diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4819 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8313847752663029,
+  "eval_steps": 500,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 29.612,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 27.4951,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5e-05,
+      "loss": 24.1786,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2e-05,
+      "loss": 19.8991,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5e-05,
+      "loss": 14.5587,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-05,
+      "loss": 10.1411,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5e-05,
+      "loss": 8.4604,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4e-05,
+      "loss": 8.2205,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5e-05,
+      "loss": 8.1131,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 8.0689,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 8.0342,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6e-05,
+      "loss": 7.9397,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 7.8807,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7e-05,
+      "loss": 7.7655,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 7.5968,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8e-05,
+      "loss": 7.4029,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.5e-05,
+      "loss": 7.204,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9e-05,
+      "loss": 6.9655,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.5e-05,
+      "loss": 6.7221,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001,
+      "loss": 6.5405,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.999972205865686e-05,
+      "loss": 6.3499,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.999888823771751e-05,
+      "loss": 6.1354,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.999749854645204e-05,
+      "loss": 5.9836,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99955530003106e-05,
+      "loss": 5.8913,
+      "step": 120
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.99930516209231e-05,
+      "loss": 5.694,
+      "step": 125
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.998999443609897e-05,
+      "loss": 5.5465,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.998638147982696e-05,
+      "loss": 5.5411,
+      "step": 135
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.998221279227467e-05,
+      "loss": 5.4117,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.997748841978812e-05,
+      "loss": 5.3572,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.997220841489122e-05,
+      "loss": 5.241,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.996637283628528e-05,
+      "loss": 5.0873,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.995998174884821e-05,
+      "loss": 5.0753,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.995303522363394e-05,
+      "loss": 5.0101,
+      "step": 165
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.99455333378715e-05,
+      "loss": 5.0237,
+      "step": 170
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.993747617496428e-05,
+      "loss": 4.9099,
+      "step": 175
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.9928863824489e-05,
+      "loss": 4.8638,
+      "step": 180
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.99196963821948e-05,
+      "loss": 4.8967,
+      "step": 185
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.990997395000217e-05,
+      "loss": 4.7744,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.989969663600169e-05,
+      "loss": 4.7181,
+      "step": 195
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.9888864554453e-05,
+      "loss": 4.7289,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.987747782578342e-05,
+      "loss": 4.6563,
+      "step": 205
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.986553657658668e-05,
+      "loss": 4.6653,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.985304093962145e-05,
+      "loss": 4.6279,
+      "step": 215
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.983999105380988e-05,
+      "loss": 4.5588,
+      "step": 220
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.982638706423608e-05,
+      "loss": 4.4593,
+      "step": 225
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.98122291221445e-05,
+      "loss": 4.5562,
+      "step": 230
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.979751738493826e-05,
+      "loss": 4.5082,
+      "step": 235
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.978225201617732e-05,
+      "loss": 4.3989,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.976643318557678e-05,
+      "loss": 4.4153,
+      "step": 245
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.975006106900495e-05,
+      "loss": 4.3993,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.973313584848132e-05,
+      "loss": 4.3516,
+      "step": 255
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.971565771217464e-05,
+      "loss": 4.4179,
+      "step": 260
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.969762685440076e-05,
+      "loss": 4.3085,
+      "step": 265
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.967904347562054e-05,
+      "loss": 4.319,
+      "step": 270
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.965990778243755e-05,
+      "loss": 4.1591,
+      "step": 275
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.964021998759577e-05,
+      "loss": 4.2833,
+      "step": 280
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.961998030997733e-05,
+      "loss": 4.2406,
+      "step": 285
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.95991889745999e-05,
+      "loss": 4.1649,
+      "step": 290
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.957784621261441e-05,
+      "loss": 4.1525,
+      "step": 295
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.955595226130226e-05,
+      "loss": 4.2209,
+      "step": 300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.953350736407282e-05,
+      "loss": 4.1854,
+      "step": 305
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.951051177046069e-05,
+      "loss": 4.0651,
+      "step": 310
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.948696573612292e-05,
+      "loss": 4.1472,
+      "step": 315
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.946286952283618e-05,
+      "loss": 4.113,
+      "step": 320
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.943822339849381e-05,
+      "loss": 4.1522,
+      "step": 325
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.941302763710288e-05,
+      "loss": 4.0967,
+      "step": 330
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.938728251878116e-05,
+      "loss": 4.15,
+      "step": 335
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.936098832975393e-05,
+      "loss": 3.9988,
+      "step": 340
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.933414536235091e-05,
+      "loss": 4.0301,
+      "step": 345
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.93067539150029e-05,
+      "loss": 4.0576,
+      "step": 350
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.927881429223853e-05,
+      "loss": 3.981,
+      "step": 355
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.925032680468085e-05,
+      "loss": 3.9577,
+      "step": 360
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.922129176904388e-05,
+      "loss": 3.9257,
+      "step": 365
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.919170950812911e-05,
+      "loss": 3.9146,
+      "step": 370
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.916158035082184e-05,
+      "loss": 4.0223,
+      "step": 375
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.913090463208763e-05,
+      "loss": 3.9243,
+      "step": 380
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.90996826929685e-05,
+      "loss": 3.9953,
+      "step": 385
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.906791488057916e-05,
+      "loss": 3.9585,
+      "step": 390
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.903560154810313e-05,
+      "loss": 3.8633,
+      "step": 395
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.900274305478887e-05,
+      "loss": 3.951,
+      "step": 400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.896933976594572e-05,
+      "loss": 3.9185,
+      "step": 405
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.893539205293989e-05,
+      "loss": 3.8792,
+      "step": 410
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.890090029319028e-05,
+      "loss": 3.9379,
+      "step": 415
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.886586487016433e-05,
+      "loss": 3.8348,
+      "step": 420
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.883028617337378e-05,
+      "loss": 3.9279,
+      "step": 425
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.879416459837022e-05,
+      "loss": 3.901,
+      "step": 430
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.875750054674082e-05,
+      "loss": 3.8083,
+      "step": 435
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.872029442610382e-05,
+      "loss": 3.8148,
+      "step": 440
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.8682546650104e-05,
+      "loss": 3.864,
+      "step": 445
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.864425763840802e-05,
+      "loss": 3.832,
+      "step": 450
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.860542781669988e-05,
+      "loss": 3.8532,
+      "step": 455
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.85660576166761e-05,
+      "loss": 3.8043,
+      "step": 460
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.852614747604093e-05,
+      "loss": 3.7628,
+      "step": 465
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.848569783850145e-05,
+      "loss": 3.7985,
+      "step": 470
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.844470915376278e-05,
+      "loss": 3.847,
+      "step": 475
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.840318187752292e-05,
+      "loss": 3.8247,
+      "step": 480
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.836111647146771e-05,
+      "loss": 3.825,
+      "step": 485
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.831851340326577e-05,
+      "loss": 3.7728,
+      "step": 490
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.82753731465633e-05,
+      "loss": 3.7376,
+      "step": 495
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.823169618097871e-05,
+      "loss": 3.7679,
+      "step": 500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.81874829920974e-05,
+      "loss": 3.711,
+      "step": 505
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.814273407146623e-05,
+      "loss": 3.7295,
+      "step": 510
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.809744991658829e-05,
+      "loss": 3.7091,
+      "step": 515
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.805163103091708e-05,
+      "loss": 3.7569,
+      "step": 520
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.800527792385112e-05,
+      "loss": 3.7087,
+      "step": 525
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.79583911107282e-05,
+      "loss": 3.6954,
+      "step": 530
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.791097111281968e-05,
+      "loss": 3.6419,
+      "step": 535
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.786301845732467e-05,
