diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5533 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.995418914791815,
+  "eval_steps": 500,
+  "global_step": 1839,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 1.611,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 1.7568,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 1.658,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.8564,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 2.0213,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 1.9062,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-06,
+      "loss": 1.828,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.9038,
+      "step": 16
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.4285714285714295e-06,
+      "loss": 1.9794,
+      "step": 18
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 1.9665,
+      "step": 20
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 1.9494,
+      "step": 22
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.962,
+      "step": 24
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.285714285714288e-06,
+      "loss": 2.195,
+      "step": 26
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 2.0743,
+      "step": 28
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 2.2212,
+      "step": 30
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 2.1635,
+      "step": 32
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2142857142857142e-05,
+      "loss": 2.3554,
+      "step": 34
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2857142857142859e-05,
+      "loss": 2.2266,
+      "step": 36
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3571428571428574e-05,
+      "loss": 2.2897,
+      "step": 38
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 2.385,
+      "step": 40
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 2.404,
+      "step": 42
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 2.5732,
+      "step": 44
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.642857142857143e-05,
+      "loss": 2.7419,
+      "step": 46
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 2.7752,
+      "step": 48
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 2.9747,
+      "step": 50
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8571428571428575e-05,
+      "loss": 1.7419,
+      "step": 52
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.928571428571429e-05,
+      "loss": 1.5881,
+      "step": 54
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 1.7405,
+      "step": 56
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9999937909292965e-05,
+      "loss": 1.7546,
+      "step": 58
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999751637942895e-05,
+      "loss": 1.8288,
+      "step": 60
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999441188262945e-05,
+      "loss": 1.8816,
+      "step": 62
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999900656410832e-05,
+      "loss": 1.7845,
+      "step": 64
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998447770876235e-05,
+      "loss": 1.8064,
+      "step": 66
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999776481550588e-05,
+      "loss": 1.7793,
+      "step": 68
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9996957706478283e-05,
+      "loss": 1.8418,
+      "step": 70
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999602645381624e-05,
+      "loss": 1.769,
+      "step": 72
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999497106908417e-05,
+      "loss": 1.9599,
+      "step": 74
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9993791565388e-05,
+      "loss": 1.822,
+      "step": 76
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999248795737497e-05,
+      "loss": 1.8957,
+      "step": 78
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9991060261233468e-05,
+      "loss": 1.9517,
+      "step": 80
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.998950849469283e-05,
+      "loss": 1.9511,
+      "step": 82
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9987832677023104e-05,
+      "loss": 2.0987,
+      "step": 84
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998603282903484e-05,
+      "loss": 2.0514,
+      "step": 86
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.99841089730788e-05,
+      "loss": 2.1349,
+      "step": 88
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.99820611330457e-05,
+      "loss": 2.1021,
+      "step": 90
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.997988933436591e-05,
+      "loss": 2.2552,
+      "step": 92
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9977593604009126e-05,
+      "loss": 2.1949,
+      "step": 94
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.997517397048406e-05,
+      "loss": 2.4357,
+      "step": 96
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9972630463838062e-05,
+      "loss": 2.1899,
+      "step": 98
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9969963115656754e-05,
+      "loss": 2.1754,
+      "step": 100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9967171959063644e-05,
+      "loss": 1.664,
+      "step": 102
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996425702871971e-05,
+      "loss": 1.5388,
+      "step": 104
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9961218360822968e-05,
+      "loss": 1.6347,
+      "step": 106
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9958055993108026e-05,
+      "loss": 1.6809,
+      "step": 108
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9954769964845614e-05,
+      "loss": 1.7729,
+      "step": 110
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9951360316842097e-05,
+      "loss": 1.7805,
+      "step": 112
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.994782709143896e-05,
+      "loss": 1.6005,
+      "step": 114
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9944170332512303e-05,
+      "loss": 1.7232,
+      "step": 116
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9940390085472274e-05,
+      "loss": 1.7918,
+      "step": 118
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9936486397262514e-05,
+      "loss": 1.7841,
+      "step": 120
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9932459316359573e-05,
+      "loss": 1.6909,
+      "step": 122
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9928308892772315e-05,
+      "loss": 1.7525,
+      "step": 124
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9924035178041284e-05,
+      "loss": 1.8016,
+      "step": 126
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9919638225238077e-05,
+      "loss": 1.8685,
+      "step": 128
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9915118088964674e-05,
+      "loss": 1.9582,
+      "step": 130
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9910474825352767e-05,
+      "loss": 1.818,
+      "step": 132
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9905708492063055e-05,
+      "loss": 2.0729,
+      "step": 134
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990081914828455e-05,
+      "loss": 2.0539,
+      "step": 136
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.98958068547338e-05,
+      "loss": 2.1301,
+      "step": 138
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9890671673654186e-05,
+      "loss": 2.1811,
+      "step": 140
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.988541366881511e-05,
+      "loss": 2.204,
+      "step": 142
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9880032905511218e-05,
+      "loss": 2.2893,
+      "step": 144
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9874529450561592e-05,
+      "loss": 2.1792,
+      "step": 146
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.986890337230891e-05,
+      "loss": 2.4459,
+      "step": 148
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9863154740618613e-05,
+      "loss": 2.3752,
+      "step": 150
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9857283626878016e-05,
+      "loss": 1.5311,
+      "step": 152
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9851290103995446e-05,
+      "loss": 1.657,
+      "step": 154
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9845174246399313e-05,
+      "loss": 1.6497,
+      "step": 156
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9838936130037202e-05,
+      "loss": 1.5877,
+      "step": 158
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9832575832374926e-05,
+      "loss": 1.6494,
+      "step": 160
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.982609343239556e-05,
+      "loss": 1.8271,
+      "step": 162
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.981948901059846e-05,
+      "loss": 1.8217,
+      "step": 164
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9812762648998278e-05,
+      "loss": 1.8643,
+      "step": 166
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9805914431123915e-05,
+      "loss": 1.7973,
+      "step": 168
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9798944442017514e-05,
+      "loss": 1.7682,
+      "step": 170
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9791852768233385e-05,
+      "loss": 1.7038,
+      "step": 172
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9784639497836934e-05,
+      "loss": 1.711,
+      "step": 174
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9777304720403574e-05,
+      "loss": 1.8575,
+      "step": 176
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9769848527017608e-05,
+      "loss": 1.8222,
+      "step": 178
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9762271010271097e-05,
+      "loss": 1.959,
+      "step": 180
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.975457226426272e-05,
+      "loss": 1.8217,
+      "step": 182
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.974675238459659e-05,
+      "loss": 2.0096,
+      "step": 184
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9738811468381086e-05,
+      "loss": 1.9606,
+      "step": 186
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9730749614227617e-05,
+      "loss": 2.0809,
+      "step": 188
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9722566922249436e-05,
+      "loss": 2.1243,
+      "step": 190
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.971426349406037e-05,
+      "loss": 1.9635,
+      "step": 192
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9705839432773555e-05,
+      "loss": 2.0439,
+      "step": 194
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9697294843000185e-05,
+      "loss": 2.2575,
+      "step": 196
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.968862983084818e-05,
+      "loss": 2.2378,
+      "step": 198
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9679844503920888e-05,
+      "loss": 1.859,
+      "step": 200
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.967093897131574e-05,
+      "loss": 1.7287,
+      "step": 202
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9661913343622894e-05,
+      "loss": 1.6376,
+      "step": 204
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.965276773292388e-05,
+      "loss": 1.7034,
+      "step": 206
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.964350225279018e-05,
+      "loss": 1.6158,
+      "step": 208
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.963411701828184e-05,
+      "loss": 1.6504,
+      "step": 210
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9624612145946024e-05,
+      "loss": 1.7069,
+      "step": 212
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9614987753815584e-05,
+      "loss": 1.7062,
+      "step": 214
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9605243961407583e-05,
+      "loss": 1.8162,
+      "step": 216
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9595380889721813e-05,
+      "loss": 1.6725,
+      "step": 218
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9585398661239287e-05,
+      "loss": 1.8527,
+      "step": 220
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9575297399920742e-05,
+      "loss": 1.6978,
+      "step": 222
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9565077231205057e-05,
+      "loss": 1.7312,
+      "step": 224
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9554738282007736e-05,
+      "loss": 1.8961,
+      "step": 226
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9544280680719316e-05,
+      "loss": 1.7643,
+      "step": 228
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.953370455720377e-05,
+      "loss": 1.8736,
+      "step": 230
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9523010042796885e-05,
+      "loss": 2.0634,
+      "step": 232
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9512197270304664e-05,
+      "loss": 1.886,
+      "step": 234
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9501266374001637e-05,
+      "loss": 1.9808,
+      "step": 236
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9490217489629226e-05,
+      "loss": 2.1238,
+      "step": 238
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.947905075439404e-05,
+      "loss": 2.2318,
+      "step": 240
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9467766306966167e-05,
+      "loss": 2.2005,
+      "step": 242
