diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5476 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 4538,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4052863436123357e-07,
+      "loss": 8.0371,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2026431718061677e-06,
+      "loss": 7.6777,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4052863436123355e-06,
+      "loss": 8.1375,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.607929515418502e-06,
+      "loss": 7.9645,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.810572687224671e-06,
+      "loss": 7.4338,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1013215859030838e-05,
+      "loss": 7.9741,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3215859030837005e-05,
+      "loss": 8.1339,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5418502202643173e-05,
+      "loss": 7.8998,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7621145374449342e-05,
+      "loss": 7.5467,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9823788546255507e-05,
+      "loss": 7.024,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2026431718061676e-05,
+      "loss": 6.8792,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.422907488986784e-05,
+      "loss": 6.7453,
+      "step": 55
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.643171806167401e-05,
+      "loss": 6.1152,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.8634361233480178e-05,
+      "loss": 6.4661,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0837004405286347e-05,
+      "loss": 6.093,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.3039647577092515e-05,
+      "loss": 5.5373,
+      "step": 75
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.5242290748898684e-05,
+      "loss": 6.0057,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.744493392070485e-05,
+      "loss": 6.0701,
+      "step": 85
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.9647577092511014e-05,
+      "loss": 5.7831,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.185022026431718e-05,
+      "loss": 5.8034,
+      "step": 95
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.405286343612335e-05,
+      "loss": 5.1736,
+      "step": 100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.625550660792951e-05,
+      "loss": 5.4872,
+      "step": 105
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.845814977973568e-05,
+      "loss": 5.2194,
+      "step": 110
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.066079295154186e-05,
+      "loss": 4.8431,
+      "step": 115
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.286343612334802e-05,
+      "loss": 4.7448,
+      "step": 120
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.506607929515418e-05,
+      "loss": 4.9598,
+      "step": 125
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.7268722466960356e-05,
+      "loss": 5.1425,
+      "step": 130
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.947136563876652e-05,
+      "loss": 4.9857,
+      "step": 135
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 6.167400881057269e-05,
+      "loss": 4.6542,
+      "step": 140
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 6.387665198237885e-05,
+      "loss": 4.8413,
+      "step": 145
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.607929515418503e-05,
+      "loss": 4.7463,
+      "step": 150
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.828193832599119e-05,
+      "loss": 4.6509,
+      "step": 155
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 7.048458149779737e-05,
+      "loss": 4.8795,
+      "step": 160
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 7.268722466960352e-05,
+      "loss": 4.4858,
+      "step": 165
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 7.48898678414097e-05,
+      "loss": 4.5239,
+      "step": 170
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 7.709251101321586e-05,
+      "loss": 4.739,
+      "step": 175
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 7.929515418502203e-05,
+      "loss": 4.5952,
+      "step": 180
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 8.14977973568282e-05,
+      "loss": 4.4029,
+      "step": 185
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 8.370044052863437e-05,
+      "loss": 4.5299,
+      "step": 190
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 8.590308370044053e-05,
+      "loss": 4.3894,
+      "step": 195
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 8.81057268722467e-05,
+      "loss": 4.588,
+      "step": 200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.030837004405287e-05,
+      "loss": 4.2276,
+      "step": 205
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.251101321585903e-05,
+      "loss": 4.2741,
+      "step": 210
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.471365638766521e-05,
+      "loss": 3.8694,
+      "step": 215
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.691629955947136e-05,
+      "loss": 4.3776,
+      "step": 220
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.911894273127755e-05,
+      "loss": 4.3895,
+      "step": 225
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00010132158590308371,
+      "loss": 4.3406,
+      "step": 230
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00010352422907488988,
+      "loss": 4.3947,
+      "step": 235
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00010572687224669604,
+      "loss": 4.3829,
+      "step": 240
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001079295154185022,
+      "loss": 4.3668,
+      "step": 245
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00011013215859030836,
+      "loss": 4.111,
+      "step": 250
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00011233480176211454,
+      "loss": 3.8334,
+      "step": 255
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00011453744493392071,
+      "loss": 4.0465,
+      "step": 260
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00011674008810572688,
+      "loss": 3.6863,
+      "step": 265
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00011894273127753304,
+      "loss": 4.3514,
+      "step": 270
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00012114537444933922,
+      "loss": 3.9275,
+      "step": 275
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00012334801762114539,
+      "loss": 4.0648,
+      "step": 280
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00012555066079295154,
+      "loss": 3.5522,
+      "step": 285
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001277533039647577,
+      "loss": 3.8612,
+      "step": 290
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001299559471365639,
+      "loss": 3.7846,
+      "step": 295
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00013215859030837006,
+      "loss": 3.9351,
+      "step": 300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00013436123348017622,
+      "loss": 3.8342,
+      "step": 305
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00013656387665198237,
+      "loss": 3.8025,
+      "step": 310
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00013876651982378855,
+      "loss": 3.7136,
+      "step": 315
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00014096916299559473,
+      "loss": 3.6827,
+      "step": 320
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001431718061674009,
+      "loss": 3.3543,
+      "step": 325
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00014537444933920705,
+      "loss": 3.5532,
+      "step": 330
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00014757709251101323,
+      "loss": 3.3296,
+      "step": 335
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001497797356828194,
+      "loss": 3.3282,
+      "step": 340
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00015198237885462556,
+      "loss": 3.7764,
+      "step": 345
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00015418502202643172,
+      "loss": 3.2611,
+      "step": 350
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00015638766519823787,
+      "loss": 3.2127,
+      "step": 355
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00015859030837004406,
+      "loss": 3.1679,
+      "step": 360
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00016079295154185024,
+      "loss": 3.4743,
+      "step": 365
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001629955947136564,
+      "loss": 3.3912,
+      "step": 370
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00016519823788546255,
+      "loss": 3.2843,
+      "step": 375
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00016740088105726873,
+      "loss": 3.6252,
+      "step": 380
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001696035242290749,
+      "loss": 3.174,
+      "step": 385
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017180616740088107,
+      "loss": 3.4322,
+      "step": 390
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017400881057268722,
+      "loss": 3.3303,
+      "step": 395
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001762114537444934,
+      "loss": 3.3238,
+      "step": 400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001784140969162996,
+      "loss": 2.9926,
+      "step": 405
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018061674008810574,
+      "loss": 3.1562,
+      "step": 410
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001828193832599119,
+      "loss": 3.0285,
+      "step": 415
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018502202643171805,
+      "loss": 3.1948,
+      "step": 420
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018722466960352423,
+      "loss": 3.2688,
+      "step": 425
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018942731277533042,
+      "loss": 3.2674,
+      "step": 430
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019162995594713657,
+      "loss": 3.136,
+      "step": 435
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019383259911894273,
+      "loss": 3.017,
+      "step": 440
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001960352422907489,
+      "loss": 3.071,
+      "step": 445
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001982378854625551,
+      "loss": 2.8048,
+      "step": 450
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019999997041318193,
+      "loss": 2.8113,
+      "step": 455
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001999989348763872,
+      "loss": 3.0145,
+      "step": 460
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001999964200161957,
+      "loss": 3.1626,
+      "step": 465
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001999924258698108,
+      "loss": 2.5559,
+      "step": 470
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019998695249631943,
+      "loss": 3.0745,
+      "step": 475
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001999799999766913,
+      "loss": 2.8221,
+      "step": 480
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001999715684137778,
+      "loss": 2.9143,
+      "step": 485
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019996165793231038,
+      "loss": 2.8685,
+      "step": 490
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001999502686788986,
+      "loss": 2.9853,
+      "step": 495
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019993740082202818,
+      "loss": 2.8111,
+      "step": 500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019992305455205836,
+      "loss": 3.3157,
+      "step": 505
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019990723008121917,
+      "loss": 2.8514,
+      "step": 510
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019988992764360814,
+      "loss": 2.8086,
+      "step": 515
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.000199871147495187,
+      "loss": 3.0512,
+      "step": 520
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001998508899137779,
+      "loss": 3.0319,
+      "step": 525
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019982915519905912,
+      "loss": 2.8135,
+      "step": 530
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019980594367256086,
+      "loss": 2.8908,
+      "step": 535
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019978125567766023,
+      "loss": 2.9204,
+      "step": 540
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001997550915795765,
+      "loss": 3.1027,
+      "step": 545
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019972745176536537,
+      "loss": 2.969,
+      "step": 550
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019969833664391343,
+      "loss": 2.8162,
+      "step": 555
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019966774664593206,
+      "loss": 2.9871,
+      "step": 560
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019963568222395111,
+      "loss": 2.6507,
+      "step": 565
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019960214385231217,
+      "loss": 2.508,
+      "step": 570
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001995671320271614,
+      "loss": 2.8594,
+      "step": 575
