0.9023,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9999396700012127e-05,
+      "loss": 0.8735,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9998927475076107e-05,
+      "loss": 0.8538,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.999832419665695e-05,
+      "loss": 0.8962,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9997586872842683e-05,
+      "loss": 0.7915,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9996715513518446e-05,
+      "loss": 0.7651,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.999571013036637e-05,
+      "loss": 0.7764,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9994570736865407e-05,
+      "loss": 0.7781,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9993297348291168e-05,
+      "loss": 0.7783,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9991889981715696e-05,
+      "loss": 0.771,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.999034865600726e-05,
+      "loss": 0.7458,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9988673391830082e-05,
+      "loss": 0.6932,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.998686421164407e-05,
+      "loss": 0.7495,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.998492113970451e-05,
+      "loss": 0.739,
+      "step": 32
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.998284420206176e-05,
+      "loss": 0.7358,
+      "step": 34
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.998063342656088e-05,
+      "loss": 0.7156,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.997828884284126e-05,
+      "loss": 0.7146,
+      "step": 38
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9975810482336234e-05,
+      "loss": 0.73,
+      "step": 40
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.997319837827265e-05,
+      "loss": 0.6868,
+      "step": 42
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9970452565670432e-05,
+      "loss": 0.6707,
+      "step": 44
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9967573081342103e-05,
+      "loss": 0.6899,
+      "step": 46
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9964559963892286e-05,
+      "loss": 0.7239,
+      "step": 48
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9961413253717214e-05,
+      "loss": 0.7026,
+      "step": 50
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.995813299300415e-05,
+      "loss": 0.6956,
+      "step": 52
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9954719225730847e-05,
+      "loss": 0.7076,
+      "step": 54
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.995117199766496e-05,
+      "loss": 0.6381,
+      "step": 56
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.994749135636341e-05,
+      "loss": 0.6985,
+      "step": 58
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9943677351171776e-05,
+      "loss": 0.6593,
+      "step": 60
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.99397300332236e-05,
+      "loss": 0.709,
+      "step": 62
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.993564945543974e-05,
+      "loss": 0.6777,
+      "step": 64
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9931435672527625e-05,
+      "loss": 0.7051,
+      "step": 66
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.992708874098054e-05,
+      "loss": 0.6799,
+      "step": 68
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9922608719076874e-05,
+      "loss": 0.6899,
+      "step": 70
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9917995666879317e-05,
+      "loss": 0.6467,
+      "step": 72
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9913249646234072e-05,
+      "loss": 0.6636,
+      "step": 74
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9908370720770027e-05,
+      "loss": 0.6821,
+      "step": 76
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9903358955897892e-05,
+      "loss": 0.6271,
+      "step": 78
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.7009,
+      "step": 80
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9892937178476044e-05,
+      "loss": 0.6855,
+      "step": 82
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9887527305648874e-05,
+      "loss": 0.6577,
+      "step": 84
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.988198487285682e-05,
+      "loss": 0.6873,
+      "step": 86
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.98763099544061e-05,
+      "loss": 0.6523,
+      "step": 88
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9870502626379127e-05,
+      "loss": 0.6362,
+      "step": 90
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9864562966633517e-05,
+      "loss": 0.6409,
+      "step": 92
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9858491054801007e-05,
+      "loss": 0.6694,
+      "step": 94
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.985228697228643e-05,
+      "loss": 0.6479,
+      "step": 96
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9845950802266584e-05,
+      "loss": 0.7373,
+      "step": 98
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9839482629689154e-05,
+      "loss": 0.6663,
+      "step": 100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9832882541271545e-05,
+      "loss": 0.6294,
+      "step": 102
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9826150625499733e-05,
+      "loss": 0.6858,
+      "step": 104
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9819286972627066e-05,
+      "loss": 0.6548,
+      "step": 106
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.981229167467308e-05,
+      "loss": 0.7283,
+      "step": 108
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.980516482542224e-05,
+      "loss": 0.6324,
+      "step": 110
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.979790652042268e-05,
+      "loss": 0.6455,
+      "step": 112
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.979051685698495e-05,
+      "loss": 0.6489,
+      "step": 114
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9782995934180687e-05,
+      "loss": 0.6376,
+      "step": 116
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9775343852841297e-05,
+      "loss": 0.679,
+      "step": 118
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.97675607155566e-05,
+      "loss": 0.6584,
+      "step": 120
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9759646626673445e-05,
+      "loss": 0.6555,
+      "step": 122
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.975160169229434e-05,
+      "loss": 0.6591,
+      "step": 124
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9743426020275996e-05,
+      "loss": 0.6335,
+      "step": 126
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.97351197202279e-05,
+      "loss": 0.6691,
+      "step": 128
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.972668290351084e-05,
+      "loss": 0.6639,
+      "step": 130
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9718115683235418e-05,
+      "loss": 0.6385,
+      "step": 132
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.6393,
+      "step": 134
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9700590493191796e-05,
+      "loss": 0.632,
+      "step": 136
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9691632758380074e-05,
+      "loss": 0.6694,
+      "step": 138
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9682545089919784e-05,
+      "loss": 0.6478,
+      "step": 140
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9673327609647367e-05,
+      "loss": 0.6167,
+      "step": 142
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9663980441139603e-05,
+      "loss": 0.6553,
+      "step": 144
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9654503709711984e-05,
+      "loss": 0.6674,
+      "step": 146
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.964489754241702e-05,
+      "loss": 0.635,
+      "step": 148
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9635162068042547e-05,
+      "loss": 0.6555,
+      "step": 150
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9625297417109982e-05,
+      "loss": 0.6583,
+      "step": 152
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9615303721872598e-05,
+      "loss": 0.6429,
+      "step": 154
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9605181116313725e-05,
+      "loss": 0.6475,
+      "step": 156
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.6976,
+      "step": 158
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.958454971880441e-05,
+      "loss": 0.6136,
+      "step": 160
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9574041203454707e-05,
+      "loss": 0.6458,
+      "step": 162
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9563404330981276e-05,
+      "loss": 0.6337,
+      "step": 164
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.95526392439904e-05,
+      "loss": 0.6094,
+      "step": 166
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9541746086807296e-05,
+      "loss": 0.6368,
+      "step": 168
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9530725005474195e-05,
+      "loss": 0.636,
+      "step": 170
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9519576147748377e-05,
+      "loss": 0.6044,
+      "step": 172
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9508299663100195e-05,
+      "loss": 0.6442,
+      "step": 174
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.949689570271107e-05,
+      "loss": 0.6149,
+      "step": 176
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9485364419471454e-05,
+      "loss": 0.6427,
+      "step": 178
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.947370596797881e-05,
+      "loss": 0.5715,
+      "step": 180
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.946192050453549e-05,
+      "loss": 0.6631,
+      "step": 182
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9450008187146685e-05,
+      "loss": 0.6294,
+      "step": 184
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9437969175518296e-05,
+      "loss": 0.6567,
+      "step": 186
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9425803631054773e-05,
+      "loss": 0.6165,
+      "step": 188
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9413511716856973e-05,
+      "loss": 0.658,
+      "step": 190
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.940109359771996e-05,
+      "loss": 0.6257,
+      "step": 192
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.93885494401308e-05,
+      "loss": 0.6348,
+      "step": 194
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9375879412266336e-05,
+      "loss": 0.6298,
+      "step": 196
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.936308368399092e-05,
+      "loss": 0.6532,
+      "step": 198
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9350162426854152e-05,
+      "loss": 0.6276,
+      "step": 200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9337115814088555e-05,
+      "loss": 0.6646,
+      "step": 202
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9323944020607282e-05,
+      "loss": 0.6339,
+      "step": 204
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9310647223001752e-05,
+      "loss": 0.6345,
+      "step": 206
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.929722559953929e-05,
+      "loss": 0.6079,
+      "step": 208
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9283679330160726e-05,
+      "loss": 0.6178,
+      "step": 210
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9270008596478008e-05,
+      "loss": 0.6501,
+      "step": 212
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.925621358177173e-05,
+      "loss": 0.6594,
+      "step": 214
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9242294470988714e-05,
+      "loss": 0.5562,
+      "step": 216
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9228251450739495e-05,
+      "loss": 0.709,
+      "step": 218
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9214084709295848e-05,
+      "loss": 0.6271,
+      "step": 220
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9199794436588244e-05,
+      "loss": 0.5695,
+      "step": 222
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.918538082420332e-05,
+      "loss": 0.6212,
+      "step": 224
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9170844065381287e-05,
+      "loss": 0.6267,
