diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,2245 +1,5953 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.9985838559579205,
-  "global_step": 1851,
+  "epoch": 4.0,
+  "global_step": 4944,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.01,
-      "learning_rate": 1.0526315789473684e-05,
-      "loss": 11.5,
+      "epoch": 0.0,
+      "learning_rate": 2.0161290322580645e-07,
+      "loss": 1.7344,
       "step": 5
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 2.105263157894737e-05,
-      "loss": 6.7393,
+      "epoch": 0.01,
+      "learning_rate": 4.032258064516129e-07,
+      "loss": 1.6586,
       "step": 10
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 3.157894736842106e-05,
-      "loss": 3.7191,
+      "epoch": 0.01,
+      "learning_rate": 6.048387096774194e-07,
+      "loss": 1.6465,
       "step": 15
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 3.999997059313686e-05,
-      "loss": 3.1877,
+      "epoch": 0.02,
+      "learning_rate": 8.064516129032258e-07,
+      "loss": 1.5832,
       "step": 20
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 3.999894136200706e-05,
-      "loss": 2.9055,
+      "epoch": 0.02,
+      "learning_rate": 1.0080645161290323e-06,
+      "loss": 1.5574,
       "step": 25
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 3.9996441874195635e-05,
-      "loss": 2.9141,
+      "epoch": 0.02,
+      "learning_rate": 1.2096774193548388e-06,
+      "loss": 1.4381,
       "step": 30
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 3.999247231345674e-05,
-      "loss": 2.7834,
+      "epoch": 0.03,
+      "learning_rate": 1.4112903225806455e-06,
+      "loss": 1.5215,
       "step": 35
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 3.998703297161948e-05,
-      "loss": 2.7901,
+      "epoch": 0.03,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.5566,
       "step": 40
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 3.9980124248566466e-05,
-      "loss": 2.8676,
+      "epoch": 0.04,
+      "learning_rate": 1.8145161290322583e-06,
+      "loss": 1.5641,
       "step": 45
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 3.9971746652204386e-05,
-      "loss": 2.6787,
+      "epoch": 0.04,
+      "learning_rate": 2.0161290322580646e-06,
+      "loss": 1.4707,
       "step": 50
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.996190079842669e-05,
-      "loss": 2.704,
+      "epoch": 0.04,
+      "learning_rate": 2.217741935483871e-06,
+      "loss": 1.4484,
       "step": 55
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.995058741106831e-05,
-      "loss": 2.6286,
+      "epoch": 0.05,
+      "learning_rate": 2.4193548387096776e-06,
+      "loss": 1.4656,
       "step": 60
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.993780732185244e-05,
-      "loss": 2.6983,
+      "epoch": 0.05,
+      "learning_rate": 2.620967741935484e-06,
+      "loss": 1.3938,
       "step": 65
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.992356147032939e-05,
-      "loss": 2.6522,
+      "epoch": 0.06,
+      "learning_rate": 2.822580645161291e-06,
+      "loss": 1.4152,
       "step": 70
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.9907850903807514e-05,
-      "loss": 2.6293,
+      "epoch": 0.06,
+      "learning_rate": 3.024193548387097e-06,
+      "loss": 1.4182,
       "step": 75
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 3.989067677727622e-05,
-      "loss": 2.5399,
+      "epoch": 0.06,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 1.5051,
       "step": 80
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 3.987204035332105e-05,
-      "loss": 2.6687,
+      "epoch": 0.07,
+      "learning_rate": 3.4274193548387097e-06,
+      "loss": 1.4918,
       "step": 85
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 3.985194300203087e-05,
-      "loss": 2.6291,
+      "epoch": 0.07,
+      "learning_rate": 3.6290322580645166e-06,
+      "loss": 1.4738,
       "step": 90
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 3.983038620089714e-05,
-      "loss": 2.6533,
+      "epoch": 0.08,
+      "learning_rate": 3.830645161290323e-06,
+      "loss": 1.4035,
       "step": 95
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 3.980737153470528e-05,
-      "loss": 2.5923,
+      "epoch": 0.08,
+      "learning_rate": 4.032258064516129e-06,
+      "loss": 1.4367,
       "step": 100
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 3.97829006954182e-05,
-      "loss": 2.5941,
+      "epoch": 0.08,
+      "learning_rate": 4.233870967741936e-06,
+      "loss": 1.4076,
       "step": 105
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 3.9756975482051855e-05,
-      "loss": 2.6612,
+      "epoch": 0.09,
+      "learning_rate": 4.435483870967742e-06,
+      "loss": 1.4902,
       "step": 110
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 3.972959780054306e-05,
-      "loss": 2.6058,
+      "epoch": 0.09,
+      "learning_rate": 4.637096774193548e-06,
+      "loss": 1.3578,
       "step": 115
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 3.9700769663609304e-05,
-      "loss": 2.5226,
+      "epoch": 0.1,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 1.4467,
       "step": 120
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 3.967049319060081e-05,
-      "loss": 2.5573,
+      "epoch": 0.1,
+      "learning_rate": 5.040322580645161e-06,
+      "loss": 1.4766,
       "step": 125
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 3.963877060734473e-05,
-      "loss": 2.603,
+      "epoch": 0.11,
+      "learning_rate": 5.241935483870968e-06,
+      "loss": 1.4828,
       "step": 130
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 3.9605604245981515e-05,
-      "loss": 2.6506,
+      "epoch": 0.11,
+      "learning_rate": 5.443548387096774e-06,
+      "loss": 1.4258,
       "step": 135
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 3.9570996544793445e-05,
-      "loss": 2.631,
+      "epoch": 0.11,
+      "learning_rate": 5.645161290322582e-06,
+      "loss": 1.4602,
       "step": 140
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 3.9534950048025396e-05,
-      "loss": 2.5668,
+      "epoch": 0.12,
+      "learning_rate": 5.846774193548388e-06,
+      "loss": 1.4902,
       "step": 145
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 3.9497467405697756e-05,
-      "loss": 2.6354,
+      "epoch": 0.12,
+      "learning_rate": 6.048387096774194e-06,
+      "loss": 1.3729,
       "step": 150
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 3.9458551373411664e-05,
-      "loss": 2.5427,
+      "epoch": 0.13,
+      "learning_rate": 6.25e-06,
+      "loss": 1.4902,
       "step": 155
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 3.941820481214637e-05,
-      "loss": 2.5586,
+      "epoch": 0.13,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 1.435,
       "step": 160
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 3.937643068804896e-05,
-      "loss": 2.5577,
+      "epoch": 0.13,
+      "learning_rate": 6.653225806451613e-06,
+      "loss": 1.4096,
       "step": 165
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 3.933323207221624e-05,
-      "loss": 2.5664,
+      "epoch": 0.14,
+      "learning_rate": 6.854838709677419e-06,
+      "loss": 1.4508,
       "step": 170
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 3.9288612140468984e-05,
-      "loss": 2.5396,
+      "epoch": 0.14,
+      "learning_rate": 7.056451612903227e-06,
+      "loss": 1.4266,
       "step": 175
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 3.924257417311846e-05,
-      "loss": 2.5558,
+      "epoch": 0.15,
+      "learning_rate": 7.258064516129033e-06,
+      "loss": 1.4648,
       "step": 180
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 3.919512155472529e-05,
-      "loss": 2.5306,
+      "epoch": 0.15,
+      "learning_rate": 7.459677419354839e-06,
+      "loss": 1.3996,
       "step": 185
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 3.9146257773850585e-05,
-      "loss": 2.4945,
+      "epoch": 0.15,
+      "learning_rate": 7.661290322580646e-06,
+      "loss": 1.4152,
       "step": 190
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 3.9095986422799506e-05,
-      "loss": 2.6086,
+      "epoch": 0.16,
+      "learning_rate": 7.862903225806451e-06,
+      "loss": 1.4648,
       "step": 195
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 3.904431119735718e-05,
-      "loss": 2.4973,
+      "epoch": 0.16,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 1.3128,
       "step": 200
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 3.899123589651695e-05,
-      "loss": 2.5872,
+      "epoch": 0.17,
+      "learning_rate": 8.266129032258065e-06,
+      "loss": 1.4242,
       "step": 205
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 3.893676442220114e-05,
-      "loss": 2.5216,
+      "epoch": 0.17,
+      "learning_rate": 8.467741935483872e-06,
+      "loss": 1.4648,
       "step": 210
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 3.888090077897418e-05,
-      "loss": 2.5367,
+      "epoch": 0.17,
+      "learning_rate": 8.669354838709677e-06,
+      "loss": 1.4527,
       "step": 215
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 3.882364907374819e-05,
-      "loss": 2.5495,
+      "epoch": 0.18,
+      "learning_rate": 8.870967741935484e-06,
+      "loss": 1.4,
       "step": 220
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 3.8765013515481065e-05,
-      "loss": 2.6037,
+      "epoch": 0.18,
+      "learning_rate": 9.072580645161291e-06,
+      "loss": 1.4984,
       "step": 225
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 3.870499841486705e-05,
-      "loss": 2.5012,
+      "epoch": 0.19,
+      "learning_rate": 9.274193548387097e-06,
+      "loss": 1.4297,
       "step": 230
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 3.864360818401982e-05,
-      "loss": 2.4773,
+      "epoch": 0.19,
+      "learning_rate": 9.475806451612905e-06,
+      "loss": 1.4645,
       "step": 235
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 3.8580847336148105e-05,
-      "loss": 2.52,
+      "epoch": 0.19,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 1.4107,
       "step": 240
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 3.851672048522395e-05,
-      "loss": 2.4718,
+      "epoch": 0.2,
+      "learning_rate": 9.879032258064517e-06,
+      "loss": 1.4328,
       "step": 245
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 3.845123234564343e-05,
-      "loss": 2.5251,
+      "epoch": 0.2,
+      "learning_rate": 9.999995524479982e-06,
+      "loss": 1.4279,
       "step": 250
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 3.838438773188014e-05,
-      "loss": 2.5547,
+      "epoch": 0.21,
+      "learning_rate": 9.999945174971776e-06,
+      "loss": 1.4738,
       "step": 255
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 3.831619155813119e-05,
-      "loss": 2.475,
+      "epoch": 0.21,
+      "learning_rate": 9.999838882120566e-06,
+      "loss": 1.3898,
       "step": 260
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 3.8246648837955965e-05,
-      "loss": 2.4957,
+      "epoch": 0.21,
+      "learning_rate": 9.999676647115646e-06,
+      "loss": 1.3346,
       "step": 265
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 3.817576468390753e-05,
-      "loss": 2.5425,
+      "epoch": 0.22,
+      "learning_rate": 9.999458471772225e-06,
+      "loss": 1.5164,
       "step": 270
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 3.810354430715678e-05,
-      "loss": 2.5495,
+      "epoch": 0.22,
+      "learning_rate": 9.999184358531422e-06,
+      "loss": 1.4148,
       "step": 275
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 3.802999301710932e-05,
-      "loss": 2.5375,
+      "epoch": 0.23,
+      "learning_rate": 9.998854310460233e-06,
+      "loss": 1.4273,
       "step": 280
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 3.795511622101516e-05,
-      "loss": 2.5151,
+      "epoch": 0.23,
+      "learning_rate": 9.998468331251499e-06,
+      "loss": 1.4592,
       "step": 285
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 3.787891942357115e-05,
-      "loss": 2.4864,
+      "epoch": 0.23,
+      "learning_rate": 9.998026425223858e-06,
+      "loss": 1.4504,
       "step": 290
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 3.780140822651633e-05,
-      "loss": 2.5396,
+      "epoch": 0.24,
+      "learning_rate": 9.997528597321704e-06,
+      "loss": 1.448,
       "step": 295
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 3.77225883282201e-05,
-      "loss": 2.5585,
+      "epoch": 0.24,
+      "learning_rate": 9.996974853115132e-06,
+      "loss": 1.4523,
       "step": 300
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 3.764246552326328e-05,
-      "loss": 2.4947,
+      "epoch": 0.25,
+      "learning_rate": 9.996365198799868e-06,
+      "loss": 1.5063,
       "step": 305
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 3.756104570201213e-05,
-      "loss": 2.5004,
+      "epoch": 0.25,
+      "learning_rate": 9.99569964119721e-06,
+      "loss": 1.4051,
       "step": 310
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 3.747833485018529e-05,
-      "loss": 2.4729,
+      "epoch": 0.25,
+      "learning_rate": 9.99497818775394e-06,
+      "loss": 1.4312,
       "step": 315
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 3.739433904841375e-05,
-      "loss": 2.4764,
+      "epoch": 0.26,
+      "learning_rate": 9.994200846542251e-06,
+      "loss": 1.4758,
       "step": 320
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 3.7309064471793794e-05,
-      "loss": 2.5369,
+      "epoch": 0.26,
+      "learning_rate": 9.993367626259652e-06,
+      "loss": 1.468,
       "step": 325
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 3.7222517389433085e-05,
-      "loss": 2.4836,
+      "epoch": 0.27,
+      "learning_rate": 9.99247853622887e-06,
+      "loss": 1.527,
       "step": 330
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 3.7134704163989705e-05,
-      "loss": 2.4558,
+      "epoch": 0.27,
+      "learning_rate": 9.991533586397751e-06,
+      "loss": 1.4969,
       "step": 335
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 3.7045631251204434e-05,
-      "loss": 2.5562,
+      "epoch": 0.28,
+      "learning_rate": 9.990532787339137e-06,
+      "loss": 1.4996,
       "step": 340
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 3.6955305199426164e-05,
-      "loss": 2.4603,
+      "epoch": 0.28,
+      "learning_rate": 9.98947615025076e-06,
+      "loss": 1.5262,
       "step": 345
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 3.6863732649130426e-05,
-      "loss": 2.4176,
+      "epoch": 0.28,
+      "learning_rate": 9.988363686955116e-06,
+      "loss": 1.5992,
       "step": 350
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 3.677092033243128e-05,
-      "loss": 2.4991,
+      "epoch": 0.29,
+      "learning_rate": 9.987195409899322e-06,
+      "loss": 1.4711,
       "step": 355
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 3.667687507258631e-05,
-      "loss": 2.4846,
+      "epoch": 0.29,
+      "learning_rate": 9.985971332154985e-06,
+      "loss": 1.45,
       "step": 360
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 3.658160378349508e-05,
-      "loss": 2.5998,
+      "epoch": 0.3,
+      "learning_rate": 9.984691467418057e-06,
+      "loss": 1.4863,
       "step": 365
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 3.648511346919079e-05,
-      "loss": 2.5061,
+      "epoch": 0.3,
+      "learning_rate": 9.983355830008678e-06,
+      "loss": 1.5219,
       "step": 370
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 3.638741122332539e-05,
-      "loss": 2.4538,
+      "epoch": 0.3,
+      "learning_rate": 9.981964434871015e-06,
+      "loss": 1.5977,
       "step": 375
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 3.628850422864807e-05,
-      "loss": 2.498,
+      "epoch": 0.31,
+      "learning_rate": 9.980517297573097e-06,
+      "loss": 1.4539,
       "step": 380
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 3.618839975647718e-05,
-      "loss": 2.4867,
+      "epoch": 0.31,
+      "learning_rate": 9.979014434306642e-06,
+      "loss": 1.3713,
       "step": 385
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 3.608710516616575e-05,
-      "loss": 2.5426,
+      "epoch": 0.32,
+      "learning_rate": 9.977455861886874e-06,
+      "loss": 1.4434,
       "step": 390
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 3.598462790456035e-05,
-      "loss": 2.4842,
+      "epoch": 0.32,
+      "learning_rate": 9.975841597752334e-06,
+      "loss": 1.5469,
       "step": 395
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 3.588097550545368e-05,
-      "loss": 2.4274,
+      "epoch": 0.32,
+      "learning_rate": 9.974171659964688e-06,
+      "loss": 1.4531,
       "step": 400
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 3.5776155589030725e-05,
-      "loss": 2.5354,
+      "epoch": 0.33,
+      "learning_rate": 9.972446067208519e-06,
+      "loss": 1.4828,
       "step": 405
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 3.5670175861308496e-05,
-      "loss": 2.455,
+      "epoch": 0.33,
+      "learning_rate": 9.970664838791126e-06,
+      "loss": 1.4512,
       "step": 410
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 3.556304411356954e-05,
-      "loss": 2.5039,
+      "epoch": 0.34,
+      "learning_rate": 9.9688279946423e-06,
+      "loss": 1.4076,
       "step": 415
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 3.545476822178915e-05,
-      "loss": 2.518,
+      "epoch": 0.34,
+      "learning_rate": 9.966935555314107e-06,
+      "loss": 1.4969,
       "step": 420
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 3.5345356146056326e-05,
-      "loss": 2.4718,
+      "epoch": 0.34,
+      "learning_rate": 9.96498754198066e-06,
+      "loss": 1.498,
       "step": 425
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 3.52348159299886e-05,
-      "loss": 2.4318,
+      "epoch": 0.35,
+      "learning_rate": 9.962983976437868e-06,
+      "loss": 1.4393,
       "step": 430
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 3.512315570014071e-05,
-      "loss": 2.5146,
+      "epoch": 0.35,
+      "learning_rate": 9.96092488110321e-06,
+      "loss": 1.5219,
       "step": 435
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 3.50103836654071e-05,
-      "loss": 2.4801,
+      "epoch": 0.36,
+      "learning_rate": 9.958810279015474e-06,
+      "loss": 1.4484,
       "step": 440
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 3.489650811641849e-05,
-      "loss": 2.4429,
+      "epoch": 0.36,
+      "learning_rate": 9.956640193834501e-06,
+      "loss": 1.4805,
       "step": 445
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 3.478153742493235e-05,
-      "loss": 2.4654,
+      "epoch": 0.36,
+      "learning_rate": 9.954414649840922e-06,