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001995306472664425,
+      "loss": 2.7387,
+      "step": 580
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001994926901098888,
+      "loss": 2.7499,
+      "step": 585
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019945326111901542,
+      "loss": 2.4711,
+      "step": 590
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019941236087711077,
+      "loss": 2.87,
+      "step": 595
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019936998998922826,
+      "loss": 2.9676,
+      "step": 600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001993261490821769,
+      "loss": 2.6321,
+      "step": 605
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001992808388045125,
+      "loss": 2.6894,
+      "step": 610
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019923405982652767,
+      "loss": 2.786,
+      "step": 615
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001991858128402422,
+      "loss": 2.6501,
+      "step": 620
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001991360985593926,
+      "loss": 2.6592,
+      "step": 625
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001990849177194217,
+      "loss": 2.4167,
+      "step": 630
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019903227107746774,
+      "loss": 2.6244,
+      "step": 635
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019897815941235307,
+      "loss": 2.7636,
+      "step": 640
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019892258352457277,
+      "loss": 2.7828,
+      "step": 645
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001988655442362827,
+      "loss": 2.8655,
+      "step": 650
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019880704239128732,
+      "loss": 2.7669,
+      "step": 655
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019874707885502745,
+      "loss": 2.5676,
+      "step": 660
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019868565451456704,
+      "loss": 2.5646,
+      "step": 665
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001986227702785805,
+      "loss": 2.8244,
+      "step": 670
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001985584270773389,
+      "loss": 2.6453,
+      "step": 675
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00019849262586269642,
+      "loss": 3.1961,
+      "step": 680
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001984253676080762,
+      "loss": 2.7,
+      "step": 685
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00019835665330845595,
+      "loss": 2.7882,
+      "step": 690
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019828648398035315,
+      "loss": 2.7521,
+      "step": 695
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001982148606618102,
+      "loss": 2.3425,
+      "step": 700
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019814178441237887,
+      "loss": 2.9472,
+      "step": 705
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019806725631310476,
+      "loss": 2.6193,
+      "step": 710
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019799127746651117,
+      "loss": 2.6507,
+      "step": 715
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.000197913848996583,
+      "loss": 2.7298,
+      "step": 720
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019783497204874985,
+      "loss": 2.7898,
+      "step": 725
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019775464778986934,
+      "loss": 2.5467,
+      "step": 730
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019767287740820965,
+      "loss": 2.615,
+      "step": 735
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019758966211343206,
+      "loss": 2.6105,
+      "step": 740
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.000197505003136573,
+      "loss": 2.6392,
+      "step": 745
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001974189017300259,
+      "loss": 2.659,
+      "step": 750
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019733135916752252,
+      "loss": 2.5102,
+      "step": 755
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019724237674411432,
+      "loss": 2.6638,
+      "step": 760
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001971519557761531,
+      "loss": 2.858,
+      "step": 765
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019706009760127164,
+      "loss": 2.6797,
+      "step": 770
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001969668035783639,
+      "loss": 2.6023,
+      "step": 775
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019687207508756486,
+      "loss": 2.5694,
+      "step": 780
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019677591353023025,
+      "loss": 2.8703,
+      "step": 785
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019667832032891554,
+      "loss": 2.5892,
+      "step": 790
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001965792969273552,
+      "loss": 2.6138,
+      "step": 795
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019647884479044123,
+      "loss": 2.7579,
+      "step": 800
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019637696540420136,
+      "loss": 2.5881,
+      "step": 805
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019627366027577726,
+      "loss": 2.4141,
+      "step": 810
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019616893093340218,
+      "loss": 2.8274,
+      "step": 815
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019606277892637823,
+      "loss": 2.5646,
+      "step": 820
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019595520582505363,
+      "loss": 2.4943,
+      "step": 825
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019584621322079942,
+      "loss": 2.3327,
+      "step": 830
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019573580272598583,
+      "loss": 2.6538,
+      "step": 835
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019562397597395857,
+      "loss": 2.9938,
+      "step": 840
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001955107346190145,
+      "loss": 2.5367,
+      "step": 845
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001953960803363774,
+      "loss": 2.3441,
+      "step": 850
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019528001482217296,
+      "loss": 2.5851,
+      "step": 855
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001951625397934037,
+      "loss": 2.4576,
+      "step": 860
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001950436569879238,
+      "loss": 2.5669,
+      "step": 865
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001949233681644131,
+      "loss": 2.3772,
+      "step": 870
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019480167510235128,
+      "loss": 2.5839,
+      "step": 875
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019467857960199142,
+      "loss": 2.5563,
+      "step": 880
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019455408348433354,
+      "loss": 2.6866,
+      "step": 885
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019442818859109737,
+      "loss": 2.5574,
+      "step": 890
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019430089678469542,
+      "loss": 2.3997,
+      "step": 895
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019417220994820514,
+      "loss": 2.6884,
+      "step": 900
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019404212998534135,
+      "loss": 2.5868,
+      "step": 905
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019391065882042786,
+      "loss": 2.4245,
+      "step": 910
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001937777983983691,
+      "loss": 2.6269,
+      "step": 915
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019364355068462126,
+      "loss": 2.4964,
+      "step": 920
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019350791766516344,
+      "loss": 2.5777,
+      "step": 925
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019337090134646787,
+      "loss": 2.4597,
+      "step": 930
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019323250375547068,
+      "loss": 2.5093,
+      "step": 935
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001930927269395416,
+      "loss": 2.3503,
+      "step": 940
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001929515729664538,
+      "loss": 2.5458,
+      "step": 945
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019280904392435328,
+      "loss": 2.6002,
+      "step": 950
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019266514192172794,
+      "loss": 2.4127,
+      "step": 955
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019251986908737646,
+      "loss": 2.4922,
+      "step": 960
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019237322757037676,
+      "loss": 2.4938,
+      "step": 965
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019222521954005424,
+      "loss": 2.6074,
+      "step": 970
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001920758471859496,
+      "loss": 2.6341,
+      "step": 975
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019192511271778656,
+      "loss": 2.4591,
+      "step": 980
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001917730183654392,
+      "loss": 2.6226,
+      "step": 985
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019161956637889872,
+      "loss": 2.6092,
+      "step": 990
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019146475902824048,
+      "loss": 2.4191,
+      "step": 995
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019130859860359026,
+      "loss": 2.4157,
+      "step": 1000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019115108741509036,
+      "loss": 2.6189,
+      "step": 1005
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001909922277928654,
+      "loss": 2.3266,
+      "step": 1010
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019083202208698805,
+      "loss": 2.36,
+      "step": 1015
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019067047266744396,
+      "loss": 2.4632,
+      "step": 1020
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019050758192409707,
+      "loss": 2.6295,
+      "step": 1025
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001903433522666538,
+      "loss": 2.5695,
+      "step": 1030
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001901777861246278,
+      "loss": 2.4457,
+      "step": 1035
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001900108859473039,
+      "loss": 2.5195,
+      "step": 1040
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001898426542037017,
+      "loss": 2.3216,
+      "step": 1045
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001896730933825393,
+      "loss": 2.6117,
+      "step": 1050
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001895022059921963,
+      "loss": 2.5559,
+      "step": 1055
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00018932999456067675,
+      "loss": 2.524,
+      "step": 1060
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00018915646163557179,
+      "loss": 2.3177,
+      "step": 1065
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00018898160978402198,
+      "loss": 2.4404,
+      "step": 1070
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00018880544159267912,
+      "loss": 2.4747,
+      "step": 1075
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00018862795966766833,
+      "loss": 2.6981,
+      "step": 1080
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001884491666345492,
+      "loss": 2.4396,
+      "step": 1085
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00018826906513827704,
+      "loss": 2.3892,
+      "step": 1090
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00018808765784316384,
+      "loss": 2.1981,
+      "step": 1095
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001879049474328387,
+      "loss": 2.5497,
+      "step": 1100
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00018772093661020825,
+      "loss": 2.3756,
+      "step": 1105
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00018753562809741673,
+      "loss": 2.4149,
+      "step": 1110
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00018734902463580542,
+      "loss": 2.483,
+      "step": 1115
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00018716112898587247,
+      "loss": 2.3027,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00018697194392723173,
+      "loss": 2.3358,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001867814722585719,
+      "loss": 2.5705,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001865897167976149,
+      "loss": 2.6456,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00018639668038107437,
+      "loss": 2.4446,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00018620236586461357,
+      "loss": 2.4226,
+      "step": 1145
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001860067761228033,
+      "loss": 2.3209,
+      "step": 1150
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00018580991404907912,
+      "loss": 2.6875,
+      "step": 1155