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.945636428747748e-05,
+      "loss": 2.334,
+      "step": 244
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9444844837519867e-05,
+      "loss": 2.1858,
+      "step": 246
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9433208100143486e-05,
+      "loss": 2.1054,
+      "step": 248
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.942145421985499e-05,
+      "loss": 2.2089,
+      "step": 250
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9409583342615718e-05,
+      "loss": 1.6846,
+      "step": 252
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.939759561583991e-05,
+      "loss": 1.5922,
+      "step": 254
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9385491188392855e-05,
+      "loss": 1.6762,
+      "step": 256
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9373270210589034e-05,
+      "loss": 1.7761,
+      "step": 258
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9360932834190285e-05,
+      "loss": 1.5138,
+      "step": 260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.934847921240389e-05,
+      "loss": 1.6493,
+      "step": 262
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.933590949988069e-05,
+      "loss": 1.7205,
+      "step": 264
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9323223852713148e-05,
+      "loss": 1.839,
+      "step": 266
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9310422428433426e-05,
+      "loss": 1.7112,
+      "step": 268
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.929750538601142e-05,
+      "loss": 1.7417,
+      "step": 270
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9284472885852787e-05,
+      "loss": 1.6419,
+      "step": 272
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9271325089796963e-05,
+      "loss": 1.7585,
+      "step": 274
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9258062161115134e-05,
+      "loss": 1.744,
+      "step": 276
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.924468426450823e-05,
+      "loss": 1.8246,
+      "step": 278
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.923119156610485e-05,
+      "loss": 2.0436,
+      "step": 280
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9217584233459245e-05,
+      "loss": 1.8373,
+      "step": 282
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9203862435549187e-05,
+      "loss": 1.8646,
+      "step": 284
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9190026342773904e-05,
+      "loss": 1.9658,
+      "step": 286
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.917607612695196e-05,
+      "loss": 2.0826,
+      "step": 288
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9162011961319095e-05,
+      "loss": 2.1591,
+      "step": 290
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.914783402052612e-05,
+      "loss": 2.154,
+      "step": 292
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9133542480636693e-05,
+      "loss": 2.0487,
+      "step": 294
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.911913751912519e-05,
+      "loss": 2.1138,
+      "step": 296
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.910461931487446e-05,
+      "loss": 2.2284,
+      "step": 298
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9089988048173612e-05,
+      "loss": 2.076,
+      "step": 300
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9075243900715786e-05,
+      "loss": 1.6895,
+      "step": 302
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9060387055595887e-05,
+      "loss": 1.6211,
+      "step": 304
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9045417697308322e-05,
+      "loss": 1.5078,
+      "step": 306
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.90303360117447e-05,
+      "loss": 1.561,
+      "step": 308
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9015142186191525e-05,
+      "loss": 1.6423,
+      "step": 310
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8999836409327868e-05,
+      "loss": 1.677,
+      "step": 312
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.898441887122303e-05,
+      "loss": 1.823,
+      "step": 314
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8968889763334183e-05,
+      "loss": 1.6767,
+      "step": 316
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8953249278503983e-05,
+      "loss": 1.6872,
+      "step": 318
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.893749761095818e-05,
+      "loss": 1.7334,
+      "step": 320
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8921634956303214e-05,
+      "loss": 1.7819,
+      "step": 322
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.890566151152377e-05,
+      "loss": 1.7185,
+      "step": 324
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.888957747498035e-05,
+      "loss": 1.8429,
+      "step": 326
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8873383046406785e-05,
+      "loss": 1.8522,
+      "step": 328
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8857078426907783e-05,
+      "loss": 1.7076,
+      "step": 330
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8840663818956413e-05,
+      "loss": 1.8622,
+      "step": 332
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8824139426391603e-05,
+      "loss": 1.9649,
+      "step": 334
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.880750545441559e-05,
+      "loss": 2.0755,
+      "step": 336
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8790762109591395e-05,
+      "loss": 2.0182,
+      "step": 338
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.877390959984024e-05,
+      "loss": 2.0193,
+      "step": 340
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8756948134438973e-05,
+      "loss": 2.0283,
+      "step": 342
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.873987792401747e-05,
+      "loss": 2.0209,
+      "step": 344
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8722699180556027e-05,
+      "loss": 2.2715,
+      "step": 346
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.870541211738269e-05,
+      "loss": 2.1141,
+      "step": 348
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8688016949170672e-05,
+      "loss": 2.0019,
+      "step": 350
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8670513891935623e-05,
+      "loss": 1.633,
+      "step": 352
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8652903163032985e-05,
+      "loss": 1.7309,
+      "step": 354
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8635184981155282e-05,
+      "loss": 1.6854,
+      "step": 356
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8617359566329397e-05,
+      "loss": 1.4624,
+      "step": 358
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8599427139913855e-05,
+      "loss": 1.7174,
+      "step": 360
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8581387924596065e-05,
+      "loss": 1.7547,
+      "step": 362
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8563242144389546e-05,
+      "loss": 1.7919,
+      "step": 364
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.854499002463117e-05,
+      "loss": 1.8004,
+      "step": 366
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.852663179197834e-05,
+      "loss": 1.7906,
+      "step": 368
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.850816767440619e-05,
+      "loss": 1.8483,
+      "step": 370
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8489597901204728e-05,
+      "loss": 1.7136,
+      "step": 372
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8470922702976034e-05,
+      "loss": 1.726,
+      "step": 374
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8452142311631364e-05,
+      "loss": 1.8759,
+      "step": 376
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8433256960388265e-05,
+      "loss": 1.8467,
+      "step": 378
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8414266883767704e-05,
+      "loss": 1.8555,
+      "step": 380
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8395172317591137e-05,
+      "loss": 1.9375,
+      "step": 382
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8375973498977586e-05,
+      "loss": 2.0685,
+      "step": 384
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8356670666340695e-05,
+      "loss": 1.967,
+      "step": 386
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.833726405938577e-05,
+      "loss": 1.9947,
+      "step": 388
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.83177539191068e-05,
+      "loss": 1.943,
+      "step": 390
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.829814048778347e-05,
+      "loss": 2.0449,
+      "step": 392
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8278424008978136e-05,
+      "loss": 2.2705,
+      "step": 394
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8258604727532827e-05,
+      "loss": 2.2501,
+      "step": 396
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.823868288956618e-05,
+      "loss": 2.0302,
+      "step": 398
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8218658742470393e-05,
+      "loss": 2.3523,
+      "step": 400
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.819853253490816e-05,
+      "loss": 1.6453,
+      "step": 402
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.817830451680957e-05,
+      "loss": 1.6131,
+      "step": 404
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8157974939369015e-05,
+      "loss": 1.6462,
+      "step": 406
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.813754405504206e-05,
+      "loss": 1.5778,
+      "step": 408
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8117012117542317e-05,
+      "loss": 1.7071,
+      "step": 410
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.809637938183829e-05,
+      "loss": 1.5606,
+      "step": 412
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8075646104150205e-05,
+      "loss": 1.8436,
+      "step": 414
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.805481254194684e-05,
+      "loss": 1.8131,
+      "step": 416
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8033878953942315e-05,
+      "loss": 1.6493,
+      "step": 418
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8012845600092886e-05,
+      "loss": 1.737,
+      "step": 420
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7991712741593715e-05,
+      "loss": 1.721,
+      "step": 422
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7970480640875626e-05,
+      "loss": 1.7574,
+      "step": 424
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7949149561601853e-05,
+      "loss": 1.8218,
+      "step": 426
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7927719768664747e-05,
+      "loss": 1.8906,
+      "step": 428
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7906191528182516e-05,
+      "loss": 1.7087,
+      "step": 430
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.788456510749589e-05,
+      "loss": 1.7493,
+      "step": 432
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7862840775164814e-05,
+      "loss": 1.7121,
+      "step": 434
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.784101880096513e-05,
+      "loss": 2.0566,
+      "step": 436
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7819099455885195e-05,
+      "loss": 1.8518,
+      "step": 438
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.779708301212253e-05,
+      "loss": 2.1761,
+      "step": 440
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7774969743080453e-05,
+      "loss": 2.1634,
+      "step": 442
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.775275992336466e-05,
+      "loss": 2.1009,
+      "step": 444
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7730453828779842e-05,
+      "loss": 2.1546,
+      "step": 446
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7708051736326232e-05,
+      "loss": 2.2202,
+      "step": 448
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.768555392419618e-05,
+      "loss": 1.9977,
+      "step": 450
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7662960671770694e-05,
+      "loss": 1.471,
+      "step": 452
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.764027225961599e-05,
+      "loss": 1.5104,
+      "step": 454
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7617488969479968e-05,
+      "loss": 1.6735,
+      "step": 456
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.759461108428875e-05,
+      "loss": 1.6133,
+      "step": 458
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.757163888814315e-05,
+      "loss": 1.6513,
+      "step": 460
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7548572666315148e-05,
+      "loss": 1.7925,
+      "step": 462
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7525412705244348e-05,
+      "loss": 1.685,
+      "step": 464
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.750215929253442e-05,
+      "loss": 1.6181,
+      "step": 466
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7478812716949533e-05,
+      "loss": 1.6964,
+      "step": 468
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7455373268410763e-05,
+      "loss": 1.7479,
+      "step": 470
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.74318412379925e-05,
+      "loss": 1.7461,