+      "loss": 1.4832,
       "step": 450
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 3.4665480043217444e-05,
-      "loss": 2.4846,
+      "epoch": 0.37,
+      "learning_rate": 9.952133671935885e-06,
+      "loss": 1.3789,
       "step": 455
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 3.454834450343245e-05,
-      "loss": 2.4371,
+      "epoch": 0.37,
+      "learning_rate": 9.949797285640771e-06,
+      "loss": 1.4773,
       "step": 460
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 3.443013941699868e-05,
-      "loss": 2.4071,
+      "epoch": 0.38,
+      "learning_rate": 9.947405517096927e-06,
+      "loss": 1.5246,
       "step": 465
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 3.431087347396702e-05,
-      "loss": 2.3886,
+      "epoch": 0.38,
+      "learning_rate": 9.944958393065343e-06,
+      "loss": 1.5094,
       "step": 470
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 3.419055544237906e-05,
-      "loss": 2.4539,
+      "epoch": 0.38,
+      "learning_rate": 9.942455940926384e-06,
+      "loss": 1.4253,
       "step": 475
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 3.40691941676225e-05,
-      "loss": 2.4406,
+      "epoch": 0.39,
+      "learning_rate": 9.939898188679465e-06,
+      "loss": 1.4396,
       "step": 480
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 3.394679857178086e-05,
-      "loss": 2.3687,
+      "epoch": 0.39,
+      "learning_rate": 9.93728516494274e-06,
+      "loss": 1.3863,
       "step": 485
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 3.382337765297756e-05,
-      "loss": 2.4244,
+      "epoch": 0.4,
+      "learning_rate": 9.934616898952787e-06,
+      "loss": 1.4965,
       "step": 490
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 3.3698940484714394e-05,
-      "loss": 2.4477,
+      "epoch": 0.4,
+      "learning_rate": 9.931893420564277e-06,
+      "loss": 1.5633,
       "step": 495
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 3.35734962152045e-05,
-      "loss": 2.5432,
+      "epoch": 0.4,
+      "learning_rate": 9.929114760249642e-06,
+      "loss": 1.3689,
       "step": 500
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 3.34470540666998e-05,
-      "loss": 2.4633,
+      "epoch": 0.41,
+      "learning_rate": 9.926280949098732e-06,
+      "loss": 1.4434,
       "step": 505
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 3.331962333481302e-05,
-      "loss": 2.4302,
+      "epoch": 0.41,
+      "learning_rate": 9.923392018818467e-06,
+      "loss": 1.4758,
       "step": 510
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 3.319121338783428e-05,
-      "loss": 2.4528,
+      "epoch": 0.42,
+      "learning_rate": 9.92044800173249e-06,
+      "loss": 1.4133,
       "step": 515
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 3.3061833666042416e-05,
-      "loss": 2.3741,
+      "epoch": 0.42,
+      "learning_rate": 9.917448930780786e-06,
+      "loss": 1.451,
       "step": 520
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 3.29314936810109e-05,
-      "loss": 2.4638,
+      "epoch": 0.42,
+      "learning_rate": 9.91439483951934e-06,
+      "loss": 1.5117,
       "step": 525
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 3.280020301490863e-05,
-      "loss": 2.4204,
+      "epoch": 0.43,
+      "learning_rate": 9.91128576211974e-06,
+      "loss": 1.4949,
       "step": 530
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 3.2667971319795473e-05,
-      "loss": 2.436,
+      "epoch": 0.43,
+      "learning_rate": 9.908121733368803e-06,
+      "loss": 1.3984,
       "step": 535
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 3.253480831691264e-05,
-      "loss": 2.4194,
+      "epoch": 0.44,
+      "learning_rate": 9.904902788668187e-06,
+      "loss": 1.6094,
       "step": 540
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 3.240072379596806e-05,
-      "loss": 2.3565,
+      "epoch": 0.44,
+      "learning_rate": 9.901628964033993e-06,
+      "loss": 1.452,
       "step": 545
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 3.226572761441666e-05,
-      "loss": 2.4421,
+      "epoch": 0.44,
+      "learning_rate": 9.89830029609636e-06,
+      "loss": 1.4293,
       "step": 550
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 3.2129829696735636e-05,
-      "loss": 2.4169,
+      "epoch": 0.45,
+      "learning_rate": 9.894916822099062e-06,
+      "loss": 1.4957,
       "step": 555
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 3.1993040033694916e-05,
-      "loss": 2.4425,
+      "epoch": 0.45,
+      "learning_rate": 9.89147857989908e-06,
+      "loss": 1.507,
       "step": 560
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 3.1855368681622584e-05,
-      "loss": 2.399,
+      "epoch": 0.46,
+      "learning_rate": 9.887985607966188e-06,
+      "loss": 1.4379,
       "step": 565
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 3.171682576166565e-05,
-      "loss": 2.3747,
+      "epoch": 0.46,
+      "learning_rate": 9.884437945382523e-06,
+      "loss": 1.4812,
       "step": 570
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 3.1577421459045905e-05,
-      "loss": 2.437,
+      "epoch": 0.47,
+      "learning_rate": 9.880835631842141e-06,
+      "loss": 1.4477,
       "step": 575
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 3.143716602231122e-05,
-      "loss": 2.4131,
+      "epoch": 0.47,
+      "learning_rate": 9.877178707650573e-06,
+      "loss": 1.4484,
       "step": 580
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 3.129606976258201e-05,
-      "loss": 2.4329,
+      "epoch": 0.47,
+      "learning_rate": 9.873467213724384e-06,
+      "loss": 1.4897,
       "step": 585
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 3.115414305279327e-05,
-      "loss": 2.4521,
+      "epoch": 0.48,
+      "learning_rate": 9.869701191590703e-06,
+      "loss": 1.4641,
       "step": 590
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 3.101139632693197e-05,
-      "loss": 2.3317,
+      "epoch": 0.48,
+      "learning_rate": 9.865880683386766e-06,
+      "loss": 1.4809,
       "step": 595
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 3.086784007926996e-05,
-      "loss": 2.4119,
+      "epoch": 0.49,
+      "learning_rate": 9.862005731859443e-06,
+      "loss": 1.4742,
       "step": 600
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 3.072348486359247e-05,
-      "loss": 2.4315,
+      "epoch": 0.49,
+      "learning_rate": 9.858076380364757e-06,
+      "loss": 1.4902,
       "step": 605
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 3.0578341292422286e-05,
-      "loss": 2.4687,
+      "epoch": 0.49,
+      "learning_rate": 9.854092672867399e-06,
+      "loss": 1.5898,
       "step": 610
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 3.043242003623947e-05,
-      "loss": 2.4703,
+      "epoch": 0.5,
+      "learning_rate": 9.850054653940247e-06,
+      "loss": 1.4984,
       "step": 615
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 3.0285731822696954e-05,
-      "loss": 2.5997,
+      "epoch": 0.5,
+      "learning_rate": 9.845962368763847e-06,
+      "loss": 1.4898,
       "step": 620
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 3.0138287435831855e-05,
-      "loss": 2.0103,
+      "epoch": 0.51,
+      "learning_rate": 9.841815863125923e-06,
+      "loss": 1.5082,
       "step": 625
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 2.9990097715272694e-05,
-      "loss": 1.9907,
+      "epoch": 0.51,
+      "learning_rate": 9.837615183420866e-06,
+      "loss": 1.4992,
       "step": 630
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 2.9841173555442463e-05,
-      "loss": 1.9292,
+      "epoch": 0.51,
+      "learning_rate": 9.8333603766492e-06,
+      "loss": 1.3744,
       "step": 635
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 2.9691525904757745e-05,
-      "loss": 1.9898,
+      "epoch": 0.52,
+      "learning_rate": 9.829051490417074e-06,
+      "loss": 1.527,
       "step": 640
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 2.954116576482378e-05,
-      "loss": 1.9234,
+      "epoch": 0.52,
+      "learning_rate": 9.824688572935713e-06,
+      "loss": 1.5008,
       "step": 645
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 2.9390104189625702e-05,
-      "loss": 1.8726,
+      "epoch": 0.53,
+      "learning_rate": 9.820271673020891e-06,
+      "loss": 1.4721,
       "step": 650
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 2.923835228471587e-05,
-      "loss": 1.9208,
+      "epoch": 0.53,
+      "learning_rate": 9.81580084009238e-06,
+      "loss": 1.5555,
       "step": 655
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 2.90859212063974e-05,
-      "loss": 1.9407,
+      "epoch": 0.53,
+      "learning_rate": 9.811276124173395e-06,
+      "loss": 1.5285,
       "step": 660
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 2.8932822160904038e-05,
-      "loss": 1.9377,
+      "epoch": 0.54,
+      "learning_rate": 9.806697575890043e-06,
+      "loss": 1.4777,
       "step": 665
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 2.877906640357628e-05,
-      "loss": 1.9665,
+      "epoch": 0.54,
+      "learning_rate": 9.802065246470738e-06,
+      "loss": 1.4322,
       "step": 670
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 2.862466523803393e-05,
-      "loss": 1.9723,
+      "epoch": 0.55,
+      "learning_rate": 9.797379187745652e-06,
+      "loss": 1.5197,
       "step": 675
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 2.846963001534507e-05,
-      "loss": 1.9876,
+      "epoch": 0.55,
+      "learning_rate": 9.792639452146116e-06,
+      "loss": 1.4703,
       "step": 680
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 2.8313972133191615e-05,
-      "loss": 1.9405,
+      "epoch": 0.55,
+      "learning_rate": 9.787846092704043e-06,
+      "loss": 1.4539,
       "step": 685
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 2.8157703035031353e-05,
-      "loss": 1.9848,
+      "epoch": 0.56,
+      "learning_rate": 9.782999163051332e-06,
+      "loss": 1.4326,
       "step": 690
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 2.8000834209256665e-05,
-      "loss": 1.9328,
+      "epoch": 0.56,
+      "learning_rate": 9.778098717419266e-06,
+      "loss": 1.3992,
       "step": 695
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 2.7843377188349962e-05,
-      "loss": 1.9343,
+      "epoch": 0.57,
+      "learning_rate": 9.773144810637908e-06,
+      "loss": 1.457,
       "step": 700
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 2.768534354803581e-05,
-      "loss": 1.9245,
+      "epoch": 0.57,
+      "learning_rate": 9.768137498135489e-06,
+      "loss": 1.477,
       "step": 705
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 2.752674490642996e-05,
-      "loss": 1.9526,
+      "epoch": 0.57,
+      "learning_rate": 9.763076835937782e-06,
+      "loss": 1.4455,
       "step": 710
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 2.7367592923185207e-05,
-      "loss": 1.93,
+      "epoch": 0.58,
+      "learning_rate": 9.75796288066748e-06,
+      "loss": 1.442,
       "step": 715
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 2.720789929863421e-05,
-      "loss": 1.9263,
+      "epoch": 0.58,
+      "learning_rate": 9.752795689543563e-06,
+      "loss": 1.5156,
       "step": 720
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 2.7047675772929328e-05,
-      "loss": 1.9432,
+      "epoch": 0.59,
+      "learning_rate": 9.747575320380652e-06,
+      "loss": 1.5129,
       "step": 725
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 2.6886934125179504e-05,
-      "loss": 1.9481,
+      "epoch": 0.59,
+      "learning_rate": 9.742301831588368e-06,
+      "loss": 1.3873,
       "step": 730
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 2.672568617258432e-05,
-      "loss": 1.909,
+      "epoch": 0.59,
+      "learning_rate": 9.736975282170678e-06,
+      "loss": 1.448,
       "step": 735
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 2.6563943769565258e-05,
-      "loss": 1.9386,
+      "epoch": 0.6,
+      "learning_rate": 9.731595731725232e-06,
+      "loss": 1.4695,
       "step": 740
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 2.6401718806894144e-05,
-      "loss": 1.9362,
+      "epoch": 0.6,
+      "learning_rate": 9.726163240442695e-06,
+      "loss": 1.4898,
       "step": 745
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 2.6239023210819027e-05,
-      "loss": 1.9494,
+      "epoch": 0.61,
+      "learning_rate": 9.720677869106077e-06,
+      "loss": 1.4619,
       "step": 750
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 2.6075868942187366e-05,
-      "loss": 1.9576,
+      "epoch": 0.61,
+      "learning_rate": 9.715139679090057e-06,
+      "loss": 1.5293,
       "step": 755
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 2.5912267995566746e-05,
-      "loss": 1.937,
+      "epoch": 0.61,
+      "learning_rate": 9.709548732360286e-06,
+      "loss": 1.4535,
       "step": 760
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 2.5748232398363044e-05,
-      "loss": 1.9889,
+      "epoch": 0.62,
+      "learning_rate": 9.703905091472698e-06,
+      "loss": 1.3992,
       "step": 765
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 2.5583774209936218e-05,
-      "loss": 1.9285,
+      "epoch": 0.62,
+      "learning_rate": 9.698208819572815e-06,
+      "loss": 1.5254,
       "step": 770
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 2.5418905520713767e-05,
-      "loss": 1.895,
+      "epoch": 0.63,
+      "learning_rate": 9.692459980395034e-06,
+      "loss": 1.351,
       "step": 775
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 2.525363845130185e-05,
-      "loss": 1.9826,
+      "epoch": 0.63,
+      "learning_rate": 9.686658638261916e-06,
+      "loss": 1.3639,
       "step": 780
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 2.5087985151594235e-05,
-      "loss": 1.9869,
+      "epoch": 0.64,
+      "learning_rate": 9.680804858083468e-06,
+      "loss": 1.3813,
       "step": 785
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 2.4921957799879076e-05,
-      "loss": 1.9325,
+      "epoch": 0.64,
+      "learning_rate": 9.674898705356413e-06,
+      "loss": 1.4016,
       "step": 790
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 2.4755568601943615e-05,
-      "loss": 1.9479,
+      "epoch": 0.64,
+      "learning_rate": 9.668940246163464e-06,
+      "loss": 1.4611,
       "step": 795
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 2.4588829790176837e-05,
-      "loss": 1.9616,
+      "epoch": 0.65,
+      "learning_rate": 9.662929547172575e-06,
+      "loss": 1.4902,
       "step": 800
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 2.4421753622670178e-05,
-      "loss": 1.9706,
+      "epoch": 0.65,
+      "learning_rate": 9.656866675636204e-06,
+      "loss": 1.5445,
       "step": 805
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 2.425435238231638e-05,
-      "loss": 1.9675,
+      "epoch": 0.66,
+      "learning_rate": 9.650751699390554e-06,
+      "loss": 1.5363,
       "step": 810
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 2.4086638375906484e-05,
-      "loss": 1.9684,
+      "epoch": 0.66,
+      "learning_rate": 9.64458468685482e-06,
+      "loss": 1.4508,
       "step": 815
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 2.3918623933225043e-05,
-      "loss": 1.9388,
+      "epoch": 0.66,
+      "learning_rate": 9.638365707030415e-06,
+      "loss": 1.4367,
       "step": 820
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 2.375032140614372e-05,
-      "loss": 1.9326,
+      "epoch": 0.67,
+      "learning_rate": 9.632094829500206e-06,
+      "loss": 1.4594,
       "step": 825
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 2.3581743167713187e-05,
-      "loss": 1.9521,
+      "epoch": 0.67,
+      "learning_rate": 9.62577212442774e-06,
+      "loss": 1.498,
       "step": 830
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 2.3412901611253524e-05,
-      "loss": 1.9704,
+      "epoch": 0.68,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 1.4816,
       "step": 835
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 2.3243809149443077e-05,
-      "loss": 1.89,
+      "epoch": 0.68,
+      "learning_rate": 9.61297151520882e-06,
+      "loss": 1.4754,
       "step": 840
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 2.3074478213405937e-05,
-      "loss": 1.9438,
+      "epoch": 0.68,
+      "learning_rate": 9.606493754285712e-06,
+      "loss": 1.4805,
       "step": 845
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 2.2904921251798052e-05,
-      "loss": 1.9682,
+      "epoch": 0.69,
+      "learning_rate": 9.599964452265434e-06,
+      "loss": 1.5539,
       "step": 850
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 2.2735150729892013e-05,
-      "loss": 2.008,
+      "epoch": 0.69,
+      "learning_rate": 9.593383682202974e-06,
+      "loss": 1.4996,
       "step": 855
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 2.2565179128660667e-05,
-      "loss": 1.9247,
+      "epoch": 0.7,
+      "learning_rate": 9.586751517729203e-06,
+      "loss": 1.484,
       "step": 860
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 2.2395018943859558e-05,
-      "loss": 1.9377,
+      "epoch": 0.7,
+      "learning_rate": 9.580068033050019e-06,
+      "loss": 1.4781,
       "step": 865
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 2.222468268510828e-05,
-      "loss": 1.9396,
+      "epoch": 0.7,
+      "learning_rate": 9.57333330294554e-06,
+      "loss": 1.5156,
       "step": 870
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 2.2054182874970808e-05,
-      "loss": 1.9848,
+      "epoch": 0.71,
+      "learning_rate": 9.566547402769255e-06,
+      "loss": 1.4734,
       "step": 875
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 2.188353204803486e-05,
-      "loss": 1.9382,
+      "epoch": 0.71,
+      "learning_rate": 9.559710408447185e-06,
+      "loss": 1.5398,
       "step": 880
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 2.1712742749990444e-05,
-      "loss": 1.9431,
+      "epoch": 0.72,
+      "learning_rate": 9.55282239647703e-06,
+      "loss": 1.5621,
       "step": 885
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 2.154182753670749e-05,
-      "loss": 1.9833,
+      "epoch": 0.72,
+      "learning_rate": 9.545883443927325e-06,
+      "loss": 1.4793,
       "step": 890
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 2.1370798973312813e-05,
-      "loss": 1.9338,
+      "epoch": 0.72,
+      "learning_rate": 9.538893628436554e-06,
+      "loss": 1.3898,
       "step": 895
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 2.1199669633266353e-05,
-      "loss": 1.9543,
+      "epoch": 0.73,
+      "learning_rate": 9.531853028212308e-06,
+      "loss": 1.4719,
       "step": 900
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 2.102845209743682e-05,