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00018561178255569879,
+      "loss": 2.2906,
+      "step": 1160
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00018541238457369904,
+      "loss": 2.1676,
+      "step": 1165
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00018521172305285236,
+      "loss": 2.2646,
+      "step": 1170
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001850098009616231,
+      "loss": 2.3756,
+      "step": 1175
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00018480662128712389,
+      "loss": 2.4612,
+      "step": 1180
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00018460218703507118,
+      "loss": 2.6361,
+      "step": 1185
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00018439650122974087,
+      "loss": 2.6342,
+      "step": 1190
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00018418956691392368,
+      "loss": 2.2129,
+      "step": 1195
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00018398138714887993,
+      "loss": 2.393,
+      "step": 1200
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00018377196501429438,
+      "loss": 2.4947,
+      "step": 1205
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00018356130360823068,
+      "loss": 2.1074,
+      "step": 1210
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001833494060470855,
+      "loss": 2.1386,
+      "step": 1215
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001831362754655424,
+      "loss": 2.4856,
+      "step": 1220
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018292191501652553,
+      "loss": 2.5449,
+      "step": 1225
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018270632787115295,
+      "loss": 2.3776,
+      "step": 1230
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018248951721868973,
+      "loss": 2.5301,
+      "step": 1235
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018227148626650072,
+      "loss": 2.3679,
+      "step": 1240
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018205223824000318,
+      "loss": 2.3381,
+      "step": 1245
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018183177638261895,
+      "loss": 2.3739,
+      "step": 1250
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018161010395572667,
+      "loss": 2.4117,
+      "step": 1255
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00018138722423861333,
+      "loss": 2.3288,
+      "step": 1260
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00018116314052842584,
+      "loss": 2.6229,
+      "step": 1265
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00018093785614012228,
+      "loss": 2.2574,
+      "step": 1270
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00018071137440642276,
+      "loss": 2.4681,
+      "step": 1275
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00018048369867776029,
+      "loss": 2.2659,
+      "step": 1280
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00018025483232223105,
+      "loss": 2.5531,
+      "step": 1285
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001800247787255447,
+      "loss": 2.323,
+      "step": 1290
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00017979354129097407,
+      "loss": 2.3548,
+      "step": 1295
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00017956112343930512,
+      "loss": 2.331,
+      "step": 1300
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00017932752860878618,
+      "loss": 2.4505,
+      "step": 1305
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00017909276025507696,
+      "loss": 2.6869,
+      "step": 1310
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001788568218511977,
+      "loss": 2.3611,
+      "step": 1315
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00017861971688747747,
+      "loss": 2.4549,
+      "step": 1320
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00017838144887150298,
+      "loss": 2.5236,
+      "step": 1325
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001781420213280662,
+      "loss": 2.2486,
+      "step": 1330
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00017790143779911256,
+      "loss": 2.5255,
+      "step": 1335
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00017765970184368835,
+      "loss": 2.339,
+      "step": 1340
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00017741681703788826,
+      "loss": 2.4522,
+      "step": 1345
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001771727869748023,
+      "loss": 2.6055,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00017692761526446268,
+      "loss": 2.317,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00017668130553379063,
+      "loss": 2.1768,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00017643386142654236,
+      "loss": 2.3331,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001761852866032554,
+      "loss": 2.4262,
+      "step": 1370
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001759355847411945,
+      "loss": 2.3873,
+      "step": 1375
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00017568475953429706,
+      "loss": 2.5427,
+      "step": 1380
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00017543281469311857,
+      "loss": 2.4491,
+      "step": 1385
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00017517975394477765,
+      "loss": 2.5059,
+      "step": 1390
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00017492558103290114,
+      "loss": 2.3825,
+      "step": 1395
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00017467029971756837,
+      "loss": 2.4185,
+      "step": 1400
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001744139137752558,
+      "loss": 2.3802,
+      "step": 1405
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00017415642699878108,
+      "loss": 2.531,
+      "step": 1410
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00017389784319724687,
+      "loss": 2.3635,
+      "step": 1415
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00017363816619598462,
+      "loss": 2.2934,
+      "step": 1420
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001733773998364979,
+      "loss": 2.184,
+      "step": 1425
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00017311554797640552,
+      "loss": 2.3262,
+      "step": 1430
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001728526144893846,
+      "loss": 2.4294,
+      "step": 1435
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00017258860326511318,
+      "loss": 2.3538,
+      "step": 1440
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00017232351820921263,
+      "loss": 2.1007,
+      "step": 1445
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00017205736324318999,
+      "loss": 2.2829,
+      "step": 1450
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00017179014230437988,
+      "loss": 2.496,
+      "step": 1455
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00017152185934588623,
+      "loss": 2.3997,
+      "step": 1460
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00017125251833652398,
+      "loss": 2.2981,
+      "step": 1465
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00017098212326076008,
+      "loss": 2.0833,
+      "step": 1470
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 2.542,
+      "step": 1475
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00017043818692580228,
+      "loss": 2.2825,
+      "step": 1480
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001701646537132716,
+      "loss": 2.4096,
+      "step": 1485
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00016989008252754655,
+      "loss": 2.119,
+      "step": 1490
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00016961447743046623,
+      "loss": 2.5423,
+      "step": 1495
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00016933784249916476,
+      "loss": 2.5358,
+      "step": 1500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00016906018182601096,
+      "loss": 1.9538,
+      "step": 1505
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001687814995185479,
+      "loss": 2.3082,
+      "step": 1510
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00016850179969943207,
+      "loss": 2.1142,
+      "step": 1515
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00016822108650637238,
+      "loss": 2.3758,
+      "step": 1520
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00016793936409206905,
+      "loss": 2.4226,
+      "step": 1525
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00016765663662415204,
+      "loss": 2.407,
+      "step": 1530
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00016737290828511948,
+      "loss": 2.2467,
+      "step": 1535
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00016708818327227574,
+      "loss": 2.4492,
+      "step": 1540
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00016680246579766944,
+      "loss": 2.4088,
+      "step": 1545
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00016651576008803112,
+      "loss": 2.3473,
+      "step": 1550
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00016622807038471052,
+      "loss": 2.3,
+      "step": 1555
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00016593940094361407,
+      "loss": 2.3139,
+      "step": 1560
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00016564975603514183,
+      "loss": 2.3906,
+      "step": 1565
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00016535913994412436,
+      "loss": 2.0856,
+      "step": 1570
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001650675569697593,
+      "loss": 2.1938,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001647750114255477,
+      "loss": 2.418,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00016448150763923036,
+      "loss": 2.2186,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00016418704995272373,
+      "loss": 2.3271,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00016389164272205566,
+      "loss": 2.4117,
+      "step": 1595
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00016359529031730093,
+      "loss": 2.1018,
+      "step": 1600
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00016329799712251674,
+      "loss": 2.199,
+      "step": 1605
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00016299976753567772,
+      "loss": 2.2997,
+      "step": 1610
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00016270060596861087,
+      "loss": 2.5793,
+      "step": 1615
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00016240051684693042,
+      "loss": 2.2919,
+      "step": 1620
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00016209950460997226,
+      "loss": 2.2764,
+      "step": 1625
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00016179757371072824,
+      "loss": 2.4133,
+      "step": 1630
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00016149472861578038,
+      "loss": 2.3608,
+      "step": 1635
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001611909738052347,
+      "loss": 2.2998,
+      "step": 1640
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001608863137726551,
+      "loss": 2.0379,
+      "step": 1645
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00016058075302499673,
+      "loss": 2.391,
+      "step": 1650
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00016027429608253934,
+      "loss": 2.3764,
+      "step": 1655
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001599669474788205,
+      "loss": 2.1648,
+      "step": 1660
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00015965871176056837,
+      "loss": 2.207,
+      "step": 1665
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00015934959348763467,
+      "loss": 2.3042,
+      "step": 1670
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00015903959723292705,
+      "loss": 2.4253,
+      "step": 1675
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00015872872758234148,
+      "loss": 2.4448,
+      "step": 1680
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001584169891346944,
+      "loss": 2.2236,
+      "step": 1685
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001581043865016547,
+      "loss": 2.1879,
+      "step": 1690
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00015779092430767556,
+      "loss": 2.3603,
+      "step": 1695
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00015747660718992598,
+      "loss": 2.3299,
+      "step": 1700
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00015716143979822207,
+      "loss": 2.1691,
+      "step": 1705
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00015684542679495847,
+      "loss": 2.4917,
+      "step": 1710
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00015652857285503938,
+      "loss": 2.4864,
+      "step": 1715
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00015621088266580904,
+      "loss": 2.3011,
+      "step": 1720
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00015589236092698286,
+      "loss": 2.0229,
+      "step": 1725
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00015557301235057767,
+      "loss": 2.3556,
+      "step": 1730