+      "loss": 3.6614,
+      "step": 540
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.781453367736418e-05,
+      "loss": 3.7198,
+      "step": 545
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.776551731197524e-05,
+      "loss": 3.7645,
+      "step": 550
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.771596990610478e-05,
+      "loss": 3.6879,
+      "step": 555
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.766589201060372e-05,
+      "loss": 3.7157,
+      "step": 560
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.761528418222077e-05,
+      "loss": 3.6085,
+      "step": 565
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.756414698359624e-05,
+      "loss": 3.6402,
+      "step": 570
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.75124809832558e-05,
+      "loss": 3.6653,
+      "step": 575
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.746028675560413e-05,
+      "loss": 3.6725,
+      "step": 580
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.740756488091861e-05,
+      "loss": 3.7271,
+      "step": 585
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.735431594534277e-05,
+      "loss": 3.7222,
+      "step": 590
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.730054054087983e-05,
+      "loss": 3.608,
+      "step": 595
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.724623926538612e-05,
+      "loss": 3.6822,
+      "step": 600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.719141272256443e-05,
+      "loss": 3.7206,
+      "step": 605
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.713606152195726e-05,
+      "loss": 3.6911,
+      "step": 610
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.708018627894011e-05,
+      "loss": 3.5546,
+      "step": 615
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.702378761471456e-05,
+      "loss": 3.6709,
+      "step": 620
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.696686615630146e-05,
+      "loss": 3.6084,
+      "step": 625
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.690942253653385e-05,
+      "loss": 3.6648,
+      "step": 630
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.685145739405002e-05,
+      "loss": 3.5705,
+      "step": 635
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.679297137328634e-05,
+      "loss": 3.6363,
+      "step": 640
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.673396512447013e-05,
+      "loss": 3.5641,
+      "step": 645
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.667443930361247e-05,
+      "loss": 3.5157,
+      "step": 650
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.661439457250076e-05,
+      "loss": 3.6104,
+      "step": 655
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.655383159869158e-05,
+      "loss": 3.5225,
+      "step": 660
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.649275105550309e-05,
+      "loss": 3.5442,
+      "step": 665
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.643115362200762e-05,
+      "loss": 3.4963,
+      "step": 670
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.636903998302409e-05,
+      "loss": 3.5083,
+      "step": 675
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.630641082911045e-05,
+      "loss": 3.5342,
+      "step": 680
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.624326685655593e-05,
+      "loss": 3.5517,
+      "step": 685
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.617960876737337e-05,
+      "loss": 3.5753,
+      "step": 690
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.611543726929134e-05,
+      "loss": 3.5652,
+      "step": 695
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.605075307574635e-05,
+      "loss": 3.4469,
+      "step": 700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.598555690587487e-05,
+      "loss": 3.4864,
+      "step": 705
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.591984948450532e-05,
+      "loss": 3.5477,
+      "step": 710
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.585363154215008e-05,
+      "loss": 3.5047,
+      "step": 715
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.578690381499728e-05,
+      "loss": 3.4961,
+      "step": 720
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.571966704490271e-05,
+      "loss": 3.4063,
+      "step": 725
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.565192197938148e-05,
+      "loss": 3.5615,
+      "step": 730
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.558366937159977e-05,
+      "loss": 3.4382,
+      "step": 735
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.551490998036646e-05,
+      "loss": 3.4905,
+      "step": 740
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.544564457012463e-05,
+      "loss": 3.4551,
+      "step": 745
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.537587391094314e-05,
+      "loss": 3.5484,
+      "step": 750
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.5305598778508e-05,
+      "loss": 3.4344,
+      "step": 755
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.52348199541138e-05,
+      "loss": 3.4585,
+      "step": 760
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.516353822465504e-05,
+      "loss": 3.5394,
+      "step": 765
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.509175438261726e-05,
+      "loss": 3.4481,
+      "step": 770
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.501946922606838e-05,
+      "loss": 3.4759,
+      "step": 775
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.494668355864973e-05,
+      "loss": 3.3645,
+      "step": 780
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.487339818956716e-05,
+      "loss": 3.4721,
+      "step": 785
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.479961393358203e-05,
+      "loss": 3.4211,
+      "step": 790
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.472533161100215e-05,
+      "loss": 3.449,
+      "step": 795
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.465055204767265e-05,
+      "loss": 3.4491,
+      "step": 800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.457527607496685e-05,
+      "loss": 3.4168,
+      "step": 805
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.44995045297769e-05,
+      "loss": 3.4716,
+      "step": 810
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.442323825450464e-05,
+      "loss": 3.5494,
+      "step": 815
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.43464780970521e-05,
+      "loss": 3.4089,
+      "step": 820
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.426922491081212e-05,
+      "loss": 3.3489,
+      "step": 825
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.419147955465888e-05,
+      "loss": 3.3706,
+      "step": 830
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.411324289293832e-05,
+      "loss": 3.4759,
+      "step": 835
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.403451579545859e-05,
+      "loss": 3.3906,
+      "step": 840
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.395529913748025e-05,
+      "loss": 3.3636,
+      "step": 845
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.387559379970672e-05,
+      "loss": 3.3434,
+      "step": 850
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.379540066827431e-05,
+      "loss": 3.3326,
+      "step": 855
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.371472063474248e-05,
+      "loss": 3.3559,
+      "step": 860
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.363355459608394e-05,
+      "loss": 3.3652,
+      "step": 865
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.355190345467457e-05,
+      "loss": 3.2626,
+      "step": 870
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.346976811828352e-05,
+      "loss": 3.3978,
+      "step": 875
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.338714950006297e-05,
+      "loss": 3.2891,
+      "step": 880
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.330404851853817e-05,
+      "loss": 3.3227,
+      "step": 885
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.3220466097597e-05,
+      "loss": 3.301,
+      "step": 890
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.313640316647991e-05,
+      "loss": 3.3792,
+      "step": 895
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.305186065976945e-05,
+      "loss": 3.2983,
+      "step": 900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.296683951737993e-05,
+      "loss": 3.4138,
+      "step": 905
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.288134068454697e-05,
+      "loss": 3.3116,
+      "step": 910
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.2795365111817e-05,
+      "loss": 3.4596,
+      "step": 915
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.270891375503665e-05,
+      "loss": 3.2271,
+      "step": 920
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.262198757534218e-05,
+      "loss": 3.2484,
+      "step": 925
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.253458753914874e-05,
+      "loss": 3.2716,
+      "step": 930
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.244671461813969e-05,
+      "loss": 3.2441,
+      "step": 935
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.235836978925572e-05,
+      "loss": 3.3062,
+      "step": 940
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.226955403468406e-05,
+      "loss": 3.2744,
+      "step": 945
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.21802683418475e-05,
+      "loss": 3.386,
+      "step": 950
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.209051370339347e-05,
+      "loss": 3.2451,
+      "step": 955
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.200029111718295e-05,
+      "loss": 3.2854,
+      "step": 960
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.190960158627941e-05,
+      "loss": 3.2094,
+      "step": 965
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.181844611893766e-05,
+      "loss": 3.2813,
+      "step": 970
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.172682572859261e-05,
+      "loss": 3.2688,
+      "step": 975
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.163474143384806e-05,
+      "loss": 3.2644,
+      "step": 980
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.154219425846528e-05,
+      "loss": 3.1576,
+      "step": 985
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.144918523135175e-05,
+      "loss": 3.2225,
+      "step": 990
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.13557153865496e-05,
+      "loss": 3.2942,
+      "step": 995
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.12617857632242e-05,
+      "loss": 3.2783,
+      "step": 1000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.116739740565259e-05,
+      "loss": 3.3015,
+      "step": 1005
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.107255136321184e-05,
+      "loss": 3.2027,
+      "step": 1010
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.09772486903674e-05,
+      "loss": 3.1369,
+      "step": 1015
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.08814904466614e-05,
+      "loss": 3.176,
+      "step": 1020
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.078527769670085e-05,
+      "loss": 3.1704,
+      "step": 1025
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.068861151014575e-05,
+      "loss": 3.1757,
+      "step": 1030
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.05914929616973e-05,