+      "step": 226
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9156184355013366e-05,
+      "loss": 0.6641,
+      "step": 228
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9141401889639167e-05,
+      "loss": 0.6367,
+      "step": 230
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.912649686744404e-05,
+      "loss": 0.5789,
+      "step": 232
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.911146948825642e-05,
+      "loss": 0.6238,
+      "step": 234
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.6227,
+      "step": 236
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9081048466416897e-05,
+      "loss": 0.6316,
+      "step": 238
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9065655231613122e-05,
+      "loss": 0.6587,
+      "step": 240
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.905014045550767e-05,
+      "loss": 0.5999,
+      "step": 242
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9034504346103825e-05,
+      "loss": 0.6345,
+      "step": 244
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9018747113031563e-05,
+      "loss": 0.6772,
+      "step": 246
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9002868967544743e-05,
+      "loss": 0.6509,
+      "step": 248
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.898687012251826e-05,
+      "loss": 0.6267,
+      "step": 250
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.897075079244522e-05,
+      "loss": 0.6625,
+      "step": 252
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.8954511193434024e-05,
+      "loss": 0.6273,
+      "step": 254
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.8938151543205513e-05,
+      "loss": 0.6298,
+      "step": 256
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.8921672061090014e-05,
+      "loss": 0.6119,
+      "step": 258
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.8905072968024424e-05,
+      "loss": 0.6257,
+      "step": 260
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.8888354486549238e-05,
+      "loss": 0.5973,
+      "step": 262
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.887151684080556e-05,
+      "loss": 0.6372,
+      "step": 264
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.6656,
+      "step": 266
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.8837484961062173e-05,
+      "loss": 0.6205,
+      "step": 268
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.8820291183320602e-05,
+      "loss": 0.6591,
+      "step": 270
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.88029791538207e-05,
+      "loss": 0.6532,
+      "step": 272
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.8785549104661135e-05,
+      "loss": 0.6185,
+      "step": 274
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.8768001269522858e-05,
+      "loss": 0.6201,
+      "step": 276
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.8750335883665948e-05,
+      "loss": 0.6368,
+      "step": 278
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.8732553183926444e-05,
+      "loss": 0.6105,
+      "step": 280
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.8714653408713204e-05,
+      "loss": 0.6622,
+      "step": 282
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.8696636798004693e-05,
+      "loss": 0.6339,
+      "step": 284
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.8678503593345755e-05,
+      "loss": 0.6726,
+      "step": 286
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.6531,
+      "step": 288
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.8641888376168483e-05,
+      "loss": 0.6283,
+      "step": 290
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.862340685454254e-05,
+      "loss": 0.5449,
+      "step": 292
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.860480972074437e-05,
+      "loss": 0.7053,
+      "step": 294
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.8586097224101767e-05,
+      "loss": 0.5902,
+      "step": 296
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.8567269615489172e-05,
+      "loss": 0.5852,
+      "step": 298
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.8548327147324316e-05,
+      "loss": 0.6201,
+      "step": 300
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.852927007356481e-05,
+      "loss": 0.6003,
+      "step": 302
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.8510098649704776e-05,
+      "loss": 0.5868,
+      "step": 304
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.8490813132771396e-05,
+      "loss": 0.6387,
+      "step": 306
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.847141378132147e-05,
+      "loss": 0.6387,
+      "step": 308
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.845190085543795e-05,
+      "loss": 0.6222,
+      "step": 310
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.843227461672647e-05,
+      "loss": 0.7063,
+      "step": 312
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.6257,
+      "step": 314
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.83926832548344e-05,
+      "loss": 0.5979,
+      "step": 316
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.837271866244674e-05,
+      "loss": 0.6379,
+      "step": 318
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.835264181880985e-05,
+      "loss": 0.6497,
+      "step": 320
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.833245299308968e-05,
+      "loss": 0.6034,
+      "step": 322
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.8312152455953505e-05,
+      "loss": 0.5946,
+      "step": 324
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.8291740479566286e-05,
+      "loss": 0.6259,
+      "step": 326
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.8271217337587023e-05,
+      "loss": 0.636,
+      "step": 328
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.8250583305165098e-05,
+      "loss": 0.6006,
+      "step": 330
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.8229838658936566e-05,
+      "loss": 0.6049,
+      "step": 332
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.820898367702047e-05,
+      "loss": 0.6079,
+      "step": 334
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.8188018639015095e-05,
+      "loss": 0.6034,
+      "step": 336
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.816694382599422e-05,
+      "loss": 0.572,
+      "step": 338
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.814575952050336e-05,
+      "loss": 0.5978,
+      "step": 340
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.8124466006555967e-05,
+      "loss": 0.6515,
+      "step": 342
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.8103063569629635e-05,
+      "loss": 0.635,
+      "step": 344
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.8081552496662256e-05,
+      "loss": 0.5991,
+      "step": 346
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.8059933076048193e-05,
+      "loss": 0.5863,
+      "step": 348
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.8038205597634392e-05,
+      "loss": 0.5939,
+      "step": 350
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.8016370352716514e-05,
+      "loss": 0.5619,
+      "step": 352
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.7994427634035016e-05,
+      "loss": 0.621,
+      "step": 354
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.7972377735771234e-05,
+      "loss": 0.6052,
+      "step": 356
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.795022095354345e-05,
+      "loss": 0.6111,
+      "step": 358
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.79279575844029e-05,
+      "loss": 0.6212,
+      "step": 360
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.7905587926829815e-05,
+      "loss": 0.5986,
+      "step": 362
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.7883112280729414e-05,
+      "loss": 0.6418,
+      "step": 364
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.7860530947427878e-05,
+      "loss": 0.5688,
+      "step": 366
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.7837844229668312e-05,
+      "loss": 0.5787,
+      "step": 368
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.7815052431606702e-05,
+      "loss": 0.6035,
+      "step": 370
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.7792155858807802e-05,
+      "loss": 0.6589,
+      "step": 372
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.776915481824107e-05,
+      "loss": 0.6351,
+      "step": 374
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.7746049618276545e-05,
+      "loss": 0.6035,
+      "step": 376
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.7722840568680704e-05,
+      "loss": 0.619,
+      "step": 378
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.7699527980612306e-05,
+      "loss": 0.5828,
+      "step": 380
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.7676112166618236e-05,
+      "loss": 0.6061,
+      "step": 382
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.7652593440629316e-05,
+      "loss": 0.5815,
+      "step": 384
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.762897211795607e-05,
+      "loss": 0.636,
+      "step": 386
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.7605248515284525e-05,
+      "loss": 0.6108,
+      "step": 388
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.7581422950671942e-05,
+      "loss": 0.6128,
+      "step": 390
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.6058,
+      "step": 392
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.7533467214683394e-05,
+      "loss": 0.6254,
+      "step": 394
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.750933768623973e-05,
+      "loss": 0.5851,
+      "step": 396
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.6285,
+      "step": 398
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.7460776925946418e-05,
+      "loss": 0.605,
+      "step": 400
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.743634634514051e-05,
+      "loss": 0.6234,
+      "step": 402
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.7411816066828852e-05,
+      "loss": 0.6101,
+      "step": 404
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.738718641988365e-05,
+      "loss": 0.6578,
+      "step": 406
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.7362457734509314e-05,
+      "loss": 0.6051,
+      "step": 408
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.733763034223804e-05,
+      "loss": 0.5908,
+      "step": 410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.7312704575925374e-05,
+      "loss": 0.6151,
+      "step": 412
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.728768076974573e-05,
+      "loss": 0.5917,
+      "step": 414
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.7262559259187936e-05,
+      "loss": 0.6049,
+      "step": 416
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.72373403810507e-05,
+      "loss": 0.5818,
+      "step": 418
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.7212024473438145e-05,
+      "loss": 0.613,
+      "step": 420
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.7186611875755227e-05,
+      "loss": 0.5859,
+      "step": 422
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.7161102928703206e-05,
+      "loss": 0.6194,
+      "step": 424
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.7135497974275088e-05,
+      "loss": 0.6182,
+      "step": 426
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.7109797355751017e-05,
+      "loss": 0.6274,
+      "step": 428
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.7084001417693702e-05,
+      "loss": 0.6072,
+      "step": 430
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.705811050594376e-05,
+      "loss": 0.5907,
+      "step": 432
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.7032124967615112e-05,
+      "loss": 0.5764,
+      "step": 434
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.7006045151090318e-05,
+      "loss": 0.575,
+      "step": 436
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.69798714060159e-05,
+      "loss": 0.568,
+      "step": 438
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.6953604083297665e-05,
+      "loss": 0.6179,
+      "step": 440
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.6927243535095995e-05,