+      "step": 472
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.740821691791882e-05,
+      "loss": 1.7439,
+      "step": 474
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.738450060155987e-05,
+      "loss": 1.6585,
+      "step": 476
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.736069258342823e-05,
+      "loss": 1.708,
+      "step": 478
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7336793159175228e-05,
+      "loss": 1.9081,
+      "step": 480
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.731280262558729e-05,
+      "loss": 2.0349,
+      "step": 482
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7288721280582263e-05,
+      "loss": 1.956,
+      "step": 484
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.726454942320569e-05,
+      "loss": 1.8301,
+      "step": 486
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7240287353627117e-05,
+      "loss": 2.0108,
+      "step": 488
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.721593537313635e-05,
+      "loss": 1.9919,
+      "step": 490
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.719149378413973e-05,
+      "loss": 2.0082,
+      "step": 492
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7166962890156366e-05,
+      "loss": 2.1452,
+      "step": 494
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7142342995814364e-05,
+      "loss": 2.1019,
+      "step": 496
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.711763440684706e-05,
+      "loss": 2.1886,
+      "step": 498
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.70928374300892e-05,
+      "loss": 2.0478,
+      "step": 500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.706795237347315e-05,
+      "loss": 1.5012,
+      "step": 502
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7042979546025063e-05,
+      "loss": 1.492,
+      "step": 504
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.701791925786104e-05,
+      "loss": 1.7338,
+      "step": 506
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6992771820183282e-05,
+      "loss": 1.5319,
+      "step": 508
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.696753754527623e-05,
+      "loss": 1.8472,
+      "step": 510
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6942216746502675e-05,
+      "loss": 1.6705,
+      "step": 512
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6916809738299875e-05,
+      "loss": 1.7004,
+      "step": 514
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6891316836175658e-05,
+      "loss": 1.7979,
+      "step": 516
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.686573835670448e-05,
+      "loss": 1.707,
+      "step": 518
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.684007461752352e-05,
+      "loss": 1.7927,
+      "step": 520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6814325937328717e-05,
+      "loss": 1.7515,
+      "step": 522
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6788492635870827e-05,
+      "loss": 1.6683,
+      "step": 524
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6762575033951438e-05,
+      "loss": 1.7289,
+      "step": 526
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6736573453418998e-05,
+      "loss": 1.8645,
+      "step": 528
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6710488217164803e-05,
+      "loss": 1.943,
+      "step": 530
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6684319649119018e-05,
+      "loss": 2.0221,
+      "step": 532
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.665806807424661e-05,
+      "loss": 1.9465,
+      "step": 534
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6631733818543352e-05,
+      "loss": 1.9296,
+      "step": 536
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6605317209031758e-05,
+      "loss": 2.1713,
+      "step": 538
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.657881857375702e-05,
+      "loss": 2.0983,
+      "step": 540
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6552238241782934e-05,
+      "loss": 2.0322,
+      "step": 542
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.652557654318782e-05,
+      "loss": 2.1081,
+      "step": 544
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.649883380906043e-05,
+      "loss": 2.2136,
+      "step": 546
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6472010371495808e-05,
+      "loss": 2.1303,
+      "step": 548
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6445106563591206e-05,
+      "loss": 1.9853,
+      "step": 550
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6418122719441903e-05,
+      "loss": 1.5762,
+      "step": 552
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.63910591741371e-05,
+      "loss": 1.5862,
+      "step": 554
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.636391626375572e-05,
+      "loss": 1.6162,
+      "step": 556
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6336694325362275e-05,
+      "loss": 1.5129,
+      "step": 558
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6309393697002637e-05,
+      "loss": 1.7277,
+      "step": 560
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6282014717699872e-05,
+      "loss": 1.7149,
+      "step": 562
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6254557727450016e-05,
+      "loss": 1.6317,
+      "step": 564
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6227023067217857e-05,
+      "loss": 1.7377,
+      "step": 566
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.61994110789327e-05,
+      "loss": 1.8907,
+      "step": 568
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6171722105484116e-05,
+      "loss": 1.7045,
+      "step": 570
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.61439564907177e-05,
+      "loss": 1.7665,
+      "step": 572
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.611611457943077e-05,
+      "loss": 1.7919,
+      "step": 574
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6088196717368133e-05,
+      "loss": 1.8146,
+      "step": 576
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6060203251217742e-05,
+      "loss": 1.7641,
+      "step": 578
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6032134528606413e-05,
+      "loss": 1.8167,
+      "step": 580
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.600399089809552e-05,
+      "loss": 1.9476,
+      "step": 582
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5975772709176638e-05,
+      "loss": 1.9411,
+      "step": 584
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5947480312267235e-05,
+      "loss": 1.9764,
+      "step": 586
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5919114058706295e-05,
+      "loss": 2.1555,
+      "step": 588
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.589067430074996e-05,
+      "loss": 2.1399,
+      "step": 590
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5862161391567172e-05,
+      "loss": 2.2399,
+      "step": 592
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.583357568523527e-05,
+      "loss": 2.2355,
+      "step": 594
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.580491753673559e-05,
+      "loss": 2.0916,
+      "step": 596
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5776187301949083e-05,
+      "loss": 2.0709,
+      "step": 598
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.574738533765186e-05,
+      "loss": 1.9886,
+      "step": 600
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.571851200151079e-05,
+      "loss": 1.5965,
+      "step": 602
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5689567652079037e-05,
+      "loss": 1.673,
+      "step": 604
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5660552648791638e-05,
+      "loss": 1.8633,
+      "step": 606
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5631467351960994e-05,
+      "loss": 1.9587,
+      "step": 608
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5602312122772443e-05,
+      "loss": 1.7769,
+      "step": 610
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.557308732327974e-05,
+      "loss": 2.2116,
+      "step": 612
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5543793316400575e-05,
+      "loss": 2.0675,
+      "step": 614
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5514430465912072e-05,
+      "loss": 1.5471,
+      "step": 616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.548499913644626e-05,
+      "loss": 1.6333,
+      "step": 618
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5455499693485547e-05,
+      "loss": 1.5103,
+      "step": 620
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5425932503358194e-05,
+      "loss": 1.6023,
+      "step": 622
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.539629793323374e-05,
+      "loss": 1.729,
+      "step": 624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.536659635111848e-05,
+      "loss": 1.5536,
+      "step": 626
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5336828125850846e-05,
+      "loss": 1.7814,
+      "step": 628
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5306993627096873e-05,
+      "loss": 1.5771,
+      "step": 630
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.527709322534559e-05,
+      "loss": 1.689,
+      "step": 632
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5247127291904415e-05,
+      "loss": 1.7706,
+      "step": 634
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5217096198894545e-05,
+      "loss": 1.7094,
+      "step": 636
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5187000319246337e-05,
+      "loss": 1.6026,
+      "step": 638
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5156840026694677e-05,
+      "loss": 1.5624,
+      "step": 640
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5126615695774352e-05,
+      "loss": 1.9486,
+      "step": 642
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5096327701815373e-05,
+      "loss": 1.8097,
+      "step": 644
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5065976420938331e-05,
+      "loss": 1.7679,
+      "step": 646
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5035562230049725e-05,
+      "loss": 1.8091,
+      "step": 648
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.500508550683728e-05,
+      "loss": 1.9126,
+      "step": 650
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4974546629765247e-05,
+      "loss": 2.1112,
+      "step": 652
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.494394597806973e-05,
+      "loss": 2.1882,
+      "step": 654
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4913283931753943e-05,
+      "loss": 1.9756,
+      "step": 656
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4882560871583514e-05,
+      "loss": 2.0534,
+      "step": 658
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4851777179081752e-05,
+      "loss": 2.14,
+      "step": 660
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4820933236524897e-05,
+      "loss": 2.1471,
+      "step": 662
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4790029426937394e-05,
+      "loss": 2.1118,
+      "step": 664
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4759066134087117e-05,
+      "loss": 1.529,
+      "step": 666
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.472804374248062e-05,
+      "loss": 1.5823,
+      "step": 668
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4696962637358347e-05,
+      "loss": 1.6219,
+      "step": 670
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4665823204689856e-05,
+      "loss": 1.6208,
+      "step": 672
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.463462583116902e-05,
+      "loss": 1.5877,
+      "step": 674
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4603370904209244e-05,
+      "loss": 1.6907,
+      "step": 676
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4572058811938625e-05,
+      "loss": 1.7026,
+      "step": 678
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4540689943195157e-05,
+      "loss": 1.7639,
+      "step": 680
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4509264687521883e-05,
+      "loss": 1.611,
+      "step": 682
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4477783435162071e-05,
+      "loss": 1.6413,
+      "step": 684
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.444624657705437e-05,
+      "loss": 1.7311,
+      "step": 686
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.441465450482794e-05,
+      "loss": 1.7851,
+      "step": 688
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4383007610797603e-05,
+      "loss": 1.8824,
+      "step": 690
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4351306287958963e-05,
+      "loss": 1.851,
+      "step": 692
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.431955092998353e-05,
+      "loss": 1.8574,
+      "step": 694
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.428774193121383e-05,
+      "loss": 1.7716,
+      "step": 696
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4255879686658508e-05,
+      "loss": 1.8968,
+      "step": 698
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4223964591987423e-05,
+      "loss": 2.0786,
+      "step": 700
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.419199704352673e-05,
