diff --git "a/checkpoint-6600/trainer_state.json" "b/checkpoint-6600/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-6600/trainer_state.json"
@@ -0,0 +1,7941 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.3915724563206577,
+  "eval_steps": 500,
+  "global_step": 6600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9999979150098865e-05,
+      "loss": 4.0418,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999989444743508e-05,
+      "loss": 3.8361,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999974458911041e-05,
+      "loss": 3.8737,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9999529575515425e-05,
+      "loss": 3.806,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9999249407210515e-05,
+      "loss": 3.4125,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9998904084925845e-05,
+      "loss": 3.6489,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999849360956141e-05,
+      "loss": 3.4719,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999811831976564e-05,
+      "loss": 3.5265,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999759057166764e-05,
+      "loss": 3.6889,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9996997673913204e-05,
+      "loss": 3.5104,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.999633962804756e-05,
+      "loss": 3.5778,
+      "step": 55
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.999561643578575e-05,
+      "loss": 3.3353,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.999482809901257e-05,
+      "loss": 3.3511,
+      "step": 65
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9993974619782625e-05,
+      "loss": 3.6355,
+      "step": 70
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999305600032027e-05,
+      "loss": 3.3915,
+      "step": 75
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999207224301965e-05,
+      "loss": 3.4926,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999102335044467e-05,
+      "loss": 3.4918,
+      "step": 85
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9989909325328996e-05,
+      "loss": 3.3984,
+      "step": 90
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.998873017057605e-05,
+      "loss": 3.4989,
+      "step": 95
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.998748588925897e-05,
+      "loss": 3.6368,
+      "step": 100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9986176484620665e-05,
+      "loss": 3.4412,
+      "step": 105
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998480196007375e-05,
+      "loss": 3.4708,
+      "step": 110
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9983362319200554e-05,
+      "loss": 3.37,
+      "step": 115
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998185756575313e-05,
+      "loss": 3.4524,
+      "step": 120
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9980287703653225e-05,
+      "loss": 3.3863,
+      "step": 125
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.997865273699227e-05,
+      "loss": 3.4414,
+      "step": 130
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.997729789123732e-05,
+      "loss": 3.3335,
+      "step": 135
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.997554574721849e-05,
+      "loss": 3.231,
+      "step": 140
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.997372851099727e-05,
+      "loss": 3.5962,
+      "step": 145
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.997184618730983e-05,
+      "loss": 3.2795,
+      "step": 150
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.996989878106192e-05,
+      "loss": 3.5384,
+      "step": 155
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.996788629732897e-05,
+      "loss": 3.3274,
+      "step": 160
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.996580874135597e-05,
+      "loss": 3.2292,
+      "step": 165
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.996366611855753e-05,
+      "loss": 3.4829,
+      "step": 170
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.996145843451785e-05,
+      "loss": 3.1998,
+      "step": 175
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9959185694990655e-05,
+      "loss": 3.3757,
+      "step": 180
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.995684790589926e-05,
+      "loss": 3.1001,
+      "step": 185
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.99544450733365e-05,
+      "loss": 3.2301,
+      "step": 190
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.995197720356471e-05,
+      "loss": 3.3187,
+      "step": 195
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.994944430301576e-05,
+      "loss": 3.527,
+      "step": 200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.994684637829098e-05,
+      "loss": 3.3808,
+      "step": 205
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.994418343616119e-05,
+      "loss": 3.2406,
+      "step": 210
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.994145548356664e-05,
+      "loss": 3.3463,
+      "step": 215
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.993866252761702e-05,
+      "loss": 3.2972,
+      "step": 220
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.993580457559145e-05,
+      "loss": 3.4135,
+      "step": 225
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9932881634938424e-05,
+      "loss": 3.4025,
+      "step": 230
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.992989371327581e-05,
+      "loss": 3.662,
+      "step": 235
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.992684081839086e-05,
+      "loss": 3.4279,
+      "step": 240
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9923722958240124e-05,
+      "loss": 3.3075,
+      "step": 245
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.992054014094951e-05,
+      "loss": 3.2826,
+      "step": 250
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.991729237481417e-05,
+      "loss": 3.4992,
+      "step": 255
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.991397966829856e-05,
+      "loss": 3.2421,
+      "step": 260
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.991060203003639e-05,
+      "loss": 3.2615,
+      "step": 265
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9907159468830585e-05,
+      "loss": 3.3299,
+      "step": 270
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.990365199365327e-05,
+      "loss": 3.3117,
+      "step": 275
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9900079613645754e-05,
+      "loss": 3.4729,
+      "step": 280
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9896442338118524e-05,
+      "loss": 3.2521,
+      "step": 285
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.989274017655117e-05,
+      "loss": 3.261,
+      "step": 290
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9888973138592414e-05,
+      "loss": 3.4403,
+      "step": 295
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.988514123406004e-05,
+      "loss": 3.3408,
+      "step": 300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9881244472940914e-05,
+      "loss": 3.3208,
+      "step": 305
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9877282865390904e-05,
+      "loss": 3.6889,
+      "step": 310
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9873256421734915e-05,
+      "loss": 3.1537,
+      "step": 315
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.986916515246681e-05,
+      "loss": 3.1837,
+      "step": 320
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.986500906824941e-05,
+      "loss": 3.23,