+      "loss": 3.2496,
+      "step": 1035
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.04939231310859e-05,
+      "loss": 3.1119,
+      "step": 1040
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.039590310305914e-05,
+      "loss": 3.1818,
+      "step": 1045
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.029743396736974e-05,
+      "loss": 3.2285,
+      "step": 1050
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.019851681876348e-05,
+      "loss": 3.2084,
+      "step": 1055
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.009915275696693e-05,
+      "loss": 3.1429,
+      "step": 1060
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.999934288667534e-05,
+      "loss": 3.2104,
+      "step": 1065
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.989908831754028e-05,
+      "loss": 3.1636,
+      "step": 1070
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.979839016415735e-05,
+      "loss": 3.1568,
+      "step": 1075
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.969724954605373e-05,
+      "loss": 3.128,
+      "step": 1080
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.959566758767581e-05,
+      "loss": 3.1145,
+      "step": 1085
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.949364541837661e-05,
+      "loss": 3.1392,
+      "step": 1090
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.939118417240329e-05,
+      "loss": 3.2515,
+      "step": 1095
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.92882849888845e-05,
+      "loss": 3.1553,
+      "step": 1100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.918494901181773e-05,
+      "loss": 3.2006,
+      "step": 1105
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.908117739005659e-05,
+      "loss": 3.1309,
+      "step": 1110
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.897697127729805e-05,
+      "loss": 3.0871,
+      "step": 1115
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.887233183206957e-05,
+      "loss": 3.1507,
+      "step": 1120
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.876726021771627e-05,
+      "loss": 3.2058,
+      "step": 1125
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.866175760238798e-05,
+      "loss": 3.1766,
+      "step": 1130
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.855582515902625e-05,
+      "loss": 3.1608,
+      "step": 1135
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.844946406535131e-05,
+      "loss": 3.1466,
+      "step": 1140
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.834267550384893e-05,
+      "loss": 3.2594,
+      "step": 1145
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.823546066175741e-05,
+      "loss": 3.1745,
+      "step": 1150
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.81278207310542e-05,
+      "loss": 3.1191,
+      "step": 1155
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.801975690844278e-05,
+      "loss": 3.1043,
+      "step": 1160
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.791127039533934e-05,
+      "loss": 3.1244,
+      "step": 1165
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.780236239785935e-05,
+      "loss": 3.1039,
+      "step": 1170
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.76930341268042e-05,
+      "loss": 3.1762,
+      "step": 1175
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.758328679764776e-05,
+      "loss": 3.1063,
+      "step": 1180
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.747312163052284e-05,
+      "loss": 3.0934,
+      "step": 1185
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.736253985020761e-05,
+      "loss": 3.1471,
+      "step": 1190
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.725154268611203e-05,
+      "loss": 3.0872,
+      "step": 1195
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.714013137226411e-05,
+      "loss": 3.1054,
+      "step": 1200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.702830714729628e-05,
+      "loss": 3.167,
+      "step": 1205
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.691607125443153e-05,
+      "loss": 3.1427,
+      "step": 1210
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.680342494146967e-05,
+      "loss": 3.1191,
+      "step": 1215
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.66903694607734e-05,
+      "loss": 3.0145,
+      "step": 1220
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.65769060692544e-05,
+      "loss": 3.0949,
+      "step": 1225
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.646303602835936e-05,
+      "loss": 3.1046,
+      "step": 1230
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.634876060405597e-05,
+      "loss": 3.0742,
+      "step": 1235
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.623408106681884e-05,
+      "loss": 2.9809,
+      "step": 1240
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.611899869161535e-05,
+      "loss": 3.0949,
+      "step": 1245
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.600351475789147e-05,
+      "loss": 3.0794,
+      "step": 1250
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.588763054955764e-05,
+      "loss": 3.1114,
+      "step": 1255
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.57713473549743e-05,
+      "loss": 2.9903,
+      "step": 1260
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.565466646693778e-05,
+      "loss": 3.1206,
+      "step": 1265
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.553758918266578e-05,
+      "loss": 3.0639,
+      "step": 1270
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.5420116803783e-05,
+      "loss": 3.1448,
+      "step": 1275
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.530225063630668e-05,
+      "loss": 3.1009,
+      "step": 1280
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.518399199063205e-05,
+      "loss": 3.0512,
+      "step": 1285
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.50653421815178e-05,
+      "loss": 3.0304,
+      "step": 1290
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.494630252807138e-05,
+      "loss": 3.0744,
+      "step": 1295
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.482687435373449e-05,
+      "loss": 3.0154,
+      "step": 1300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.470705898626817e-05,
+      "loss": 3.0791,
+      "step": 1305
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.458685775773822e-05,
+      "loss": 3.0064,
+      "step": 1310
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.446627200450025e-05,
+      "loss": 3.0958,
+      "step": 1315
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.434530306718493e-05,
+      "loss": 2.9886,
+      "step": 1320
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.4223952290683e-05,
+      "loss": 2.9943,
+      "step": 1325
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.41022210241304e-05,
+      "loss": 3.0203,
+      "step": 1330
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.398011062089316e-05,
+      "loss": 2.9917,
+      "step": 1335
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.385762243855249e-05,
+      "loss": 3.0321,
+      "step": 1340
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.373475783888958e-05,
+      "loss": 3.0035,
+      "step": 1345
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.36115181878705e-05,
+      "loss": 2.9577,
+      "step": 1350
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.348790485563101e-05,
+      "loss": 3.01,
+      "step": 1355
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.336391921646134e-05,
+      "loss": 3.0572,
+      "step": 1360
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.323956264879089e-05,
+      "loss": 3.0401,
+      "step": 1365
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.311483653517294e-05,
+      "loss": 3.0022,
+      "step": 1370
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.298974226226919e-05,
+      "loss": 3.0551,
+      "step": 1375
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.28642812208345e-05,
+      "loss": 2.966,
+      "step": 1380
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.273845480570123e-05,
+      "loss": 2.9784,
+      "step": 1385
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.26122644157639e-05,
+      "loss": 2.9453,
+      "step": 1390
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.248571145396362e-05,
+      "loss": 3.0228,
+      "step": 1395
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.235879732727236e-05,
+      "loss": 3.0858,
+      "step": 1400
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.223152344667745e-05,
+      "loss": 2.9494,
+      "step": 1405
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.21038912271658e-05,
+      "loss": 3.0174,
+      "step": 1410
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.197590208770824e-05,
+      "loss": 3.0147,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.184755745124371e-05,
+      "loss": 2.9655,
+      "step": 1420
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.171885874466342e-05,
+      "loss": 3.0138,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.158980739879507e-05,
+      "loss": 2.9664,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.146040484838677e-05,
+      "loss": 3.029,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.133065253209132e-05,
+      "loss": 2.976,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.120055189245e-05,
+      "loss": 2.9549,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.10701043758767e-05,
+      "loss": 2.8932,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.093931143264174e-05,
+      "loss": 2.9468,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.080817451685576e-05,
+      "loss": 3.0075,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.067669508645356e-05,
+      "loss": 2.8236,
+      "step": 1465
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.054487460317797e-05,
+      "loss": 3.0143,
+      "step": 1470
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.041271453256345e-05,
+      "loss": 3.0258,
+      "step": 1475
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.02802163439199e-05,
+      "loss": 2.9206,
+      "step": 1480
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.01473815103163e-05,
+      "loss": 2.9468,
+      "step": 1485
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.001421150856434e-05,
+      "loss": 2.864,
+      "step": 1490
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.988070781920197e-05,
+      "loss": 2.9844,
+      "step": 1495
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.9746871926477e-05,
+      "loss": 2.9748,
+      "step": 1500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.961270531833052e-05,
+      "loss": 2.9793,
+      "step": 1505
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.947820948638045e-05,
+      "loss": 2.9883,
+      "step": 1510
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.934338592590486e-05,
+      "loss": 3.057,
+      "step": 1515
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.92082361358254e-05,
+      "loss": 2.8892,
+      "step": 1520
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.907276161869065e-05,
+      "loss": 3.0292,
+      "step": 1525
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.893696388065936e-05,
+      "loss": 2.9495,
+      "step": 1530
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.88008444314838e-05,