+      "loss": 1.8951,
+      "step": 702
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4159977438253971e-05,
+      "loss": 1.985,
+      "step": 704
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.412790617379313e-05,
+      "loss": 2.0482,
+      "step": 706
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4095783648409703e-05,
+      "loss": 2.1777,
+      "step": 708
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4063610261005753e-05,
+      "loss": 1.9897,
+      "step": 710
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4031386411114958e-05,
+      "loss": 2.1258,
+      "step": 712
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3999112498897638e-05,
+      "loss": 2.1569,
+      "step": 714
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.39667889251358e-05,
+      "loss": 1.6332,
+      "step": 716
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3934416091228158e-05,
+      "loss": 1.4974,
+      "step": 718
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3901994399185134e-05,
+      "loss": 1.629,
+      "step": 720
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3869524251623889e-05,
+      "loss": 1.6409,
+      "step": 722
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3837006051763306e-05,
+      "loss": 1.7071,
+      "step": 724
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3804440203418991e-05,
+      "loss": 1.7401,
+      "step": 726
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.377182711099825e-05,
+      "loss": 1.6579,
+      "step": 728
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.373916717949508e-05,
+      "loss": 1.7399,
+      "step": 730
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3706460814485127e-05,
+      "loss": 1.7,
+      "step": 732
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3673708422120658e-05,
+      "loss": 1.7109,
+      "step": 734
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.364091040912551e-05,
+      "loss": 1.6352,
+      "step": 736
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3608067182790046e-05,
+      "loss": 1.5825,
+      "step": 738
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.35751791509661e-05,
+      "loss": 1.7866,
+      "step": 740
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3542246722061897e-05,
+      "loss": 1.8663,
+      "step": 742
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3509270305036999e-05,
+      "loss": 1.7922,
+      "step": 744
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3476250309397214e-05,
+      "loss": 1.8688,
+      "step": 746
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3443187145189517e-05,
+      "loss": 1.98,
+      "step": 748
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.341008122299696e-05,
+      "loss": 2.1442,
+      "step": 750
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3376932953933562e-05,
+      "loss": 2.0696,
+      "step": 752
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3343742749639214e-05,
+      "loss": 2.0473,
+      "step": 754
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3310511022274572e-05,
+      "loss": 2.0118,
+      "step": 756
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3277238184515924e-05,
+      "loss": 2.0139,
+      "step": 758
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3243924649550073e-05,
+      "loss": 2.0929,
+      "step": 760
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3210570831069207e-05,
+      "loss": 2.2703,
+      "step": 762
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.317717714326576e-05,
+      "loss": 2.0147,
+      "step": 764
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.314374400082727e-05,
+      "loss": 1.5511,
+      "step": 766
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3110271818931226e-05,
+      "loss": 1.554,
+      "step": 768
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3076761013239921e-05,
+      "loss": 1.6176,
+      "step": 770
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3043211999895273e-05,
+      "loss": 1.6345,
+      "step": 772
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3009625195513678e-05,
+      "loss": 1.7802,
+      "step": 774
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.2976001017180817e-05,
+      "loss": 1.7645,
+      "step": 776
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2942339882446497e-05,
+      "loss": 1.7157,
+      "step": 778
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2908642209319446e-05,
+      "loss": 1.8843,
+      "step": 780
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2874908416262137e-05,
+      "loss": 1.7453,
+      "step": 782
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2841138922185577e-05,
+      "loss": 1.8738,
+      "step": 784
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2807334146444124e-05,
+      "loss": 1.6823,
+      "step": 786
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.277349450883026e-05,
+      "loss": 1.7135,
+      "step": 788
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.273962042956939e-05,
+      "loss": 1.7416,
+      "step": 790
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2705712329314623e-05,
+      "loss": 1.6427,
+      "step": 792
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2671770629141542e-05,
+      "loss": 1.7839,
+      "step": 794
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.263779575054298e-05,
+      "loss": 1.8602,
+      "step": 796
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.260378811542378e-05,
+      "loss": 1.9448,
+      "step": 798
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2569748146095569e-05,
+      "loss": 1.9093,
+      "step": 800
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2535676265271495e-05,
+      "loss": 1.9411,
+      "step": 802
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2501572896060994e-05,
+      "loss": 2.0018,
+      "step": 804
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.246743846196453e-05,
+      "loss": 2.1838,
+      "step": 806
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2433273386868327e-05,
+      "loss": 2.3012,
+      "step": 808
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2399078095039124e-05,
+      "loss": 2.0348,
+      "step": 810
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2364853011118887e-05,
+      "loss": 2.0092,
+      "step": 812
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2330598560119547e-05,
+      "loss": 1.8901,
+      "step": 814
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2296315167417728e-05,
+      "loss": 1.5123,
+      "step": 816
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.226200325874944e-05,
+      "loss": 1.5888,
+      "step": 818
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2227663260204818e-05,
+      "loss": 1.504,
+      "step": 820
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2193295598222825e-05,
+      "loss": 1.5482,
+      "step": 822
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2158900699585943e-05,
+      "loss": 1.6201,
+      "step": 824
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2124478991414885e-05,
+      "loss": 1.6641,
+      "step": 826
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2090030901163297e-05,
+      "loss": 1.7015,
+      "step": 828
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2055556856612429e-05,
+      "loss": 1.7003,
+      "step": 830
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2021057285865845e-05,
+      "loss": 1.6527,
+      "step": 832
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.198653261734409e-05,
+      "loss": 1.7511,
+      "step": 834
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1951983279779382e-05,
+      "loss": 1.6885,
+      "step": 836
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1917409702210283e-05,
+      "loss": 1.7034,
+      "step": 838
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1882812313976362e-05,
+      "loss": 1.6924,
+      "step": 840
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1848191544712883e-05,
+      "loss": 1.7866,
+      "step": 842
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1813547824345453e-05,
+      "loss": 1.7805,
+      "step": 844
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1778881583084689e-05,
+      "loss": 1.868,
+      "step": 846
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1744193251420877e-05,
+      "loss": 1.9408,
+      "step": 848
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.170948326011863e-05,
+      "loss": 1.9899,
+      "step": 850
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1674752040211521e-05,
+      "loss": 2.0072,
+      "step": 852
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1640000022996755e-05,
+      "loss": 1.9553,
+      "step": 854
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1605227640029795e-05,
+      "loss": 2.1992,
+      "step": 856
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1570435323119006e-05,
+      "loss": 2.0417,
+      "step": 858
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1535623504320308e-05,
+      "loss": 2.2629,
+      "step": 860
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1500792615931781e-05,
+      "loss": 2.0293,
+      "step": 862
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1465943090488326e-05,
+      "loss": 2.0421,
+      "step": 864
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1431075360756277e-05,
+      "loss": 1.5993,
+      "step": 866
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1396189859728032e-05,
+      "loss": 1.5874,
+      "step": 868
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1361287020616675e-05,
+      "loss": 1.6249,
+      "step": 870
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1326367276850605e-05,
+      "loss": 1.5156,
+      "step": 872
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1291431062068128e-05,
+      "loss": 1.7634,
+      "step": 874
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1256478810112105e-05,
+      "loss": 1.6592,
+      "step": 876
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1221510955024542e-05,
+      "loss": 1.6603,
+      "step": 878
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1186527931041205e-05,
+      "loss": 1.6741,
+      "step": 880
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1151530172586238e-05,
+      "loss": 1.8044,
+      "step": 882
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1116518114266757e-05,
+      "loss": 1.7933,
+      "step": 884
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1081492190867445e-05,
+      "loss": 1.7928,
+      "step": 886
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1046452837345174e-05,
+      "loss": 1.6921,
+      "step": 888
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1011400488823594e-05,
+      "loss": 1.8955,
+      "step": 890
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.097633558058772e-05,
+      "loss": 1.6873,
+      "step": 892
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.094125854807855e-05,
+      "loss": 1.701,
+      "step": 894
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0906169826887625e-05,
+      "loss": 1.9367,
+      "step": 896
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0871069852751653e-05,
+      "loss": 2.0128,
+      "step": 898
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.083595906154707e-05,
+      "loss": 1.8403,
+      "step": 900
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.080083788928465e-05,
+      "loss": 1.8517,
+      "step": 902
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0765706772104076e-05,
+      "loss": 1.9253,
+      "step": 904
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0730566146268533e-05,
+      "loss": 2.1538,
+      "step": 906
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0695416448159274e-05,
+      "loss": 2.1409,
+      "step": 908
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0660258114270227e-05,
+      "loss": 2.2853,
+      "step": 910
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.062509158120255e-05,
+      "loss": 2.0466,
+      "step": 912
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0589917285659222e-05,
+      "loss": 1.807,
+      "step": 914
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0554735664439623e-05,
+      "loss": 1.795,
+      "step": 916
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0519547154434105e-05,
+      "loss": 1.5502,
+      "step": 918
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.048435219261855e-05,
+      "loss": 1.6967,
+      "step": 920
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0449151216048976e-05,
+      "loss": 1.6531,
+      "step": 922
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0413944661856084e-05,
+      "loss": 1.6669,
+      "step": 924
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.037873296723985e-05,
+      "loss": 1.8114,
+      "step": 926
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0343516569464076e-05,
+      "loss": 1.6277,
+      "step": 928
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0308295905850963e-05,
+      "loss": 1.6846,
+      "step": 930
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0273071413775695e-05,