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00015525284166084194,
+      "loss": 2.3913,
+      "step": 1735
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001549318535941861,
+      "loss": 2.5327,
+      "step": 1740
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001546100528991123,
+      "loss": 1.9177,
+      "step": 1745
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00015428744433614415,
+      "loss": 2.2646,
+      "step": 1750
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00015396403267775647,
+      "loss": 2.2946,
+      "step": 1755
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001536398227083046,
+      "loss": 2.3676,
+      "step": 1760
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00015331481922395347,
+      "loss": 2.272,
+      "step": 1765
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00015298902703260692,
+      "loss": 2.3888,
+      "step": 1770
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001526624509538364,
+      "loss": 2.1347,
+      "step": 1775
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00015233509581880973,
+      "loss": 2.3202,
+      "step": 1780
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00015200696647021955,
+      "loss": 2.4352,
+      "step": 1785
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00015167806776221178,
+      "loss": 2.2928,
+      "step": 1790
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00015134840456031384,
+      "loss": 2.3575,
+      "step": 1795
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00015101798174136247,
+      "loss": 2.1881,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00015068680419343187,
+      "loss": 2.1145,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001503548768157612,
+      "loss": 2.1411,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00015002220451868203,
+      "loss": 2.2795,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00014968879222354597,
+      "loss": 2.2624,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001493546448626516,
+      "loss": 2.4409,
+      "step": 1825
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001490197673791717,
+      "loss": 2.4049,
+      "step": 1830
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014868416472707993,
+      "loss": 2.2198,
+      "step": 1835
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014834784187107785,
+      "loss": 2.4333,
+      "step": 1840
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014801080378652103,
+      "loss": 2.2905,
+      "step": 1845
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014767305545934588,
+      "loss": 2.0903,
+      "step": 1850
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014733460188599563,
+      "loss": 2.2965,
+      "step": 1855
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001469954480733465,
+      "loss": 2.3256,
+      "step": 1860
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014665559903863365,
+      "loss": 2.1393,
+      "step": 1865
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014631505980937688,
+      "loss": 2.4199,
+      "step": 1870
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014597383542330636,
+      "loss": 2.2842,
+      "step": 1875
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014563193092828803,
+      "loss": 2.428,
+      "step": 1880
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.000145289351382249,
+      "loss": 2.2509,
+      "step": 1885
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014494610185310252,
+      "loss": 2.4455,
+      "step": 1890
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00014460218741867338,
+      "loss": 2.3054,
+      "step": 1895
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00014425761316662241,
+      "loss": 2.5075,
+      "step": 1900
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00014391238419437144,
+      "loss": 2.2115,
+      "step": 1905
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001435665056090278,
+      "loss": 2.4274,
+      "step": 1910
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001432199825273089,
+      "loss": 2.2174,
+      "step": 1915
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014287282007546627,
+      "loss": 2.2667,
+      "step": 1920
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014252502338921007,
+      "loss": 2.2216,
+      "step": 1925
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001421765976136328,
+      "loss": 2.2736,
+      "step": 1930
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014182754790313342,
+      "loss": 2.4801,
+      "step": 1935
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014147787942134089,
+      "loss": 2.2007,
+      "step": 1940
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014112759734103808,
+      "loss": 2.5219,
+      "step": 1945
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014077670684408485,
+      "loss": 2.2544,
+      "step": 1950
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001404252131213418,
+      "loss": 2.2832,
+      "step": 1955
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014007312137259307,
+      "loss": 2.2898,
+      "step": 1960
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00013972043680646987,
+      "loss": 2.3556,
+      "step": 1965
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.000139367164640373,
+      "loss": 2.3909,
+      "step": 1970
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.000139013310100396,
+      "loss": 2.389,
+      "step": 1975
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00013865887842124755,
+      "loss": 1.815,
+      "step": 1980
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00013830387484617434,
+      "loss": 2.2869,
+      "step": 1985
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001379483046268832,
+      "loss": 2.2448,
+      "step": 1990
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00013759217302346358,
+      "loss": 2.3517,
+      "step": 1995
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00013723548530430974,
+      "loss": 2.1264,
+      "step": 2000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001368782467460428,
+      "loss": 2.3207,
+      "step": 2005
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00013652046263343262,
+      "loss": 2.2458,
+      "step": 2010
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00013616213825931963,
+      "loss": 2.1549,
+      "step": 2015
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001358032789245366,
+      "loss": 2.331,
+      "step": 2020
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00013544388993783023,
+      "loss": 2.2386,
+      "step": 2025
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00013508397661578242,
+      "loss": 2.4434,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00013472354428273193,
+      "loss": 2.2391,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00013436259827069534,
+      "loss": 2.1787,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001340011439192884,
+      "loss": 1.9544,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001336391865756468,
+      "loss": 2.1705,
+      "step": 2050
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00013327673159434724,
+      "loss": 2.4587,
+      "step": 2055
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00013291378433732818,
+      "loss": 2.2278,
+      "step": 2060
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001325503501738105,
+      "loss": 2.3232,
+      "step": 2065
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001321864344802181,
+      "loss": 2.2155,
+      "step": 2070
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001318220426400983,
+      "loss": 2.2087,
+      "step": 2075
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00013145718004404223,
+      "loss": 2.2386,
+      "step": 2080
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001310918520896052,
+      "loss": 2.3636,
+      "step": 2085
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00013072606418122667,
+      "loss": 1.9883,
+      "step": 2090
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00013035982173015034,
+      "loss": 2.2703,
+      "step": 2095
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001299931301543442,
+      "loss": 2.0428,
+      "step": 2100
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001296259948784203,
+      "loss": 2.2012,
+      "step": 2105
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00012925842133355454,
+      "loss": 2.1794,
+      "step": 2110
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001288904149574063,
+      "loss": 2.1578,
+      "step": 2115
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00012852198119403798,
+      "loss": 2.0021,
+      "step": 2120
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001281531254938345,
+      "loss": 2.2277,
+      "step": 2125
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001277838533134226,
+      "loss": 2.4797,
+      "step": 2130
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001274141701155903,
+      "loss": 2.1755,
+      "step": 2135
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00012704408136920585,
+      "loss": 2.4891,
+      "step": 2140
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00012667359254913698,
+      "loss": 2.2988,
+      "step": 2145
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00012630270913616985,
+      "loss": 2.5418,
+      "step": 2150
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00012593143661692808,
+      "loss": 2.35,
+      "step": 2155
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00012555978048379133,
+      "loss": 2.0868,
+      "step": 2160
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00012518774623481433,
+      "loss": 2.4799,
+      "step": 2165
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001248153393736454,
+      "loss": 2.2664,
+      "step": 2170
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.000124442565409445,
+      "loss": 2.2299,
+      "step": 2175
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00012406942985680437,
+      "loss": 2.2555,
+      "step": 2180
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00012369593823566383,
+      "loss": 2.1997,
+      "step": 2185
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00012332209607123117,
+      "loss": 2.2222,
+      "step": 2190
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00012294790889389994,
+      "loss": 2.3091,
+      "step": 2195
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001225733822391675,
+      "loss": 2.0859,
+      "step": 2200
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001221985216475533,
+      "loss": 2.3683,
+      "step": 2205
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00012182333266451684,
+      "loss": 2.1938,
+      "step": 2210
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00012144782084037571,
+      "loss": 2.0238,
+      "step": 2215
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00012107199173022327,
+      "loss": 2.345,
+      "step": 2220
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00012069585089384674,
+      "loss": 2.1972,
+      "step": 2225
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00012031940389564478,
+      "loss": 2.1552,
+      "step": 2230
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00011994265630454524,
+      "loss": 2.2012,
+      "step": 2235
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00011956561369392274,
+      "loss": 2.265,
+      "step": 2240
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001191882816415162,
+      "loss": 2.2672,
+      "step": 2245
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00011881066572934644,
+      "loss": 2.2446,
+      "step": 2250
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00011843277154363353,
+      "loss": 2.2362,
+      "step": 2255
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001180546046747141,
+      "loss": 2.1197,
+      "step": 2260
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00011767617071695867,
+      "loss": 1.9836,
+      "step": 2265
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.000117297475268689,
+      "loss": 2.1665,
+      "step": 2270
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00011691852393209508,
+      "loss": 2.4497,
+      "step": 2275
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00011653932231315245,
+      "loss": 2.0536,
+      "step": 2280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0001161598760215391,
+      "loss": 2.3549,
+      "step": 2285
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0001157801906705526,
+      "loss": 2.1831,
+      "step": 2290
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00011540027187702698,
+      "loss": 2.1216,
+      "step": 2295
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00011502012526124978,
+      "loss": 2.2128,
+      "step": 2300
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00011463975644687867,
+      "loss": 2.1899,
+      "step": 2305
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00011425917106085844,