+      "loss": 2.9831,
+      "step": 1535
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.866440478449283e-05,
+      "loss": 2.9176,
+      "step": 1540
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.852764645657522e-05,
+      "loss": 2.8421,
+      "step": 1545
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.839057096816271e-05,
+      "loss": 2.9493,
+      "step": 1550
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.82531798432131e-05,
+      "loss": 2.8956,
+      "step": 1555
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.811547460919333e-05,
+      "loss": 2.9113,
+      "step": 1560
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.797745679706254e-05,
+      "loss": 2.8989,
+      "step": 1565
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.783912794125496e-05,
+      "loss": 2.9093,
+      "step": 1570
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.770048957966291e-05,
+      "loss": 2.8537,
+      "step": 1575
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.756154325361967e-05,
+      "loss": 2.8785,
+      "step": 1580
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.74222905078824e-05,
+      "loss": 2.8461,
+      "step": 1585
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.728273289061489e-05,
+      "loss": 2.9334,
+      "step": 1590
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.714287195337044e-05,
+      "loss": 2.929,
+      "step": 1595
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.700270925107448e-05,
+      "loss": 2.9332,
+      "step": 1600
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.686224634200742e-05,
+      "loss": 2.9095,
+      "step": 1605
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.672148478778722e-05,
+      "loss": 2.9405,
+      "step": 1610
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.658042615335212e-05,
+      "loss": 2.8898,
+      "step": 1615
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.643907200694318e-05,
+      "loss": 2.9084,
+      "step": 1620
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.629742392008684e-05,
+      "loss": 2.8985,
+      "step": 1625
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.615548346757749e-05,
+      "loss": 2.9177,
+      "step": 1630
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.60132522274599e-05,
+      "loss": 2.8662,
+      "step": 1635
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.587073178101178e-05,
+      "loss": 2.9329,
+      "step": 1640
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.572792371272609e-05,
+      "loss": 2.8451,
+      "step": 1645
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.55848296102935e-05,
+      "loss": 2.9332,
+      "step": 1650
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.544145106458465e-05,
+      "loss": 2.839,
+      "step": 1655
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.529778966963259e-05,
+      "loss": 2.9092,
+      "step": 1660
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.515384702261496e-05,
+      "loss": 2.9198,
+      "step": 1665
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.500962472383627e-05,
+      "loss": 2.7638,
+      "step": 1670
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.486512437671011e-05,
+      "loss": 2.9139,
+      "step": 1675
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.472034758774128e-05,
+      "loss": 2.8766,
+      "step": 1680
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.457529596650797e-05,
+      "loss": 2.8897,
+      "step": 1685
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.442997112564392e-05,
+      "loss": 2.7989,
+      "step": 1690
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.428437468082037e-05,
+      "loss": 2.8649,
+      "step": 1695
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.413850825072817e-05,
+      "loss": 2.9498,
+      "step": 1700
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.39923734570598e-05,
+      "loss": 2.8839,
+      "step": 1705
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.384597192449126e-05,
+      "loss": 2.8805,
+      "step": 1710
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.369930528066412e-05,
+      "loss": 2.8905,
+      "step": 1715
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.355237515616732e-05,
+      "loss": 2.8185,
+      "step": 1720
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.340518318451914e-05,
+      "loss": 2.851,
+      "step": 1725
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.325773100214893e-05,
+      "loss": 2.8218,
+      "step": 1730
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.311002024837899e-05,
+      "loss": 2.8796,
+      "step": 1735
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.296205256540633e-05,
+      "loss": 2.8326,
+      "step": 1740
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.281382959828443e-05,
+      "loss": 2.8062,
+      "step": 1745
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.26653529949049e-05,
+      "loss": 2.8478,
+      "step": 1750
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.25166244059792e-05,
+      "loss": 2.8806,
+      "step": 1755
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.236764548502029e-05,
+      "loss": 2.8154,
+      "step": 1760
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.221841788832421e-05,
+      "loss": 2.8577,
+      "step": 1765
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.206894327495173e-05,
+      "loss": 2.798,
+      "step": 1770
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.191922330670982e-05,
+      "loss": 2.8287,
+      "step": 1775
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.176925964813326e-05,
+      "loss": 2.9241,
+      "step": 1780
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.161905396646607e-05,
+      "loss": 2.8994,
+      "step": 1785
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.146860793164299e-05,
+      "loss": 2.7951,
+      "step": 1790
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.131792321627098e-05,
+      "loss": 2.9515,
+      "step": 1795
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.116700149561048e-05,
+      "loss": 2.8669,
+      "step": 1800
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.101584444755696e-05,
+      "loss": 2.9084,
+      "step": 1805
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.086445375262212e-05,
+      "loss": 2.8364,
+      "step": 1810
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.071283109391528e-05,
+      "loss": 2.8491,
+      "step": 1815
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.056097815712466e-05,
+      "loss": 2.8712,
+      "step": 1820
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.040889663049862e-05,
+      "loss": 2.8808,
+      "step": 1825
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.025658820482693e-05,
+      "loss": 2.8055,
+      "step": 1830
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.010405457342192e-05,
+      "loss": 2.7946,
+      "step": 1835
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 6.995129743209967e-05,
+      "loss": 2.839,
+      "step": 1840
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 6.97983184791612e-05,
+      "loss": 2.749,
+      "step": 1845
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 6.964511941537355e-05,
+      "loss": 2.776,
+      "step": 1850
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.949170194395083e-05,
+      "loss": 2.8686,
+      "step": 1855
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.933806777053536e-05,
+      "loss": 2.8789,
+      "step": 1860
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.918421860317872e-05,
+      "loss": 2.7964,
+      "step": 1865
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.903015615232263e-05,
+      "loss": 2.8079,
+      "step": 1870
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.887588213078012e-05,
+      "loss": 2.7974,
+      "step": 1875
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.87213982537163e-05,
+      "loss": 2.8404,
+      "step": 1880
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.856670623862943e-05,
+      "loss": 2.7915,
+      "step": 1885
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.841180780533179e-05,
+      "loss": 2.8744,
+      "step": 1890
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.82567046759305e-05,
+      "loss": 2.7534,
+      "step": 1895
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.810139857480844e-05,
+      "loss": 2.8097,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.794589122860509e-05,
+      "loss": 2.7996,
+      "step": 1905
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.779018436619725e-05,
+      "loss": 2.9034,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.763427971867992e-05,
+      "loss": 2.724,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.747817901934699e-05,
+      "loss": 2.812,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.732188400367197e-05,
+      "loss": 2.8425,
+      "step": 1925
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.716539640928871e-05,
+      "loss": 2.9424,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.70087179759721e-05,
+      "loss": 2.6963,
+      "step": 1935
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.685185044561874e-05,
+      "loss": 2.788,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.669479556222747e-05,
+      "loss": 2.8517,
+      "step": 1945
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.653755507188013e-05,
+      "loss": 2.8291,
+      "step": 1950
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.638013072272205e-05,
+      "loss": 2.8426,
+      "step": 1955
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.622252426494259e-05,
+      "loss": 2.7857,
+      "step": 1960
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.606473745075581e-05,
+      "loss": 2.7813,
+      "step": 1965
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.590677203438084e-05,
+      "loss": 2.7745,
+      "step": 1970
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.574862977202252e-05,
+      "loss": 2.7734,
+      "step": 1975
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.559031242185174e-05,
+      "loss": 2.7842,
+      "step": 1980
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.543182174398597e-05,
+      "loss": 2.7615,
+      "step": 1985
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.52731595004697e-05,
+      "loss": 2.8467,
+      "step": 1990
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.51143274552548e-05,
+      "loss": 2.7726,
+      "step": 1995
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.495532737418098e-05,
+      "loss": 2.7732,
+      "step": 2000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.479616102495605e-05,
+      "loss": 2.8283,
+      "step": 2005
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.463683017713638e-05,
+      "loss": 2.8062,
+      "step": 2010
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.447733660210715e-05,
+      "loss": 2.7754,
+      "step": 2015
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.431768207306272e-05,
+      "loss": 2.7978,
+      "step": 2020
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.415786836498684e-05,
+      "loss": 2.8643,
+      "step": 2025
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.399789725463298e-05,
+      "loss": 2.8335,