+      "loss": 1.647,
+      "step": 932
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0237843530660996e-05,
+      "loss": 1.8631,
+      "step": 934
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.02026126939717e-05,
+      "loss": 1.6792,
+      "step": 936
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0167379341209323e-05,
+      "loss": 1.7413,
+      "step": 938
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0132143909906617e-05,
+      "loss": 1.6664,
+      "step": 940
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.009690683762215e-05,
+      "loss": 1.7221,
+      "step": 942
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0061668561934869e-05,
+      "loss": 1.8029,
+      "step": 944
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0026429520438665e-05,
+      "loss": 1.9112,
+      "step": 946
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.991190150736941e-06,
+      "loss": 2.031,
+      "step": 948
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.955950890437174e-06,
+      "loss": 1.9326,
+      "step": 950
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.920712177145475e-06,
+      "loss": 1.8636,
+      "step": 952
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.88547444846117e-06,
+      "loss": 2.1533,
+      "step": 954
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.850238141971363e-06,
+      "loss": 1.9499,
+      "step": 956
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.815003695245482e-06,
+      "loss": 2.0289,
+      "step": 958
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.779771545829878e-06,
+      "loss": 1.9591,
+      "step": 960
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.744542131242359e-06,
+      "loss": 2.0738,
+      "step": 962
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.709315888966773e-06,
+      "loss": 1.7359,
+      "step": 964
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.67409325644759e-06,
+      "loss": 1.6213,
+      "step": 966
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.638874671084428e-06,
+      "loss": 1.6477,
+      "step": 968
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.603660570226673e-06,
+      "loss": 1.5659,
+      "step": 970
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.568451391167995e-06,
+      "loss": 1.5428,
+      "step": 972
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.533247571140971e-06,
+      "loss": 1.5364,
+      "step": 974
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.498049547311611e-06,
+      "loss": 1.6608,
+      "step": 976
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.46285775677395e-06,
+      "loss": 1.6773,
+      "step": 978
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.427672636544624e-06,
+      "loss": 1.7897,
+      "step": 980
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.392494623557425e-06,
+      "loss": 1.72,
+      "step": 982
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.3573241546579e-06,
+      "loss": 1.5833,
+      "step": 984
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.322161666597903e-06,
+      "loss": 1.7148,
+      "step": 986
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.287007596030178e-06,
+      "loss": 1.6607,
+      "step": 988
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.251862379502952e-06,
+      "loss": 1.8589,
+      "step": 990
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.216726453454486e-06,
+      "loss": 1.9797,
+      "step": 992
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.181600254207685e-06,
+      "loss": 1.7569,
+      "step": 994
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.146484217964656e-06,
+      "loss": 1.7409,
+      "step": 996
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.111378780801298e-06,
+      "loss": 1.843,
+      "step": 998
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.076284378661904e-06,
+      "loss": 1.9225,
+      "step": 1000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.04120144735371e-06,
+      "loss": 1.9379,
+      "step": 1002
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.00613042254153e-06,
+      "loss": 2.0141,
+      "step": 1004
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.971071739742301e-06,
+      "loss": 2.096,
+      "step": 1006
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.936025834319707e-06,
+      "loss": 2.203,
+      "step": 1008
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.900993141478759e-06,
+      "loss": 2.1307,
+      "step": 1010
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.865974096260383e-06,
+      "loss": 1.9568,
+      "step": 1012
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.830969133536046e-06,
+      "loss": 1.9975,
+      "step": 1014
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.795978688002313e-06,
+      "loss": 1.7385,
+      "step": 1016
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.761003194175494e-06,
+      "loss": 1.5292,
+      "step": 1018
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.726043086386215e-06,
+      "loss": 1.6952,
+      "step": 1020
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.691098798774033e-06,
+      "loss": 1.5524,
+      "step": 1022
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.656170765282062e-06,
+      "loss": 1.5926,
+      "step": 1024
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.621259419651552e-06,
+      "loss": 1.707,
+      "step": 1026
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.586365195416532e-06,
+      "loss": 1.6256,
+      "step": 1028
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.55148852589842e-06,
+      "loss": 1.6525,
+      "step": 1030
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.516629844200618e-06,
+      "loss": 1.6297,
+      "step": 1032
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.481789583203176e-06,
+      "loss": 1.7687,
+      "step": 1034
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.446968175557373e-06,
+      "loss": 1.6962,
+      "step": 1036
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.412166053680377e-06,
+      "loss": 1.7744,
+      "step": 1038
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.377383649749862e-06,
+      "loss": 1.7246,
+      "step": 1040
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.342621395698634e-06,
+      "loss": 1.9096,
+      "step": 1042
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.307879723209281e-06,
+      "loss": 1.815,
+      "step": 1044
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.2731590637088e-06,
+      "loss": 1.7844,
+      "step": 1046
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.238459848363262e-06,
+      "loss": 1.9144,
+      "step": 1048
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.203782508072417e-06,
+      "loss": 1.9194,
+      "step": 1050
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.169127473464387e-06,
+      "loss": 1.9421,
+      "step": 1052
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.134495174890295e-06,
+      "loss": 2.1128,
+      "step": 1054
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.099886042418914e-06,
+      "loss": 2.0412,
+      "step": 1056
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.065300505831353e-06,
+      "loss": 2.0167,
+      "step": 1058
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.030738994615687e-06,
+      "loss": 2.101,
+      "step": 1060
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.996201937961659e-06,
+      "loss": 2.1433,
+      "step": 1062
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.96168976475532e-06,
+      "loss": 1.8789,
+      "step": 1064
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.92720290357371e-06,
+      "loss": 1.5519,
+      "step": 1066
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.89274178267956e-06,
+      "loss": 1.5451,
+      "step": 1068
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.858306830015929e-06,
+      "loss": 1.4921,
+      "step": 1070
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.823898473200935e-06,
+      "loss": 1.6422,
+      "step": 1072
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.789517139522424e-06,
+      "loss": 1.7339,
+      "step": 1074
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.755163255932654e-06,
+      "loss": 1.7226,
+      "step": 1076
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.720837249043008e-06,
+      "loss": 1.6448,
+      "step": 1078
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.686539545118694e-06,
+      "loss": 1.5942,
+      "step": 1080
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.652270570073447e-06,
+      "loss": 1.6787,
+      "step": 1082
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.618030749464255e-06,
+      "loss": 1.8133,
+      "step": 1084
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.5838205084860415e-06,
+      "loss": 1.7191,
+      "step": 1086
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.5496402719664245e-06,
+      "loss": 1.7413,
+      "step": 1088
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.5154904643604045e-06,
+      "loss": 1.7457,
+      "step": 1090
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.481371509745128e-06,
+      "loss": 1.7851,
+      "step": 1092
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.447283831814603e-06,
+      "loss": 1.8332,
+      "step": 1094
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.413227853874425e-06,
+      "loss": 1.8734,
+      "step": 1096
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.379203998836551e-06,
+      "loss": 1.9721,
+      "step": 1098
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.345212689214014e-06,
+      "loss": 1.8037,
+      "step": 1100
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.311254347115717e-06,
+      "loss": 2.178,
+      "step": 1102
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.277329394241141e-06,
+      "loss": 2.1488,
+      "step": 1104
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.243438251875157e-06,
+      "loss": 2.1007,
+      "step": 1106
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.209581340882763e-06,
+      "loss": 2.2559,
+      "step": 1108
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.175759081703863e-06,
+      "loss": 2.1401,
+      "step": 1110
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.14197189434806e-06,
+      "loss": 2.0104,
+      "step": 1112
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.108220198389415e-06,
+      "loss": 1.7817,
+      "step": 1114
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.074504412961267e-06,
+      "loss": 1.5753,
+      "step": 1116
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.040824956751012e-06,
+      "loss": 1.6232,
+      "step": 1118
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.007182247994893e-06,
+      "loss": 1.6046,
+      "step": 1120
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.973576704472829e-06,
+      "loss": 1.6074,
+      "step": 1122
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.9400087435032025e-06,
+      "loss": 1.5879,
+      "step": 1124
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.906478781937708e-06,
+      "loss": 1.5187,
+      "step": 1126
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.872987236156151e-06,
+      "loss": 1.7352,
+      "step": 1128
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.839534522061278e-06,
+      "loss": 1.7378,
+      "step": 1130
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.806121055073626e-06,
+      "loss": 1.7823,
+      "step": 1132
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.772747250126349e-06,
+      "loss": 1.6373,
+      "step": 1134
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.739413521660079e-06,
+      "loss": 1.7084,
+      "step": 1136
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.706120283617773e-06,
+      "loss": 1.7693,
+      "step": 1138
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.6728679494395624e-06,
+      "loss": 1.9544,
+      "step": 1140
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.639656932057641e-06,
+      "loss": 1.7807,
+      "step": 1142
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.6064876438911154e-06,
+      "loss": 1.7377,
+      "step": 1144
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.573360496840899e-06,
+      "loss": 2.0219,
+      "step": 1146
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.540275902284583e-06,
+      "loss": 1.974,
+      "step": 1148
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.507234271071345e-06,
+      "loss": 1.8626,
+      "step": 1150
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.474236013516837e-06,
+      "loss": 1.9131,
+      "step": 1152
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.441281539398082e-06,
+      "loss": 2.1233,
+      "step": 1154
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.408371257948405e-06,
+      "loss": 2.0239,
+      "step": 1156
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.375505577852326e-06,
+      "loss": 2.1019,
+      "step": 1158
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.342684907240513e-06,