+      "loss": 0.6506,
+      "step": 442
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.6900790114821122e-05,
+      "loss": 0.6372,
+      "step": 444
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.6874244177128395e-05,
+      "loss": 0.5818,
+      "step": 446
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.6847606077913527e-05,
+      "loss": 0.623,
+      "step": 448
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.682087617430782e-05,
+      "loss": 0.5952,
+      "step": 450
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.679405482467338e-05,
+      "loss": 0.5507,
+      "step": 452
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.676714238859831e-05,
+      "loss": 0.562,
+      "step": 454
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.674013922689188e-05,
+      "loss": 0.5887,
+      "step": 456
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.6713045701579705e-05,
+      "loss": 0.6136,
+      "step": 458
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.6685862175898892e-05,
+      "loss": 0.5789,
+      "step": 460
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.665858901429316e-05,
+      "loss": 0.5686,
+      "step": 462
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.6631226582407954e-05,
+      "loss": 0.6134,
+      "step": 464
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.660377524708555e-05,
+      "loss": 0.5704,
+      "step": 466
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.6576235376360128e-05,
+      "loss": 0.5771,
+      "step": 468
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.5836,
+      "step": 470
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.6520891506766905e-05,
+      "loss": 0.6,
+      "step": 472
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.6493088249882533e-05,
+      "loss": 0.6404,
+      "step": 474
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.6465197941552054e-05,
+      "loss": 0.6133,
+      "step": 476
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.643722095569487e-05,
+      "loss": 0.5649,
+      "step": 478
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.6409157667392456e-05,
+      "loss": 0.6,
+      "step": 480
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.638100845288331e-05,
+      "loss": 0.5974,
+      "step": 482
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.6352773689557938e-05,
+      "loss": 0.6078,
+      "step": 484
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.6324453755953772e-05,
+      "loss": 0.6965,
+      "step": 486
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.6296049031750114e-05,
+      "loss": 0.5884,
+      "step": 488
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.626755989776303e-05,
+      "loss": 0.6278,
+      "step": 490
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.623898673594024e-05,
+      "loss": 0.5658,
+      "step": 492
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.6210329929356017e-05,
+      "loss": 0.6034,
+      "step": 494
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.6181589862206053e-05,
+      "loss": 0.6543,
+      "step": 496
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.6152766919802275e-05,
+      "loss": 0.6089,
+      "step": 498
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.612386148856771e-05,
+      "loss": 0.6199,
+      "step": 500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.60948739560313e-05,
+      "loss": 0.585,
+      "step": 502
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.606580471082269e-05,
+      "loss": 0.587,
+      "step": 504
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.6036654142667043e-05,
+      "loss": 0.6616,
+      "step": 506
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.600742264237979e-05,
+      "loss": 0.5929,
+      "step": 508
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.5978110601861408e-05,
+      "loss": 0.6448,
+      "step": 510
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.5948718414092163e-05,
+      "loss": 0.6243,
+      "step": 512
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.5919246473126825e-05,
+      "loss": 0.6438,
+      "step": 514
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.588969517408941e-05,
+      "loss": 0.6562,
+      "step": 516
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.5860064913167863e-05,
+      "loss": 0.6133,
+      "step": 518
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.5830356087608763e-05,
+      "loss": 0.6266,
+      "step": 520
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.5800569095711983e-05,
+      "loss": 0.5811,
+      "step": 522
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.5770704336825354e-05,
+      "loss": 0.6057,
+      "step": 524
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.5740762211339314e-05,
+      "loss": 0.6555,
+      "step": 526
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.571074312068154e-05,
+      "loss": 0.6107,
+      "step": 528
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.6567,
+      "step": 530
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.5650475654715367e-05,
+      "loss": 0.6681,
+      "step": 532
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.5620228087400004e-05,
+      "loss": 0.6376,
+      "step": 534
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.5589905170888136e-05,
+      "loss": 0.6646,
+      "step": 536
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.5559507311712626e-05,
+      "loss": 0.6548,
+      "step": 538
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.552903491741107e-05,
+      "loss": 0.6615,
+      "step": 540
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.549848839652035e-05,
+      "loss": 0.6372,
+      "step": 542
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.5467868158571138e-05,
+      "loss": 0.6307,
+      "step": 544
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.5437174614082418e-05,
+      "loss": 0.6758,
+      "step": 546
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.6561,
+      "step": 548
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.5375569252470897e-05,
+      "loss": 0.6129,
+      "step": 550
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.5344658261278013e-05,
+      "loss": 0.5748,
+      "step": 552
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.5313675615394373e-05,
+      "loss": 0.6631,
+      "step": 554
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.528262173019769e-05,
+      "loss": 0.6227,
+      "step": 556
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.5251497022020764e-05,
+      "loss": 0.6296,
+      "step": 558
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.5220301908145905e-05,
+      "loss": 0.5913,
+      "step": 560
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.5189036806799333e-05,
+      "loss": 0.604,
+      "step": 562
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.5157702137145575e-05,
+      "loss": 0.6252,
+      "step": 564
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.5126298319281859e-05,
+      "loss": 0.6234,
+      "step": 566
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.5094825774232447e-05,
+      "loss": 0.6163,
+      "step": 568
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.506328492394303e-05,
+      "loss": 0.6284,
+      "step": 570
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.503167619127504e-05,
+      "loss": 0.6534,
+      "step": 572
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.6293,
+      "step": 574
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.4968256774793839e-05,
+      "loss": 0.6257,
+      "step": 576
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.4936446941231186e-05,
+      "loss": 0.6289,
+      "step": 578
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.4904570925779682e-05,
+      "loss": 0.6068,
+      "step": 580
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.4872629155794251e-05,
+      "loss": 0.6091,
+      "step": 582
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.4840622059511376e-05,
+      "loss": 4.1061,
+      "step": 584
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.4808550066043353e-05,
+      "loss": 0.7163,
+      "step": 586
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.4776413605372544e-05,
+      "loss": 0.6737,
+      "step": 588
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.4744213108345605e-05,
+      "loss": 0.6829,
+      "step": 590
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.4711949006667718e-05,
+      "loss": 0.6444,
+      "step": 592
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.4679621732896796e-05,
+      "loss": 0.6283,
+      "step": 594
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.4647231720437687e-05,
+      "loss": 0.7153,
+      "step": 596
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.4614779403536366e-05,
+      "loss": 0.6687,
+      "step": 598
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.4582265217274105e-05,
+      "loss": 0.6541,
+      "step": 600
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.4549689597561652e-05,
+      "loss": 0.6589,
+      "step": 602
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.4517052981133375e-05,
+      "loss": 0.6356,
+      "step": 604
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4484355805541413e-05,
+      "loss": 0.6328,
+      "step": 606
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4451598509149809e-05,
+      "loss": 0.6481,
+      "step": 608
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4418781531128636e-05,
+      "loss": 0.6239,
+      "step": 610
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4385905311448098e-05,
+      "loss": 0.6473,
+      "step": 612
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.435297029087265e-05,
+      "loss": 0.6244,
+      "step": 614
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4319976910955067e-05,
+      "loss": 0.5803,
+      "step": 616
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4286925614030542e-05,
+      "loss": 0.6458,
+      "step": 618
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4253816843210751e-05,
+      "loss": 0.6193,
+      "step": 620
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.42206510423779e-05,
+      "loss": 0.6089,
+      "step": 622
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4187428656178796e-05,
+      "loss": 0.585,
+      "step": 624
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.6609,
+      "step": 626
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4120815910056196e-05,
+      "loss": 0.6121,
+      "step": 628
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4087426443195549e-05,
+      "loss": 0.6002,
+      "step": 630
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4053982177082369e-05,
+      "loss": 0.5796,
+      "step": 632
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.402048356009678e-05,
+      "loss": 0.5991,
+      "step": 634
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.398693104134758e-05,
+      "loss": 0.5928,
+      "step": 636
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3953325070666215e-05,
+      "loss": 0.5884,
+      "step": 638
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3919666098600753e-05,
+      "loss": 0.5908,
+      "step": 640
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.388595457640983e-05,
+      "loss": 0.5999,
+      "step": 642
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3852190956056623e-05,
+      "loss": 0.6143,
+      "step": 644
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3818375690202774e-05,
+      "loss": 0.5918,
+      "step": 646
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3784509232202317e-05,
+      "loss": 0.5693,
+      "step": 648
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.375059203609562e-05,
+      "loss": 0.6007,
+      "step": 650
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3716624556603275e-05,
+      "loss": 0.6082,
+      "step": 652
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3682607249120024e-05,
+      "loss": 0.5833,
+      "step": 654
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3648540569708637e-05,
+      "loss": 0.5956,
+      "step": 656