+      "step": 2030
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.383777052050458e-05,
+      "loss": 2.8009,
+      "step": 2035
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.367748994283518e-05,
+      "loss": 2.7856,
+      "step": 2040
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.351705730356877e-05,
+      "loss": 2.7322,
+      "step": 2045
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.335647438633987e-05,
+      "loss": 2.7436,
+      "step": 2050
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.319574297645374e-05,
+      "loss": 2.7457,
+      "step": 2055
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.303486486086654e-05,
+      "loss": 2.718,
+      "step": 2060
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.287384182816546e-05,
+      "loss": 2.7478,
+      "step": 2065
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.271267566854883e-05,
+      "loss": 2.813,
+      "step": 2070
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.255136817380618e-05,
+      "loss": 2.8196,
+      "step": 2075
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.23899211372984e-05,
+      "loss": 2.7727,
+      "step": 2080
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.222833635393772e-05,
+      "loss": 2.7919,
+      "step": 2085
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.206661562016782e-05,
+      "loss": 2.7548,
+      "step": 2090
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.190476073394382e-05,
+      "loss": 2.8187,
+      "step": 2095
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.17427734947123e-05,
+      "loss": 2.7466,
+      "step": 2100
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.158065570339127e-05,
+      "loss": 2.7092,
+      "step": 2105
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.141840916235021e-05,
+      "loss": 2.8063,
+      "step": 2110
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.125603567539001e-05,
+      "loss": 2.6874,
+      "step": 2115
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.109353704772284e-05,
+      "loss": 2.7834,
+      "step": 2120
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.0930915085952164e-05,
+      "loss": 2.7841,
+      "step": 2125
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.076817159805267e-05,
+      "loss": 2.7499,
+      "step": 2130
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.06053083933501e-05,
+      "loss": 2.8277,
+      "step": 2135
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.044232728250116e-05,
+      "loss": 2.7049,
+      "step": 2140
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.027923007747339e-05,
+      "loss": 2.742,
+      "step": 2145
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.011601859152506e-05,
+      "loss": 2.7525,
+      "step": 2150
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.995269463918495e-05,
+      "loss": 2.7663,
+      "step": 2155
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.97892600362322e-05,
+      "loss": 2.7436,
+      "step": 2160
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.962571659967614e-05,
+      "loss": 2.7198,
+      "step": 2165
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.946206614773606e-05,
+      "loss": 2.7422,
+      "step": 2170
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.929831049982103e-05,
+      "loss": 2.7783,
+      "step": 2175
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.9134451476509633e-05,
+      "loss": 2.7628,
+      "step": 2180
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.897049089952974e-05,
+      "loss": 2.7292,
+      "step": 2185
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.880643059173826e-05,
+      "loss": 2.6699,
+      "step": 2190
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.864227237710093e-05,
+      "loss": 2.7718,
+      "step": 2195
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.847801808067189e-05,
+      "loss": 2.6958,
+      "step": 2200
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.831366952857357e-05,
+      "loss": 2.7366,
+      "step": 2205
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.814922854797622e-05,
+      "loss": 2.7661,
+      "step": 2210
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.798469696707775e-05,
+      "loss": 2.7467,
+      "step": 2215
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.782007661508331e-05,
+      "loss": 2.7477,
+      "step": 2220
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.765536932218495e-05,
+      "loss": 2.7503,
+      "step": 2225
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.7490576919541315e-05,
+      "loss": 2.6846,
+      "step": 2230
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.732570123925729e-05,
+      "loss": 2.6627,
+      "step": 2235
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7160744114363593e-05,
+      "loss": 2.7512,
+      "step": 2240
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.699570737879641e-05,
+      "loss": 2.7775,
+      "step": 2245
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.683059286737702e-05,
+      "loss": 2.667,
+      "step": 2250
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.666540241579139e-05,
+      "loss": 2.6693,
+      "step": 2255
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.6500137860569766e-05,
+      "loss": 2.7436,
+      "step": 2260
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.633480103906624e-05,
+      "loss": 2.7099,
+      "step": 2265
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.616939378943834e-05,
+      "loss": 2.7656,
+      "step": 2270
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.6003917950626595e-05,
+      "loss": 2.733,
+      "step": 2275
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.583837536233407e-05,
+      "loss": 2.7934,
+      "step": 2280
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.567276786500596e-05,
+      "loss": 2.7978,
+      "step": 2285
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.5507097299809054e-05,
+      "loss": 2.6622,
+      "step": 2290
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.534136550861133e-05,
+      "loss": 2.7831,
+      "step": 2295
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.5175574333961465e-05,
+      "loss": 2.738,
+      "step": 2300
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.500972561906832e-05,
+      "loss": 2.7164,
+      "step": 2305
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.484382120778048e-05,
+      "loss": 2.7275,
+      "step": 2310
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.467786294456575e-05,
+      "loss": 2.7034,
+      "step": 2315
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.451185267449061e-05,
+      "loss": 2.6887,
+      "step": 2320
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.43457922431998e-05,
+      "loss": 2.697,
+      "step": 2325
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.417968349689566e-05,
+      "loss": 2.6939,
+      "step": 2330
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.401352828231772e-05,
+      "loss": 2.6987,
+      "step": 2335
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.384732844672211e-05,
+      "loss": 2.6785,
+      "step": 2340
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.368108583786107e-05,
+      "loss": 2.677,
+      "step": 2345
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.3514802303962344e-05,
+      "loss": 2.6854,
+      "step": 2350
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.334847969370868e-05,
+      "loss": 2.7325,
+      "step": 2355
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.3182119856217284e-05,
+      "loss": 2.6739,
+      "step": 2360
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.3015724641019214e-05,
+      "loss": 2.7009,
+      "step": 2365
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.284929589803884e-05,
+      "loss": 2.6944,
+      "step": 2370
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.2682835477573336e-05,
+      "loss": 2.7387,
+      "step": 2375
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.2516345230271965e-05,
+      "loss": 2.6848,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.234982700711569e-05,
+      "loss": 2.7023,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.218328265939643e-05,
+      "loss": 2.7108,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.201671403869657e-05,
+      "loss": 2.786,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.1850122996868366e-05,
+      "loss": 2.7416,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.168351138601334e-05,
+      "loss": 2.7202,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.1516881058461675e-05,
+      "loss": 2.7309,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.135023386675166e-05,
+      "loss": 2.6839,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.118357166360906e-05,
+      "loss": 2.6701,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.101689630192655e-05,
+      "loss": 2.7146,
+      "step": 2425
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.085020963474307e-05,
+      "loss": 2.7234,
+      "step": 2430
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.068351351522329e-05,
+      "loss": 2.6783,
+      "step": 2435
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.0516809796636935e-05,
+      "loss": 2.7627,
+      "step": 2440
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.035010033233821e-05,
+      "loss": 2.6553,
+      "step": 2445
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.018338697574523e-05,
+      "loss": 2.7378,
+      "step": 2450
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.0016671580319354e-05,
+      "loss": 2.7161,
+      "step": 2455
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.984995599954461e-05,
+      "loss": 2.7021,
+      "step": 2460
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.968324208690712e-05,
+      "loss": 2.7127,
+      "step": 2465
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.951653169587441e-05,
+      "loss": 2.6836,
+      "step": 2470
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.93498266798749e-05,
+      "loss": 2.6705,
+      "step": 2475
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.918312889227722e-05,
+      "loss": 2.6761,
+      "step": 2480
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.901644018636966e-05,
+      "loss": 2.6141,
+      "step": 2485
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.8849762415339526e-05,
+      "loss": 2.759,
+      "step": 2490
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.868309743225256e-05,
+      "loss": 2.7271,
+      "step": 2495
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.851644709003233e-05,
+      "loss": 2.7086,
+      "step": 2500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.834981324143964e-05,
+      "loss": 2.6388,
+      "step": 2505
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.818319773905191e-05,
+      "loss": 2.715,
+      "step": 2510
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.801660243524261e-05,
+      "loss": 2.7208,
+      "step": 2515
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.7850029182160626e-05,
+      "loss": 2.6289,
+      "step": 2520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.768347983170973e-05,
+      "loss": 2.6604,
+      "step": 2525
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.7516956235527884e-05,