-      "loss": 1.9455,
+      "epoch": 0.73,
+      "learning_rate": 9.52476172203039e-06,
+      "loss": 1.5215,
       "step": 905
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 2.085715895317679e-05,
-      "loss": 1.9533,
+      "epoch": 0.74,
+      "learning_rate": 9.51761978923395e-06,
+      "loss": 1.4264,
       "step": 910
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 2.0685802793397317e-05,
-      "loss": 2.0128,
+      "epoch": 0.74,
+      "learning_rate": 9.51042730973258e-06,
+      "loss": 1.502,
       "step": 915
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 2.051439621564216e-05,
-      "loss": 1.9471,
+      "epoch": 0.74,
+      "learning_rate": 9.503184364001432e-06,
+      "loss": 1.4287,
       "step": 920
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 2.0342951821161648e-05,
-      "loss": 1.9474,
+      "epoch": 0.75,
+      "learning_rate": 9.495891033080315e-06,
+      "loss": 1.4693,
       "step": 925
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 2.017148221398625e-05,
-      "loss": 1.9946,
+      "epoch": 0.75,
+      "learning_rate": 9.488547398572787e-06,
+      "loss": 1.5344,
       "step": 930
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 2e-05,
-      "loss": 1.913,
+      "epoch": 0.76,
+      "learning_rate": 9.48115354264524e-06,
+      "loss": 1.3703,
       "step": 935
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 1.9828517786013752e-05,
-      "loss": 1.981,
+      "epoch": 0.76,
+      "learning_rate": 9.473709548025987e-06,
+      "loss": 1.6094,
       "step": 940
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 1.965704817883836e-05,
-      "loss": 1.9809,
+      "epoch": 0.76,
+      "learning_rate": 9.466215498004328e-06,
+      "loss": 1.4773,
       "step": 945
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 1.948560378435784e-05,
-      "loss": 1.9793,
+      "epoch": 0.77,
+      "learning_rate": 9.458671476429624e-06,
+      "loss": 1.4602,
       "step": 950
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 1.9314197206602693e-05,
-      "loss": 1.9207,
+      "epoch": 0.77,
+      "learning_rate": 9.45107756771036e-06,
+      "loss": 1.4798,
       "step": 955
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 1.914284104682322e-05,
-      "loss": 1.8926,
+      "epoch": 0.78,
+      "learning_rate": 9.443433856813197e-06,
+      "loss": 1.516,
       "step": 960
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 1.897154790256319e-05,
-      "loss": 2.0005,
+      "epoch": 0.78,
+      "learning_rate": 9.435740429262016e-06,
+      "loss": 1.4941,
       "step": 965
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 1.8800330366733654e-05,
-      "loss": 1.9432,
+      "epoch": 0.78,
+      "learning_rate": 9.427997371136976e-06,
+      "loss": 1.402,
       "step": 970
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 1.862920102668719e-05,
-      "loss": 1.8667,
+      "epoch": 0.79,
+      "learning_rate": 9.420204769073538e-06,
+      "loss": 1.5305,
       "step": 975
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 1.8458172463292516e-05,
-      "loss": 1.9405,
+      "epoch": 0.79,
+      "learning_rate": 9.4123627102615e-06,
+      "loss": 1.4604,
       "step": 980
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 1.828725725000956e-05,
-      "loss": 1.9617,
+      "epoch": 0.8,
+      "learning_rate": 9.404471282444019e-06,
+      "loss": 1.5129,
       "step": 985
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 1.8116467951965145e-05,
-      "loss": 1.9447,
+      "epoch": 0.8,
+      "learning_rate": 9.396530573916636e-06,
+      "loss": 1.5453,
       "step": 990
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 1.79458171250292e-05,
-      "loss": 1.9093,
+      "epoch": 0.81,
+      "learning_rate": 9.38854067352628e-06,
+      "loss": 1.4834,
       "step": 995
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 1.7775317314891724e-05,
-      "loss": 1.9051,
+      "epoch": 0.81,
+      "learning_rate": 9.38050167067028e-06,
+      "loss": 1.4387,
       "step": 1000
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 1.7604981056140446e-05,
-      "loss": 1.916,
+      "epoch": 0.81,
+      "learning_rate": 9.372413655295362e-06,
+      "loss": 1.4383,
       "step": 1005
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 1.7434820871339336e-05,
-      "loss": 1.8569,
+      "epoch": 0.82,
+      "learning_rate": 9.364276717896639e-06,
+      "loss": 1.602,
       "step": 1010
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 1.7264849270107994e-05,
-      "loss": 1.9163,
+      "epoch": 0.82,
+      "learning_rate": 9.356090949516608e-06,
+      "loss": 1.4193,
       "step": 1015
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 1.709507874820195e-05,
-      "loss": 1.9342,
+      "epoch": 0.83,
+      "learning_rate": 9.347856441744122e-06,
+      "loss": 1.5031,
       "step": 1020
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 1.6925521786594067e-05,
-      "loss": 1.8947,
+      "epoch": 0.83,
+      "learning_rate": 9.339573286713369e-06,
+      "loss": 1.4828,
       "step": 1025
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 1.675619085055693e-05,
-      "loss": 1.9396,
+      "epoch": 0.83,
+      "learning_rate": 9.331241577102841e-06,
+      "loss": 1.5191,
       "step": 1030
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 1.6587098388746486e-05,
-      "loss": 1.9416,
+      "epoch": 0.84,
+      "learning_rate": 9.322861406134302e-06,
+      "loss": 1.4305,
       "step": 1035
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 1.6418256832286816e-05,
-      "loss": 1.9382,
+      "epoch": 0.84,
+      "learning_rate": 9.314432867571732e-06,
+      "loss": 1.4625,
       "step": 1040
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 1.6249678593856288e-05,
-      "loss": 1.9747,
+      "epoch": 0.85,
+      "learning_rate": 9.30595605572029e-06,
+      "loss": 1.5246,
       "step": 1045
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 1.6081376066774964e-05,
-      "loss": 1.8799,
+      "epoch": 0.85,
+      "learning_rate": 9.297431065425257e-06,
+      "loss": 1.5227,
       "step": 1050
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 1.591336162409352e-05,
-      "loss": 1.8957,
+      "epoch": 0.85,
+      "learning_rate": 9.28885799207097e-06,
+      "loss": 1.5367,
       "step": 1055
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 1.5745647617683627e-05,
-      "loss": 1.8921,
+      "epoch": 0.86,
+      "learning_rate": 9.280236931579754e-06,
+      "loss": 1.4879,
       "step": 1060
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 1.557824637732983e-05,
-      "loss": 1.9406,
+      "epoch": 0.86,
+      "learning_rate": 9.271567980410859e-06,
+      "loss": 1.5137,
       "step": 1065
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 1.5411170209823177e-05,
-      "loss": 1.9282,
+      "epoch": 0.87,
+      "learning_rate": 9.26285123555937e-06,
+      "loss": 1.4449,
       "step": 1070
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 1.5244431398056392e-05,
-      "loss": 1.8621,
+      "epoch": 0.87,
+      "learning_rate": 9.254086794555121e-06,
+      "loss": 1.4602,
       "step": 1075
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 1.5078042200120933e-05,
-      "loss": 1.9375,
+      "epoch": 0.87,
+      "learning_rate": 9.245274755461621e-06,
+      "loss": 1.4187,
       "step": 1080
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 1.4912014848405771e-05,
-      "loss": 1.8779,
+      "epoch": 0.88,
+      "learning_rate": 9.23641521687493e-06,
+      "loss": 1.5391,
       "step": 1085
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 1.4746361548698151e-05,
-      "loss": 1.9353,
+      "epoch": 0.88,
+      "learning_rate": 9.227508277922579e-06,
+      "loss": 1.3988,
       "step": 1090
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 1.4581094479286234e-05,
-      "loss": 1.9255,
+      "epoch": 0.89,
+      "learning_rate": 9.218554038262448e-06,
+      "loss": 1.5984,
       "step": 1095
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 1.4416225790063784e-05,
-      "loss": 1.9163,
+      "epoch": 0.89,
+      "learning_rate": 9.209552598081657e-06,
+      "loss": 1.5109,
       "step": 1100
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 1.4251767601636965e-05,
-      "loss": 1.9314,
+      "epoch": 0.89,
+      "learning_rate": 9.200504058095439e-06,
+      "loss": 1.5418,
       "step": 1105
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 1.4087732004433258e-05,
-      "loss": 1.8751,
+      "epoch": 0.9,
+      "learning_rate": 9.191408519546022e-06,
+      "loss": 1.4275,
       "step": 1110
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 1.3924131057812642e-05,
-      "loss": 1.8934,
+      "epoch": 0.9,
+      "learning_rate": 9.182266084201486e-06,
+      "loss": 1.4074,
       "step": 1115
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 1.376097678918098e-05,
-      "loss": 1.9148,
+      "epoch": 0.91,
+      "learning_rate": 9.173076854354634e-06,
+      "loss": 1.5016,
       "step": 1120
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 1.3598281193105858e-05,
-      "loss": 1.8754,
+      "epoch": 0.91,
+      "learning_rate": 9.16384093282184e-06,
+      "loss": 1.5188,
       "step": 1125
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 1.3436056230434747e-05,
-      "loss": 1.9183,
+      "epoch": 0.91,
+      "learning_rate": 9.154558422941901e-06,
+      "loss": 1.4738,
       "step": 1130
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 1.3274313827415678e-05,
-      "loss": 1.9236,
+      "epoch": 0.92,
+      "learning_rate": 9.145229428574886e-06,
+      "loss": 1.4049,
       "step": 1135
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 1.3113065874820506e-05,
-      "loss": 1.889,
+      "epoch": 0.92,
+      "learning_rate": 9.135854054100961e-06,
+      "loss": 1.4363,
       "step": 1140
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 1.295232422707068e-05,
-      "loss": 1.8898,
+      "epoch": 0.93,
+      "learning_rate": 9.126432404419239e-06,
+      "loss": 1.5211,
       "step": 1145
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 1.2792100701365794e-05,
-      "loss": 1.8991,
+      "epoch": 0.93,
+      "learning_rate": 9.11696458494659e-06,
+      "loss": 1.457,
       "step": 1150
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 1.2632407076814794e-05,
-      "loss": 1.9559,
+      "epoch": 0.93,
+      "learning_rate": 9.107450701616469e-06,
+      "loss": 1.4543,
       "step": 1155
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 1.2473255093570039e-05,
-      "loss": 1.9048,
+      "epoch": 0.94,
+      "learning_rate": 9.097890860877732e-06,
+      "loss": 1.534,
       "step": 1160
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 1.2314656451964196e-05,
-      "loss": 1.859,
+      "epoch": 0.94,
+      "learning_rate": 9.088285169693442e-06,
+      "loss": 1.5254,
       "step": 1165
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 1.2156622811650043e-05,
-      "loss": 1.8825,
+      "epoch": 0.95,
+      "learning_rate": 9.078633735539673e-06,
+      "loss": 1.5371,
       "step": 1170
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 1.1999165790743338e-05,
-      "loss": 1.9094,
+      "epoch": 0.95,
+      "learning_rate": 9.068936666404307e-06,
+      "loss": 1.475,
       "step": 1175
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 1.1842296964968652e-05,
-      "loss": 1.937,
+      "epoch": 0.95,
+      "learning_rate": 9.059194070785823e-06,
+      "loss": 1.457,
       "step": 1180
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 1.1686027866808394e-05,
-      "loss": 1.8838,
+      "epoch": 0.96,
+      "learning_rate": 9.049406057692097e-06,
+      "loss": 1.4891,
       "step": 1185
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 1.1530369984654936e-05,
-      "loss": 1.9023,
+      "epoch": 0.96,
+      "learning_rate": 9.03957273663916e-06,
+      "loss": 1.5648,
       "step": 1190
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 1.1375334761966074e-05,
-      "loss": 1.9099,
+      "epoch": 0.97,
+      "learning_rate": 9.02969421764999e-06,
+      "loss": 1.5855,
       "step": 1195
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 1.122093359642372e-05,
-      "loss": 1.9058,
+      "epoch": 0.97,
+      "learning_rate": 9.019770611253272e-06,
+      "loss": 1.4734,
       "step": 1200
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 1.1067177839095957e-05,
-      "loss": 1.9359,
+      "epoch": 0.97,
+      "learning_rate": 9.009802028482169e-06,
+      "loss": 1.4867,
       "step": 1205
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 1.0914078793602601e-05,
-      "loss": 1.8897,
+      "epoch": 0.98,
+      "learning_rate": 8.999788580873074e-06,
+      "loss": 1.5094,
       "step": 1210
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 1.0761647715284139e-05,
-      "loss": 1.9341,
+      "epoch": 0.98,
+      "learning_rate": 8.989730380464362e-06,
+      "loss": 1.4965,
       "step": 1215
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 1.0609895810374304e-05,
-      "loss": 1.876,
+      "epoch": 0.99,
+      "learning_rate": 8.979627539795136e-06,
+      "loss": 1.4887,
       "step": 1220
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 1.0458834235176225e-05,
-      "loss": 1.8287,
+      "epoch": 0.99,
+      "learning_rate": 8.969480171903973e-06,
+      "loss": 1.4398,
       "step": 1225
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 1.0308474095242267e-05,
-      "loss": 1.8523,
+      "epoch": 1.0,
+      "learning_rate": 8.959288390327656e-06,
+      "loss": 1.4301,
       "step": 1230
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 1.0128983382202781e-05,
-      "loss": 2.0887,
+      "epoch": 1.0,
+      "learning_rate": 8.949052309099897e-06,
+      "loss": 1.5309,
       "step": 1235
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 9.980205236069665e-06,
-      "loss": 1.4855,
+      "epoch": 1.0,
+      "learning_rate": 8.938772042750078e-06,
+      "loss": 1.3054,
       "step": 1240
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 9.832163712437392e-06,
-      "loss": 1.4915,
+      "epoch": 1.01,
+      "learning_rate": 8.928447706301951e-06,
+      "loss": 1.2152,
       "step": 1245
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 9.684869694834003e-06,
-      "loss": 1.4679,
+      "epoch": 1.01,
+      "learning_rate": 8.91807941527236e-06,
+      "loss": 1.2238,
       "step": 1250
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 9.538334011833363e-06,
-      "loss": 1.4298,
+      "epoch": 1.02,
+      "learning_rate": 8.907667285669955e-06,
+      "loss": 1.1881,
       "step": 1255
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 9.392567436259034e-06,
-      "loss": 1.4018,
+      "epoch": 1.02,
+      "learning_rate": 8.897211433993873e-06,
+      "loss": 1.234,
       "step": 1260
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 9.247580684392345e-06,
-      "loss": 1.4642,
+      "epoch": 1.02,
+      "learning_rate": 8.886711977232463e-06,
+      "loss": 1.1724,
       "step": 1265
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 9.10338441518453e-06,
-      "loss": 1.4434,
+      "epoch": 1.03,
+      "learning_rate": 8.87616903286195e-06,
+      "loss": 1.3313,
       "step": 1270
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 8.959989229473125e-06,
-      "loss": 1.4574,
+      "epoch": 1.03,
+      "learning_rate": 8.865582718845142e-06,
+      "loss": 1.2277,
       "step": 1275
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 8.817405669202619e-06,
-      "loss": 1.4256,
+      "epoch": 1.04,
+      "learning_rate": 8.854953153630097e-06,
+      "loss": 1.2145,
       "step": 1280
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 8.675644216649478e-06,
-      "loss": 1.4539,
+      "epoch": 1.04,
+      "learning_rate": 8.844280456148799e-06,
+      "loss": 1.243,
       "step": 1285
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 8.534715293651492e-06,
-      "loss": 1.5016,
+      "epoch": 1.04,
+      "learning_rate": 8.833564745815835e-06,
+      "loss": 1.1816,
       "step": 1290
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 8.39462926084159e-06,
-      "loss": 1.4738,
+      "epoch": 1.05,
+      "learning_rate": 8.82280614252705e-06,
+      "loss": 1.1965,
       "step": 1295
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 8.255396416886194e-06,
-      "loss": 1.4265,
+      "epoch": 1.05,
+      "learning_rate": 8.81200476665821e-06,
+      "loss": 1.2035,
       "step": 1300
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 8.117026997728079e-06,
-      "loss": 1.4235,
+      "epoch": 1.06,
+      "learning_rate": 8.801160739063657e-06,
+      "loss": 1.2477,
       "step": 1305
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 7.979531175833828e-06,
-      "loss": 1.5084,
+      "epoch": 1.06,
+      "learning_rate": 8.790274181074951e-06,
+      "loss": 1.1686,
       "step": 1310
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 7.842919059446046e-06,
-      "loss": 1.4426,
+      "epoch": 1.06,
+      "learning_rate": 8.779345214499517e-06,
+      "loss": 1.1877,
       "step": 1315
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 7.707200691840173e-06,
-      "loss": 1.4797,
+      "epoch": 1.07,
+      "learning_rate": 8.768373961619283e-06,
+      "loss": 1.2209,
       "step": 1320
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 7.572386050586196e-06,
-      "loss": 1.4309,
+      "epoch": 1.07,
+      "learning_rate": 8.757360545189308e-06,
+      "loss": 1.2066,
       "step": 1325
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 7.438485046815078e-06,
-      "loss": 1.4505,
+      "epoch": 1.08,
+      "learning_rate": 8.746305088436406e-06,
+      "loss": 1.2484,
       "step": 1330
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 7.305507524490145e-06,
-      "loss": 1.4734,
+      "epoch": 1.08,
+      "learning_rate": 8.735207715057779e-06,
+      "loss": 1.2068,
       "step": 1335
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 7.1734632596834106e-06,
-      "loss": 1.397,
+      "epoch": 1.08,
+      "learning_rate": 8.724068549219618e-06,
+      "loss": 1.1803,
       "step": 1340
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 7.042361959856825e-06,
-      "loss": 1.4341,
+      "epoch": 1.09,
+      "learning_rate": 8.712887715555728e-06,
+      "loss": 1.234,
       "step": 1345
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 6.912213263148673e-06,
-      "loss": 1.4599,
+      "epoch": 1.09,
+      "learning_rate": 8.701665339166122e-06,
+      "loss": 1.2441,
       "step": 1350
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 6.783026737664942e-06,