+      "loss": 1.9495,
+      "step": 1160
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.3099096536847e-06,
+      "loss": 2.0837,
+      "step": 1162
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.277180224192613e-06,
+      "loss": 1.9115,
+      "step": 1164
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.24449702520294e-06,
+      "loss": 1.5175,
+      "step": 1166
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.211860462580261e-06,
+      "loss": 1.5625,
+      "step": 1168
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.179270941610031e-06,
+      "loss": 1.4726,
+      "step": 1170
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.146728866993535e-06,
+      "loss": 1.7119,
+      "step": 1172
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.114234642842847e-06,
+      "loss": 1.8043,
+      "step": 1174
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.081788672675845e-06,
+      "loss": 1.677,
+      "step": 1176
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.049391359411171e-06,
+      "loss": 1.6549,
+      "step": 1178
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.017043105363241e-06,
+      "loss": 1.8129,
+      "step": 1180
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.9847443122372525e-06,
+      "loss": 1.7374,
+      "step": 1182
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.952495381124181e-06,
+      "loss": 1.6513,
+      "step": 1184
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.920296712495819e-06,
+      "loss": 1.6845,
+      "step": 1186
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.888148706199782e-06,
+      "loss": 1.6281,
+      "step": 1188
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.856051761454562e-06,
+      "loss": 1.7846,
+      "step": 1190
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.8240062768445486e-06,
+      "loss": 1.7218,
+      "step": 1192
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.792012650315112e-06,
+      "loss": 1.8823,
+      "step": 1194
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.760071279167629e-06,
+      "loss": 1.7295,
+      "step": 1196
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.728182560054556e-06,
+      "loss": 1.8422,
+      "step": 1198
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.6963468889745265e-06,
+      "loss": 1.8647,
+      "step": 1200
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.6645646612673934e-06,
+      "loss": 1.9415,
+      "step": 1202
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.6328362716093645e-06,
+      "loss": 2.1005,
+      "step": 1204
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.60116211400807e-06,
+      "loss": 1.9825,
+      "step": 1206
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.569542581797667e-06,
+      "loss": 2.136,
+      "step": 1208
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.537978067633993e-06,
+      "loss": 2.1115,
+      "step": 1210
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.506468963489631e-06,
+      "loss": 2.0926,
+      "step": 1212
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.475015660649106e-06,
+      "loss": 2.0519,
+      "step": 1214
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.443618549703979e-06,
+      "loss": 1.4451,
+      "step": 1216
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.412278020548003e-06,
+      "loss": 1.6499,
+      "step": 1218
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.380994462372315e-06,
+      "loss": 1.7575,
+      "step": 1220
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.349768263660551e-06,
+      "loss": 1.768,
+      "step": 1222
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.318599812184071e-06,
+      "loss": 1.9914,
+      "step": 1224
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.287489494997111e-06,
+      "loss": 2.0054,
+      "step": 1226
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.256437698431986e-06,
+      "loss": 1.9918,
+      "step": 1228
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.225444808094305e-06,
+      "loss": 1.6525,
+      "step": 1230
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.1945112088581485e-06,
+      "loss": 1.5217,
+      "step": 1232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.1636372848613395e-06,
+      "loss": 1.5475,
+      "step": 1234
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.132823419500626e-06,
+      "loss": 1.6077,
+      "step": 1236
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.102069995426946e-06,
+      "loss": 1.6792,
+      "step": 1238
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.071377394540673e-06,
+      "loss": 1.5353,
+      "step": 1240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.040745997986853e-06,
+      "loss": 1.6658,
+      "step": 1242
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.010176186150515e-06,
+      "loss": 1.8029,
+      "step": 1244
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.979668338651891e-06,
+      "loss": 1.5903,
+      "step": 1246
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.9492228343417545e-06,
+      "loss": 1.6274,
+      "step": 1248
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.918840051296686e-06,
+      "loss": 1.574,
+      "step": 1250
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.888520366814369e-06,
+      "loss": 1.5849,
+      "step": 1252
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.858264157408948e-06,
+      "loss": 1.6375,
+      "step": 1254
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.828071798806294e-06,
+      "loss": 1.7622,
+      "step": 1256
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.797943665939398e-06,
+      "loss": 1.623,
+      "step": 1258
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.767880132943671e-06,
+      "loss": 1.8606,
+      "step": 1260
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.737881573152317e-06,
+      "loss": 1.882,
+      "step": 1262
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.707948359091694e-06,
+      "loss": 1.8452,
+      "step": 1264
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.678080862476679e-06,
+      "loss": 1.909,
+      "step": 1266
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.648279454206079e-06,
+      "loss": 2.0825,
+      "step": 1268
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.618544504357992e-06,
+      "loss": 1.9816,
+      "step": 1270
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.588876382185231e-06,
+      "loss": 2.1092,
+      "step": 1272
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.559275456110736e-06,
+      "loss": 2.0248,
+      "step": 1274
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.529742093722978e-06,
+      "loss": 2.1858,
+      "step": 1276
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.5002766617714444e-06,
+      "loss": 1.8722,
+      "step": 1278
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.470879526162028e-06,
+      "loss": 1.5387,
+      "step": 1280
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.441551051952518e-06,
+      "loss": 1.6771,
+      "step": 1282
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.412291603348055e-06,
+      "loss": 1.6703,
+      "step": 1284
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.383101543696603e-06,
+      "loss": 1.6778,
+      "step": 1286
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.353981235484461e-06,
+      "loss": 1.56,
+      "step": 1288
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.324931040331722e-06,
+      "loss": 1.5857,
+      "step": 1290
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.295951318987826e-06,
+      "loss": 1.6695,
+      "step": 1292
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.267042431327049e-06,
+      "loss": 1.6566,
+      "step": 1294
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.238204736344045e-06,
+      "loss": 1.6524,
+      "step": 1296
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.209438592149392e-06,
+      "loss": 1.7873,
+      "step": 1298
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.180744355965126e-06,
+      "loss": 1.7025,
+      "step": 1300
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.152122384120339e-06,
+      "loss": 1.6507,
+      "step": 1302
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.123573032046723e-06,
+      "loss": 1.9534,
+      "step": 1304
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.0950966542741675e-06,
+      "loss": 1.7546,
+      "step": 1306
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.066693604426364e-06,
+      "loss": 1.7409,
+      "step": 1308
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.038364235216389e-06,
+      "loss": 1.9443,
+      "step": 1310
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.010108898442368e-06,
+      "loss": 1.764,
+      "step": 1312
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.981927944983063e-06,
+      "loss": 1.8742,
+      "step": 1314
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.953821724793539e-06,
+      "loss": 1.9115,
+      "step": 1316
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.92579058690082e-06,
+      "loss": 1.9487,
+      "step": 1318
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.897834879399526e-06,
+      "loss": 2.1663,
+      "step": 1320
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.869954949447596e-06,
+      "loss": 2.0637,
+      "step": 1322
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.842151143261943e-06,
+      "loss": 2.1487,
+      "step": 1324
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.8144238061141593e-06,
+      "loss": 2.0735,
+      "step": 1326
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.7867732823262417e-06,
+      "loss": 1.9353,
+      "step": 1328
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7591999152663027e-06,
+      "loss": 1.5337,
+      "step": 1330
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7317040473443157e-06,
+      "loss": 1.6132,
+      "step": 1332
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7042860200078545e-06,
+      "loss": 1.4843,
+      "step": 1334
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6769461737378597e-06,
+      "loss": 1.5181,
+      "step": 1336
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6496848480444104e-06,
+      "loss": 1.7481,
+      "step": 1338
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6225023814625025e-06,
+      "loss": 1.7613,
+      "step": 1340
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.5953991115478526e-06,
+      "loss": 1.6056,
+      "step": 1342
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.568375374872691e-06,
+      "loss": 1.6225,
+      "step": 1344
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.54143150702161e-06,
+      "loss": 1.6991,
+      "step": 1346
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.5145678425873663e-06,
+      "loss": 1.6515,
+      "step": 1348
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.4877847151667452e-06,
+      "loss": 1.6193,
+      "step": 1350
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.461082457356414e-06,
+      "loss": 1.591,
+      "step": 1352
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.4344614007487742e-06,
+      "loss": 1.708,
+      "step": 1354
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.407921875927883e-06,
+      "loss": 1.7268,
+      "step": 1356
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.3814642124653086e-06,
+      "loss": 1.9632,
+      "step": 1358
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.355088738916056e-06,
+      "loss": 1.7395,
+      "step": 1360
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.328795782814487e-06,
+      "loss": 1.8749,
+      "step": 1362
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3025856706702477e-06,
+      "loss": 1.802,
+      "step": 1364
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.276458727964218e-06,
+      "loss": 2.1502,
+      "step": 1366
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2504152791444656e-06,
+      "loss": 1.9775,
+      "step": 1368
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.224455647622221e-06,
+      "loss": 2.1675,
+      "step": 1370
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.1985801557678606e-06,
+      "loss": 2.0203,
+      "step": 1372
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1727891249069e-06,
+      "loss": 2.0424,
+      "step": 1374
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.147082875316008e-06,
+      "loss": 2.1101,
+      "step": 1376
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.121461726219026e-06,
+      "loss": 1.9402,
+      "step": 1378
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.095925995783008e-06,
+      "loss": 1.5451,
+      "step": 1380
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.0704760011142653e-06,
+      "loss": 1.5736,
+      "step": 1382
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.0451120582544303e-06,
+      "loss": 1.5506,
+      "step": 1384
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0198344821765346e-06,
+      "loss": 1.5911,
+      "step": 1386
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.994643586781083e-06,