+      "loss": 2.1396,
+      "step": 2310
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00011387837473333774,
+      "loss": 2.5057,
+      "step": 2315
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00011349737309758572,
+      "loss": 2.0298,
+      "step": 2320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00011311617178990879,
+      "loss": 2.2519,
+      "step": 2325
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0001127347764495671,
+      "loss": 2.1852,
+      "step": 2330
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00011235319271869129,
+      "loss": 2.2552,
+      "step": 2335
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00011197142624219887,
+      "loss": 2.1774,
+      "step": 2340
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00011158948266771086,
+      "loss": 2.2011,
+      "step": 2345
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00011120736764546799,
+      "loss": 2.2846,
+      "step": 2350
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00011082508682824748,
+      "loss": 2.1659,
+      "step": 2355
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001104426458712791,
+      "loss": 2.1766,
+      "step": 2360
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00011006005043216172,
+      "loss": 1.9432,
+      "step": 2365
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00010967730617077938,
+      "loss": 2.1685,
+      "step": 2370
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00010929441874921789,
+      "loss": 2.3591,
+      "step": 2375
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00010891139383168072,
+      "loss": 2.3673,
+      "step": 2380
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00010852823708440551,
+      "loss": 2.3499,
+      "step": 2385
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00010814495417557997,
+      "loss": 2.2054,
+      "step": 2390
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00010776155077525827,
+      "loss": 2.2129,
+      "step": 2395
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00010737803255527702,
+      "loss": 2.1971,
+      "step": 2400
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00010699440518917145,
+      "loss": 2.1717,
+      "step": 2405
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00010661067435209135,
+      "loss": 1.9133,
+      "step": 2410
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00010622684572071724,
+      "loss": 2.1657,
+      "step": 2415
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00010584292497317633,
+      "loss": 2.2476,
+      "step": 2420
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00010545891778895862,
+      "loss": 2.1151,
+      "step": 2425
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00010507482984883268,
+      "loss": 1.9689,
+      "step": 2430
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00010469066683476181,
+      "loss": 2.1355,
+      "step": 2435
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00010430643442981986,
+      "loss": 2.0989,
+      "step": 2440
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00010392213831810733,
+      "loss": 2.113,
+      "step": 2445
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00010353778418466697,
+      "loss": 2.1521,
+      "step": 2450
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0001031533777154,
+      "loss": 2.1597,
+      "step": 2455
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00010276892459698182,
+      "loss": 2.1667,
+      "step": 2460
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00010238443051677798,
+      "loss": 1.9723,
+      "step": 2465
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00010199990116275988,
+      "loss": 2.1176,
+      "step": 2470
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00010161534222342083,
+      "loss": 2.1394,
+      "step": 2475
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00010123075938769187,
+      "loss": 2.1762,
+      "step": 2480
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00010084615834485733,
+      "loss": 2.1752,
+      "step": 2485
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00010046154478447114,
+      "loss": 2.1837,
+      "step": 2490
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00010007692439627224,
+      "loss": 2.2389,
+      "step": 2495
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.96923028701006e-05,
+      "loss": 2.1724,
+      "step": 2500
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.93076858958131e-05,
+      "loss": 2.2438,
+      "step": 2505
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.892307916319919e-05,
+      "loss": 2.4101,
+      "step": 2510
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.853848836189691e-05,
+      "loss": 2.134,
+      "step": 2515
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.815391918130848e-05,
+      "loss": 2.2203,
+      "step": 2520
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.776937731051641e-05,
+      "loss": 2.2268,
+      "step": 2525
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.738486843819919e-05,
+      "loss": 2.4168,
+      "step": 2530
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.700039825254703e-05,
+      "loss": 2.2746,
+      "step": 2535
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.661597244117802e-05,
+      "loss": 2.216,
+      "step": 2540
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.623159669105364e-05,
+      "loss": 2.1087,
+      "step": 2545
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.584727668839487e-05,
+      "loss": 2.2089,
+      "step": 2550
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.5463018118598e-05,
+      "loss": 2.2306,
+      "step": 2555
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.507882666615049e-05,
+      "loss": 2.2674,
+      "step": 2560
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.469470801454688e-05,
+      "loss": 2.1774,
+      "step": 2565
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.431066784620486e-05,
+      "loss": 1.9788,
+      "step": 2570
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.392671184238092e-05,
+      "loss": 2.1459,
+      "step": 2575
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 9.354284568308665e-05,
+      "loss": 1.9707,
+      "step": 2580
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 9.31590750470044e-05,
+      "loss": 2.1999,
+      "step": 2585
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 9.277540561140342e-05,
+      "loss": 2.2249,
+      "step": 2590
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 9.239184305205594e-05,
+      "loss": 2.0383,
+      "step": 2595
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 9.2008393043153e-05,
+      "loss": 2.0296,
+      "step": 2600
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 9.162506125722082e-05,
+      "loss": 2.4062,
+      "step": 2605
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 9.124185336503656e-05,
+      "loss": 2.253,
+      "step": 2610
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 9.085877503554455e-05,
+      "loss": 2.0702,
+      "step": 2615
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 9.04758319357726e-05,
+      "loss": 2.1399,
+      "step": 2620
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 9.009302973074783e-05,
+      "loss": 2.2487,
+      "step": 2625
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.971037408341319e-05,
+      "loss": 2.0532,
+      "step": 2630
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.932787065454352e-05,
+      "loss": 2.2577,
+      "step": 2635
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.894552510266172e-05,
+      "loss": 2.3149,
+      "step": 2640
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.856334308395536e-05,
+      "loss": 1.9829,
+      "step": 2645
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.818133025219258e-05,
+      "loss": 2.1338,
+      "step": 2650
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.77994922586388e-05,
+      "loss": 2.1417,
+      "step": 2655
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.741783475197301e-05,
+      "loss": 2.1236,
+      "step": 2660
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.703636337820407e-05,
+      "loss": 1.9999,
+      "step": 2665
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.665508378058737e-05,
+      "loss": 2.5837,
+      "step": 2670
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.627400159954136e-05,
+      "loss": 2.1618,
+      "step": 2675
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.589312247256385e-05,
+      "loss": 2.2525,
+      "step": 2680
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.551245203414901e-05,
+      "loss": 2.0536,
+      "step": 2685
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.51319959157036e-05,
+      "loss": 2.0885,
+      "step": 2690
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.475175974546394e-05,
+      "loss": 1.9987,
+      "step": 2695
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.437174914841261e-05,
+      "loss": 2.0031,
+      "step": 2700
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.399196974619516e-05,
+      "loss": 2.2783,
+      "step": 2705
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.36124271570369e-05,
+      "loss": 2.1375,
+      "step": 2710
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.323312699566008e-05,
+      "loss": 2.1599,
+      "step": 2715
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.285407487320042e-05,
+      "loss": 2.274,
+      "step": 2720
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.24752763971244e-05,
+      "loss": 2.0749,
+      "step": 2725
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.209673717114618e-05,
+      "loss": 2.3841,
+      "step": 2730
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.171846279514476e-05,
+      "loss": 2.1772,
+      "step": 2735
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.134045886508108e-05,
+      "loss": 2.008,
+      "step": 2740
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.096273097291526e-05,
+      "loss": 2.2219,
+      "step": 2745
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.058528470652396e-05,
+      "loss": 2.1922,
+      "step": 2750
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.020812564961754e-05,
+      "loss": 2.1488,
+      "step": 2755
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.983125938165758e-05,
+      "loss": 2.2547,
+      "step": 2760
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.94546914777744e-05,
+      "loss": 2.0456,
+      "step": 2765
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.907842750868441e-05,
+      "loss": 2.0717,
+      "step": 2770
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.870247304060776e-05,
+      "loss": 2.1157,
+      "step": 2775
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.832683363518621e-05,
+      "loss": 2.2051,
+      "step": 2780
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.795151484940047e-05,
+      "loss": 2.2932,
+      "step": 2785
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.757652223548836e-05,
+      "loss": 2.2374,
+      "step": 2790
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.720186134086241e-05,
+      "loss": 1.9577,
+      "step": 2795
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.682753770802791e-05,
+      "loss": 2.2365,
+      "step": 2800
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.645355687450095e-05,
+      "loss": 2.2679,
+      "step": 2805
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.607992437272642e-05,
+      "loss": 2.1363,
+      "step": 2810
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.570664572999616e-05,
+      "loss": 2.2174,
+      "step": 2815
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.533372646836736e-05,
+      "loss": 2.3122,
+      "step": 2820
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.496117210458063e-05,
+      "loss": 2.3869,
+      "step": 2825
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.458898814997852e-05,
+      "loss": 2.0335,
+      "step": 2830
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.421718011042406e-05,
+      "loss": 1.883,
+      "step": 2835
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.384575348621909e-05,
+      "loss": 2.2373,
+      "step": 2840
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.347471377202314e-05,
+      "loss": 2.3576,
+      "step": 2845
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.31040664567719e-05,
+      "loss": 2.2499,
+      "step": 2850
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.273381702359632e-05,
+      "loss": 2.2617,
+      "step": 2855
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.236397094974119e-05,
+      "loss": 2.1201,
+      "step": 2860
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.199453370648427e-05,
+      "loss": 2.0605,
+      "step": 2865
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.162551075905538e-05,
+      "loss": 2.0576,
+      "step": 2870