-      "loss": 1.4466,
+      "epoch": 1.1,
+      "learning_rate": 8.690401545615626e-06,
+      "loss": 1.2082,
       "step": 1355
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 6.654811880775973e-06,
-      "loss": 1.4435,
+      "epoch": 1.1,
+      "learning_rate": 8.679096460932477e-06,
+      "loss": 1.2176,
       "step": 1360
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 6.527578118418187e-06,
-      "loss": 1.4597,
+      "epoch": 1.1,
+      "learning_rate": 8.667750211606906e-06,
+      "loss": 1.2516,
       "step": 1365
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 6.401334804401171e-06,
-      "loss": 1.4217,
+      "epoch": 1.11,
+      "learning_rate": 8.65636292458973e-06,
+      "loss": 1.1766,
       "step": 1370
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 6.276091219719984e-06,
-      "loss": 1.4477,
+      "epoch": 1.11,
+      "learning_rate": 8.644934727290927e-06,
+      "loss": 1.2277,
       "step": 1375
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 6.151856571872854e-06,
-      "loss": 1.4716,
+      "epoch": 1.12,
+      "learning_rate": 8.63346574757821e-06,
+      "loss": 1.2773,
       "step": 1380
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 6.028639994184277e-06,
-      "loss": 1.4398,
+      "epoch": 1.12,
+      "learning_rate": 8.621956113775601e-06,
+      "loss": 1.2162,
       "step": 1385
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 5.906450545133564e-06,
-      "loss": 1.4442,
+      "epoch": 1.12,
+      "learning_rate": 8.610405954661988e-06,
+      "loss": 1.2551,
       "step": 1390
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 5.785297207688905e-06,
-      "loss": 1.4506,
+      "epoch": 1.13,
+      "learning_rate": 8.598815399469694e-06,
+      "loss": 1.2625,
       "step": 1395
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 5.665188888646935e-06,
-      "loss": 1.4123,
+      "epoch": 1.13,
+      "learning_rate": 8.587184577883018e-06,
+      "loss": 1.2465,
       "step": 1400
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 5.546134417977984e-06,
-      "loss": 1.456,
+      "epoch": 1.14,
+      "learning_rate": 8.5755136200368e-06,
+      "loss": 1.2008,
       "step": 1405
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 5.428142548176876e-06,
-      "loss": 1.4274,
+      "epoch": 1.14,
+      "learning_rate": 8.563802656514946e-06,
+      "loss": 1.1623,
       "step": 1410
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 5.311221953619514e-06,
-      "loss": 1.4062,
+      "epoch": 1.14,
+      "learning_rate": 8.552051818348986e-06,
+      "loss": 1.1625,
       "step": 1415
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 5.195381229925156e-06,
-      "loss": 1.427,
+      "epoch": 1.15,
+      "learning_rate": 8.540261237016597e-06,
+      "loss": 1.1723,
       "step": 1420
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 5.080628893324475e-06,
-      "loss": 1.4783,
+      "epoch": 1.15,
+      "learning_rate": 8.528431044440127e-06,
+      "loss": 1.268,
       "step": 1425
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 4.9669733800334955e-06,
-      "loss": 1.4356,
+      "epoch": 1.16,
+      "learning_rate": 8.516561372985137e-06,
+      "loss": 1.2488,
       "step": 1430
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 4.854423045633392e-06,
-      "loss": 1.4809,
+      "epoch": 1.16,
+      "learning_rate": 8.504652355458901e-06,
+      "loss": 1.298,
       "step": 1435
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 4.742986164456196e-06,
-      "loss": 1.4079,
+      "epoch": 1.17,
+      "learning_rate": 8.492704125108933e-06,
+      "loss": 1.2168,
       "step": 1440
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 4.632670928976501e-06,
-      "loss": 1.4884,
+      "epoch": 1.17,
+      "learning_rate": 8.480716815621486e-06,
+      "loss": 1.2166,
       "step": 1445
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 4.523485449209195e-06,
-      "loss": 1.4499,
+      "epoch": 1.17,
+      "learning_rate": 8.468690561120064e-06,
+      "loss": 1.201,
       "step": 1450
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 4.415437752113223e-06,
-      "loss": 1.4065,
+      "epoch": 1.18,
+      "learning_rate": 8.456625496163921e-06,
+      "loss": 1.2266,
       "step": 1455
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 4.308535781001457e-06,
-      "loss": 1.4888,
+      "epoch": 1.18,
+      "learning_rate": 8.444521755746547e-06,
+      "loss": 1.1812,
       "step": 1460
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 4.202787394956769e-06,
-      "loss": 1.4707,
+      "epoch": 1.19,
+      "learning_rate": 8.43237947529417e-06,
+      "loss": 1.1762,
       "step": 1465
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 4.0982003682542146e-06,
-      "loss": 1.4426,
+      "epoch": 1.19,
+      "learning_rate": 8.420198790664232e-06,
+      "loss": 1.2473,
       "step": 1470
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 3.994782389789535e-06,
-      "loss": 1.3991,
+      "epoch": 1.19,
+      "learning_rate": 8.407979838143869e-06,
+      "loss": 1.1887,
       "step": 1475
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 3.892541062513853e-06,
-      "loss": 1.4187,
+      "epoch": 1.2,
+      "learning_rate": 8.395722754448392e-06,
+      "loss": 1.2277,
       "step": 1480
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 3.7914839028747507e-06,
-      "loss": 1.4248,
+      "epoch": 1.2,
+      "learning_rate": 8.38342767671975e-06,
+      "loss": 1.2418,
       "step": 1485
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 3.691618340263701e-06,
-      "loss": 1.447,
+      "epoch": 1.21,
+      "learning_rate": 8.371094742525006e-06,
+      "loss": 1.2081,
       "step": 1490
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 3.5929517164698436e-06,
-      "loss": 1.4394,
+      "epoch": 1.21,
+      "learning_rate": 8.358724089854784e-06,
+      "loss": 1.3461,
       "step": 1495
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 3.495491285140282e-06,
-      "loss": 1.4359,
+      "epoch": 1.21,
+      "learning_rate": 8.346315857121732e-06,
+      "loss": 1.1977,
       "step": 1500
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 3.399244211246779e-06,
-      "loss": 1.4752,
+      "epoch": 1.22,
+      "learning_rate": 8.33387018315898e-06,
+      "loss": 1.2336,
       "step": 1505
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 3.304217570559052e-06,
-      "loss": 1.4508,
+      "epoch": 1.22,
+      "learning_rate": 8.321387207218578e-06,
+      "loss": 1.249,
       "step": 1510
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 3.2104183491245466e-06,
-      "loss": 1.4718,
+      "epoch": 1.23,
+      "learning_rate": 8.308867068969933e-06,
+      "loss": 1.2188,
       "step": 1515
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 3.117853442754879e-06,
-      "loss": 1.4514,
+      "epoch": 1.23,
+      "learning_rate": 8.296309908498264e-06,
+      "loss": 1.1823,
       "step": 1520
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 3.026529656518864e-06,
-      "loss": 1.399,
+      "epoch": 1.23,
+      "learning_rate": 8.283715866303016e-06,
+      "loss": 1.2462,
       "step": 1525
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.936453704242215e-06,
-      "loss": 1.4136,
+      "epoch": 1.24,
+      "learning_rate": 8.271085083296295e-06,
+      "loss": 1.218,
       "step": 1530
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.8476322080139862e-06,
-      "loss": 1.4474,
+      "epoch": 1.24,
+      "learning_rate": 8.258417700801301e-06,
+      "loss": 1.249,
       "step": 1535
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.760071697699729e-06,
-      "loss": 1.4542,
+      "epoch": 1.25,
+      "learning_rate": 8.245713860550734e-06,
+      "loss": 1.2629,
       "step": 1540
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.673778610461448e-06,
-      "loss": 1.4176,
+      "epoch": 1.25,
+      "learning_rate": 8.232973704685208e-06,
+      "loss": 1.2605,
       "step": 1545
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.588759290284337e-06,
-      "loss": 1.4471,
+      "epoch": 1.25,
+      "learning_rate": 8.220197375751667e-06,
+      "loss": 1.2232,
       "step": 1550
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.505019987510426e-06,
-      "loss": 1.4217,
+      "epoch": 1.26,
+      "learning_rate": 8.207385016701792e-06,
+      "loss": 1.2242,
       "step": 1555
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.4225668583790474e-06,
-      "loss": 1.4194,
+      "epoch": 1.26,
+      "learning_rate": 8.194536770890392e-06,
+      "loss": 1.1824,
       "step": 1560
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.3414059645742504e-06,
-      "loss": 1.3959,
+      "epoch": 1.27,
+      "learning_rate": 8.181652782073808e-06,
+      "loss": 1.275,
       "step": 1565
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.261543272779192e-06,
-      "loss": 1.4689,
+      "epoch": 1.27,
+      "learning_rate": 8.168733194408302e-06,
+      "loss": 1.2164,
       "step": 1570
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.1829846542374565e-06,
-      "loss": 1.4568,
+      "epoch": 1.27,
+      "learning_rate": 8.155778152448443e-06,
+      "loss": 1.2207,
       "step": 1575
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.105735884321436e-06,
-      "loss": 1.451,
+      "epoch": 1.28,
+      "learning_rate": 8.142787801145495e-06,
+      "loss": 1.2266,
       "step": 1580
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.029802642107734e-06,
-      "loss": 1.4418,
+      "epoch": 1.28,
+      "learning_rate": 8.129762285845784e-06,
+      "loss": 1.1971,
       "step": 1585
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 1.9551905099596813e-06,
-      "loss": 1.4619,
+      "epoch": 1.29,
+      "learning_rate": 8.116701752289084e-06,
+      "loss": 1.2107,
       "step": 1590
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 1.8819049731169059e-06,
-      "loss": 1.4182,
+      "epoch": 1.29,
+      "learning_rate": 8.103606346606978e-06,
+      "loss": 1.1904,
       "step": 1595
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 1.809951419292104e-06,
-      "loss": 1.4095,
+      "epoch": 1.29,
+      "learning_rate": 8.090476215321226e-06,
+      "loss": 1.2895,
       "step": 1600
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 1.7393351382749424e-06,
-      "loss": 1.4397,
+      "epoch": 1.3,
+      "learning_rate": 8.07731150534213e-06,
+      "loss": 1.235,
       "step": 1605
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 1.6700613215431549e-06,
-      "loss": 1.4747,
+      "epoch": 1.3,
+      "learning_rate": 8.064112363966877e-06,
+      "loss": 1.2238,
       "step": 1610
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 1.6021350618809184e-06,
-      "loss": 1.4356,
+      "epoch": 1.31,
+      "learning_rate": 8.050878938877908e-06,
+      "loss": 1.2535,
       "step": 1615
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 1.5355613530044089e-06,
-      "loss": 1.4381,
+      "epoch": 1.31,
+      "learning_rate": 8.037611378141257e-06,
+      "loss": 1.2336,
       "step": 1620
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 1.470345089194709e-06,
-      "loss": 1.4444,
+      "epoch": 1.31,
+      "learning_rate": 8.024309830204888e-06,
+      "loss": 1.2285,
       "step": 1625
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 1.4064910649379803e-06,
-      "loss": 1.469,
+      "epoch": 1.32,
+      "learning_rate": 8.010974443897046e-06,
+      "loss": 1.25,
       "step": 1630
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 1.3440039745729894e-06,
-      "loss": 1.4427,
+      "epoch": 1.32,
+      "learning_rate": 7.997605368424585e-06,
+      "loss": 1.2492,
       "step": 1635
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 1.2828884119460105e-06,
-      "loss": 1.3941,
+      "epoch": 1.33,
+      "learning_rate": 7.9842027533713e-06,
+      "loss": 1.1783,
       "step": 1640
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 1.2231488700730742e-06,
-      "loss": 1.4452,
+      "epoch": 1.33,
+      "learning_rate": 7.970766748696254e-06,
+      "loss": 1.171,
       "step": 1645
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 1.1647897408096886e-06,
-      "loss": 1.4236,
+      "epoch": 1.33,
+      "learning_rate": 7.9572975047321e-06,
+      "loss": 1.2336,
       "step": 1650
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 1.107815314527929e-06,
-      "loss": 1.4538,
+      "epoch": 1.34,
+      "learning_rate": 7.943795172183394e-06,
+      "loss": 1.204,
       "step": 1655
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 1.0522297798010594e-06,
-      "loss": 1.4112,
+      "epoch": 1.34,
+      "learning_rate": 7.93025990212492e-06,
+      "loss": 1.2342,
       "step": 1660
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 9.980372230955693e-07,
-      "loss": 1.4808,
+      "epoch": 1.35,
+      "learning_rate": 7.916691845999986e-06,
+      "loss": 1.1936,
       "step": 1665
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 9.452416284707743e-07,
-      "loss": 1.4509,
+      "epoch": 1.35,
+      "learning_rate": 7.903091155618747e-06,
+      "loss": 1.1992,
       "step": 1670
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 8.938468772859132e-07,
-      "loss": 1.4414,
+      "epoch": 1.36,
+      "learning_rate": 7.889457983156484e-06,
+      "loss": 1.1707,
       "step": 1675
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 8.438567479147975e-07,
-      "loss": 1.4203,
+      "epoch": 1.36,
+      "learning_rate": 7.875792481151916e-06,
+      "loss": 1.1835,
       "step": 1680
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 7.952749154680405e-07,
-      "loss": 1.4294,
+      "epoch": 1.36,
+      "learning_rate": 7.862094802505498e-06,
+      "loss": 1.2359,
       "step": 1685
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 7.481049515228811e-07,
-      "loss": 1.4136,
+      "epoch": 1.37,
+      "learning_rate": 7.848365100477695e-06,
+      "loss": 1.2754,
       "step": 1690
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 7.023503238606122e-07,
-      "loss": 1.4316,
+      "epoch": 1.37,
+      "learning_rate": 7.834603528687277e-06,
+      "loss": 1.2664,
       "step": 1695
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 6.580143962116281e-07,
-      "loss": 1.4645,
+      "epoch": 1.38,
+      "learning_rate": 7.8208102411096e-06,
+      "loss": 1.2535,
       "step": 1700
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 6.151004280081574e-07,
-      "loss": 1.4692,
+      "epoch": 1.38,
+      "learning_rate": 7.806985392074877e-06,
+      "loss": 1.2158,
       "step": 1705
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 5.736115741446146e-07,
-      "loss": 1.4408,
+      "epoch": 1.38,
+      "learning_rate": 7.793129136266464e-06,
+      "loss": 1.2504,
       "step": 1710
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 5.335508847456794e-07,
-      "loss": 1.4552,
+      "epoch": 1.39,
+      "learning_rate": 7.779241628719108e-06,
+      "loss": 1.1505,
       "step": 1715
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 4.949213049420576e-07,
-      "loss": 1.4657,
+      "epoch": 1.39,
+      "learning_rate": 7.765323024817237e-06,
+      "loss": 1.2695,
       "step": 1720
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 4.577256746539638e-07,
-      "loss": 1.4189,
+      "epoch": 1.4,
+      "learning_rate": 7.751373480293205e-06,
+      "loss": 1.2059,
       "step": 1725
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 4.2196672838233257e-07,
-      "loss": 1.4573,
+      "epoch": 1.4,
+      "learning_rate": 7.737393151225555e-06,
+      "loss": 1.2547,
       "step": 1730
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 3.876470950078037e-07,
-      "loss": 1.4382,
+      "epoch": 1.4,
+      "learning_rate": 7.723382194037266e-06,
+      "loss": 1.2127,
       "step": 1735
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 3.5476929759743927e-07,
-      "loss": 1.4272,
+      "epoch": 1.41,
+      "learning_rate": 7.709340765494017e-06,
+      "loss": 1.1734,
       "step": 1740
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 3.233357532192494e-07,
-      "loss": 1.4866,
+      "epoch": 1.41,
+      "learning_rate": 7.695269022702425e-06,
+      "loss": 1.2037,
       "step": 1745
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.933487727644813e-07,
-      "loss": 1.4132,
+      "epoch": 1.42,
+      "learning_rate": 7.681167123108277e-06,
+      "loss": 1.2891,
       "step": 1750
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.648105607777507e-07,
-      "loss": 1.4498,
+      "epoch": 1.42,
+      "learning_rate": 7.667035224494787e-06,
+      "loss": 1.2285,
       "step": 1755
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.3772321529494712e-07,
-      "loss": 1.4505,
+      "epoch": 1.42,
+      "learning_rate": 7.65287348498082e-06,
+      "loss": 1.217,
       "step": 1760
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.1208872768901713e-07,
-      "loss": 1.4338,
+      "epoch": 1.43,
+      "learning_rate": 7.63868206301912e-06,
+      "loss": 1.1856,
       "step": 1765
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 1.8790898252354583e-07,
-      "loss": 1.4299,
+      "epoch": 1.43,
+      "learning_rate": 7.62446111739455e-06,
+      "loss": 1.2613,
       "step": 1770
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 1.6518575741421904e-07,
-      "loss": 1.4378,
+      "epoch": 1.44,
+      "learning_rate": 7.6102108072223e-06,
+      "loss": 1.1617,
       "step": 1775
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 1.4392072289814319e-07,
-      "loss": 1.4323,
+      "epoch": 1.44,
+      "learning_rate": 7.595931291946116e-06,
+      "loss": 1.2006,
       "step": 1780
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.241154423110169e-07,
-      "loss": 1.4144,
+      "epoch": 1.44,
+      "learning_rate": 7.581622731336515e-06,
+      "loss": 1.2543,
       "step": 1785
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.0577137167221863e-07,
-      "loss": 1.4343,
+      "epoch": 1.45,
+      "learning_rate": 7.567285285488994e-06,
+      "loss": 1.2498,
       "step": 1790
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 8.88898595777543e-08,
-      "loss": 1.4625,
+      "epoch": 1.45,
+      "learning_rate": 7.552919114822246e-06,
+      "loss": 1.2484,
       "step": 1795
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 7.347214710111239e-08,
-      "loss": 1.3614,
+      "epoch": 1.46,
+      "learning_rate": 7.5385243800763505e-06,
+      "loss": 1.2543,
       "step": 1800
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 5.951936770202782e-08,
-      "loss": 1.4099,