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3614424975093794e-05,
+      "loss": 0.5646,
+      "step": 658
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3580260922655986e-05,
+      "loss": 0.5762,
+      "step": 660
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.5704,
+      "step": 662
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3511789277075582e-05,
+      "loss": 0.594,
+      "step": 664
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.34774826019177e-05,
+      "loss": 0.6062,
+      "step": 666
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.3443129304893974e-05,
+      "loss": 0.5741,
+      "step": 668
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.3408729846571716e-05,
+      "loss": 0.5793,
+      "step": 670
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.3374284688137105e-05,
+      "loss": 0.5925,
+      "step": 672
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.3339794291389015e-05,
+      "loss": 0.5909,
+      "step": 674
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.3305259118732827e-05,
+      "loss": 0.5466,
+      "step": 676
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.3270679633174219e-05,
+      "loss": 0.5699,
+      "step": 678
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.3236056298312957e-05,
+      "loss": 0.584,
+      "step": 680
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.320138957833669e-05,
+      "loss": 0.6599,
+      "step": 682
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.3166679938014728e-05,
+      "loss": 0.6227,
+      "step": 684
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.3131927842691793e-05,
+      "loss": 0.614,
+      "step": 686
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.3097133758281795e-05,
+      "loss": 0.5942,
+      "step": 688
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.3062298151261592e-05,
+      "loss": 0.5587,
+      "step": 690
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.3027421488664723e-05,
+      "loss": 0.66,
+      "step": 692
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.299250423807514e-05,
+      "loss": 0.599,
+      "step": 694
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2957546867620956e-05,
+      "loss": 0.5881,
+      "step": 696
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2922549845968174e-05,
+      "loss": 0.5818,
+      "step": 698
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2887513642314373e-05,
+      "loss": 0.5358,
+      "step": 700
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2852438726382448e-05,
+      "loss": 0.635,
+      "step": 702
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.5869,
+      "step": 704
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2782174639164528e-05,
+      "loss": 0.5635,
+      "step": 706
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.274698640989413e-05,
+      "loss": 0.6047,
+      "step": 708
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2711761352364172e-05,
+      "loss": 0.5842,
+      "step": 710
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2676499938829469e-05,
+      "loss": 0.5934,
+      "step": 712
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2641202642032257e-05,
+      "loss": 0.5756,
+      "step": 714
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2605869935195844e-05,
+      "loss": 0.6052,
+      "step": 716
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2570502292018281e-05,
+      "loss": 0.5312,
+      "step": 718
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2535100186666e-05,
+      "loss": 0.6018,
+      "step": 720
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2499664093767458e-05,
+      "loss": 0.6255,
+      "step": 722
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2464194488406775e-05,
+      "loss": 0.5896,
+      "step": 724
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.2428691846117371e-05,
+      "loss": 0.5957,
+      "step": 726
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.2393156642875579e-05,
+      "loss": 0.5859,
+      "step": 728
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.2357589355094275e-05,
+      "loss": 0.5979,
+      "step": 730
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.232199045961648e-05,
+      "loss": 0.6124,
+      "step": 732
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.2286360433708976e-05,
+      "loss": 0.5422,
+      "step": 734
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.2250699755055905e-05,
+      "loss": 0.641,
+      "step": 736
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.2215008901752361e-05,
+      "loss": 0.6781,
+      "step": 738
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.2179288352297985e-05,
+      "loss": 0.6317,
+      "step": 740
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.2143538585590548e-05,
+      "loss": 0.6423,
+      "step": 742
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.210776008091953e-05,
+      "loss": 0.594,
+      "step": 744
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.2071953317959692e-05,
+      "loss": 0.6129,
+      "step": 746
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.2036118776764653e-05,
+      "loss": 0.5791,
+      "step": 748
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.2000256937760446e-05,
+      "loss": 0.5807,
+      "step": 750
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1964368281739078e-05,
+      "loss": 0.6272,
+      "step": 752
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.192845328985209e-05,
+      "loss": 0.6018,
+      "step": 754
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1892512443604103e-05,
+      "loss": 0.5743,
+      "step": 756
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1856546224846354e-05,
+      "loss": 0.5873,
+      "step": 758
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1820555115770256e-05,
+      "loss": 0.5973,
+      "step": 760
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1784539598900905e-05,
+      "loss": 0.5819,
+      "step": 762
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1748500157090645e-05,
+      "loss": 0.6498,
+      "step": 764
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1712437273512562e-05,
+      "loss": 0.5624,
+      "step": 766
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.167635143165403e-05,
+      "loss": 0.5734,
+      "step": 768
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1640243115310219e-05,
+      "loss": 0.765,
+      "step": 770
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1604112808577603e-05,
+      "loss": 0.5706,
+      "step": 772
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1567960995847485e-05,
+      "loss": 0.6354,
+      "step": 774
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.153178816179948e-05,
+      "loss": 0.5659,
+      "step": 776
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.149559479139505e-05,
+      "loss": 0.5551,
+      "step": 778
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1459381369870974e-05,
+      "loss": 0.6296,
+      "step": 780
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.5815,
+      "step": 782
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1386896315748597e-05,
+      "loss": 0.5753,
+      "step": 784
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1350625654941919e-05,
+      "loss": 0.582,
+      "step": 786
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1314336886585818e-05,
+      "loss": 0.5764,
+      "step": 788
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.127803049719605e-05,
+      "loss": 0.6,
+      "step": 790
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1241706973524613e-05,
+      "loss": 0.6129,
+      "step": 792
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.5731,
+      "step": 794
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1169010471486807e-05,
+      "loss": 0.6465,
+      "step": 796
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1132638467746892e-05,
+      "loss": 0.6089,
+      "step": 798
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1096251278965173e-05,
+      "loss": 0.5923,
+      "step": 800
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1059849392976912e-05,
+      "loss": 0.5652,
+      "step": 802
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1023433297814403e-05,
+      "loss": 0.5728,
+      "step": 804
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0987003481700456e-05,
+      "loss": 0.5605,
+      "step": 806
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0950560433041825e-05,
+      "loss": 0.5948,
+      "step": 808
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.091410464042268e-05,
+      "loss": 1.1218,
+      "step": 810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.087763659259803e-05,
+      "loss": 0.5941,
+      "step": 812
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0841156778487199e-05,
+      "loss": 0.6121,
+      "step": 814
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0804665687167262e-05,
+      "loss": 0.6055,
+      "step": 816
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.076816380786647e-05,
+      "loss": 0.6215,
+      "step": 818
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0731651629957722e-05,
+      "loss": 0.6392,
+      "step": 820
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.069512964295198e-05,
+      "loss": 0.6127,
+      "step": 822
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0658598336491715e-05,
+      "loss": 0.6002,
+      "step": 824
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0622058200344345e-05,
+      "loss": 0.6313,
+      "step": 826
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0585509724395662e-05,
+      "loss": 0.6331,
+      "step": 828
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0548953398643276e-05,
+      "loss": 0.624,
+      "step": 830
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0512389713190025e-05,
+      "loss": 0.5737,
+      "step": 832
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0475819158237426e-05,
+      "loss": 0.5935,
+      "step": 834
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.043924222407909e-05,
+      "loss": 0.5836,
+      "step": 836
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0402659401094154e-05,
+      "loss": 0.6315,
+      "step": 838
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0366071179740706e-05,
+      "loss": 0.5624,
+      "step": 840
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0329478050549208e-05,
+      "loss": 0.5681,
+      "step": 842
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0292880504115924e-05,
+      "loss": 0.5962,
+      "step": 844
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0256279031096327e-05,
+      "loss": 0.6051,
+      "step": 846
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0219674122198548e-05,
+      "loss": 0.6156,
+      "step": 848
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0183066268176775e-05,
+      "loss": 0.575,
+      "step": 850
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.014645595982468e-05,
+      "loss": 0.606,
+      "step": 852
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0109843687968837e-05,
+      "loss": 0.5729,
+      "step": 854
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0073229943462152e-05,
+      "loss": 0.5778,
+      "step": 856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0036615217177273e-05,
+      "loss": 0.61,
+      "step": 858
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.4856,
+      "step": 860
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.963384782822732e-06,
+      "loss": 0.4084,
+      "step": 862
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.92677005653785e-06,
+      "loss": 0.4366,
+      "step": 864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.890156312031165e-06,
+      "loss": 0.3928,
+      "step": 866
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.853544040175327e-06,
+      "loss": 0.4016,
+      "step": 868
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.81693373182323e-06,
+      "loss": 0.3881,
+      "step": 870
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.780325877801455e-06,
+      "loss": 0.3711,