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.125690756655545e-05,
+      "loss": 2.0743,
+      "step": 2875
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.088872958187578e-05,
+      "loss": 2.2857,
+      "step": 2880
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.052098225161756e-05,
+      "loss": 1.9972,
+      "step": 2885
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.015367101601091e-05,
+      "loss": 2.0777,
+      "step": 2890
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 6.978680130883488e-05,
+      "loss": 2.3187,
+      "step": 2895
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 6.942037855733661e-05,
+      "loss": 2.165,
+      "step": 2900
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 6.905440818215133e-05,
+      "loss": 2.0252,
+      "step": 2905
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 6.868889559722213e-05,
+      "loss": 2.1558,
+      "step": 2910
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 6.832384620971977e-05,
+      "loss": 2.2264,
+      "step": 2915
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 6.795926541996273e-05,
+      "loss": 2.1026,
+      "step": 2920
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 6.759515862133745e-05,
+      "loss": 1.9323,
+      "step": 2925
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 6.723153120021833e-05,
+      "loss": 2.2133,
+      "step": 2930
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 6.686838853588815e-05,
+      "loss": 2.2056,
+      "step": 2935
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 6.65057360004586e-05,
+      "loss": 2.2081,
+      "step": 2940
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 6.614357895879057e-05,
+      "loss": 2.2055,
+      "step": 2945
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 6.578192276841501e-05,
+      "loss": 2.2554,
+      "step": 2950
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 6.542077277945353e-05,
+      "loss": 2.0585,
+      "step": 2955
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 6.506013433453926e-05,
+      "loss": 2.3229,
+      "step": 2960
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 6.470001276873803e-05,
+      "loss": 2.1304,
+      "step": 2965
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 6.434041340946909e-05,
+      "loss": 2.1815,
+      "step": 2970
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 6.398134157642659e-05,
+      "loss": 1.9168,
+      "step": 2975
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 6.362280258150074e-05,
+      "loss": 2.332,
+      "step": 2980
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.326480172869921e-05,
+      "loss": 2.2381,
+      "step": 2985
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.29073443140689e-05,
+      "loss": 2.2526,
+      "step": 2990
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.255043562561719e-05,
+      "loss": 2.2872,
+      "step": 2995
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.219408094323415e-05,
+      "loss": 2.364,
+      "step": 3000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.183828553861409e-05,
+      "loss": 1.9686,
+      "step": 3005
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.148305467517768e-05,
+      "loss": 2.3009,
+      "step": 3010
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.112839360799422e-05,
+      "loss": 2.1697,
+      "step": 3015
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.077430758370376e-05,
+      "loss": 2.1991,
+      "step": 3020
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.042080184043941e-05,
+      "loss": 2.223,
+      "step": 3025
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.0067881607750134e-05,
+      "loss": 2.0575,
+      "step": 3030
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.971555210652304e-05,
+      "loss": 2.2195,
+      "step": 3035
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.936381854890646e-05,
+      "loss": 2.1538,
+      "step": 3040
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.9012686138232585e-05,
+      "loss": 2.2694,
+      "step": 3045
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.8662160068940655e-05,
+      "loss": 2.2146,
+      "step": 3050
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.831224552650003e-05,
+      "loss": 2.0831,
+      "step": 3055
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.796294768733362e-05,
+      "loss": 2.3443,
+      "step": 3060
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.761427171874103e-05,
+      "loss": 2.2026,
+      "step": 3065
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.726622277882243e-05,
+      "loss": 2.3855,
+      "step": 3070
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.691880601640204e-05,
+      "loss": 2.1517,
+      "step": 3075
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.657202657095206e-05,
+      "loss": 1.9031,
+      "step": 3080
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.622588957251665e-05,
+      "loss": 2.0287,
+      "step": 3085
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.588040014163585e-05,
+      "loss": 2.0336,
+      "step": 3090
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.5535563389270196e-05,
+      "loss": 2.3351,
+      "step": 3095
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.519138441672471e-05,
+      "loss": 2.2632,
+      "step": 3100
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.484786831557369e-05,
+      "loss": 2.2373,
+      "step": 3105
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.4505020167585396e-05,
+      "loss": 2.3092,
+      "step": 3110
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.416284504464657e-05,
+      "loss": 1.9927,
+      "step": 3115
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.3821348008687967e-05,
+      "loss": 2.2656,
+      "step": 3120
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.3480534111608805e-05,
+      "loss": 2.1348,
+      "step": 3125
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.314040839520253e-05,
+      "loss": 2.2229,
+      "step": 3130
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.2800975891082126e-05,
+      "loss": 2.1983,
+      "step": 3135
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.2462241620605366e-05,
+      "loss": 2.1517,
+      "step": 3140
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.21242105948011e-05,
+      "loss": 2.4824,
+      "step": 3145
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.178688781429455e-05,
+      "loss": 2.0474,
+      "step": 3150
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.145027826923371e-05,
+      "loss": 2.3389,
+      "step": 3155
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.111438693921536e-05,
+      "loss": 2.1822,
+      "step": 3160
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.0779218793211524e-05,
+      "loss": 2.0217,
+      "step": 3165
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.044477878949571e-05,
+      "loss": 2.3117,
+      "step": 3170
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.011107187556989e-05,
+      "loss": 1.9003,
+      "step": 3175
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.97781029880911e-05,
+      "loss": 2.1841,
+      "step": 3180
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.944587705279846e-05,
+      "loss": 2.2673,
+      "step": 3185
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.911439898444036e-05,
+      "loss": 1.9417,
+      "step": 3190
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.8783673686701514e-05,
+      "loss": 2.0593,
+      "step": 3195
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.845370605213091e-05,
+      "loss": 2.3553,
+      "step": 3200
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.812450096206884e-05,
+      "loss": 2.3565,
+      "step": 3205
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.779606328657513e-05,
+      "loss": 2.1688,
+      "step": 3210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.7468397884356966e-05,
+      "loss": 2.2295,
+      "step": 3215
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.71415096026968e-05,
+      "loss": 2.3155,
+      "step": 3220
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.681540327738113e-05,
+      "loss": 2.0892,
+      "step": 3225
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.64900837326284e-05,
+      "loss": 2.245,
+      "step": 3230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.6165555781018046e-05,
+      "loss": 2.3306,
+      "step": 3235
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.584182422341915e-05,
+      "loss": 2.2678,
+      "step": 3240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.5518893848919286e-05,
+      "loss": 1.9734,
+      "step": 3245
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.519676943475408e-05,
+      "loss": 2.0935,
+      "step": 3250
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.4875455746236004e-05,
+      "loss": 2.1692,
+      "step": 3255
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.455495753668428e-05,
+      "loss": 2.2828,
+      "step": 3260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.4235279547354414e-05,
+      "loss": 2.2103,
+      "step": 3265
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.391642650736811e-05,
+      "loss": 2.3463,
+      "step": 3270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.3598403133643165e-05,
+      "loss": 2.1223,
+      "step": 3275
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.328121413082388e-05,
+      "loss": 1.9965,
+      "step": 3280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.296486419121137e-05,
+      "loss": 2.1909,
+      "step": 3285
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.264935799469417e-05,
+      "loss": 2.2577,
+      "step": 3290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.233470020867899e-05,
+      "loss": 2.1195,
+      "step": 3295
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.202089548802157e-05,
+      "loss": 1.8477,
+      "step": 3300
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.170794847495814e-05,
+      "loss": 2.0003,
+      "step": 3305
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.139586379903629e-05,
+      "loss": 2.0966,
+      "step": 3310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.1084646077046864e-05,
+      "loss": 1.9883,
+      "step": 3315
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.077429991295549e-05,
+      "loss": 2.0205,
+      "step": 3320
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.046482989783441e-05,
+      "loss": 2.17,
+      "step": 3325
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.015624060979486e-05,
+      "loss": 2.114,
+      "step": 3330
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.984853661391888e-05,
+      "loss": 2.0522,
+      "step": 3335
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.9541722462192196e-05,
+      "loss": 2.2381,
+      "step": 3340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.923580269343673e-05,
+      "loss": 2.0224,
+      "step": 3345
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.893078183324329e-05,
+      "loss": 2.3678,
+      "step": 3350
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.8626664393905024e-05,
+      "loss": 1.8977,
+      "step": 3355
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.832345487435019e-05,
+      "loss": 2.1492,
+      "step": 3360
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.8021157760075945e-05,
+      "loss": 2.0719,
+      "step": 3365
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.7719777523081864e-05,
+      "loss": 2.1618,
+      "step": 3370
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.741931862180384e-05,
+      "loss": 2.1497,
+      "step": 3375
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.7119785501047977e-05,
+      "loss": 1.8983,
+      "step": 3380
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.682118259192502e-05,
+      "loss": 2.2605,
+      "step": 3385
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.652351431178473e-05,
+      "loss": 2.1666,
+      "step": 3390
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.6226785064150524e-05,
+      "loss": 2.2626,
+      "step": 3395
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.593099923865438e-05,
+      "loss": 2.0513,
+      "step": 3400
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.5636161210971705e-05,
+      "loss": 2.1085,
+      "step": 3405
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.5342275342757046e-05,
+      "loss": 1.9572,
+      "step": 3410
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.5049345981578994e-05,
+      "loss": 2.1665,
+      "step": 3415
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.475737746085631e-05,
+      "loss": 2.2385,
+      "step": 3420
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.4466374099793696e-05,
+      "loss": 2.1853,
+      "step": 3425
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.417634020331769e-05,
+      "loss": 2.0076,
+      "step": 3430
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.3887280062013425e-05,