+      "epoch": 1.46,
+      "learning_rate": 7.524101242310993e-06,
+      "loss": 1.2621,
       "step": 1805
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 4.7032547143155417e-08,
-      "loss": 1.4601,
+      "epoch": 1.46,
+      "learning_rate": 7.509649862903652e-06,
+      "loss": 1.2176,
       "step": 1810
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 3.60126034146524e-08,
-      "loss": 1.4231,
+      "epoch": 1.47,
+      "learning_rate": 7.495170403547797e-06,
+      "loss": 1.2189,
       "step": 1815
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 2.6460346666696835e-08,
-      "loss": 1.4549,
+      "epoch": 1.47,
+      "learning_rate": 7.480663026251073e-06,
+      "loss": 1.2503,
       "step": 1820
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8376479149926353e-08,
-      "loss": 1.4122,
+      "epoch": 1.48,
+      "learning_rate": 7.466127893333498e-06,
+      "loss": 1.2186,
       "step": 1825
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.176159516380837e-08,
-      "loss": 1.3961,
+      "epoch": 1.48,
+      "learning_rate": 7.451565167425642e-06,
+      "loss": 1.2805,
       "step": 1830
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 6.616181012955025e-09,
-      "loss": 1.4652,
+      "epoch": 1.48,
+      "learning_rate": 7.436975011466805e-06,
+      "loss": 1.2347,
       "step": 1835
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 2.9406149713628874e-09,
-      "loss": 1.425,
+      "epoch": 1.49,
+      "learning_rate": 7.422357588703195e-06,
+      "loss": 1.266,
       "step": 1840
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 7.351672546129785e-10,
-      "loss": 1.4916,
+      "epoch": 1.49,
+      "learning_rate": 7.407713062686107e-06,
+      "loss": 1.2496,
       "step": 1845
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 0.0,
-      "loss": 1.3883,
+      "epoch": 1.5,
+      "learning_rate": 7.393041597270085e-06,
+      "loss": 1.2902,
       "step": 1850
     },
     {
-      "epoch": 3.0,
-      "step": 1851,
-      "total_flos": 1.8340911973547377e+18,
-      "train_loss": 2.0081952639620475,
-      "train_runtime": 43810.4873,
-      "train_samples_per_second": 5.415,
-      "train_steps_per_second": 0.042
+      "epoch": 1.5,
+      "learning_rate": 7.378343356611093e-06,
+      "loss": 1.2367,
+      "step": 1855
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 7.363618505164678e-06,
+      "loss": 1.274,
+      "step": 1860
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 7.348867207684132e-06,
+      "loss": 1.2242,
+      "step": 1865
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 7.334089629218639e-06,
+      "loss": 1.2844,
+      "step": 1870
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 7.319285935111444e-06,
+      "loss": 1.2672,
+      "step": 1875
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 7.304456290997991e-06,
+      "loss": 1.1542,
+      "step": 1880
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.289600862804069e-06,
+      "loss": 1.15,
+      "step": 1885
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.274719816743967e-06,
+      "loss": 1.2385,
+      "step": 1890
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.259813319318601e-06,
+      "loss": 1.2348,
+      "step": 1895
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.244881537313664e-06,
+      "loss": 1.2578,
+      "step": 1900
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.229924637797742e-06,
+      "loss": 1.2191,
+      "step": 1905
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.214942788120466e-06,
+      "loss": 1.251,
+      "step": 1910
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.1999361559106225e-06,
+      "loss": 1.2031,
+      "step": 1915
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.184904909074293e-06,
+      "loss": 1.2766,
+      "step": 1920
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.169849215792955e-06,
+      "loss": 1.2299,
+      "step": 1925
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.15476924452162e-06,
+      "loss": 1.2355,
+      "step": 1930
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 7.139665163986938e-06,
+      "loss": 1.2336,
+      "step": 1935
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 7.124537143185317e-06,
+      "loss": 1.3566,
+      "step": 1940
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 7.109385351381022e-06,
+      "loss": 1.1423,
+      "step": 1945
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 7.09420995810429e-06,
+      "loss": 1.2576,
+      "step": 1950
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 7.079011133149427e-06,
+      "loss": 1.2563,
+      "step": 1955
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 7.0637890465729165e-06,
+      "loss": 1.2695,
+      "step": 1960
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 7.048543868691506e-06,
+      "loss": 1.1986,
+      "step": 1965
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 7.033275770080309e-06,
+      "loss": 1.25,
+      "step": 1970
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 7.017984921570895e-06,
+      "loss": 1.2025,
+      "step": 1975
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 7.002671494249376e-06,
+      "loss": 1.2465,
+      "step": 1980
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 6.987335659454493e-06,
+      "loss": 1.2336,
+      "step": 1985
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 6.971977588775703e-06,
+      "loss": 1.2436,
+      "step": 1990
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 6.956597454051253e-06,
+      "loss": 1.2429,
+      "step": 1995
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 6.941195427366259e-06,
+      "loss": 1.2574,
+      "step": 2000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 6.925771681050784e-06,
+      "loss": 1.2465,
+      "step": 2005
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 6.910326387677906e-06,
+      "loss": 1.2805,
+      "step": 2010
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 6.89485972006179e-06,
+      "loss": 1.2664,
+      "step": 2015
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 6.879371851255747e-06,
+      "loss": 1.1826,
+      "step": 2020
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 6.863862954550315e-06,
+      "loss": 1.2441,
+      "step": 2025
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 6.8483332034713006e-06,
+      "loss": 1.191,
+      "step": 2030
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 6.832782771777846e-06,
+      "loss": 1.2574,
+      "step": 2035
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 6.817211833460484e-06,
+      "loss": 1.2865,
+      "step": 2040
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 6.801620562739197e-06,
+      "loss": 1.2504,
+      "step": 2045
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 6.7860091340614575e-06,
+      "loss": 1.2084,
+      "step": 2050
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 6.770377722100284e-06,
+      "loss": 1.2609,
+      "step": 2055
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 6.75472650175228e-06,
+      "loss": 1.2723,
+      "step": 2060
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 6.739055648135685e-06,
+      "loss": 1.1243,
+      "step": 2065
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 6.723365336588409e-06,
+      "loss": 1.2529,
+      "step": 2070
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 6.707655742666074e-06,
+      "loss": 1.3047,
+      "step": 2075
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 6.691927042140044e-06,
+      "loss": 1.257,
+      "step": 2080
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 6.6761794109954714e-06,
+      "loss": 1.2086,
+      "step": 2085
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 6.660413025429312e-06,
+      "loss": 1.2711,
+      "step": 2090
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 6.644628061848363e-06,
+      "loss": 1.1157,
+      "step": 2095
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 6.628824696867286e-06,
+      "loss": 1.2309,
+      "step": 2100
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 6.613003107306637e-06,
+      "loss": 1.2363,
+      "step": 2105
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 6.597163470190877e-06,
+      "loss": 1.207,
+      "step": 2110
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 6.5813059627464e-06,
+      "loss": 1.2641,
+      "step": 2115
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 6.565430762399546e-06,
+      "loss": 1.252,
+      "step": 2120
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 6.549538046774621e-06,
+      "loss": 1.2586,
+      "step": 2125
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 6.533627993691901e-06,
+      "loss": 1.3012,
+      "step": 2130
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 6.517700781165649e-06,
+      "loss": 1.1842,
+      "step": 2135
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 6.501756587402124e-06,
+      "loss": 1.2016,
+      "step": 2140
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 6.485795590797579e-06,
+      "loss": 1.2988,
+      "step": 2145
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 6.469817969936277e-06,
+      "loss": 1.2547,
+      "step": 2150
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 6.453823903588481e-06,
+      "loss": 1.2309,
+      "step": 2155
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 6.437813570708463e-06,
+      "loss": 1.2855,
+      "step": 2160
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 6.421787150432493e-06,
+      "loss": 1.1488,
+      "step": 2165
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 6.405744822076845e-06,
+      "loss": 1.2115,
+      "step": 2170
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 6.389686765135782e-06,
+      "loss": 1.2336,
+      "step": 2175
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 6.3736131592795525e-06,
+      "loss": 1.2746,
+      "step": 2180
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 6.357524184352375e-06,
+      "loss": 1.201,
+      "step": 2185
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 6.341420020370435e-06,
+      "loss": 1.2703,
+      "step": 2190
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 6.325300847519859e-06,
+      "loss": 1.2441,
+      "step": 2195
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 6.309166846154713e-06,
+      "loss": 1.2684,
+      "step": 2200
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 6.293018196794964e-06,
+      "loss": 1.2449,
+      "step": 2205
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 6.276855080124483e-06,
+      "loss": 1.268,
+      "step": 2210
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 6.260677676989008e-06,
+      "loss": 1.2906,
+      "step": 2215
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 6.24448616839413e-06,
+      "loss": 1.307,
+      "step": 2220
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 6.228280735503254e-06,
+      "loss": 1.2699,
+      "step": 2225
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 6.212061559635588e-06,
+      "loss": 1.24,
+      "step": 2230
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 6.195828822264107e-06,
+      "loss": 1.2605,
+      "step": 2235
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 6.179582705013519e-06,
+      "loss": 1.3457,
+      "step": 2240
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 6.163323389658242e-06,
+      "loss": 1.2301,
+      "step": 2245
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 6.147051058120359e-06,
+      "loss": 1.257,
+      "step": 2250
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 6.130765892467595e-06,
+      "loss": 1.1584,
+      "step": 2255
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.114468074911265e-06,
+      "loss": 1.2537,
+      "step": 2260
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.098157787804252e-06,
+      "loss": 1.2559,
+      "step": 2265
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.081835213638951e-06,
+      "loss": 1.2307,
+      "step": 2270
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.0655005350452414e-06,
+      "loss": 1.1664,
+      "step": 2275
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.049153934788429e-06,
+      "loss": 1.2146,
+      "step": 2280
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.032795595767214e-06,
+      "loss": 1.2498,
+      "step": 2285
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.016425701011637e-06,
+      "loss": 1.2379,
+      "step": 2290
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.000044433681034e-06,
+      "loss": 1.2584,
+      "step": 2295
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5.9836519770619865e-06,
+      "loss": 1.2805,
+      "step": 2300
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5.967248514566271e-06,
+      "loss": 1.2348,
+      "step": 2305
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.9508342297288035e-06,
+      "loss": 1.2572,
+      "step": 2310
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.934409306205593e-06,
+      "loss": 1.2018,
+      "step": 2315
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.917973927771678e-06,
+      "loss": 1.2641,
+      "step": 2320
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.901528278319083e-06,
+      "loss": 1.2293,
+      "step": 2325
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.885072541854742e-06,
+      "loss": 1.3113,
+      "step": 2330
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.868606902498457e-06,
+      "loss": 1.26,
+      "step": 2335
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.852131544480831e-06,
+      "loss": 1.2092,
+      "step": 2340
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.835646652141208e-06,
+      "loss": 1.2535,
+      "step": 2345
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.8191524099256035e-06,
+      "loss": 1.2535,
+      "step": 2350
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.802649002384655e-06,
+      "loss": 1.2629,
+      "step": 2355
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.786136614171542e-06,
+      "loss": 1.233,
+      "step": 2360
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.769615430039931e-06,
+      "loss": 1.2375,
+      "step": 2365
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.753085634841903e-06,
+      "loss": 1.2312,
+      "step": 2370
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.736547413525888e-06,
+      "loss": 1.1715,
+      "step": 2375
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.72000095113459e-06,
+      "loss": 1.2695,
+      "step": 2380
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.703446432802924e-06,
+      "loss": 1.1672,
+      "step": 2385
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.686884043755942e-06,
+      "loss": 1.2637,
+      "step": 2390
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.6703139693067554e-06,
+      "loss": 1.1591,
+      "step": 2395
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.653736394854471e-06,
+      "loss": 1.2343,
+      "step": 2400
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.637151505882109e-06,
+      "loss": 1.2172,
+      "step": 2405
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.620559487954531e-06,
+      "loss": 1.2121,
+      "step": 2410
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.603960526716361e-06,
+      "loss": 1.178,
+      "step": 2415
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.587354807889913e-06,
+      "loss": 1.2256,
+      "step": 2420
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.570742517273109e-06,
+      "loss": 1.241,
+      "step": 2425
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.554123840737402e-06,
+      "loss": 1.2773,
+      "step": 2430
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.537498964225694e-06,
+      "loss": 1.2383,
+      "step": 2435
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.520868073750261e-06,
+      "loss": 1.177,
+      "step": 2440
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.50423135539066e-06,
+      "loss": 1.1607,
+      "step": 2445
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.487588995291666e-06,
+      "loss": 1.2957,
+      "step": 2450
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.47094117966117e-06,
+      "loss": 1.164,
+      "step": 2455
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.454288094768108e-06,
+      "loss": 1.225,
+      "step": 2460
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.437629926940367e-06,
+      "loss": 1.2602,
+      "step": 2465
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.420966862562718e-06,
+      "loss": 1.2434,
+      "step": 2470
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.404299088074702e-06,
+      "loss": 1.0836,
+      "step": 2475
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.387626789968574e-06,
+      "loss": 1.0635,
+      "step": 2480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.370950154787195e-06,
+      "loss": 1.032,
+      "step": 2485
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.354269369121958e-06,
+      "loss": 1.0236,
+      "step": 2490
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.337584619610691e-06,
+      "loss": 1.0402,
+      "step": 2495
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.320896092935575e-06,
+      "loss": 1.0713,
+      "step": 2500
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.304203975821048e-06,
+      "loss": 1.0443,
+      "step": 2505
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.287508455031729e-06,
+      "loss": 1.0523,
+      "step": 2510
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.270809717370314e-06,
+      "loss": 1.0072,
+      "step": 2515
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.254107949675493e-06,
+      "loss": 1.0473,
+      "step": 2520
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.237403338819859e-06,
+      "loss": 1.0189,
+      "step": 2525
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.220696071707816e-06,
+      "loss": 1.027,
+      "step": 2530
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.20398633527349e-06,
+      "loss": 0.9773,
+      "step": 2535
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.187274316478632e-06,
+      "loss": 0.9916,
+      "step": 2540
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.170560202310536e-06,
+      "loss": 1.0252,
+      "step": 2545
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.153844179779932e-06,
+      "loss": 1.0508,
+      "step": 2550
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.137126435918912e-06,
+      "loss": 1.0217,
+      "step": 2555
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.12040715777882e-06,