+      "step": 872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.743720968903676e-06,
+      "loss": 0.3812,
+      "step": 874
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.707119495884081e-06,
+      "loss": 0.3527,
+      "step": 876
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.670521949450793e-06,
+      "loss": 0.3604,
+      "step": 878
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.633928820259296e-06,
+      "loss": 0.3759,
+      "step": 880
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.597340598905851e-06,
+      "loss": 0.3853,
+      "step": 882
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.560757775920917e-06,
+      "loss": 0.3712,
+      "step": 884
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.524180841762577e-06,
+      "loss": 0.3702,
+      "step": 886
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.487610286809978e-06,
+      "loss": 0.5617,
+      "step": 888
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.451046601356725e-06,
+      "loss": 0.371,
+      "step": 890
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.414490275604338e-06,
+      "loss": 0.3474,
+      "step": 892
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.377941799655656e-06,
+      "loss": 0.3607,
+      "step": 894
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.34140166350829e-06,
+      "loss": 0.3926,
+      "step": 896
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.304870357048025e-06,
+      "loss": 0.368,
+      "step": 898
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.268348370042281e-06,
+      "loss": 0.3925,
+      "step": 900
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.231836192133532e-06,
+      "loss": 0.359,
+      "step": 902
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.195334312832742e-06,
+      "loss": 0.3605,
+      "step": 904
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.1588432215128e-06,
+      "loss": 0.3659,
+      "step": 906
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.122363407401971e-06,
+      "loss": 0.3752,
+      "step": 908
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.085895359577324e-06,
+      "loss": 0.4451,
+      "step": 910
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.049439566958176e-06,
+      "loss": 0.3895,
+      "step": 912
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.012996518299547e-06,
+      "loss": 0.376,
+      "step": 914
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 8.976566702185599e-06,
+      "loss": 0.3794,
+      "step": 916
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 8.940150607023091e-06,
+      "loss": 0.3634,
+      "step": 918
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 8.903748721034827e-06,
+      "loss": 0.3684,
+      "step": 920
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 8.867361532253108e-06,
+      "loss": 0.3752,
+      "step": 922
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 8.8309895285132e-06,
+      "loss": 0.3888,
+      "step": 924
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 0.38,
+      "step": 926
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 8.75829302647539e-06,
+      "loss": 0.3776,
+      "step": 928
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 8.721969502803954e-06,
+      "loss": 0.3522,
+      "step": 930
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 8.685663113414186e-06,
+      "loss": 0.3983,
+      "step": 932
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 8.649374345058083e-06,
+      "loss": 0.3816,
+      "step": 934
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 8.613103684251404e-06,
+      "loss": 0.3323,
+      "step": 936
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.3547,
+      "step": 938
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 8.540618630129028e-06,
+      "loss": 0.3724,
+      "step": 940
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 8.504405208604951e-06,
+      "loss": 0.391,
+      "step": 942
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 8.468211838200523e-06,
+      "loss": 0.3806,
+      "step": 944
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 8.43203900415252e-06,
+      "loss": 0.366,
+      "step": 946
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 8.395887191422397e-06,
+      "loss": 0.3686,
+      "step": 948
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 8.359756884689785e-06,
+      "loss": 0.3481,
+      "step": 950
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 8.32364856834597e-06,
+      "loss": 0.3672,
+      "step": 952
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 8.28756272648744e-06,
+      "loss": 0.3602,
+      "step": 954
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 8.251499842909358e-06,
+      "loss": 0.3627,
+      "step": 956
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 8.215460401099098e-06,
+      "loss": 0.3608,
+      "step": 958
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 8.179444884229746e-06,
+      "loss": 0.3683,
+      "step": 960
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 8.143453775153646e-06,
+      "loss": 0.3809,
+      "step": 962
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 8.107487556395902e-06,
+      "loss": 0.3653,
+      "step": 964
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 8.071546710147911e-06,
+      "loss": 0.52,
+      "step": 966
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.035631718260923e-06,
+      "loss": 0.3899,
+      "step": 968
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 7.999743062239557e-06,
+      "loss": 0.3678,
+      "step": 970
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 7.96388122323535e-06,
+      "loss": 0.3861,
+      "step": 972
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 7.928046682040311e-06,
+      "loss": 0.4042,
+      "step": 974
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 7.89223991908047e-06,
+      "loss": 0.401,
+      "step": 976
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 7.856461414409458e-06,
+      "loss": 0.3823,
+      "step": 978
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 7.820711647702017e-06,
+      "loss": 0.3906,
+      "step": 980
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 7.784991098247642e-06,
+      "loss": 0.3682,
+      "step": 982
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 7.749300244944099e-06,
+      "loss": 0.3862,
+      "step": 984
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 7.713639566291028e-06,
+      "loss": 0.3754,
+      "step": 986
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 7.678009540383525e-06,
+      "loss": 0.4198,
+      "step": 988
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 7.642410644905726e-06,
+      "loss": 0.3751,
+      "step": 990
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 7.606843357124426e-06,
+      "loss": 0.6063,
+      "step": 992
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 7.571308153882634e-06,
+      "loss": 0.5238,
+      "step": 994
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 7.535805511593227e-06,
+      "loss": 0.3877,
+      "step": 996
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 7.500335906232544e-06,
+      "loss": 0.4056,
+      "step": 998
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 7.4648998133340016e-06,
+      "loss": 0.3794,
+      "step": 1000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 7.42949770798172e-06,
+      "loss": 0.4066,
+      "step": 1002
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 7.394130064804157e-06,
+      "loss": 0.3899,
+      "step": 1004
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 7.358797357967749e-06,
+      "loss": 0.3882,
+      "step": 1006
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 7.323500061170535e-06,
+      "loss": 0.3674,
+      "step": 1008
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 7.2882386476358304e-06,
+      "loss": 0.3856,
+      "step": 1010
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 7.2530135901058726e-06,
+      "loss": 0.4198,
+      "step": 1012
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 7.217825360835475e-06,
+      "loss": 0.3961,
+      "step": 1014
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.3899,
+      "step": 1016
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 7.147561273617555e-06,
+      "loss": 0.3766,
+      "step": 1018
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 7.112486357685632e-06,
+      "loss": 0.3393,
+      "step": 1020
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 7.0774501540318305e-06,
+      "loss": 0.3729,
+      "step": 1022
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 7.042453132379046e-06,
+      "loss": 0.3547,
+      "step": 1024
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 7.007495761924863e-06,
+      "loss": 0.3856,
+      "step": 1026
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.97257851133528e-06,
+      "loss": 0.3521,
+      "step": 1028
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.937701848738407e-06,
+      "loss": 0.3774,
+      "step": 1030
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.9028662417182054e-06,
+      "loss": 0.3746,
+      "step": 1032
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.868072157308213e-06,
+      "loss": 0.359,
+      "step": 1034
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 6.833320061985278e-06,
+      "loss": 0.3578,
+      "step": 1036
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 6.798610421663313e-06,
+      "loss": 0.3409,
+      "step": 1038
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 6.763943701687046e-06,
+      "loss": 0.3843,
+      "step": 1040
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 6.729320366825785e-06,
+      "loss": 0.3712,
+      "step": 1042
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 6.6947408812671745e-06,
+      "loss": 0.3851,
+      "step": 1044
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 6.660205708610987e-06,
+      "loss": 0.3577,
+      "step": 1046
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 6.625715311862901e-06,
+      "loss": 0.3759,
+      "step": 1048
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 6.591270153428288e-06,
+      "loss": 0.3604,
+      "step": 1050
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 6.556870695106028e-06,
+      "loss": 0.3416,
+      "step": 1052
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.522517398082303e-06,
+      "loss": 0.3917,
+      "step": 1054
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.488210722924419e-06,
+      "loss": 0.3888,
+      "step": 1056
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.453951129574644e-06,
+      "loss": 0.3428,
+      "step": 1058
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.419739077344016e-06,
+      "loss": 0.3687,
+      "step": 1060
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.3855750249062096e-06,
+      "loss": 0.3547,
+      "step": 1062
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.351459430291369e-06,
+      "loss": 0.3755,
+      "step": 1064
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.317392750879978e-06,
+      "loss": 0.365,
+      "step": 1066
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.283375443396726e-06,
+      "loss": 0.3441,
+      "step": 1068
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 6.249407963904381e-06,
+      "loss": 0.3534,
+      "step": 1070
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 6.215490767797682e-06,
+      "loss": 0.3519,
+      "step": 1072
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 6.181624309797228e-06,
+      "loss": 0.3824,
+      "step": 1074
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 6.1478090439433776e-06,
+      "loss": 0.3638,
+      "step": 1076
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 6.1140454235901714e-06,
+      "loss": 0.3932,
+      "step": 1078
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 6.080333901399252e-06,
+      "loss": 0.3781,
+      "step": 1080
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 6.046674929333787e-06,
+      "loss": 0.3297,
+      "step": 1082
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 6.013068958652422e-06,
+      "loss": 0.3705,