+      "loss": 2.6866,
+      "step": 2530
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.735046024496682e-05,
+      "loss": 2.7408,
+      "step": 2535
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.7183993711071286e-05,
+      "loss": 2.7182,
+      "step": 2540
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.7017558484558554e-05,
+      "loss": 2.6496,
+      "step": 2545
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.6851156415797844e-05,
+      "loss": 2.7255,
+      "step": 2550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.6684789354789746e-05,
+      "loss": 2.6329,
+      "step": 2555
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.651845915114563e-05,
+      "loss": 2.7216,
+      "step": 2560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.6352167654067095e-05,
+      "loss": 2.6893,
+      "step": 2565
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.618591671232544e-05,
+      "loss": 2.6181,
+      "step": 2570
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.601970817424106e-05,
+      "loss": 2.693,
+      "step": 2575
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.585354388766292e-05,
+      "loss": 2.6225,
+      "step": 2580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.568742569994802e-05,
+      "loss": 2.7006,
+      "step": 2585
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.552135545794086e-05,
+      "loss": 2.6518,
+      "step": 2590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.535533500795288e-05,
+      "loss": 2.6686,
+      "step": 2595
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.5189366195741953e-05,
+      "loss": 2.7356,
+      "step": 2600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.502345086649186e-05,
+      "loss": 2.6798,
+      "step": 2605
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.485759086479179e-05,
+      "loss": 2.7112,
+      "step": 2610
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.469178803461579e-05,
+      "loss": 2.615,
+      "step": 2615
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.4526044219302326e-05,
+      "loss": 2.6756,
+      "step": 2620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.4360361261533745e-05,
+      "loss": 2.6108,
+      "step": 2625
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.419474100331579e-05,
+      "loss": 2.6449,
+      "step": 2630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.402918528595715e-05,
+      "loss": 2.6129,
+      "step": 2635
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.386369595004896e-05,
+      "loss": 2.7304,
+      "step": 2640
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.3698274835444354e-05,
+      "loss": 2.6224,
+      "step": 2645
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.3532923781238e-05,
+      "loss": 2.6844,
+      "step": 2650
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.336764462574566e-05,
+      "loss": 2.6738,
+      "step": 2655
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.320243920648376e-05,
+      "loss": 2.6619,
+      "step": 2660
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.303730936014894e-05,
+      "loss": 2.5888,
+      "step": 2665
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.287225692259765e-05,
+      "loss": 2.7353,
+      "step": 2670
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.270728372882575e-05,
+      "loss": 2.6994,
+      "step": 2675
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.254239161294804e-05,
+      "loss": 2.642,
+      "step": 2680
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.237758240817802e-05,
+      "loss": 2.6403,
+      "step": 2685
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.2212857946807336e-05,
+      "loss": 2.6637,
+      "step": 2690
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.2048220060185516e-05,
+      "loss": 2.6056,
+      "step": 2695
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.188367057869957e-05,
+      "loss": 2.6637,
+      "step": 2700
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.171921133175365e-05,
+      "loss": 2.7226,
+      "step": 2705
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.155484414774872e-05,
+      "loss": 2.6303,
+      "step": 2710
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.139057085406221e-05,
+      "loss": 2.6647,
+      "step": 2715
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1226393277027726e-05,
+      "loss": 2.6761,
+      "step": 2720
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.106231324191471e-05,
+      "loss": 2.6062,
+      "step": 2725
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.089833257290817e-05,
+      "loss": 2.6803,
+      "step": 2730
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.073445309308842e-05,
+      "loss": 2.6647,
+      "step": 2735
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.0570676624410756e-05,
+      "loss": 2.662,
+      "step": 2740
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.040700498768525e-05,
+      "loss": 2.7006,
+      "step": 2745
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.024344000255648e-05,
+      "loss": 2.6133,
+      "step": 2750
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.0079983487483313e-05,
+      "loss": 2.654,
+      "step": 2755
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.9916637259718683e-05,
+      "loss": 2.5982,
+      "step": 2760
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.9753403135289396e-05,
+      "loss": 2.6992,
+      "step": 2765
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.9590282928975914e-05,
+      "loss": 2.6935,
+      "step": 2770
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.942727845429221e-05,
+      "loss": 2.6968,
+      "step": 2775
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.926439152346558e-05,
+      "loss": 2.667,
+      "step": 2780
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.910162394741653e-05,
+      "loss": 2.641,
+      "step": 2785
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.893897753573861e-05,
+      "loss": 2.5928,
+      "step": 2790
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.877645409667829e-05,
+      "loss": 2.6669,
+      "step": 2795
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.861405543711491e-05,
+      "loss": 2.6599,
+      "step": 2800
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.8451783362540507e-05,
+      "loss": 2.5583,
+      "step": 2805
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.828963967703983e-05,
+      "loss": 2.6561,
+      "step": 2810
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.8127626183270223e-05,
+      "loss": 2.6415,
+      "step": 2815
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.796574468244161e-05,
+      "loss": 2.6902,
+      "step": 2820
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.7803996974296444e-05,
+      "loss": 2.6692,
+      "step": 2825
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.7642384857089776e-05,
+      "loss": 2.6657,
+      "step": 2830
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.748091012756915e-05,
+      "loss": 2.6734,
+      "step": 2835
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.731957458095467e-05,
+      "loss": 2.6369,
+      "step": 2840
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.71583800109191e-05,
+      "loss": 2.5924,
+      "step": 2845
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.699732820956784e-05,
+      "loss": 2.6948,
+      "step": 2850
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.6836420967419057e-05,
+      "loss": 2.6539,
+      "step": 2855
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.6675660073383745e-05,
+      "loss": 2.5948,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.6515047314745856e-05,
+      "loss": 2.6483,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.6354584477142437e-05,
+      "loss": 2.6971,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.6194273344543736e-05,
+      "loss": 2.6479,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.6034115699233425e-05,
+      "loss": 2.6546,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.5874113321788736e-05,
+      "loss": 2.6514,
+      "step": 2885
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.571426799106071e-05,
+      "loss": 2.554,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.555458148415437e-05,
+      "loss": 2.6514,
+      "step": 2895
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.539505557640901e-05,
+      "loss": 2.579,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.523569204137843e-05,
+      "loss": 2.652,
+      "step": 2905
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.5076492650811246e-05,
+      "loss": 2.6642,
+      "step": 2910
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.491745917463113e-05,
+      "loss": 2.5986,
+      "step": 2915
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.475859338091721e-05,
+      "loss": 2.6119,
+      "step": 2920
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.4599897035884374e-05,
+      "loss": 2.5645,
+      "step": 2925
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.444137190386363e-05,
+      "loss": 2.6185,
+      "step": 2930
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.4283019747282514e-05,
+      "loss": 2.5903,
+      "step": 2935
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.412484232664545e-05,
+      "loss": 2.6423,
+      "step": 2940
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.396684140051424e-05,
+      "loss": 2.623,
+      "step": 2945
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.3809018725488466e-05,
+      "loss": 2.6353,
+      "step": 2950
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.365137605618598e-05,
+      "loss": 2.6697,
+      "step": 2955
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.3493915145223395e-05,
+      "loss": 2.6555,
+      "step": 2960
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.3336637743196584e-05,
+      "loss": 2.6587,
+      "step": 2965
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.317954559866126e-05,
+      "loss": 2.7183,
+      "step": 2970
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.302264045811344e-05,
+      "loss": 2.6083,
+      "step": 2975
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.286592406597021e-05,
+      "loss": 2.6429,
+      "step": 2980
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.270939816455012e-05,
+      "loss": 2.578,
+      "step": 2985
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.255306449405395e-05,
+      "loss": 2.6455,
+      "step": 2990
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.2396924792545304e-05,
+      "loss": 2.6339,
+      "step": 2995
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.224098079593132e-05,
+      "loss": 2.6968,
+      "step": 3000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.2085234237943354e-05,
+      "loss": 2.6141,
+      "step": 3005
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.19296868501177e-05,
+      "loss": 2.6455,
+      "step": 3010
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.177434036177636e-05,
+      "loss": 2.5694,
+      "step": 3015
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.1619196500007804e-05,
+      "loss": 2.6539,
+      "step": 3020
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.146425698964776e-05,