+      "loss": 1.0367,
+      "step": 2560
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.1036865324281716e-06,
+      "loss": 1.0121,
+      "step": 2565
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.08696474695055e-06,
+      "loss": 0.9992,
+      "step": 2570
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.070241988442528e-06,
+      "loss": 1.0778,
+      "step": 2575
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.053518444011557e-06,
+      "loss": 1.0703,
+      "step": 2580
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.036794300773887e-06,
+      "loss": 1.017,
+      "step": 2585
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.020069745852463e-06,
+      "loss": 0.9813,
+      "step": 2590
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.003344966374843e-06,
+      "loss": 1.0287,
+      "step": 2595
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.9866201494710934e-06,
+      "loss": 1.0617,
+      "step": 2600
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.969895482271695e-06,
+      "loss": 1.1227,
+      "step": 2605
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.953171151905466e-06,
+      "loss": 1.0496,
+      "step": 2610
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.936447345497443e-06,
+      "loss": 1.0287,
+      "step": 2615
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.919724250166808e-06,
+      "loss": 1.0656,
+      "step": 2620
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.903002053024782e-06,
+      "loss": 1.0287,
+      "step": 2625
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.886280941172539e-06,
+      "loss": 1.0293,
+      "step": 2630
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.869561101699113e-06,
+      "loss": 1.0805,
+      "step": 2635
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.852842721679293e-06,
+      "loss": 1.0068,
+      "step": 2640
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.836125988171547e-06,
+      "loss": 1.0056,
+      "step": 2645
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.8194110882159175e-06,
+      "loss": 1.0256,
+      "step": 2650
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.802698208831929e-06,
+      "loss": 1.0551,
+      "step": 2655
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.785987537016504e-06,
+      "loss": 1.002,
+      "step": 2660
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 4.769279259741858e-06,
+      "loss": 1.0378,
+      "step": 2665
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 4.752573563953422e-06,
+      "loss": 1.0088,
+      "step": 2670
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 4.735870636567736e-06,
+      "loss": 0.9963,
+      "step": 2675
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 4.719170664470371e-06,
+      "loss": 0.9977,
+      "step": 2680
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 4.702473834513826e-06,
+      "loss": 1.0533,
+      "step": 2685
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 4.685780333515449e-06,
+      "loss": 1.0148,
+      "step": 2690
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 4.669090348255338e-06,
+      "loss": 1.0023,
+      "step": 2695
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 4.652404065474257e-06,
+      "loss": 1.0227,
+      "step": 2700
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 4.6357216718715375e-06,
+      "loss": 1.0236,
+      "step": 2705
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 4.619043354103002e-06,
+      "loss": 1.01,
+      "step": 2710
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 4.602369298778866e-06,
+      "loss": 1.0625,
+      "step": 2715
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 4.585699692461655e-06,
+      "loss": 1.0154,
+      "step": 2720
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 4.569034721664114e-06,
+      "loss": 1.0547,
+      "step": 2725
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 4.552374572847122e-06,
+      "loss": 0.981,
+      "step": 2730
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 4.535719432417612e-06,
+      "loss": 1.0691,
+      "step": 2735
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 4.519069486726468e-06,
+      "loss": 1.0451,
+      "step": 2740
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 4.502424922066462e-06,
+      "loss": 0.9773,
+      "step": 2745
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 4.485785924670151e-06,
+      "loss": 0.9898,
+      "step": 2750
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 4.469152680707804e-06,
+      "loss": 1.0496,
+      "step": 2755
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 4.452525376285319e-06,
+      "loss": 1.0211,
+      "step": 2760
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 4.435904197442131e-06,
+      "loss": 1.0961,
+      "step": 2765
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 4.419289330149145e-06,
+      "loss": 1.0279,
+      "step": 2770
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 4.4026809603066375e-06,
+      "loss": 1.0081,
+      "step": 2775
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 4.386079273742199e-06,
+      "loss": 1.0764,
+      "step": 2780
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 4.3694844562086325e-06,
+      "loss": 1.0342,
+      "step": 2785
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.3528966933818865e-06,
+      "loss": 1.0707,
+      "step": 2790
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.33631617085898e-06,
+      "loss": 1.0127,
+      "step": 2795
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.319743074155916e-06,
+      "loss": 1.0658,
+      "step": 2800
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.3031775887056176e-06,
+      "loss": 1.0881,
+      "step": 2805
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.2866198998558404e-06,
+      "loss": 1.043,
+      "step": 2810
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.2700701928671105e-06,
+      "loss": 1.008,
+      "step": 2815
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.253528652910647e-06,
+      "loss": 1.0571,
+      "step": 2820
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.236995465066287e-06,
+      "loss": 1.0859,
+      "step": 2825
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.220470814320417e-06,
+      "loss": 1.0085,
+      "step": 2830
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.203954885563909e-06,
+      "loss": 1.0146,
+      "step": 2835
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.187447863590039e-06,
+      "loss": 1.0562,
+      "step": 2840
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.170949933092432e-06,
+      "loss": 1.1096,
+      "step": 2845
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.154461278662989e-06,
+      "loss": 1.0555,
+      "step": 2850
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.137982084789823e-06,
+      "loss": 0.9902,
+      "step": 2855
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.121512535855193e-06,
+      "loss": 1.06,
+      "step": 2860
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.105052816133448e-06,
+      "loss": 1.0412,
+      "step": 2865
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.0886031097889556e-06,
+      "loss": 1.0354,
+      "step": 2870
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.072163600874045e-06,
+      "loss": 1.0928,
+      "step": 2875
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.0557344733269505e-06,
+      "loss": 1.0645,
+      "step": 2880
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.039315910969754e-06,
+      "loss": 0.9994,
+      "step": 2885
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.02290809750632e-06,
+      "loss": 1.003,
+      "step": 2890
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.006511216520251e-06,
+      "loss": 1.0512,
+      "step": 2895
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 3.9901254514728225e-06,
+      "loss": 1.06,
+      "step": 2900
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 3.973750985700943e-06,
+      "loss": 1.0541,
+      "step": 2905
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 3.957388002415093e-06,
+      "loss": 1.0078,
+      "step": 2910
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 3.941036684697274e-06,
+      "loss": 1.0104,
+      "step": 2915
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 3.924697215498971e-06,
+      "loss": 1.0465,
+      "step": 2920
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 3.908369777639091e-06,
+      "loss": 0.9527,
+      "step": 2925
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 3.892054553801931e-06,
+      "loss": 1.0559,
+      "step": 2930
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 3.875751726535124e-06,
+      "loss": 1.041,
+      "step": 2935
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 3.8594614782476024e-06,
+      "loss": 1.0352,
+      "step": 2940
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 3.843183991207551e-06,
+      "loss": 1.0175,
+      "step": 2945
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 3.82691944754038e-06,
+      "loss": 0.9959,
+      "step": 2950
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 3.8106680292266717e-06,
+      "loss": 1.0094,
+      "step": 2955
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 3.7944299181001544e-06,
+      "loss": 1.0367,
+      "step": 2960
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3.778205295845663e-06,
+      "loss": 1.0443,
+      "step": 2965
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3.7619943439971107e-06,
+      "loss": 1.0074,
+      "step": 2970
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.7457972439354526e-06,
+      "loss": 1.0396,
+      "step": 2975
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.7296141768866635e-06,
+      "loss": 1.0506,
+      "step": 2980
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.7134453239196987e-06,
+      "loss": 1.0268,
+      "step": 2985
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.6972908659444828e-06,
+      "loss": 1.0101,
+      "step": 2990
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.6811509837098756e-06,
+      "loss": 1.0076,
+      "step": 2995
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.6650258578016474e-06,
+      "loss": 1.0602,
+      "step": 3000
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.6489156686404683e-06,
+      "loss": 1.0418,
+      "step": 3005
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.6328205964798822e-06,
+      "loss": 1.0498,
+      "step": 3010
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.616740821404292e-06,
+      "loss": 1.0277,
+      "step": 3015
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.600676523326946e-06,
+      "loss": 1.0979,
+      "step": 3020
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.5846278819879197e-06,
+      "loss": 1.0467,
+      "step": 3025
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.568595076952113e-06,
+      "loss": 1.0344,
+      "step": 3030
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.552578287607237e-06,
+      "loss": 0.9874,
+      "step": 3035
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.536577693161801e-06,
+      "loss": 1.0688,
+      "step": 3040
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.520593472643122e-06,
+      "loss": 1.0023,
+      "step": 3045
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.504625804895302e-06,
+      "loss": 1.0315,
+      "step": 3050
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.488674868577246e-06,
+      "loss": 1.0318,
+      "step": 3055
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.472740842160649e-06,
+      "loss": 1.057,
+      "step": 3060
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.4568239039280094e-06,
+      "loss": 1.041,
+      "step": 3065
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.4409242319706225e-06,
+      "loss": 1.126,
+      "step": 3070
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.4250420041866057e-06,
+      "loss": 1.0151,
+      "step": 3075
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.4091773982788867e-06,
+      "loss": 1.0395,
+      "step": 3080
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.393330591753231e-06,
+      "loss": 1.0207,
+      "step": 3085
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.377501761916249e-06,
+      "loss": 0.9663,
+      "step": 3090
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.3616910858734143e-06,
+      "loss": 1.055,
+      "step": 3095
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.3458987405270803e-06,
+      "loss": 1.017,
+      "step": 3100
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.330124902574505e-06,
+      "loss": 1.0034,
+      "step": 3105
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.3143697485058666e-06,
+      "loss": 1.0262,
+      "step": 3110
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.2986334546022964e-06,
+      "loss": 1.0723,
+      "step": 3115
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.282916196933904e-06,
+      "loss": 1.0314,
+      "step": 3120
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.2672181513578038e-06,
+      "loss": 1.0613,
+      "step": 3125
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.251539493516152e-06,
+      "loss": 1.0641,
+      "step": 3130
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.2358803988341776e-06,
+      "loss": 1.0283,
+      "step": 3135
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.220241042518223e-06,
+      "loss": 1.0502,
+      "step": 3140
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.2046215995537837e-06,
+      "loss": 1.0416,
+      "step": 3145
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 3.1890222447035444e-06,
+      "loss": 1.0549,
+      "step": 3150
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 3.173443152505431e-06,
+      "loss": 1.034,
+      "step": 3155
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 3.157884497270658e-06,
+      "loss": 1.0594,
+      "step": 3160
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 3.1423464530817673e-06,
+      "loss": 1.0637,
+      "step": 3165
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 3.1268291937906957e-06,
+      "loss": 1.0402,
+      "step": 3170
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 3.1113328930168153e-06,
+      "loss": 1.0236,
+      "step": 3175
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 3.095857724145004e-06,
+      "loss": 1.0414,
+      "step": 3180
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 3.0804038603236943e-06,
+      "loss": 1.0465,
+      "step": 3185
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 3.0649714744629454e-06,
+      "loss": 1.0561,
+      "step": 3190
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 3.0495607392324987e-06,
+      "loss": 1.0414,
+      "step": 3195
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 3.0341718270598557e-06,
+      "loss": 1.0492,
+      "step": 3200
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 3.0188049101283433e-06,
+      "loss": 1.0053,
+      "step": 3205
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 3.003460160375189e-06,
+      "loss": 1.0193,
+      "step": 3210
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.9881377494895925e-06,
+      "loss": 1.093,
+      "step": 3215
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.9728378489108135e-06,
+      "loss": 1.0285,
+      "step": 3220
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.957560629826244e-06,
+      "loss": 1.0982,
+      "step": 3225
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.942306263169502e-06,
+      "loss": 1.0438,
+      "step": 3230
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.9270749196185095e-06,
+      "loss": 1.0695,
+      "step": 3235
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.911866769593592e-06,
+      "loss": 1.0139,
+      "step": 3240
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.896681983255565e-06,
+      "loss": 1.1477,
+      "step": 3245
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.881520730503837e-06,
+      "loss": 1.0437,
+      "step": 3250
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.866383180974498e-06,
+      "loss": 1.0455,
+      "step": 3255
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.8512695040384287e-06,
+      "loss": 1.0014,
+      "step": 3260
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.8361798687994097e-06,
+      "loss": 1.0016,
+      "step": 3265
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.8211144440922176e-06,
+      "loss": 0.9983,
+      "step": 3270
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.8060733984807466e-06,
+      "loss": 1.0927,
+      "step": 3275
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.7910569002561137e-06,
+      "loss": 1.0424,
+      "step": 3280
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.7760651174347854e-06,
+      "loss": 1.0555,
+      "step": 3285
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.7610982177566926e-06,
+      "loss": 0.983,
+      "step": 3290
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.7461563686833504e-06,
+      "loss": 0.9712,
+      "step": 3295
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.7312397373959894e-06,
+      "loss": 1.04,
+      "step": 3300
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.716348490793681e-06,
+      "loss": 1.092,
+      "step": 3305
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.7014827954914814e-06,
+      "loss": 0.9855,
+      "step": 3310
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.686642817818548e-06,
+      "loss": 1.0319,
+      "step": 3315
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.6718287238162963e-06,
+      "loss": 0.9938,
+      "step": 3320
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.6570406792365268e-06,
+      "loss": 1.0662,
+      "step": 3325
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.6422788495395912e-06,
+      "loss": 1.0263,
+      "step": 3330
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.6275433998925176e-06,
+      "loss": 1.0584,
+      "step": 3335
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.612834495167177e-06,
+      "loss": 1.0334,
+      "step": 3340
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.5981522999384323e-06,
+      "loss": 1.0426,
+      "step": 3345
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.583496978482305e-06,
+      "loss": 1.0199,
+      "step": 3350
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.568868694774127e-06,
+      "loss": 1.0363,
+      "step": 3355
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.5542676124867103e-06,
+      "loss": 0.9959,
+      "step": 3360