+      "step": 1084
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.979516439903221e-06,
+      "loss": 0.4174,
+      "step": 1086
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.946017822917632e-06,
+      "loss": 0.5333,
+      "step": 1088
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.912573556804453e-06,
+      "loss": 0.3457,
+      "step": 1090
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.879184089943805e-06,
+      "loss": 0.3784,
+      "step": 1092
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.3651,
+      "step": 1094
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.812571343821207e-06,
+      "loss": 0.3596,
+      "step": 1096
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.7793489576220994e-06,
+      "loss": 0.402,
+      "step": 1098
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.746183156789252e-06,
+      "loss": 0.3871,
+      "step": 1100
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.713074385969457e-06,
+      "loss": 0.3372,
+      "step": 1102
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.680023089044935e-06,
+      "loss": 0.3779,
+      "step": 1104
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.647029709127355e-06,
+      "loss": 0.3624,
+      "step": 1106
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.614094688551903e-06,
+      "loss": 0.3813,
+      "step": 1108
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.581218468871365e-06,
+      "loss": 0.3568,
+      "step": 1110
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.548401490850193e-06,
+      "loss": 0.369,
+      "step": 1112
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.515644194458588e-06,
+      "loss": 0.4028,
+      "step": 1114
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.482947018866628e-06,
+      "loss": 0.3574,
+      "step": 1116
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.450310402438353e-06,
+      "loss": 0.3611,
+      "step": 1118
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.417734782725896e-06,
+      "loss": 0.3724,
+      "step": 1120
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.385220596463638e-06,
+      "loss": 0.4346,
+      "step": 1122
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.352768279562315e-06,
+      "loss": 0.353,
+      "step": 1124
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.320378267103206e-06,
+      "loss": 0.3557,
+      "step": 1126
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.288050993332286e-06,
+      "loss": 0.3644,
+      "step": 1128
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.2557868916543996e-06,
+      "loss": 0.3555,
+      "step": 1130
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.223586394627459e-06,
+      "loss": 0.376,
+      "step": 1132
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.1914499339566514e-06,
+      "loss": 0.388,
+      "step": 1134
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.1593779404886255e-06,
+      "loss": 0.3589,
+      "step": 1136
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.127370844205753e-06,
+      "loss": 0.3448,
+      "step": 1138
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.09542907422032e-06,
+      "loss": 0.4056,
+      "step": 1140
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.063553058768814e-06,
+      "loss": 0.3397,
+      "step": 1142
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.031743225206167e-06,
+      "loss": 0.3713,
+      "step": 1144
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.3369,
+      "step": 1146
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.968323808724962e-06,
+      "loss": 0.3643,
+      "step": 1148
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.9367150760569746e-06,
+      "loss": 0.3633,
+      "step": 1150
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.905174225767555e-06,
+      "loss": 0.3848,
+      "step": 1152
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.873701680718146e-06,
+      "loss": 0.3956,
+      "step": 1154
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.842297862854426e-06,
+      "loss": 0.3606,
+      "step": 1156
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.810963193200673e-06,
+      "loss": 0.3436,
+      "step": 1158
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.779698091854098e-06,
+      "loss": 0.3824,
+      "step": 1160
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.748502977979239e-06,
+      "loss": 0.3252,
+      "step": 1162
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.717378269802311e-06,
+      "loss": 0.5341,
+      "step": 1164
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.686324384605629e-06,
+      "loss": 0.4489,
+      "step": 1166
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.655341738721989e-06,
+      "loss": 0.3633,
+      "step": 1168
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.6244307475291025e-06,
+      "loss": 0.3739,
+      "step": 1170
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.4054,
+      "step": 1172
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.562825385917586e-06,
+      "loss": 0.4028,
+      "step": 1174
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.532131841428864e-06,
+      "loss": 0.4413,
+      "step": 1176
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.501511603479653e-06,
+      "loss": 0.3989,
+      "step": 1178
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.470965082588929e-06,
+      "loss": 0.3848,
+      "step": 1180
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.440492688287377e-06,
+      "loss": 0.3552,
+      "step": 1182
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.410094829111865e-06,
+      "loss": 0.3857,
+      "step": 1184
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.379771912600001e-06,
+      "loss": 0.4124,
+      "step": 1186
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.349524345284638e-06,
+      "loss": 0.3977,
+      "step": 1188
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.319352532688444e-06,
+      "loss": 0.3724,
+      "step": 1190
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.289256879318462e-06,
+      "loss": 0.3196,
+      "step": 1192
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.259237788660689e-06,
+      "loss": 0.3562,
+      "step": 1194
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.2292956631746475e-06,
+      "loss": 0.4104,
+      "step": 1196
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.19943090428802e-06,
+      "loss": 0.3323,
+      "step": 1198
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.169643912391241e-06,
+      "loss": 0.4098,
+      "step": 1200
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.13993508683214e-06,
+      "loss": 0.3248,
+      "step": 1202
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.110304825910597e-06,
+      "loss": 0.3563,
+      "step": 1204
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.08075352687318e-06,
+      "loss": 0.3807,
+      "step": 1206
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.051281585907841e-06,
+      "loss": 0.3665,
+      "step": 1208
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.0218893981385935e-06,
+      "loss": 0.3785,
+      "step": 1210
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.99257735762021e-06,
+      "loss": 0.3455,
+      "step": 1212
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.96334585733296e-06,
+      "loss": 0.4025,
+      "step": 1214
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.934195289177314e-06,
+      "loss": 0.3899,
+      "step": 1216
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.905126043968705e-06,
+      "loss": 0.377,
+      "step": 1218
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.8761385114322905e-06,
+      "loss": 0.37,
+      "step": 1220
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.847233080197728e-06,
+      "loss": 0.3577,
+      "step": 1222
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.818410137793947e-06,
+      "loss": 0.3442,
+      "step": 1224
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.7896700706439826e-06,
+      "loss": 0.359,
+      "step": 1226
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.761013264059766e-06,
+      "loss": 0.3689,
+      "step": 1228
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.732440102236975e-06,
+      "loss": 0.3818,
+      "step": 1230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.7039509682498887e-06,
+      "loss": 0.3458,
+      "step": 1232
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.6755462440462288e-06,
+      "loss": 0.3678,
+      "step": 1234
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.6472263104420635e-06,
+      "loss": 0.3617,
+      "step": 1236
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.6189915471166927e-06,
+      "loss": 0.3533,
+      "step": 1238
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.590842332607546e-06,
+      "loss": 0.3677,
+      "step": 1240
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.5627790443051303e-06,
+      "loss": 0.3401,
+      "step": 1242
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.5348020584479492e-06,
+      "loss": 0.355,
+      "step": 1244
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.506911750117469e-06,
+      "loss": 0.3799,
+      "step": 1246
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.4791084932330976e-06,
+      "loss": 0.3821,
+      "step": 1248
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.3491,
+      "step": 1250
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.4237646236398727e-06,
+      "loss": 0.3768,
+      "step": 1252
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.3962247529144544e-06,
+      "loss": 0.3667,
+      "step": 1254
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.3687734175920505e-06,
+      "loss": 0.4319,
+      "step": 1256
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.341410985706842e-06,
+      "loss": 0.3751,
+      "step": 1258
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.3141378241011113e-06,
+      "loss": 0.3627,
+      "step": 1260
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.2869542984202974e-06,
+      "loss": 0.3948,
+      "step": 1262
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.259860773108123e-06,
+      "loss": 0.3398,
+      "step": 1264
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.232857611401693e-06,
+      "loss": 0.3462,
+      "step": 1266
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.205945175326617e-06,
+      "loss": 0.4498,
+      "step": 1268
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.1791238256921785e-06,
+      "loss": 0.4011,
+      "step": 1270
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.1523939220864743e-06,
+      "loss": 0.3678,
+      "step": 1272
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.125755822871607e-06,
+      "loss": 0.343,
+      "step": 1274
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.099209885178882e-06,
+      "loss": 0.3712,
+      "step": 1276
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.0727564649040066e-06,
+      "loss": 0.342,
+      "step": 1278
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.0463959167023336e-06,
+      "loss": 0.3579,
+      "step": 1280
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.0201285939841006e-06,
+      "loss": 0.3507,
+      "step": 1282
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.993954848909686e-06,
+      "loss": 0.3497,
+      "step": 1284
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.9678750323848893e-06,
+      "loss": 0.3254,
+      "step": 1286
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.9418894940562438e-06,
+      "loss": 0.3467,
+      "step": 1288
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.9159985823062997e-06,
+      "loss": 0.3241,
+      "step": 1290
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.890202644248983e-06,
+      "loss": 0.3513,
+      "step": 1292
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.8645020257249136e-06,
+      "loss": 0.3755,
+      "step": 1294
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.838897071296796e-06,
+      "loss": 0.4177,
+      "step": 1296
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.813388124244778e-06,