+      "loss": 2.1504,
+      "step": 3435
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.359919795206065e-05,
+      "loss": 1.9365,
+      "step": 3440
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.331209813517082e-05,
+      "loss": 2.2387,
+      "step": 3445
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.302598485852401e-05,
+      "loss": 1.9987,
+      "step": 3450
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.2740862354705834e-05,
+      "loss": 2.0303,
+      "step": 3455
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.245673484164521e-05,
+      "loss": 2.2559,
+      "step": 3460
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.2173606522551636e-05,
+      "loss": 2.0267,
+      "step": 3465
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.1891481585853224e-05,
+      "loss": 2.0388,
+      "step": 3470
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.1610364205134615e-05,
+      "loss": 1.9447,
+      "step": 3475
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.133025853907531e-05,
+      "loss": 2.2349,
+      "step": 3480
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.105116873138806e-05,
+      "loss": 2.0192,
+      "step": 3485
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.077309891075766e-05,
+      "loss": 2.1557,
+      "step": 3490
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.0496053190779872e-05,
+      "loss": 2.2377,
+      "step": 3495
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.0220035669900493e-05,
+      "loss": 2.2612,
+      "step": 3500
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.9945050431354814e-05,
+      "loss": 2.0673,
+      "step": 3505
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.9671101543107037e-05,
+      "loss": 1.9976,
+      "step": 3510
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.9398193057790456e-05,
+      "loss": 2.1429,
+      "step": 3515
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.9126329012647048e-05,
+      "loss": 2.1853,
+      "step": 3520
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.8855513429468107e-05,
+      "loss": 2.2061,
+      "step": 3525
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.8585750314534633e-05,
+      "loss": 2.4006,
+      "step": 3530
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.8317043658557906e-05,
+      "loss": 2.084,
+      "step": 3535
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.8049397436620817e-05,
+      "loss": 2.0911,
+      "step": 3540
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.778281560811864e-05,
+      "loss": 2.0431,
+      "step": 3545
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.751730211670075e-05,
+      "loss": 2.2465,
+      "step": 3550
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.725286089021223e-05,
+      "loss": 2.1271,
+      "step": 3555
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.6989495840635615e-05,
+      "loss": 2.152,
+      "step": 3560
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.6727210864033235e-05,
+      "loss": 2.2119,
+      "step": 3565
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.6466009840489436e-05,
+      "loss": 1.9971,
+      "step": 3570
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.620589663405323e-05,
+      "loss": 1.9645,
+      "step": 3575
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.5946875092681134e-05,
+      "loss": 2.0142,
+      "step": 3580
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.5688949048180245e-05,
+      "loss": 2.1484,
+      "step": 3585
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.5432122316151463e-05,
+      "loss": 2.2044,
+      "step": 3590
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.5176398695933233e-05,
+      "loss": 2.3567,
+      "step": 3595
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4921781970545178e-05,
+      "loss": 2.0796,
+      "step": 3600
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4668275906632232e-05,
+      "loss": 2.0059,
+      "step": 3605
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.441588425440886e-05,
+      "loss": 2.2845,
+      "step": 3610
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4164610747603524e-05,
+      "loss": 2.1271,
+      "step": 3615
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3914459103403696e-05,
+      "loss": 2.2532,
+      "step": 3620
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3665433022400517e-05,
+      "loss": 2.1088,
+      "step": 3625
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3417536188534327e-05,
+      "loss": 2.0666,
+      "step": 3630
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3170772269040075e-05,
+      "loss": 1.994,
+      "step": 3635
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.292514491439297e-05,
+      "loss": 2.1057,
+      "step": 3640
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.268065775825473e-05,
+      "loss": 2.2502,
+      "step": 3645
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.2437314417419518e-05,
+      "loss": 2.0369,
+      "step": 3650
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.2195118491760668e-05,
+      "loss": 2.13,
+      "step": 3655
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.195407356417737e-05,
+      "loss": 2.1939,
+      "step": 3660
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.171418320054155e-05,
+      "loss": 2.3154,
+      "step": 3665
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.1475450949645325e-05,
+      "loss": 2.0751,
+      "step": 3670
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.1237880343148352e-05,
+      "loss": 2.0632,
+      "step": 3675
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.100147489552562e-05,
+      "loss": 2.2255,
+      "step": 3680
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.076623810401549e-05,
+      "loss": 2.2219,
+      "step": 3685
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.0532173448567936e-05,
+      "loss": 2.2883,
+      "step": 3690
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.0299284391793006e-05,
+      "loss": 2.0587,
+      "step": 3695
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.0067574378909726e-05,
+      "loss": 2.1494,
+      "step": 3700
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.9837046837695072e-05,
+      "loss": 2.3035,
+      "step": 3705
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.9607705178433124e-05,
+      "loss": 2.0228,
+      "step": 3710
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.9379552793864962e-05,
+      "loss": 1.9292,
+      "step": 3715
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.9152593059138036e-05,
+      "loss": 2.1265,
+      "step": 3720
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.892682933175667e-05,
+      "loss": 1.9264,
+      "step": 3725
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.870226495153199e-05,
+      "loss": 2.3023,
+      "step": 3730
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.847890324053283e-05,
+      "loss": 2.213,
+      "step": 3735
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.8256747503036465e-05,
+      "loss": 2.0093,
+      "step": 3740
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.8035801025479615e-05,
+      "loss": 2.1466,
+      "step": 3745
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.7816067076410138e-05,
+      "loss": 2.2077,
+      "step": 3750
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.759754890643829e-05,
+      "loss": 2.1645,
+      "step": 3755
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.738024974818896e-05,
+      "loss": 2.1975,
+      "step": 3760
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.71641728162537e-05,
+      "loss": 2.0768,
+      "step": 3765
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.6949321307143096e-05,
+      "loss": 2.1729,
+      "step": 3770
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.6735698399239687e-05,
+      "loss": 2.1934,
+      "step": 3775
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.6523307252750787e-05,
+      "loss": 2.0382,
+      "step": 3780
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.6312151009661792e-05,
+      "loss": 2.1809,
+      "step": 3785
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.6102232793689652e-05,
+      "loss": 2.2001,
+      "step": 3790
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.5893555710236784e-05,
+      "loss": 2.2799,
+      "step": 3795
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.5686122846344932e-05,
+      "loss": 2.2641,
+      "step": 3800
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.547993727064968e-05,
+      "loss": 2.126,
+      "step": 3805
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.5275002033335016e-05,
+      "loss": 2.1068,
+      "step": 3810
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.5071320166088066e-05,
+      "loss": 2.0277,
+      "step": 3815
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.4868894682054535e-05,
+      "loss": 2.1387,
+      "step": 3820
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.4667728575793772e-05,
+      "loss": 2.3061,
+      "step": 3825
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.4467824823234843e-05,
+      "loss": 2.0555,
+      "step": 3830
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.426918638163216e-05,
+      "loss": 2.206,
+      "step": 3835
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.407181618952199e-05,
+      "loss": 2.0053,
+      "step": 3840
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.3875717166678892e-05,
+      "loss": 2.0465,
+      "step": 3845
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.3680892214072405e-05,
+      "loss": 2.0906,
+      "step": 3850
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.3487344213824437e-05,
+      "loss": 2.049,
+      "step": 3855
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.3295076029166265e-05,
+      "loss": 2.0334,
+      "step": 3860
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.310409050439645e-05,
+      "loss": 2.2541,
+      "step": 3865
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.2914390464838655e-05,
+      "loss": 2.339,
+      "step": 3870
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.2725978716799791e-05,
+      "loss": 2.1708,
+      "step": 3875
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.2538858047528646e-05,
+      "loss": 2.1017,
+      "step": 3880
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.235303122517455e-05,
+      "loss": 2.1563,
+      "step": 3885
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.2168500998746435e-05,
+      "loss": 1.9978,
+      "step": 3890
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1985270098072188e-05,
+      "loss": 2.0827,
+      "step": 3895
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1803341233758291e-05,
+      "loss": 2.2082,
+      "step": 3900
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1622717097149627e-05,
+      "loss": 2.08,
+      "step": 3905
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.144340036028978e-05,
+      "loss": 2.0822,
+      "step": 3910
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1265393675881487e-05,
+      "loss": 2.1352,
+      "step": 3915
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1088699677247238e-05,
+      "loss": 2.0985,
+      "step": 3920
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.0913320978290642e-05,
+      "loss": 2.011,
+      "step": 3925
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.0739260173457355e-05,
+      "loss": 1.9461,
+      "step": 3930
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.056651983769712e-05,
+      "loss": 2.0538,
+      "step": 3935
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.0395102526425282e-05,
+      "loss": 1.8053,
+      "step": 3940
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.0225010775485277e-05,
+      "loss": 2.1013,
+      "step": 3945
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.0056247101110972e-05,
+      "loss": 2.0568,
+      "step": 3950
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 9.888813999889446e-06,
+      "loss": 2.1572,
+      "step": 3955
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 9.72271394872416e-06,
+      "loss": 2.263,
+      "step": 3960
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 9.557949404798183e-06,
+      "loss": 2.0593,
+      "step": 3965
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 9.394522805537931e-06,
+      "loss": 2.0521,
+      "step": 3970
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 9.232436568577085e-06,
+      "loss": 2.2315,
+      "step": 3975
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 9.0716930917208e-06,
+      "loss": 2.0703,
+      "step": 3980
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.912294752910254e-06,
+      "loss": 2.4333,
+      "step": 3985
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.754243910187498e-06,
+      "loss": 2.213,
+      "step": 3990
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.597542901660538e-06,