+      "loss": 2.5823,
+      "step": 3025
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.1309523553260046e-05,
+      "loss": 2.6408,
+      "step": 3030
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.115499791111743e-05,
+      "loss": 2.6327,
+      "step": 3035
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.10006817811825e-05,
+      "loss": 2.605,
+      "step": 3040
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.084657687908855e-05,
+      "loss": 2.6639,
+      "step": 3045
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.069268491812052e-05,
+      "loss": 2.6458,
+      "step": 3050
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.0539007609195934e-05,
+      "loss": 2.5843,
+      "step": 3055
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.0385546660845908e-05,
+      "loss": 2.6201,
+      "step": 3060
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.0232303779196132e-05,
+      "loss": 2.6606,
+      "step": 3065
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.0079280667947885e-05,
+      "loss": 2.6143,
+      "step": 3070
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.9926479028359132e-05,
+      "loss": 2.645,
+      "step": 3075
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.97739005592256e-05,
+      "loss": 2.6509,
+      "step": 3080
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.962154695686187e-05,
+      "loss": 2.5686,
+      "step": 3085
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.9469419915082536e-05,
+      "loss": 2.6267,
+      "step": 3090
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.9317521125183368e-05,
+      "loss": 2.6627,
+      "step": 3095
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.9165852275922524e-05,
+      "loss": 2.6267,
+      "step": 3100
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.901441505350174e-05,
+      "loss": 2.5988,
+      "step": 3105
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.886321114154762e-05,
+      "loss": 2.5767,
+      "step": 3110
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.87122422210929e-05,
+      "loss": 2.5978,
+      "step": 3115
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8561509970557736e-05,
+      "loss": 2.5346,
+      "step": 3120
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8411016065731146e-05,
+      "loss": 2.588,
+      "step": 3125
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.826076217975222e-05,
+      "loss": 2.463,
+      "step": 3130
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8110749983091632e-05,
+      "loss": 2.5405,
+      "step": 3135
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.7960981143533053e-05,
+      "loss": 2.6099,
+      "step": 3140
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.781145732615457e-05,
+      "loss": 2.5535,
+      "step": 3145
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.7662180193310218e-05,
+      "loss": 2.5949,
+      "step": 3150
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.751315140461145e-05,
+      "loss": 2.5891,
+      "step": 3155
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.7364372616908744e-05,
+      "loss": 2.5942,
+      "step": 3160
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.7215845484273152e-05,
+      "loss": 2.5888,
+      "step": 3165
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.7067571657977893e-05,
+      "loss": 2.5798,
+      "step": 3170
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.691955278648003e-05,
+      "loss": 2.5523,
+      "step": 3175
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6771790515402112e-05,
+      "loss": 2.5862,
+      "step": 3180
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6624286487513916e-05,
+      "loss": 2.6262,
+      "step": 3185
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6477042342714137e-05,
+      "loss": 2.5153,
+      "step": 3190
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.633005971801219e-05,
+      "loss": 2.6274,
+      "step": 3195
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6183340247510013e-05,
+      "loss": 2.6365,
+      "step": 3200
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6036885562383856e-05,
+      "loss": 2.6049,
+      "step": 3205
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.5890697290866206e-05,
+      "loss": 2.625,
+      "step": 3210
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.5744777058227642e-05,
+      "loss": 2.6267,
+      "step": 3215
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.5599126486758777e-05,
+      "loss": 2.6139,
+      "step": 3220
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.5453747195752243e-05,
+      "loss": 2.6867,
+      "step": 3225
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.530864080148464e-05,
+      "loss": 2.5539,
+      "step": 3230
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.5163808917198615e-05,
+      "loss": 2.5324,
+      "step": 3235
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.501925315308492e-05,
+      "loss": 2.6232,
+      "step": 3240
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.4874975116264477e-05,
+      "loss": 2.6176,
+      "step": 3245
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.4730976410770534e-05,
+      "loss": 2.6331,
+      "step": 3250
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.458725863753084e-05,
+      "loss": 2.5987,
+      "step": 3255
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.4443823394349834e-05,
+      "loss": 2.6177,
+      "step": 3260
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.430067227589088e-05,
+      "loss": 2.5567,
+      "step": 3265
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.4157806873658517e-05,
+      "loss": 2.5992,
+      "step": 3270
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.401522877598087e-05,
+      "loss": 2.6343,
+      "step": 3275
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.3872939567991827e-05,
+      "loss": 2.5103,
+      "step": 3280
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.373094083161353e-05,
+      "loss": 2.555,
+      "step": 3285
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.358923414553877e-05,
+      "loss": 2.5686,
+      "step": 3290
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.3447821085213405e-05,
+      "loss": 2.5766,
+      "step": 3295
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.3306703222818878e-05,
+      "loss": 2.5535,
+      "step": 3300
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.3165882127254705e-05,
+      "loss": 2.5784,
+      "step": 3305
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.302535936412108e-05,
+      "loss": 2.665,
+      "step": 3310
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.2885136495701415e-05,
+      "loss": 2.5652,
+      "step": 3315
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.274521508094501e-05,
+      "loss": 2.6428,
+      "step": 3320
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.2605596675449698e-05,
+      "loss": 2.6415,
+      "step": 3325
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.246628283144457e-05,
+      "loss": 2.5523,
+      "step": 3330
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.232727509777269e-05,
+      "loss": 2.5768,
+      "step": 3335
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.2188575019873932e-05,
+      "loss": 2.594,
+      "step": 3340
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.2050184139767704e-05,
+      "loss": 2.6318,
+      "step": 3345
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.191210399603591e-05,
+      "loss": 2.5975,
+      "step": 3350
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.1774336123805772e-05,
+      "loss": 2.6264,
+      "step": 3355
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.1636882054732776e-05,
+      "loss": 2.5242,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.1499743316983684e-05,
+      "loss": 2.5668,
+      "step": 3365
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.1362921435219473e-05,
+      "loss": 2.5771,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.1226417930578464e-05,
+      "loss": 2.5857,
+      "step": 3375
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.109023432065935e-05,
+      "loss": 2.5728,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.095437211950434e-05,
+      "loss": 2.6298,
+      "step": 3385
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.0818832837582352e-05,
+      "loss": 2.5506,
+      "step": 3390
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.068361798177218e-05,
+      "loss": 2.591,
+      "step": 3395
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.0548729055345778e-05,
+      "loss": 2.5689,
+      "step": 3400
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.0414167557951514e-05,
+      "loss": 2.5466,
+      "step": 3405
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.0279934985597527e-05,
+      "loss": 2.6142,
+      "step": 3410
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.0146032830635054e-05,
+      "loss": 2.5777,
+      "step": 3415
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.001246258174192e-05,
+      "loss": 2.6231,
+      "step": 3420
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9879225723905886e-05,
+      "loss": 2.5642,
+      "step": 3425
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9746323738408203e-05,
+      "loss": 2.5353,
+      "step": 3430
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9613758102807117e-05,
+      "loss": 2.6072,
+      "step": 3435
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9481530290921474e-05,
+      "loss": 2.4539,
+      "step": 3440
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.934964177281428e-05,
+      "loss": 2.6172,
+      "step": 3445
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9218094014776434e-05,
+      "loss": 2.547,
+      "step": 3450
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9086888479310333e-05,
+      "loss": 2.612,
+      "step": 3455
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.895602662511371e-05,
+      "loss": 2.5996,
+      "step": 3460
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.8825509907063327e-05,
+      "loss": 2.57,
+      "step": 3465
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.8695339776198872e-05,
+      "loss": 2.5647,
+      "step": 3470
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.8565517679706783e-05,
+      "loss": 2.6039,
+      "step": 3475
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.8436045060904174e-05,
+      "loss": 2.563,
+      "step": 3480
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.830692335922279e-05,
+      "loss": 2.6022,
+      "step": 3485
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.8178154010192994e-05,
+      "loss": 2.5485,
+      "step": 3490
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.8049738445427822e-05,
+      "loss": 2.4907,
+      "step": 3495
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7921678092607052e-05,
+      "loss": 2.5985,
+      "step": 3500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7793974375461352e-05,
+      "loss": 2.5648,
+      "step": 3505
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7666628713756417e-05,
+      "loss": 2.5742,
+      "step": 3510
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7539642523277228e-05,
+      "loss": 2.522,
+      "step": 3515