+      "loss": 0.3722,
+      "step": 1298
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.7879755265618558e-06,
+      "loss": 0.3538,
+      "step": 1300
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.7626596189492983e-06,
+      "loss": 0.3346,
+      "step": 1302
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.7374407408120685e-06,
+      "loss": 0.3699,
+      "step": 1304
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.71231923025427e-06,
+      "loss": 0.3445,
+      "step": 1306
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.6872954240746287e-06,
+      "loss": 0.3614,
+      "step": 1308
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.662369657761963e-06,
+      "loss": 0.3663,
+      "step": 1310
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.637542265490689e-06,
+      "loss": 0.3667,
+      "step": 1312
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.6128135801163547e-06,
+      "loss": 0.3448,
+      "step": 1314
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5881839331711524e-06,
+      "loss": 0.348,
+      "step": 1316
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.563653654859495e-06,
+      "loss": 0.3696,
+      "step": 1318
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5392230740535852e-06,
+      "loss": 0.3519,
+      "step": 1320
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.514892518288988e-06,
+      "loss": 0.3373,
+      "step": 1322
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4906623137602736e-06,
+      "loss": 0.4167,
+      "step": 1324
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4665327853166076e-06,
+      "loss": 0.3593,
+      "step": 1326
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.3651,
+      "step": 1328
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.418577049328058e-06,
+      "loss": 0.3472,
+      "step": 1330
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.394751484715481e-06,
+      "loss": 0.3667,
+      "step": 1332
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.3710278820439313e-06,
+      "loss": 0.3281,
+      "step": 1334
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.347406559370686e-06,
+      "loss": 0.363,
+      "step": 1336
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.323887833381766e-06,
+      "loss": 0.3533,
+      "step": 1338
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.3004720193876972e-06,
+      "loss": 0.3636,
+      "step": 1340
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.2771594313193012e-06,
+      "loss": 0.3656,
+      "step": 1342
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.2539503817234553e-06,
+      "loss": 0.355,
+      "step": 1344
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.230845181758928e-06,
+      "loss": 0.3658,
+      "step": 1346
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.2078441411922004e-06,
+      "loss": 0.3291,
+      "step": 1348
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.1849475683932996e-06,
+      "loss": 0.3726,
+      "step": 1350
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.1621557703316876e-06,
+      "loss": 0.3622,
+      "step": 1352
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.1394690525721275e-06,
+      "loss": 0.353,
+      "step": 1354
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.1168877192705906e-06,
+      "loss": 0.3915,
+      "step": 1356
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.09441207317019e-06,
+      "loss": 0.3882,
+      "step": 1358
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.072042415597104e-06,
+      "loss": 0.378,
+      "step": 1360
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.0497790464565527e-06,
+      "loss": 0.3557,
+      "step": 1362
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.027622264228768e-06,
+      "loss": 0.3198,
+      "step": 1364
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.0055723659649907e-06,
+      "loss": 0.334,
+      "step": 1366
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.9836296472834905e-06,
+      "loss": 0.3842,
+      "step": 1368
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.961794402365611e-06,
+      "loss": 0.3658,
+      "step": 1370
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.940066923951808e-06,
+      "loss": 0.3443,
+      "step": 1372
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.9184475033377414e-06,
+      "loss": 0.3729,
+      "step": 1374
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.8969364303703664e-06,
+      "loss": 0.3389,
+      "step": 1376
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.8755339934440331e-06,
+      "loss": 0.3563,
+      "step": 1378
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.854240479496643e-06,
+      "loss": 0.3437,
+      "step": 1380
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.8330561740057839e-06,
+      "loss": 0.3618,
+      "step": 1382
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.811981360984908e-06,
+      "loss": 0.3593,
+      "step": 1384
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.791016322979533e-06,
+      "loss": 0.3584,
+      "step": 1386
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.7701613410634367e-06,
+      "loss": 0.3549,
+      "step": 1388
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.7494166948349057e-06,
+      "loss": 0.3573,
+      "step": 1390
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.7287826624129799e-06,
+      "loss": 0.3569,
+      "step": 1392
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.7082595204337183e-06,
+      "loss": 0.3437,
+      "step": 1394
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.687847544046497e-06,
+      "loss": 0.3577,
+      "step": 1396
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.6675470069103228e-06,
+      "loss": 0.3783,
+      "step": 1398
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.6473581811901529e-06,
+      "loss": 0.3477,
+      "step": 1400
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.6272813375532637e-06,
+      "loss": 0.3549,
+      "step": 1402
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.6073167451656001e-06,
+      "loss": 0.3854,
+      "step": 1404
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.3286,
+      "step": 1406
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.567725383273535e-06,
+      "loss": 0.4374,
+      "step": 1408
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.5480991445620541e-06,
+      "loss": 0.3781,
+      "step": 1410
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.528586218678535e-06,
+      "loss": 0.3622,
+      "step": 1412
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.509186867228607e-06,
+      "loss": 0.3342,
+      "step": 1414
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.4899013502952242e-06,
+      "loss": 0.3362,
+      "step": 1416
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.4707299264351914e-06,
+      "loss": 0.3718,
+      "step": 1418
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.4516728526756873e-06,
+      "loss": 0.3544,
+      "step": 1420
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.4327303845108299e-06,
+      "loss": 0.3402,
+      "step": 1422
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.413902775898236e-06,
+      "loss": 0.3322,
+      "step": 1424
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.3951902792556337e-06,
+      "loss": 0.3408,
+      "step": 1426
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.376593145457461e-06,
+      "loss": 0.3418,
+      "step": 1428
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.3581116238315194e-06,
+      "loss": 0.3701,
+      "step": 1430
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.3401,
+      "step": 1432
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.3214964066542458e-06,
+      "loss": 0.3661,
+      "step": 1434
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.3033632019953113e-06,
+      "loss": 0.3905,
+      "step": 1436
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2853465912867979e-06,
+      "loss": 0.3414,
+      "step": 1438
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2674468160735586e-06,
+      "loss": 0.36,
+      "step": 1440
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2496641163340562e-06,
+      "loss": 0.3502,
+      "step": 1442
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2319987304771397e-06,
+      "loss": 0.3519,
+      "step": 1444
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2144508953388655e-06,
+      "loss": 0.3534,
+      "step": 1446
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.197020846179303e-06,
+      "loss": 0.3525,
+      "step": 1448
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.1797088166794002e-06,
+      "loss": 0.3346,
+      "step": 1450
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.1625150389378326e-06,
+      "loss": 0.3602,
+      "step": 1452
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.1454397434679022e-06,
+      "loss": 0.3068,
+      "step": 1454
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.1284831591944456e-06,
+      "loss": 0.3585,
+      "step": 1456
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1116455134507665e-06,
+      "loss": 0.3495,
+      "step": 1458
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.0949270319755768e-06,
+      "loss": 0.3629,
+      "step": 1460
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.0783279389099887e-06,
+      "loss": 0.4005,
+      "step": 1462
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.061848456794492e-06,
+      "loss": 0.3597,
+      "step": 1464
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.0454888065659775e-06,
+      "loss": 0.3357,
+      "step": 1466
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.0292492075547845e-06,
+      "loss": 0.3607,
+      "step": 1468
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.013129877481741e-06,
+      "loss": 0.3535,
+      "step": 1470
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 9.971310324552597e-07,
+      "loss": 0.3049,
+      "step": 1472
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 9.81252886968439e-07,
+      "loss": 0.3677,
+      "step": 1474
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 9.65495653896179e-07,
+      "loss": 0.3474,
+      "step": 1476
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 9.49859544492332e-07,
+      "loss": 0.3484,
+      "step": 1478
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 9.3434476838688e-07,
+      "loss": 0.3561,
+      "step": 1480
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 9.189515335831045e-07,
+      "loss": 0.3712,
+      "step": 1482
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.3495,
+      "step": 1484
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.885305117435794e-07,
+      "loss": 0.3364,
+      "step": 1486
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.73503132555964e-07,
+      "loss": 0.319,
+      "step": 1488
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.585981103608343e-07,
+      "loss": 0.3513,
+      "step": 1490
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.438156449866353e-07,
+      "loss": 0.3513,
+      "step": 1492
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.291559346187161e-07,
+      "loss": 0.3411,
+      "step": 1494
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.146191757966859e-07,
+      "loss": 0.3439,
+      "step": 1496
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.002055634117578e-07,
+      "loss": 0.3308,
+      "step": 1498
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.859152907041545e-07,
+      "loss": 0.3425,
+      "step": 1500
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.71748549260507e-07,
+      "loss": 0.3688,
+      "step": 1502
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.5770552901129e-07,
+      "loss": 0.3354,
+      "step": 1504
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.437864182282706e-07,
+      "loss": 0.3186,
+      "step": 1506
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.299914035219957e-07,
+      "loss": 0.3707,
+      "step": 1508
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.163206698392744e-07,