+      "loss": 1.6752,
+      "step": 1388
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.9695396848921875e-06,
+      "loss": 1.6838,
+      "step": 1390
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.944523088253648e-06,
+      "loss": 1.7638,
+      "step": 1392
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9195941075250976e-06,
+      "loss": 1.836,
+      "step": 1394
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.8947530522781455e-06,
+      "loss": 1.6376,
+      "step": 1396
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.870000230992528e-06,
+      "loss": 1.7293,
+      "step": 1398
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8453359510522803e-06,
+      "loss": 1.9062,
+      "step": 1400
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8207605187419194e-06,
+      "loss": 1.8306,
+      "step": 1402
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7962742392426377e-06,
+      "loss": 1.7754,
+      "step": 1404
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7718774166285157e-06,
+      "loss": 1.8018,
+      "step": 1406
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7475703538627485e-06,
+      "loss": 1.7589,
+      "step": 1408
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.723353352793878e-06,
+      "loss": 1.7515,
+      "step": 1410
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.6992267141520467e-06,
+      "loss": 2.0015,
+      "step": 1412
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.675190737545267e-06,
+      "loss": 2.0211,
+      "step": 1414
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6512457214556943e-06,
+      "loss": 1.9215,
+      "step": 1416
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6273919632359247e-06,
+      "loss": 1.9639,
+      "step": 1418
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6036297591052993e-06,
+      "loss": 2.0678,
+      "step": 1420
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5799594041462318e-06,
+      "loss": 2.1676,
+      "step": 1422
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5563811923005355e-06,
+      "loss": 2.0718,
+      "step": 1424
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.532895416365779e-06,
+      "loss": 2.0191,
+      "step": 1426
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.5095023679916508e-06,
+      "loss": 1.6374,
+      "step": 1428
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.4862023376763324e-06,
+      "loss": 1.4945,
+      "step": 1430
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.462995614762894e-06,
+      "loss": 1.5496,
+      "step": 1432
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4398824874357042e-06,
+      "loss": 1.6265,
+      "step": 1434
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.416863242716845e-06,
+      "loss": 1.6466,
+      "step": 1436
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.393938166462554e-06,
+      "loss": 1.6862,
+      "step": 1438
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3711075433596675e-06,
+      "loss": 1.5553,
+      "step": 1440
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.348371656922095e-06,
+      "loss": 1.6354,
+      "step": 1442
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3257307894872873e-06,
+      "loss": 1.7651,
+      "step": 1444
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.303185222212737e-06,
+      "loss": 1.7357,
+      "step": 1446
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.2807352350724874e-06,
+      "loss": 1.7449,
+      "step": 1448
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.2583811068536556e-06,
+      "loss": 1.6902,
+      "step": 1450
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.236123115152963e-06,
+      "loss": 1.7208,
+      "step": 1452
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2139615363733015e-06,
+      "loss": 1.86,
+      "step": 1454
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1918966457202873e-06,
+      "loss": 1.7434,
+      "step": 1456
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1699287171988546e-06,
+      "loss": 1.8137,
+      "step": 1458
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1480580236098448e-06,
+      "loss": 1.9533,
+      "step": 1460
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.126284836546626e-06,
+      "loss": 1.981,
+      "step": 1462
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1046094263917115e-06,
+      "loss": 1.9583,
+      "step": 1464
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0830320623134105e-06,
+      "loss": 1.9106,
+      "step": 1466
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0615530122624816e-06,
+      "loss": 1.945,
+      "step": 1468
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.040172542968806e-06,
+      "loss": 2.1462,
+      "step": 1470
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0188909199380734e-06,
+      "loss": 2.1872,
+      "step": 1472
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.9977084074484888e-06,
+      "loss": 1.9743,
+      "step": 1474
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.9766252685474887e-06,
+      "loss": 2.101,
+      "step": 1476
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9556417650484716e-06,
+      "loss": 1.8562,
+      "step": 1478
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.934758157527552e-06,
+      "loss": 1.5153,
+      "step": 1480
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.913974705320322e-06,
+      "loss": 1.6029,
+      "step": 1482
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8932916665186275e-06,
+      "loss": 1.6023,
+      "step": 1484
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8727092979673722e-06,
+      "loss": 1.5499,
+      "step": 1486
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.852227855261317e-06,
+      "loss": 1.7338,
+      "step": 1488
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8318475927419143e-06,
+      "loss": 1.6786,
+      "step": 1490
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.811568763494147e-06,
+      "loss": 1.7163,
+      "step": 1492
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.7913916193433822e-06,
+      "loss": 1.7118,
+      "step": 1494
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7713164108522518e-06,
+      "loss": 1.764,
+      "step": 1496
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7513433873175313e-06,
+      "loss": 1.5847,
+      "step": 1498
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.731472796767053e-06,
+      "loss": 1.7989,
+      "step": 1500
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7117048859566188e-06,
+      "loss": 1.647,
+      "step": 1502
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.6920399003669418e-06,
+      "loss": 1.8616,
+      "step": 1504
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.6724780842005928e-06,
+      "loss": 1.7615,
+      "step": 1506
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6530196803789711e-06,
+      "loss": 1.8888,
+      "step": 1508
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6336649305392883e-06,
+      "loss": 1.7978,
+      "step": 1510
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.614414075031564e-06,
+      "loss": 1.9045,
+      "step": 1512
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5952673529156426e-06,
+      "loss": 1.9761,
+      "step": 1514
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5762250019582293e-06,
+      "loss": 2.0572,
+      "step": 1516
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5572872586299304e-06,
+      "loss": 2.0228,
+      "step": 1518
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5384543581023192e-06,
+      "loss": 2.1219,
+      "step": 1520
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5197265342450184e-06,
+      "loss": 2.0404,
+      "step": 1522
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5011040196227932e-06,
+      "loss": 2.0621,
+      "step": 1524
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.482587045492664e-06,
+      "loss": 1.9169,
+      "step": 1526
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.464175841801032e-06,
+      "loss": 1.7185,
+      "step": 1528
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4458706371808306e-06,
+      "loss": 1.6368,
+      "step": 1530
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4276716589486784e-06,
+      "loss": 1.5826,
+      "step": 1532
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4095791331020592e-06,
+      "loss": 1.5946,
+      "step": 1534
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.3915932843165192e-06,
+      "loss": 1.7053,
+      "step": 1536
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3737143359428706e-06,
+      "loss": 1.6754,
+      "step": 1538
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3559425100044233e-06,
+      "loss": 1.6797,
+      "step": 1540
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3382780271942253e-06,
+      "loss": 1.788,
+      "step": 1542
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3207211068723203e-06,
+      "loss": 1.6312,
+      "step": 1544
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.30327196706303e-06,
+      "loss": 1.8689,
+      "step": 1546
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2859308244522384e-06,
+      "loss": 1.7198,
+      "step": 1548
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.268697894384706e-06,
+      "loss": 1.63,
+      "step": 1550
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.251573390861397e-06,
+      "loss": 1.6771,
+      "step": 1552
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.234557526536817e-06,
+      "loss": 1.7827,
+      "step": 1554
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2176505127163752e-06,
+      "loss": 1.8489,
+      "step": 1556
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2008525593537601e-06,
+      "loss": 1.7908,
+      "step": 1558
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1841638750483308e-06,
+      "loss": 1.9689,
+      "step": 1560
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1675846670425317e-06,
+      "loss": 1.966,
+      "step": 1562
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.151115141219309e-06,
+      "loss": 1.8851,
+      "step": 1564
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.134755502099566e-06,
+      "loss": 1.8687,
+      "step": 1566
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1185059528396125e-06,
+      "loss": 1.9745,
+      "step": 1568
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1023666952286504e-06,
+      "loss": 2.0723,
+      "step": 1570
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.086337929686261e-06,
+      "loss": 2.0465,
+      "step": 1572
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0704198552599265e-06,
+      "loss": 2.1794,
+      "step": 1574
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0546126696225412e-06,
+      "loss": 2.138,
+      "step": 1576
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.038916569069972e-06,
+      "loss": 2.0238,
+      "step": 1578
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0233317485186167e-06,
+      "loss": 1.7425,
+      "step": 1580
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.007858401502979e-06,
+      "loss": 1.434,
+      "step": 1582
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.924967201732726e-07,
+      "loss": 1.625,
+      "step": 1584
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.772468952930259e-07,
+      "loss": 1.5343,
+      "step": 1586
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.621091162367225e-07,
+      "loss": 1.7319,
+      "step": 1588
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.470835709874437e-07,
+      "loss": 1.808,
+      "step": 1590
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.321704461345326e-07,
+      "loss": 1.657,
+      "step": 1592
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.173699268712855e-07,
+      "loss": 1.7956,
+      "step": 1594
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.026821969926414e-07,
+      "loss": 1.6254,
+      "step": 1596
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.881074388929078e-07,
+      "loss": 1.6897,
+      "step": 1598
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.736458335634912e-07,
+      "loss": 1.7652,
+      "step": 1600
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.592975605906528e-07,
+      "loss": 1.6281,
+      "step": 1602
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.450627981532733e-07,
+      "loss": 1.6301,
+      "step": 1604
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.309417230206462e-07,
+      "loss": 1.866,
+      "step": 1606
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.169345105502824e-07,
+      "loss": 1.8391,
+      "step": 1608
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.030413346857236e-07,
+      "loss": 1.7652,
+      "step": 1610
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.892623679543932e-07,
+      "loss": 1.8361,
+      "step": 1612
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.75597781465448e-07,
+      "loss": 2.0671,
+      "step": 1614