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.5396938949885163e-06,
+      "loss": 1.0357,
+      "step": 3365
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.52514770534183e-06,
+      "loss": 1.0444,
+      "step": 3370
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.510629206300933e-06,
+      "loss": 1.0627,
+      "step": 3375
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.4961385603102794e-06,
+      "loss": 1.0535,
+      "step": 3380
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.481675929502682e-06,
+      "loss": 1.0276,
+      "step": 3385
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.467241475697498e-06,
+      "loss": 1.0057,
+      "step": 3390
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.45283536039882e-06,
+      "loss": 1.0055,
+      "step": 3395
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.438457744793665e-06,
+      "loss": 1.0001,
+      "step": 3400
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.4241087897501703e-06,
+      "loss": 1.1129,
+      "step": 3405
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.409788655815802e-06,
+      "loss": 0.9816,
+      "step": 3410
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.395497503215551e-06,
+      "loss": 1.008,
+      "step": 3415
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.3812354918501397e-06,
+      "loss": 1.0068,
+      "step": 3420
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.3670027812942353e-06,
+      "loss": 1.0779,
+      "step": 3425
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.3527995307946655e-06,
+      "loss": 1.0264,
+      "step": 3430
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.338625899268638e-06,
+      "loss": 1.0395,
+      "step": 3435
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.3244820453019566e-06,
+      "loss": 1.0604,
+      "step": 3440
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.3103681271472516e-06,
+      "loss": 1.0236,
+      "step": 3445
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.296284302722205e-06,
+      "loss": 1.0918,
+      "step": 3450
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.28223072960779e-06,
+      "loss": 1.0504,
+      "step": 3455
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2682075650465063e-06,
+      "loss": 1.0361,
+      "step": 3460
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2542149659406126e-06,
+      "loss": 1.0268,
+      "step": 3465
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.2402530888503783e-06,
+      "loss": 1.0434,
+      "step": 3470
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.226322089992336e-06,
+      "loss": 1.0348,
+      "step": 3475
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.2124221252375215e-06,
+      "loss": 1.0135,
+      "step": 3480
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.1985533501097407e-06,
+      "loss": 1.0488,
+      "step": 3485
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.1847159197838213e-06,
+      "loss": 0.9809,
+      "step": 3490
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.1709099890838846e-06,
+      "loss": 1.0627,
+      "step": 3495
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.1571357124816107e-06,
+      "loss": 1.0373,
+      "step": 3500
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.1433932440945028e-06,
+      "loss": 1.0068,
+      "step": 3505
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.129682737684171e-06,
+      "loss": 1.0604,
+      "step": 3510
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.11600434665461e-06,
+      "loss": 1.0337,
+      "step": 3515
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.1023582240504836e-06,
+      "loss": 1.0668,
+      "step": 3520
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.088744522555409e-06,
+      "loss": 1.0088,
+      "step": 3525
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.0751633944902487e-06,
+      "loss": 1.0436,
+      "step": 3530
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.061614991811414e-06,
+      "loss": 1.0138,
+      "step": 3535
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.0480994661091507e-06,
+      "loss": 1.1406,
+      "step": 3540
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.0346169686058586e-06,
+      "loss": 1.0391,
+      "step": 3545
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.0211676501543866e-06,
+      "loss": 1.0592,
+      "step": 3550
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.00775166123635e-06,
+      "loss": 0.9783,
+      "step": 3555
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.9943691519604523e-06,
+      "loss": 1.0473,
+      "step": 3560
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.9810202720607945e-06,
+      "loss": 1.0555,
+      "step": 3565
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.967705170895208e-06,
+      "loss": 1.0691,
+      "step": 3570
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.9544239974435797e-06,
+      "loss": 1.026,
+      "step": 3575
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.9411769003061874e-06,
+      "loss": 1.0588,
+      "step": 3580
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.9279640277020396e-06,
+      "loss": 1.0635,
+      "step": 3585
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.9147855274672073e-06,
+      "loss": 0.9919,
+      "step": 3590
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.9016415470531773e-06,
+      "loss": 1.0053,
+      "step": 3595
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.8885322335252076e-06,
+      "loss": 1.0461,
+      "step": 3600
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.8754577335606689e-06,
+      "loss": 1.0051,
+      "step": 3605
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.8624181934474117e-06,
+      "loss": 1.0521,
+      "step": 3610
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.8494137590821282e-06,
+      "loss": 0.9926,
+      "step": 3615
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.8364445759687233e-06,
+      "loss": 1.0264,
+      "step": 3620
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.823510789216676e-06,
+      "loss": 1.0475,
+      "step": 3625
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.8106125435394312e-06,
+      "loss": 1.012,
+      "step": 3630
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.7977499832527655e-06,
+      "loss": 1.0269,
+      "step": 3635
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.7849232522731797e-06,
+      "loss": 1.0463,
+      "step": 3640
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.7721324941162933e-06,
+      "loss": 1.025,
+      "step": 3645
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.7593778518952275e-06,
+      "loss": 1.0326,
+      "step": 3650
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.7466594683190107e-06,
+      "loss": 1.0389,
+      "step": 3655
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.7339774856909851e-06,
+      "loss": 1.0609,
+      "step": 3660
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.7213320459072047e-06,
+      "loss": 0.9949,
+      "step": 3665
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.7087232904548595e-06,
+      "loss": 1.0083,
+      "step": 3670
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.69615136041068e-06,
+      "loss": 1.0377,
+      "step": 3675
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.6836163964393664e-06,
+      "loss": 1.0514,
+      "step": 3680
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.6711185387920176e-06,
+      "loss": 0.99,
+      "step": 3685
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.6586579273045529e-06,
+      "loss": 1.0146,
+      "step": 3690
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.6462347013961526e-06,
+      "loss": 1.0445,
+      "step": 3695
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.6338490000676987e-06,
+      "loss": 1.0674,
+      "step": 3700
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.6215009619002197e-06,
+      "loss": 1.0215,
+      "step": 3705
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.609190725053335e-06,
+      "loss": 0.9832,
+      "step": 3710
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.5969184272637184e-06,
+      "loss": 0.9313,
+      "step": 3715
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.5846842058435457e-06,
+      "loss": 1.0244,
+      "step": 3720
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.5724881976789696e-06,
+      "loss": 0.9002,
+      "step": 3725
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.5603305392285785e-06,
+      "loss": 0.957,
+      "step": 3730
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.548211366521875e-06,
+      "loss": 0.9404,
+      "step": 3735
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.5361308151577526e-06,
+      "loss": 0.9199,
+      "step": 3740
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.5240890203029813e-06,
+      "loss": 0.9224,
+      "step": 3745
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.5120861166906869e-06,
+      "loss": 0.9822,
+      "step": 3750
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.5001222386188573e-06,
+      "loss": 0.9063,
+      "step": 3755
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.4881975199488247e-06,
+      "loss": 0.9455,
+      "step": 3760
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.4763120941037757e-06,
+      "loss": 0.8986,
+      "step": 3765
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.9297,
+      "step": 3770
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.4526596523817066e-06,
+      "loss": 0.9889,
+      "step": 3775
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.4408929011469175e-06,
+      "loss": 0.9387,
+      "step": 3780
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.4291659720186218e-06,
+      "loss": 0.8889,
+      "step": 3785
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.4174789962069808e-06,
+      "loss": 0.9965,
+      "step": 3790
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.4058321044751255e-06,
+      "loss": 0.9279,
+      "step": 3795
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.3942254271377004e-06,
+      "loss": 0.9621,
+      "step": 3800
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.3826590940593926e-06,
+      "loss": 0.9081,
+      "step": 3805
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.3711332346534916e-06,
+      "loss": 0.9201,
+      "step": 3810
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.3596479778804312e-06,
+      "loss": 0.9013,
+      "step": 3815
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.3482034522463522e-06,
+      "loss": 0.9255,
+      "step": 3820
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.3367997858016619e-06,
+      "loss": 0.9678,
+      "step": 3825
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.325437106139607e-06,
+      "loss": 0.9334,
+      "step": 3830
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.3141155403948358e-06,
+      "loss": 0.9455,
+      "step": 3835
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.3028352152419876e-06,
+      "loss": 0.9025,
+      "step": 3840
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.291596256894263e-06,
+      "loss": 0.8933,
+      "step": 3845
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.2803987911020239e-06,
+      "loss": 0.999,
+      "step": 3850
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.269242943151377e-06,
+      "loss": 0.8996,
+      "step": 3855
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.2581288378627759e-06,
+      "loss": 0.9594,
+      "step": 3860
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.2470565995896244e-06,
+      "loss": 0.9385,
+      "step": 3865
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.236026352216888e-06,
+      "loss": 0.9508,
+      "step": 3870
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.2250382191597015e-06,
+      "loss": 0.9479,
+      "step": 3875
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.21409232336199e-06,
+      "loss": 0.8861,
+      "step": 3880
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.2031887872951004e-06,
+      "loss": 0.9539,
+      "step": 3885
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.1923277329564192e-06,
+      "loss": 0.8969,
+      "step": 3890
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.181509281868019e-06,
+      "loss": 0.9248,
+      "step": 3895
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.1707335550752901e-06,
+      "loss": 0.8923,
+      "step": 3900
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.1600006731455888e-06,
+      "loss": 0.8534,
+      "step": 3905
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.1493107561668943e-06,
+      "loss": 0.9193,
+      "step": 3910
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.1386639237464542e-06,
+      "loss": 0.9688,
+      "step": 3915
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.1280602950094532e-06,
+      "loss": 0.8982,
+      "step": 3920
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.1174999885976834e-06,
+      "loss": 0.9001,
+      "step": 3925
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.106983122668206e-06,
+      "loss": 0.9189,
+      "step": 3930
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.0965098148920422e-06,
+      "loss": 0.9842,
+      "step": 3935
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.0860801824528443e-06,
+      "loss": 0.9438,
+      "step": 3940
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.0756943420455934e-06,
+      "loss": 0.9412,
+      "step": 3945
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.0653524098752894e-06,
+      "loss": 0.9695,
+      "step": 3950
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.055054501655654e-06,
+      "loss": 0.9145,
+      "step": 3955
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.0448007326078336e-06,
+      "loss": 0.9602,
+      "step": 3960
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.0345912174591071e-06,
+      "loss": 0.9009,
+      "step": 3965
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.0244260704416104e-06,
+      "loss": 0.9375,
+      "step": 3970
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.0143054052910534e-06,
+      "loss": 0.9402,
+      "step": 3975
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.0042293352454446e-06,
+      "loss": 0.9182,
+      "step": 3980
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 9.94197973043829e-07,
+      "loss": 0.909,
+      "step": 3985
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 9.842114309250222e-07,
+      "loss": 0.9285,
+      "step": 3990
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 9.74269820626364e-07,
+      "loss": 0.9264,
+      "step": 3995
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 9.643732533824545e-07,
+      "loss": 0.9205,
+      "step": 4000
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 9.545218399239186e-07,
+      "loss": 0.96,
+      "step": 4005
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 9.447156904761668e-07,
+      "loss": 0.9473,
+      "step": 4010
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 9.349549147581571e-07,
+      "loss": 0.9281,
+      "step": 4015
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 9.252396219811737e-07,
+      "loss": 0.9311,
+      "step": 4020
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 9.155699208475988e-07,
+      "loss": 0.9789,
+      "step": 4025
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 9.059459195496989e-07,
+      "loss": 0.8984,
+      "step": 4030
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 8.963677257684184e-07,
+      "loss": 0.9564,
+      "step": 4035
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 8.868354466721668e-07,
+      "loss": 0.9293,
+      "step": 4040
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 8.773491889156254e-07,
+      "loss": 0.9678,
+      "step": 4045
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 8.679090586385519e-07,
+      "loss": 0.9275,
+      "step": 4050
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 8.585151614645942e-07,
+      "loss": 0.966,
+      "step": 4055
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 8.491676025001083e-07,
+      "loss": 0.9049,
+      "step": 4060
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 8.398664863329792e-07,
+      "loss": 0.9385,
+      "step": 4065
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 8.306119170314553e-07,
+      "loss": 0.9529,
+      "step": 4070
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 8.214039981429789e-07,
+      "loss": 0.9412,
+      "step": 4075
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 8.122428326930348e-07,
+      "loss": 0.9852,
+      "step": 4080
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 8.031285231839908e-07,
+      "loss": 0.9223,
+      "step": 4085
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 7.940611715939522e-07,
+      "loss": 0.9592,
+      "step": 4090
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 7.850408793756242e-07,
+      "loss": 0.9758,
+      "step": 4095
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 7.760677474551759e-07,
+      "loss": 0.842,
+      "step": 4100
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 7.67141876231105e-07,
+      "loss": 0.9406,
+      "step": 4105
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 7.582633655731231e-07,
+      "loss": 0.9397,
+      "step": 4110
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 7.494323148210303e-07,
+      "loss": 0.9193,
+      "step": 4115
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 7.406488227836139e-07,
+      "loss": 0.9529,
+      "step": 4120
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 7.319129877375314e-07,
+      "loss": 0.973,
+      "step": 4125
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 7.232249074262176e-07,
+      "loss": 0.9596,
+      "step": 4130
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 7.145846790587891e-07,
+      "loss": 0.9477,
+      "step": 4135
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 7.059923993089585e-07,
+      "loss": 0.9809,
+      "step": 4140
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 6.974481643139514e-07,
+      "loss": 0.9863,
+      "step": 4145
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 6.889520696734297e-07,
+      "loss": 0.9666,
+      "step": 4150
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 6.805042104484216e-07,
+      "loss": 0.9328,
+      "step": 4155
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 6.721046811602622e-07,
+      "loss": 0.8867,
+      "step": 4160
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 6.63753575789532e-07,
+      "loss": 0.9635,