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7413017215812273e-05,
+      "loss": 2.5161,
+      "step": 3520
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.728675419913788e-05,
+      "loss": 2.5744,
+      "step": 3525
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.716085487700253e-05,
+      "loss": 2.5203,
+      "step": 3530
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.703532064911131e-05,
+      "loss": 2.5437,
+      "step": 3535
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6910152911110283e-05,
+      "loss": 2.5465,
+      "step": 3540
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6785353054571024e-05,
+      "loss": 2.6576,
+      "step": 3545
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.666092246697512e-05,
+      "loss": 2.6085,
+      "step": 3550
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6536862531698766e-05,
+      "loss": 2.566,
+      "step": 3555
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6413174627997328e-05,
+      "loss": 2.5171,
+      "step": 3560
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6289860130990147e-05,
+      "loss": 2.5731,
+      "step": 3565
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6166920411645064e-05,
+      "loss": 2.5133,
+      "step": 3570
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6044356836763315e-05,
+      "loss": 2.5862,
+      "step": 3575
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.5922170768964285e-05,
+      "loss": 2.5072,
+      "step": 3580
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5800363566670362e-05,
+      "loss": 2.523,
+      "step": 3585
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5678936584091852e-05,
+      "loss": 2.6213,
+      "step": 3590
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5557891171211892e-05,
+      "loss": 2.5779,
+      "step": 3595
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5437228673771465e-05,
+      "loss": 2.5765,
+      "step": 3600
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5316950433254445e-05,
+      "loss": 2.6403,
+      "step": 3605
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5197057786872649e-05,
+      "loss": 2.4515,
+      "step": 3610
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5077552067551015e-05,
+      "loss": 2.4518,
+      "step": 3615
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4958434603912747e-05,
+      "loss": 2.5988,
+      "step": 3620
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4839706720264546e-05,
+      "loss": 2.6563,
+      "step": 3625
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4721369736581924e-05,
+      "loss": 2.5724,
+      "step": 3630
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4603424968494484e-05,
+      "loss": 2.6055,
+      "step": 3635
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.448587372727132e-05,
+      "loss": 2.5468,
+      "step": 3640
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4368717319806419e-05,
+      "loss": 2.5653,
+      "step": 3645
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4251957048604152e-05,
+      "loss": 2.5951,
+      "step": 3650
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.413559421176479e-05,
+      "loss": 2.6623,
+      "step": 3655
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4019630102970056e-05,
+      "loss": 2.4612,
+      "step": 3660
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3904066011468753e-05,
+      "loss": 2.5215,
+      "step": 3665
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3788903222062433e-05,
+      "loss": 2.5728,
+      "step": 3670
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3674143015091118e-05,
+      "loss": 2.4948,
+      "step": 3675
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.355978666641905e-05,
+      "loss": 2.5785,
+      "step": 3680
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3445835447420507e-05,
+      "loss": 2.5608,
+      "step": 3685
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3332290624965688e-05,
+      "loss": 2.5159,
+      "step": 3690
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3219153461406609e-05,
+      "loss": 2.5324,
+      "step": 3695
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3106425214563078e-05,
+      "loss": 2.5425,
+      "step": 3700
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2994107137708716e-05,
+      "loss": 2.5266,
+      "step": 3705
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2882200479556988e-05,
+      "loss": 2.5088,
+      "step": 3710
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2770706484247397e-05,
+      "loss": 2.517,
+      "step": 3715
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2659626391331564e-05,
+      "loss": 2.5517,
+      "step": 3720
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2548961435759493e-05,
+      "loss": 2.4874,
+      "step": 3725
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2438712847865846e-05,
+      "loss": 2.5048,
+      "step": 3730
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2328881853356244e-05,
+      "loss": 2.5472,
+      "step": 3735
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.221946967329365e-05,
+      "loss": 2.5503,
+      "step": 3740
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2110477524084796e-05,
+      "loss": 2.5793,
+      "step": 3745
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2001906617466657e-05,
+      "loss": 2.5316,
+      "step": 3750
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1893758160492978e-05,
+      "loss": 2.5308,
+      "step": 3755
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1786033355520859e-05,
+      "loss": 2.5244,
+      "step": 3760
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1678733400197373e-05,
+      "loss": 2.4888,
+      "step": 3765
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1571859487446263e-05,
+      "loss": 2.4566,
+      "step": 3770
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1465412805454695e-05,
+      "loss": 2.5082,
+      "step": 3775
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1359394537660011e-05,
+      "loss": 2.5853,
+      "step": 3780
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.125380586273661e-05,
+      "loss": 2.4946,
+      "step": 3785
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1148647954582808e-05,
+      "loss": 2.5282,
+      "step": 3790
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1043921982307819e-05,
+      "loss": 2.5895,
+      "step": 3795
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0939629110218735e-05,
+      "loss": 2.5703,
+      "step": 3800
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0835770497807596e-05,
+      "loss": 2.4824,
+      "step": 3805
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0732347299738493e-05,
+      "loss": 2.5405,
+      "step": 3810
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0629360665834732e-05,
+      "loss": 2.5766,
+      "step": 3815
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.052681174106604e-05,
+      "loss": 2.5603,
+      "step": 3820
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0424701665535852e-05,
+      "loss": 2.5756,
+      "step": 3825
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0323031574468638e-05,
+      "loss": 2.5578,
+      "step": 3830
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0221802598197261e-05,
+      "loss": 2.5898,
+      "step": 3835
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0121015862150423e-05,
+      "loss": 2.4904,
+      "step": 3840
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0020672486840154e-05,
+      "loss": 2.56,
+      "step": 3845
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.920773587849364e-06,
+      "loss": 2.5242,
+      "step": 3850
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.821320275819401e-06,
+      "loss": 2.5581,
+      "step": 3855
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.72231365643777e-06,
+      "loss": 2.5445,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.623754830425779e-06,
+      "loss": 2.5413,
+      "step": 3865
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.52564489352632e-06,
+      "loss": 2.5475,
+      "step": 3870
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.427984936491702e-06,
+      "loss": 2.5661,
+      "step": 3875
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.330776045071509e-06,
+      "loss": 2.5114,
+      "step": 3880
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.23401930000054e-06,
+      "loss": 2.5738,
+      "step": 3885
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.137715776986772e-06,
+      "loss": 2.5187,
+      "step": 3890
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.041866546699434e-06,
+      "loss": 2.5497,
+      "step": 3895
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.946472674757078e-06,
+      "loss": 2.4964,
+      "step": 3900
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.851535221715735e-06,
+      "loss": 2.4861,
+      "step": 3905
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.757055243057132e-06,
+      "loss": 2.5917,
+      "step": 3910
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.663033789176967e-06,
+      "loss": 2.49,
+      "step": 3915
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.5694719053732e-06,
+      "loss": 2.5959,
+      "step": 3920
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.476370631834458e-06,
+      "loss": 2.5676,
+      "step": 3925
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.383731003628452e-06,
+      "loss": 2.5383,
+      "step": 3930
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.291554050690508e-06,
+      "loss": 2.5179,
+      "step": 3935
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.199840797812058e-06,
+      "loss": 2.4902,
+      "step": 3940
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.108592264629295e-06,
+      "loss": 2.5481,
+      "step": 3945
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.017809465611803e-06,
+      "loss": 2.5816,
+      "step": 3950
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.927493410051324e-06,
+      "loss": 2.583,
+      "step": 3955
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.837645102050473e-06,
+      "loss": 2.5129,
+      "step": 3960
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.748265540511635e-06,
+      "loss": 2.586,
+      "step": 3965
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.65935571912582e-06,
+      "loss": 2.5749,
+      "step": 3970
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.5709166263616405e-06,
+      "loss": 2.4871,
+      "step": 3975
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.482949245454302e-06,
+      "loss": 2.4824,
+      "step": 3980
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.3954545543946876e-06,
+      "loss": 2.5759,
+      "step": 3985
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.308433525918468e-06,
+      "loss": 2.5121,
+      "step": 3990
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.221887127495313e-06,
+      "loss": 2.5489,
+      "step": 3995
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.1358163213181114e-06,
+      "loss": 2.5184,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 4811,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 1.1658574908358656e+19,
+  "trial_name": null,
+  "trial_params": null
+}