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.620477449076546e-07,
+      "loss": 1.9406,
+      "step": 1616
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.486124265472871e-07,
+      "loss": 2.0041,
+      "step": 1618
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.35291993226025e-07,
+      "loss": 1.9721,
+      "step": 1620
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.22086610358893e-07,
+      "loss": 2.0757,
+      "step": 1622
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.089964419322049e-07,
+      "loss": 2.1608,
+      "step": 1624
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.960216505015216e-07,
+      "loss": 2.1365,
+      "step": 1626
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.831623971896384e-07,
+      "loss": 1.8488,
+      "step": 1628
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.704188416845814e-07,
+      "loss": 1.5193,
+      "step": 1630
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.577911422376238e-07,
+      "loss": 1.5308,
+      "step": 1632
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.452794556613251e-07,
+      "loss": 1.5771,
+      "step": 1634
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.328839373275775e-07,
+      "loss": 1.5246,
+      "step": 1636
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.20604741165679e-07,
+      "loss": 1.5939,
+      "step": 1638
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.084420196604246e-07,
+      "loss": 1.635,
+      "step": 1640
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.963959238502137e-07,
+      "loss": 1.5927,
+      "step": 1642
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.844666033251634e-07,
+      "loss": 1.7587,
+      "step": 1644
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.726542062252626e-07,
+      "loss": 1.7935,
+      "step": 1646
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.609588792385301e-07,
+      "loss": 1.7692,
+      "step": 1648
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.493807675991891e-07,
+      "loss": 1.778,
+      "step": 1650
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.37920015085871e-07,
+      "loss": 1.6862,
+      "step": 1652
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.265767640198171e-07,
+      "loss": 1.8197,
+      "step": 1654
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.15351155263124e-07,
+      "loss": 1.8365,
+      "step": 1656
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.042433282169901e-07,
+      "loss": 1.7107,
+      "step": 1658
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 4.93253420819979e-07,
+      "loss": 1.8031,
+      "step": 1660
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.823815695463208e-07,
+      "loss": 1.865,
+      "step": 1662
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.7162790940419643e-07,
+      "loss": 1.9976,
+      "step": 1664
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.6099257393407926e-07,
+      "loss": 1.9547,
+      "step": 1666
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.50475695207071e-07,
+      "loss": 2.1711,
+      "step": 1668
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.400774038232569e-07,
+      "loss": 1.9739,
+      "step": 1670
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.297978289100924e-07,
+      "loss": 2.0658,
+      "step": 1672
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.1963709812078756e-07,
+      "loss": 2.1135,
+      "step": 1674
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.0959533763274106e-07,
+      "loss": 1.9309,
+      "step": 1676
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 3.9967267214594874e-07,
+      "loss": 1.8501,
+      "step": 1678
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.8986922488147614e-07,
+      "loss": 1.4997,
+      "step": 1680
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.801851175799176e-07,
+      "loss": 1.7059,
+      "step": 1682
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.7062047049988505e-07,
+      "loss": 1.6233,
+      "step": 1684
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.611754024165226e-07,
+      "loss": 1.5627,
+      "step": 1686
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.5185003062001676e-07,
+      "loss": 1.5997,
+      "step": 1688
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.426444709141563e-07,
+      "loss": 1.7561,
+      "step": 1690
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.335588376148824e-07,
+      "loss": 1.6633,
+      "step": 1692
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.24593243548873e-07,
+      "loss": 1.6728,
+      "step": 1694
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.1574780005214633e-07,
+      "loss": 1.6645,
+      "step": 1696
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.070226169686663e-07,
+      "loss": 1.7514,
+      "step": 1698
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.984178026489937e-07,
+      "loss": 1.6465,
+      "step": 1700
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.8993346394892706e-07,
+      "loss": 1.7203,
+      "step": 1702
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.8156970622818416e-07,
+      "loss": 1.6954,
+      "step": 1704
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.733266333490947e-07,
+      "loss": 1.6774,
+      "step": 1706
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.6520434767529856e-07,
+      "loss": 1.8674,
+      "step": 1708
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.57202950070492e-07,
+      "loss": 1.903,
+      "step": 1710
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.4932253989715883e-07,
+      "loss": 1.9868,
+      "step": 1712
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.415632150153469e-07,
+      "loss": 1.9943,
+      "step": 1714
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.3392507178145229e-07,
+      "loss": 1.8895,
+      "step": 1716
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.264082050470151e-07,
+      "loss": 2.0532,
+      "step": 1718
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.1901270815755215e-07,
+      "loss": 2.0355,
+      "step": 1720
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.1173867295138816e-07,
+      "loss": 2.0338,
+      "step": 1722
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0458618975851996e-07,
+      "loss": 2.0172,
+      "step": 1724
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.9755534739949733e-07,
+      "loss": 1.9194,
+      "step": 1726
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.9064623318431175e-07,
+      "loss": 1.9948,
+      "step": 1728
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.8385893291132494e-07,
+      "loss": 1.7778,
+      "step": 1730
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.7719353086618763e-07,
+      "loss": 1.4658,
+      "step": 1732
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.7065010982080688e-07,
+      "loss": 1.6131,
+      "step": 1734
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.6422875103230928e-07,
+      "loss": 1.5365,
+      "step": 1736
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.5792953424203727e-07,
+      "loss": 1.6563,
+      "step": 1738
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.5175253767455656e-07,
+      "loss": 1.6242,
+      "step": 1740
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4569783803668135e-07,
+      "loss": 1.5964,
+      "step": 1742
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.3976551051653187e-07,
+      "loss": 1.5942,
+      "step": 1744
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.3395562878258606e-07,
+      "loss": 1.5532,
+      "step": 1746
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2826826498277823e-07,
+      "loss": 1.6633,
+      "step": 1748
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.227034897435986e-07,
+      "loss": 1.7018,
+      "step": 1750
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.172613721692084e-07,
+      "loss": 1.6879,
+      "step": 1752
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1194197984059852e-07,
+      "loss": 1.7878,
+      "step": 1754
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.0674537881473102e-07,
+      "loss": 1.8444,
+      "step": 1756
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.0167163362373333e-07,
+      "loss": 1.8548,
+      "step": 1758
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.672080727409217e-08,
+      "loss": 1.7185,
+      "step": 1760
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.189296124586744e-08,
+      "loss": 2.0053,
+      "step": 1762
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.718815549193516e-08,
+      "loss": 1.9904,
+      "step": 1764
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.260644843723686e-08,
+      "loss": 1.9012,
+      "step": 1766
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.814789697806025e-08,
+      "loss": 2.1227,
+      "step": 1768
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.381255648132746e-08,
+      "loss": 2.1608,
+      "step": 1770
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.960048078390902e-08,
+      "loss": 2.1626,
+      "step": 1772
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.551172219195767e-08,
+      "loss": 2.1162,
+      "step": 1774
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.154633148025446e-08,
+      "loss": 2.094,
+      "step": 1776
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.770435789158369e-08,
+      "loss": 1.8872,
+      "step": 1778
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.398584913611449e-08,
+      "loss": 1.6439,
+      "step": 1780
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.039085139081579e-08,
+      "loss": 1.484,
+      "step": 1782
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.691940929887895e-08,
+      "loss": 1.5765,
+      "step": 1784
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.357156596916046e-08,
+      "loss": 1.5862,
+      "step": 1786
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.0347362975654556e-08,
+      "loss": 1.5702,
+      "step": 1788
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.724684035696813e-08,
+      "loss": 1.6757,
+      "step": 1790
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.427003661582995e-08,
+      "loss": 1.7416,
+      "step": 1792
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.1416988718609986e-08,
+      "loss": 1.7207,
+      "step": 1794
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.8687732094859755e-08,
+      "loss": 1.6749,
+      "step": 1796
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.6082300636876e-08,
+      "loss": 1.6948,
+      "step": 1798
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.360072669927327e-08,
+      "loss": 1.6434,
+      "step": 1800
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.1243041098587548e-08,
+      "loss": 1.7158,
+      "step": 1802
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.9009273112893245e-08,
+      "loss": 1.6931,
+      "step": 1804
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.689945048143571e-08,
+      "loss": 1.7745,
+      "step": 1806
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.491359940429149e-08,
+      "loss": 1.8563,
+      "step": 1808
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.3051744542039724e-08,
+      "loss": 1.8418,
+      "step": 1810
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.1313909015457925e-08,
+      "loss": 1.8236,
+      "step": 1812
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.70011440523222e-09,
+      "loss": 1.7733,
+      "step": 1814
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.210380751694225e-09,
+      "loss": 1.9181,
+      "step": 1816
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 6.844726554565695e-09,
+      "loss": 2.1189,
+      "step": 1818
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5.603168772733147e-09,
+      "loss": 1.9684,
+      "step": 1820
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.4857228240380305e-09,
+      "loss": 2.0925,
+      "step": 1822
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.4924025850813225e-09,
+      "loss": 1.8737,
+      "step": 1824
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.6232203910547817e-09,
+      "loss": 2.1593,
+      "step": 1826
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.87818703558551e-09,
+      "loss": 1.9364,
+      "step": 1828
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.2573117706027317e-09,
+      "loss": 1.5308,
+      "step": 1830
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 7.606023062234346e-10,
+      "loss": 1.6164,
+      "step": 1832
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.880648106557861e-10,
+      "loss": 1.739,
+      "step": 1834
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.3970391012363415e-10,
+      "loss": 1.7537,
+      "step": 1836
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.552268880766761e-11,
+      "loss": 1.8015,
+      "step": 1838
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1839,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 6.817382226043699e+16,
+  "trial_name": null,
+  "trial_params": null
+}