+      "step": 4165
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 6.554509877750042e-07,
+      "loss": 0.9605,
+      "step": 4170
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 6.471970100126035e-07,
+      "loss": 0.989,
+      "step": 4175
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 6.389917348543651e-07,
+      "loss": 0.9393,
+      "step": 4180
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 6.308352541074014e-07,
+      "loss": 0.9385,
+      "step": 4185
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 6.227276590328713e-07,
+      "loss": 0.9325,
+      "step": 4190
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 6.146690403449646e-07,
+      "loss": 0.9801,
+      "step": 4195
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 6.066594882098831e-07,
+      "loss": 0.976,
+      "step": 4200
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5.98699092244835e-07,
+      "loss": 0.9523,
+      "step": 4205
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5.907879415170287e-07,
+      "loss": 0.8773,
+      "step": 4210
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5.829261245426793e-07,
+      "loss": 0.8939,
+      "step": 4215
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5.751137292860126e-07,
+      "loss": 0.9383,
+      "step": 4220
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5.673508431582936e-07,
+      "loss": 0.9797,
+      "step": 4225
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5.596375530168329e-07,
+      "loss": 0.932,
+      "step": 4230
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5.519739451640238e-07,
+      "loss": 0.9015,
+      "step": 4235
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5.443601053463743e-07,
+      "loss": 0.966,
+      "step": 4240
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5.367961187535504e-07,
+      "loss": 0.9252,
+      "step": 4245
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5.292820700174189e-07,
+      "loss": 0.925,
+      "step": 4250
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5.218180432111026e-07,
+      "loss": 0.9445,
+      "step": 4255
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5.144041218480389e-07,
+      "loss": 0.9461,
+      "step": 4260
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5.070403888810471e-07,
+      "loss": 0.926,
+      "step": 4265
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 4.997269267013993e-07,
+      "loss": 0.9242,
+      "step": 4270
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 4.924638171378976e-07,
+      "loss": 0.9514,
+      "step": 4275
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 4.852511414559575e-07,
+      "loss": 0.9877,
+      "step": 4280
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 4.780889803567018e-07,
+      "loss": 0.9541,
+      "step": 4285
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 4.7097741397605754e-07,
+      "loss": 0.9449,
+      "step": 4290
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 4.639165218838559e-07,
+      "loss": 0.9361,
+      "step": 4295
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 4.569063830829445e-07,
+      "loss": 0.9908,
+      "step": 4300
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 4.49947076008303e-07,
+      "loss": 0.9355,
+      "step": 4305
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 4.4303867852616755e-07,
+      "loss": 0.9096,
+      "step": 4310
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 4.361812679331551e-07,
+      "loss": 0.9555,
+      "step": 4315
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 4.2937492095540043e-07,
+      "loss": 0.9221,
+      "step": 4320
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 4.2261971374769893e-07,
+      "loss": 0.9594,
+      "step": 4325
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 4.159157218926557e-07,
+      "loss": 0.914,
+      "step": 4330
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 4.09263020399836e-07,
+      "loss": 0.9935,
+      "step": 4335
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 4.02661683704928e-07,
+      "loss": 0.9467,
+      "step": 4340
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 3.9611178566890894e-07,
+      "loss": 0.943,
+      "step": 4345
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 3.896133995772233e-07,
+      "loss": 0.9232,
+      "step": 4350
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 3.8316659813895597e-07,
+      "loss": 0.9545,
+      "step": 4355
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 3.767714534860223e-07,
+      "loss": 0.9242,
+      "step": 4360
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 3.704280371723601e-07,
+      "loss": 0.9379,
+      "step": 4365
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 3.6413642017313233e-07,
+      "loss": 0.9506,
+      "step": 4370
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 3.5789667288392784e-07,
+      "loss": 0.9465,
+      "step": 4375
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 3.517088651199768e-07,
+      "loss": 0.9365,
+      "step": 4380
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 3.455730661153672e-07,
+      "loss": 0.9195,
+      "step": 4385
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 3.394893445222752e-07,
+      "loss": 0.9746,
+      "step": 4390
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 3.334577684101925e-07,
+      "loss": 0.9289,
+      "step": 4395
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 3.2747840526516414e-07,
+      "loss": 0.9038,
+      "step": 4400
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 3.215513219890365e-07,
+      "loss": 0.9098,
+      "step": 4405
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 3.15676584898707e-07,
+      "loss": 0.9435,
+      "step": 4410
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 3.0985425972538343e-07,
+      "loss": 0.9098,
+      "step": 4415
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 3.040844116138475e-07,
+      "loss": 0.9318,
+      "step": 4420
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 2.9836710512172353e-07,
+      "loss": 0.9592,
+      "step": 4425
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 2.9270240421876204e-07,
+      "loss": 0.9756,
+      "step": 4430
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 2.8709037228611903e-07,
+      "loss": 0.9189,
+      "step": 4435
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 2.815310721156489e-07,
+      "loss": 0.9139,
+      "step": 4440
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 2.7602456590920034e-07,
+      "loss": 0.9127,
+      "step": 4445
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 2.7057091527792125e-07,
+      "loss": 0.9602,
+      "step": 4450
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 2.6517018124157137e-07,
+      "loss": 0.9787,
+      "step": 4455
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 2.598224242278369e-07,
+      "loss": 0.916,
+      "step": 4460
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 2.545277040716537e-07,
+      "loss": 0.9846,
+      "step": 4465
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 2.492860800145408e-07,
+      "loss": 0.9484,
+      "step": 4470
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 2.4409761070393614e-07,
+      "loss": 0.9191,
+      "step": 4475
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 2.389623541925407e-07,
+      "loss": 0.9266,
+      "step": 4480
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 2.3388036793766723e-07,
+      "loss": 0.9034,
+      "step": 4485
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 2.2885170880059758e-07,
+      "loss": 0.896,
+      "step": 4490
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 2.2387643304595196e-07,
+      "loss": 0.9574,
+      "step": 4495
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 2.189545963410511e-07,
+      "loss": 0.9387,
+      "step": 4500
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 2.1408625375529845e-07,
+      "loss": 0.9322,
+      "step": 4505
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 2.0927145975956297e-07,
+      "loss": 0.9088,
+      "step": 4510
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 2.0451026822556952e-07,
+      "loss": 0.9168,
+      "step": 4515
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.9980273242529825e-07,
+      "loss": 0.951,
+      "step": 4520
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.951489050303834e-07,
+      "loss": 0.916,
+      "step": 4525
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.9054883811152837e-07,
+      "loss": 0.8936,
+      "step": 4530
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.8600258313792142e-07,
+      "loss": 0.9279,
+      "step": 4535
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.8151019097666146e-07,
+      "loss": 0.9666,
+      "step": 4540
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1.7707171189218663e-07,
+      "loss": 0.9555,
+      "step": 4545
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1.7268719554571157e-07,
+      "loss": 0.945,
+      "step": 4550
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.683566909946771e-07,
+      "loss": 0.9357,
+      "step": 4555
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.640802466921926e-07,
+      "loss": 0.9528,
+      "step": 4560
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.5985791048650223e-07,
+      "loss": 0.8418,
+      "step": 4565
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.5568972962044405e-07,
+      "loss": 0.9797,
+      "step": 4570
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.515757507309229e-07,
+      "loss": 0.9197,
+      "step": 4575
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.4751601984839159e-07,
+      "loss": 1.0133,
+      "step": 4580
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.4351058239633065e-07,
+      "loss": 0.9518,
+      "step": 4585
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.3955948319074374e-07,
+      "loss": 0.881,
+      "step": 4590
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.3566276643965538e-07,
+      "loss": 0.9238,
+      "step": 4595
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.3182047574261557e-07,
+      "loss": 0.9002,
+      "step": 4600
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.2803265409021436e-07,
+      "loss": 0.948,
+      "step": 4605
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.2429934386359643e-07,
+      "loss": 0.9025,
+      "step": 4610
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.2062058683399048e-07,
+      "loss": 0.9354,
+      "step": 4615
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.1699642416224233e-07,
+      "loss": 0.9582,
+      "step": 4620
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.1342689639835036e-07,
+      "loss": 0.9734,
+      "step": 4625
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.0991204348101692e-07,
+      "loss": 0.9267,
+      "step": 4630
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.0645190473719647e-07,
+      "loss": 0.9705,
+      "step": 4635
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.0304651888166039e-07,
+      "loss": 0.9285,
+      "step": 4640
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 9.969592401655903e-08,
+      "loss": 0.9494,
+      "step": 4645
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 9.640015763100031e-08,
+      "loss": 0.8965,
+      "step": 4650
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 9.315925660062619e-08,
+      "loss": 0.9922,
+      "step": 4655
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 8.997325718720085e-08,
+      "loss": 0.9295,
+      "step": 4660
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 8.684219503820756e-08,
+      "loss": 0.9564,
+      "step": 4665
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 8.376610518644746e-08,
+      "loss": 0.9201,
+      "step": 4670
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 8.074502204964696e-08,
+      "loss": 0.9303,
+      "step": 4675
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 7.777897943007595e-08,
+      "loss": 0.9636,
+      "step": 4680
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 7.486801051416525e-08,
+      "loss": 0.9542,
+      "step": 4685
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 7.201214787213862e-08,
+      "loss": 0.9684,
+      "step": 4690
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 6.921142345764798e-08,
+      "loss": 0.924,
+      "step": 4695
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 6.646586860741322e-08,
+      "loss": 0.9271,
+      "step": 4700
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 6.377551404087467e-08,
+      "loss": 0.9333,
+      "step": 4705
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 6.114038985984894e-08,
+      "loss": 0.9413,
+      "step": 4710
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5.856052554818969e-08,
+      "loss": 0.9223,
+      "step": 4715
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5.603594997145967e-08,
+      "loss": 0.9301,
+      "step": 4720
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5.3566691376609744e-08,
+      "loss": 0.9072,
+      "step": 4725
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5.115277739165703e-08,
+      "loss": 0.9152,
+      "step": 4730
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 4.8794235025383386e-08,
+      "loss": 0.9234,
+      "step": 4735
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 4.6491090667025176e-08,
+      "loss": 0.943,
+      "step": 4740
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 4.4243370085985114e-08,
+      "loss": 0.8847,
+      "step": 4745
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 4.2051098431539764e-08,
+      "loss": 1.0156,
+      "step": 4750
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 3.991430023255804e-08,
+      "loss": 0.866,
+      "step": 4755
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 3.783299939722984e-08,
+      "loss": 0.9083,
+      "step": 4760
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 3.580721921279562e-08,
+      "loss": 0.9077,
+      "step": 4765
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 3.383698234528665e-08,
+      "loss": 0.9351,
+      "step": 4770
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 3.1922310839272444e-08,
+      "loss": 0.9322,
+      "step": 4775
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 3.006322611761314e-08,
+      "loss": 0.9379,
+      "step": 4780
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 2.8259748981219194e-08,
+      "loss": 0.9136,
+      "step": 4785
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 2.651189960882039e-08,
+      "loss": 0.9764,
+      "step": 4790
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 2.4819697556737742e-08,
+      "loss": 0.9348,
+      "step": 4795
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 2.318316175866697e-08,
+      "loss": 0.9345,
+      "step": 4800
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 2.1602310525466464e-08,
+      "loss": 0.8879,
+      "step": 4805
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 2.007716154494965e-08,
+      "loss": 0.9619,
+      "step": 4810
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.8607731881690737e-08,
+      "loss": 0.9516,
+      "step": 4815
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.7194037976831502e-08,
+      "loss": 0.9471,
+      "step": 4820
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.583609564789812e-08,
+      "loss": 0.9197,
+      "step": 4825
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.4533920088623533e-08,
+      "loss": 0.8611,
+      "step": 4830
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.3287525868778128e-08,
+      "loss": 0.9449,
+      "step": 4835
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.2096926934007103e-08,
+      "loss": 0.9418,
+      "step": 4840
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.0962136605673357e-08,
+      "loss": 0.9337,
+      "step": 4845
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 9.883167580709285e-09,
+      "loss": 0.9118,
+      "step": 4850
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 8.860031931473555e-09,
+      "loss": 0.9563,
+      "step": 4855
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 7.892741105617329e-09,
+      "loss": 0.9342,
+      "step": 4860
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 6.981305925956583e-09,
+      "loss": 0.9553,
+      "step": 4865
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 6.1257365903488745e-09,
+      "loss": 0.9455,
+      "step": 4870
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5.326042671580655e-09,
+      "loss": 0.8813,
+      "step": 4875
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 4.582233117260693e-09,
+      "loss": 0.8929,
+      "step": 4880
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 3.894316249717922e-09,
+      "loss": 0.9463,
+      "step": 4885
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 3.2622997659120802e-09,
+      "loss": 0.9428,
+      "step": 4890
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 2.6861907373432193e-09,
+      "loss": 0.866,
+      "step": 4895
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 2.165995609973992e-09,
+      "loss": 0.94,
+      "step": 4900
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.7017202041602621e-09,
+      "loss": 0.9525,
+      "step": 4905
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.293369714582271e-09,
+      "loss": 0.9548,
+      "step": 4910
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 9.409487101880167e-10,
+      "loss": 0.9668,
+      "step": 4915
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 6.444611341432927e-10,
+      "loss": 0.9349,
+      "step": 4920
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 4.0391030378561513e-10,
+      "loss": 0.974,
+      "step": 4925
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 2.1929891058758424e-10,
+      "loss": 0.9563,
+      "step": 4930
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 9.0629020127464e-11,
+      "loss": 0.9373,
+      "step": 4935
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.790207206586736e-11,
+      "loss": 0.9326,
+      "step": 4940
+    },
+    {
+      "epoch": 4.0,
+      "step": 4944,
+      "total_flos": 2.446826463366742e+18,
+      "train_loss": 1.1695684537918436,
+      "train_runtime": 57751.42,
+      "train_samples_per_second": 5.478,
+      "train_steps_per_second": 0.086
     }
   ],
-  "max_steps": 1851,
-  "num_train_epochs": 3,
-  "total_flos": 1.8340911973547377e+18,
+  "max_steps": 4944,
+  "num_train_epochs": 4,
+  "total_flos": 2.446826463366742e+18,
   "trial_name": null,
   "trial_params": null
 }