+      "loss": 0.3398,
+      "step": 1510
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.02774400460714e-07,
+      "loss": 0.389,
+      "step": 1512
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 6.8935277699825e-07,
+      "loss": 0.3394,
+      "step": 1514
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 6.760559793927202e-07,
+      "loss": 0.3566,
+      "step": 1516
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 6.628841859114499e-07,
+      "loss": 0.35,
+      "step": 1518
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 6.498375731458529e-07,
+      "loss": 0.3633,
+      "step": 1520
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 6.369163160090808e-07,
+      "loss": 0.3339,
+      "step": 1522
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 6.241205877336665e-07,
+      "loss": 0.3441,
+      "step": 1524
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 6.114505598692011e-07,
+      "loss": 0.3307,
+      "step": 1526
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.989064022800428e-07,
+      "loss": 0.3586,
+      "step": 1528
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.864882831430274e-07,
+      "loss": 0.3602,
+      "step": 1530
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.741963689452268e-07,
+      "loss": 0.3755,
+      "step": 1532
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.620308244817041e-07,
+      "loss": 0.3668,
+      "step": 1534
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.499918128533155e-07,
+      "loss": 0.3542,
+      "step": 1536
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.380794954645141e-07,
+      "loss": 0.3481,
+      "step": 1538
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.262940320211951e-07,
+      "loss": 0.3582,
+      "step": 1540
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.146355805285452e-07,
+      "loss": 0.3443,
+      "step": 1542
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.031042972889311e-07,
+      "loss": 0.3414,
+      "step": 1544
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.917003368998063e-07,
+      "loss": 0.3232,
+      "step": 1546
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.804238522516235e-07,
+      "loss": 0.3316,
+      "step": 1548
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.6927499452580574e-07,
+      "loss": 0.3458,
+      "step": 1550
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 4.582539131927044e-07,
+      "loss": 0.3307,
+      "step": 1552
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 4.473607560096005e-07,
+      "loss": 0.3302,
+      "step": 1554
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 4.365956690187256e-07,
+      "loss": 0.3612,
+      "step": 1556
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 4.2595879654529605e-07,
+      "loss": 0.3726,
+      "step": 1558
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 4.154502811955907e-07,
+      "loss": 0.3563,
+      "step": 1560
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.3745,
+      "step": 1562
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.9481888368627764e-07,
+      "loss": 0.317,
+      "step": 1564
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.846962781274055e-07,
+      "loss": 0.3605,
+      "step": 1566
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.747025828900208e-07,
+      "loss": 0.3605,
+      "step": 1568
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.6483793195745686e-07,
+      "loss": 0.3834,
+      "step": 1570
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.55102457582982e-07,
+      "loss": 0.3582,
+      "step": 1572
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.454962902880199e-07,
+      "loss": 0.3677,
+      "step": 1574
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.360195588603987e-07,
+      "loss": 0.3364,
+      "step": 1576
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.266723903526359e-07,
+      "loss": 0.3564,
+      "step": 1578
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.1745491008021603e-07,
+      "loss": 0.3298,
+      "step": 1580
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.083672416199301e-07,
+      "loss": 0.373,
+      "step": 1582
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.9940950680820704e-07,
+      "loss": 0.3654,
+      "step": 1584
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.905818257394799e-07,
+      "loss": 0.3601,
+      "step": 1586
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.818843167645835e-07,
+      "loss": 0.3425,
+      "step": 1588
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.733170964891607e-07,
+      "loss": 0.3262,
+      "step": 1590
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.6488027977210175e-07,
+      "loss": 0.3197,
+      "step": 1592
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.5657397972400523e-07,
+      "loss": 0.3304,
+      "step": 1594
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.4839830770566175e-07,
+      "loss": 0.3629,
+      "step": 1596
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.4035337332655504e-07,
+      "loss": 0.3683,
+      "step": 1598
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.3243928444340426e-07,
+      "loss": 0.3733,
+      "step": 1600
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.24656147158705e-07,
+      "loss": 0.379,
+      "step": 1602
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.1700406581931398e-07,
+      "loss": 0.3646,
+      "step": 1604
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.0948314301505346e-07,
+      "loss": 0.375,
+      "step": 1606
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.0209347957732328e-07,
+      "loss": 0.3199,
+      "step": 1608
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.9483517457776436e-07,
+      "loss": 0.3439,
+      "step": 1610
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.8770832532691964e-07,
+      "loss": 0.3662,
+      "step": 1612
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.8071302737293294e-07,
+      "loss": 0.3395,
+      "step": 1614
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.7384937450026895e-07,
+      "loss": 0.3418,
+      "step": 1616
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.6711745872845518e-07,
+      "loss": 0.3563,
+      "step": 1618
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.6051737031084537e-07,
+      "loss": 0.3369,
+      "step": 1620
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.5404919773341576e-07,
+      "loss": 0.3486,
+      "step": 1622
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.47713027713573e-07,
+      "loss": 0.3358,
+      "step": 1624
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.4150894519899262e-07,
+      "loss": 0.3711,
+      "step": 1626
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.354370333664845e-07,
+      "loss": 0.3468,
+      "step": 1628
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2949737362087156e-07,
+      "loss": 0.3352,
+      "step": 1630
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2369004559390273e-07,
+      "loss": 0.3628,
+      "step": 1632
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.1801512714318286e-07,
+      "loss": 0.3474,
+      "step": 1634
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.1247269435113006e-07,
+      "loss": 0.3416,
+      "step": 1636
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.0706282152395776e-07,
+      "loss": 0.3542,
+      "step": 1638
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.3331,
+      "step": 1640
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 9.664104410210928e-08,
+      "loss": 0.3805,
+      "step": 1642
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 9.162927922997445e-08,
+      "loss": 0.3625,
+      "step": 1644
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.675035376593088e-08,
+      "loss": 0.3535,
+      "step": 1646
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.200433312068657e-08,
+      "loss": 0.3511,
+      "step": 1648
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.73912809231292e-08,
+      "loss": 0.7921,
+      "step": 1650
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.291125901946027e-08,
+      "loss": 0.3314,
+      "step": 1652
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.856432747237684e-08,
+      "loss": 0.3265,
+      "step": 1654
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.435054456026213e-08,
+      "loss": 0.344,
+      "step": 1656
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.026996677640062e-08,
+      "loss": 0.3174,
+      "step": 1658
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.632264882822758e-08,
+      "loss": 0.3363,
+      "step": 1660
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.2508643636590694e-08,
+      "loss": 0.3284,
+      "step": 1662
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 4.882800233504292e-08,
+      "loss": 0.3647,
+      "step": 1664
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 4.528077426915412e-08,
+      "loss": 0.3502,
+      "step": 1666
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 4.186700699585377e-08,
+      "loss": 0.3358,
+      "step": 1668
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.858674628278825e-08,
+      "loss": 0.3434,
+      "step": 1670
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.5440036107714516e-08,
+      "loss": 0.3378,
+      "step": 1672
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.242691865790071e-08,
+      "loss": 0.3734,
+      "step": 1674
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.9547434329568747e-08,
+      "loss": 0.3683,
+      "step": 1676
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.6801621727350347e-08,
+      "loss": 0.3898,
+      "step": 1678
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.4189517663767425e-08,
+      "loss": 0.3619,
+      "step": 1680
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.171115715874139e-08,
+      "loss": 0.3333,
+      "step": 1682
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.9366573439121296e-08,
+      "loss": 0.3604,
+      "step": 1684
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7155797938239737e-08,
+      "loss": 0.3399,
+      "step": 1686
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.5078860295490995e-08,
+      "loss": 0.3696,
+      "step": 1688
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.3135788355934652e-08,
+      "loss": 0.3613,
+      "step": 1690
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.1326608169920373e-08,
+      "loss": 0.3888,
+      "step": 1692
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 9.651343992740369e-09,
+      "loss": 0.3966,
+      "step": 1694
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 8.110018284304132e-09,
+      "loss": 0.3521,
+      "step": 1696
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.702651708834218e-09,
+      "loss": 0.3424,
+      "step": 1698
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.429263134594242e-09,
+      "loss": 0.3356,
+      "step": 1700
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 4.2898696336335364e-09,
+      "loss": 0.3508,
+      "step": 1702
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.284486481556215e-09,
+      "loss": 0.3591,
+      "step": 1704
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.4131271573191172e-09,
+      "loss": 0.3525,
+      "step": 1706
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.6758033430508414e-09,
+      "loss": 0.3348,
+      "step": 1708
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.0725249238940916e-09,
+      "loss": 0.3365,
+      "step": 1710
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.032999878735624e-10,
+      "loss": 0.3418,
+      "step": 1712
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.681348257882466e-10,
+      "loss": 0.3535,
+      "step": 1714
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.703393112483802e-11,
+      "loss": 0.3329,
+      "step": 1716
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.0,
+      "loss": 0.3544,
+      "step": 1718
+    }
+  ],
+  "max_steps": 1718,
+  "num_train_epochs": 2,
+  "total_flos": 192008473460736.0,
+  "trial_name": null,
+  "trial_params": null
+}