+      "loss": 1.992,
+      "step": 3995
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.442194045468733e-06,
+      "loss": 2.0252,
+      "step": 4000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.28819963974855e-06,
+      "loss": 1.9801,
+      "step": 4005
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.135561962599514e-06,
+      "loss": 2.1347,
+      "step": 4010
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.984283272050553e-06,
+      "loss": 2.1591,
+      "step": 4015
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.834365806026578e-06,
+      "loss": 1.9528,
+      "step": 4020
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.68581178231531e-06,
+      "loss": 1.9756,
+      "step": 4025
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.538623398534661e-06,
+      "loss": 2.093,
+      "step": 4030
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.392802832099965e-06,
+      "loss": 1.9631,
+      "step": 4035
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.248352240192002e-06,
+      "loss": 2.135,
+      "step": 4040
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.105273759724884e-06,
+      "loss": 2.1906,
+      "step": 4045
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 6.963569507314627e-06,
+      "loss": 2.3328,
+      "step": 4050
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 6.82324157924773e-06,
+      "loss": 2.2979,
+      "step": 4055
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 6.684292051450147e-06,
+      "loss": 1.9134,
+      "step": 4060
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 6.546722979456643e-06,
+      "loss": 2.1358,
+      "step": 4065
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 6.410536398380385e-06,
+      "loss": 2.2415,
+      "step": 4070
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 6.275734322882776e-06,
+      "loss": 2.333,
+      "step": 4075
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 6.142318747143716e-06,
+      "loss": 2.001,
+      "step": 4080
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 6.010291644832044e-06,
+      "loss": 2.1209,
+      "step": 4085
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.8796549690763645e-06,
+      "loss": 1.9222,
+      "step": 4090
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.750410652436178e-06,
+      "loss": 2.1773,
+      "step": 4095
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.622560606873262e-06,
+      "loss": 2.2215,
+      "step": 4100
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.496106723723416e-06,
+      "loss": 1.8445,
+      "step": 4105
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.371050873668437e-06,
+      "loss": 2.3022,
+      "step": 4110
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.247394906708469e-06,
+      "loss": 1.9745,
+      "step": 4115
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.125140652134652e-06,
+      "loss": 2.0829,
+      "step": 4120
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.004289918502047e-06,
+      "loss": 1.9314,
+      "step": 4125
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 4.884844493602847e-06,
+      "loss": 2.0076,
+      "step": 4130
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 4.766806144440039e-06,
+      "loss": 2.1293,
+      "step": 4135
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 4.650176617201074e-06,
+      "loss": 2.0244,
+      "step": 4140
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.5349576372322845e-06,
+      "loss": 1.9962,
+      "step": 4145
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.421150909013094e-06,
+      "loss": 2.1399,
+      "step": 4150
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.308758116131029e-06,
+      "loss": 1.8508,
+      "step": 4155
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.197780921256678e-06,
+      "loss": 1.9079,
+      "step": 4160
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 4.088220966119105e-06,
+      "loss": 2.0411,
+      "step": 4165
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.9800798714816566e-06,
+      "loss": 2.1043,
+      "step": 4170
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.873359237117868e-06,
+      "loss": 2.1293,
+      "step": 4175
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.768060641787874e-06,
+      "loss": 2.07,
+      "step": 4180
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.6641856432150232e-06,
+      "loss": 2.0227,
+      "step": 4185
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.561735778062847e-06,
+      "loss": 2.1693,
+      "step": 4190
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.4607125619123048e-06,
+      "loss": 1.9553,
+      "step": 4195
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.3611174892393848e-06,
+      "loss": 2.1492,
+      "step": 4200
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.2629520333930075e-06,
+      "loss": 2.3135,
+      "step": 4205
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.1662176465731776e-06,
+      "loss": 2.1186,
+      "step": 4210
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.070915759809567e-06,
+      "loss": 2.2169,
+      "step": 4215
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.97704778294029e-06,
+      "loss": 1.9161,
+      "step": 4220
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.884615104591093e-06,
+      "loss": 2.0842,
+      "step": 4225
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.793619092154787e-06,
+      "loss": 2.0509,
+      "step": 4230
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.7040610917709597e-06,
+      "loss": 2.0457,
+      "step": 4235
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.6159424283062507e-06,
+      "loss": 2.0591,
+      "step": 4240
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.529264405334486e-06,
+      "loss": 2.0041,
+      "step": 4245
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.4440283051176405e-06,
+      "loss": 2.227,
+      "step": 4250
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.3602353885866956e-06,
+      "loss": 1.8759,
+      "step": 4255
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.277886895323078e-06,
+      "loss": 2.0249,
+      "step": 4260
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.1969840435402953e-06,
+      "loss": 2.1304,
+      "step": 4265
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.117528030065907e-06,
+      "loss": 2.1629,
+      "step": 4270
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.039520030323816e-06,
+      "loss": 2.1015,
+      "step": 4275
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.962961198316937e-06,
+      "loss": 2.0901,
+      "step": 4280
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.8878526666100128e-06,
+      "loss": 2.2332,
+      "step": 4285
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.8141955463129912e-06,
+      "loss": 2.1336,
+      "step": 4290
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.7419909270644853e-06,
+      "loss": 2.0399,
+      "step": 4295
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.6712398770156734e-06,
+      "loss": 1.9857,
+      "step": 4300
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.6019434428145686e-06,
+      "loss": 2.0271,
+      "step": 4305
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.5341026495904409e-06,
+      "loss": 2.1896,
+      "step": 4310
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.4677185009387085e-06,
+      "loss": 2.2911,
+      "step": 4315
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.4027919789060818e-06,
+      "loss": 2.082,
+      "step": 4320
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.339324043976009e-06,
+      "loss": 1.829,
+      "step": 4325
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.27731563505451e-06,
+      "loss": 2.3073,
+      "step": 4330
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.2167676694562535e-06,
+      "loss": 2.22,
+      "step": 4335
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.1576810428910012e-06,
+      "loss": 2.1452,
+      "step": 4340
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.1000566294503523e-06,
+      "loss": 1.9852,
+      "step": 4345
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.04389528159482e-06,
+      "loss": 2.1192,
+      "step": 4350
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 9.891978301412308e-07,
+      "loss": 2.1154,
+      "step": 4355
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 9.359650842503565e-07,
+      "loss": 1.9588,
+      "step": 4360
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.841978314150901e-07,
+      "loss": 2.2275,
+      "step": 4365
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.338968374486555e-07,
+      "loss": 2.1793,
+      "step": 4370
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.850628464733501e-07,
+      "loss": 2.1219,
+      "step": 4375
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.376965809095193e-07,
+      "loss": 2.1371,
+      "step": 4380
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.917987414648885e-07,
+      "loss": 1.8774,
+      "step": 4385
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.473700071241484e-07,
+      "loss": 2.173,
+      "step": 4390
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.044110351389743e-07,
+      "loss": 2.0139,
+      "step": 4395
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.629224610182671e-07,
+      "loss": 2.012,
+      "step": 4400
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.229048985187279e-07,
+      "loss": 2.0772,
+      "step": 4405
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 4.843589396358427e-07,
+      "loss": 1.9894,
+      "step": 4410
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 4.472851545950563e-07,
+      "loss": 2.0555,
+      "step": 4415
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 4.116840918434006e-07,
+      "loss": 1.9658,
+      "step": 4420
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.7755627804134664e-07,
+      "loss": 2.1685,
+      "step": 4425
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.449022180549766e-07,
+      "loss": 2.1805,
+      "step": 4430
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.1372239494860123e-07,
+      "loss": 2.0124,
+      "step": 4435
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.840172699775656e-07,
+      "loss": 2.0179,
+      "step": 4440
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.5578728258138783e-07,
+      "loss": 2.1432,
+      "step": 4445
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.290328503773309e-07,
+      "loss": 2.1226,
+      "step": 4450
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.0375436915420764e-07,
+      "loss": 2.0209,
+      "step": 4455
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7995221286645215e-07,
+      "loss": 2.1942,
+      "step": 4460
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.5762673362871293e-07,
+      "loss": 2.3072,
+      "step": 4465
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.367782617105351e-07,
+      "loss": 2.15,
+      "step": 4470
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.1740710553153067e-07,
+      "loss": 2.0169,
+      "step": 4475
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 9.951355165678244e-08,
+      "loss": 2.1183,
+      "step": 4480
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 8.309786479264726e-08,
+      "loss": 2.1434,
+      "step": 4485
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.816028778281469e-08,
+      "loss": 2.0947,
+      "step": 4490
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.4701041604710014e-08,
+      "loss": 2.0434,
+      "step": 4495
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 4.272032536621895e-08,
+      "loss": 2.0497,
+      "step": 4500
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.221831630277894e-08,
+      "loss": 2.207,
+      "step": 4505
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.3195169774714586e-08,
+      "loss": 2.1715,
+      "step": 4510
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.565101926497281e-08,
+      "loss": 2.2098,
+      "step": 4515
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 9.585976377124439e-09,
+      "loss": 2.1455,
+      "step": 4520
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.0001308337210835e-09,
+      "loss": 2.1535,
+      "step": 4525
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8935504749628684e-09,
+      "loss": 1.9348,
+      "step": 4530
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6628125773253686e-10,
+      "loss": 2.1515,
+      "step": 4535
+    },
+    {
+      "epoch": 2.0,
+      "step": 4538,
+      "total_flos": 1.080061286023168e+16,
+      "train_loss": 2.5026560972944236,
+      "train_runtime": 41380.2373,
+      "train_samples_per_second": 0.439,
+      "train_steps_per_second": 0.11
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 4538,
+  "num_train_epochs": 2,
+  "save_steps": 100,
+  "total_flos": 1.080061286023168e+16,
+  "trial_name": null,
+  "trial_params": null
+}