+      "step": 325
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9860788179914456e-05,
+      "loss": 3.4343,
+      "step": 330
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.985650249846258e-05,
+      "loss": 3.3317,
+      "step": 335
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9852152035063294e-05,
+      "loss": 3.155,
+      "step": 340
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.984773680105493e-05,
+      "loss": 3.3527,
+      "step": 345
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.984415798673684e-05,
+      "loss": 3.3783,
+      "step": 350
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9839626194743575e-05,
+      "loss": 3.2567,
+      "step": 355
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.983502966478656e-05,
+      "loss": 3.1278,
+      "step": 360
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9830368408845425e-05,
+      "loss": 3.2344,
+      "step": 365
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.982564243906852e-05,
+      "loss": 3.3458,
+      "step": 370
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9820851767772844e-05,
+      "loss": 3.1467,
+      "step": 375
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.981599640744402e-05,
+      "loss": 3.217,
+      "step": 380
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.981107637073628e-05,
+      "loss": 3.4011,
+      "step": 385
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.98060916704724e-05,
+      "loss": 3.4098,
+      "step": 390
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9801042319643686e-05,
+      "loss": 3.3332,
+      "step": 395
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.979592833140996e-05,
+      "loss": 3.4042,
+      "step": 400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.979074971909948e-05,
+      "loss": 3.3127,
+      "step": 405
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.978550649620894e-05,
+      "loss": 3.1726,
+      "step": 410
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.978019867640342e-05,
+      "loss": 3.3029,
+      "step": 415
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.977482627351637e-05,
+      "loss": 3.3519,
+      "step": 420
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.976938930154954e-05,
+      "loss": 3.2507,
+      "step": 425
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.976388777467297e-05,
+      "loss": 3.1322,
+      "step": 430
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9758321707224954e-05,
+      "loss": 3.1306,
+      "step": 435
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.975269111371197e-05,
+      "loss": 3.189,
+      "step": 440
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.974699600880869e-05,
+      "loss": 3.2598,
+      "step": 445
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.974123640735791e-05,
+      "loss": 3.3483,
+      "step": 450
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.973541232437052e-05,
+      "loss": 3.4678,
+      "step": 455
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.972952377502545e-05,
+      "loss": 2.9668,
+      "step": 460
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9723570774669687e-05,
+      "loss": 3.372,
+      "step": 465
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.971755333881814e-05,
+      "loss": 3.2808,
+      "step": 470
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.971147148315367e-05,
+      "loss": 3.2235,
+      "step": 475
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9705325223527057e-05,
+      "loss": 3.0441,
+      "step": 480
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.96991145759569e-05,
+      "loss": 3.4724,
+      "step": 485
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.969283955662963e-05,
+      "loss": 3.389,
+      "step": 490
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.968650018189943e-05,
+      "loss": 3.3817,
+      "step": 495
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.968009646828822e-05,
+      "loss": 3.2597,
+      "step": 500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.967362843248559e-05,
+      "loss": 3.253,
+      "step": 505
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.966709609134878e-05,
+      "loss": 3.1903,
+      "step": 510
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.9660499461902645e-05,
+      "loss": 3.1286,
+      "step": 515
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.965383856133953e-05,
+      "loss": 3.292,
+      "step": 520
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.964711340701935e-05,
+      "loss": 3.4217,
+      "step": 525
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.964032401646944e-05,
+      "loss": 3.2587,
+      "step": 530
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.963347040738457e-05,
+      "loss": 3.3325,
+      "step": 535
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.962655259762687e-05,
+      "loss": 3.3475,
+      "step": 540
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.961957060522581e-05,
+      "loss": 3.2882,
+      "step": 545
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.961252444837809e-05,
+      "loss": 3.3631,
+      "step": 550
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.9605414145447696e-05,
+      "loss": 3.2589,
+      "step": 555
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.959823971496574e-05,
+      "loss": 3.4855,
+      "step": 560
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.959100117563052e-05,
+      "loss": 3.1354,
+      "step": 565
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.958369854630737e-05,
+      "loss": 3.3037,
+      "step": 570
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.957633184602867e-05,
+      "loss": 3.1023,
+      "step": 575
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.956890109399381e-05,
+      "loss": 3.1933,
+      "step": 580
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.956140630956908e-05,
+      "loss": 3.3994,
+      "step": 585
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.955384751228767e-05,
+      "loss": 3.1178,
+      "step": 590
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.954622472184961e-05,
+      "loss": 3.2513,
+      "step": 595
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.9538537958121686e-05,
+      "loss": 3.5002,
+      "step": 600
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.953078724113744e-05,
+      "loss": 3.1867,
+      "step": 605
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.9522972591097074e-05,
+      "loss": 3.3883,
+      "step": 610
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.9515094028367425e-05,
+      "loss": 3.3708,
+      "step": 615
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.950715157348191e-05,
+      "loss": 3.2425,
+      "step": 620
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.9499145247140434e-05,
+      "loss": 3.1046,
+      "step": 625
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.9491075070209385e-05,
+      "loss": 3.3236,
+      "step": 630
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.9482941063721576e-05,
+      "loss": 3.4957,
+      "step": 635
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.947474324887614e-05,
+      "loss": 3.2225,
+      "step": 640
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.9466481647038554e-05,
+      "loss": 3.5314,
+      "step": 645
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.9458156279740484e-05,
+      "loss": 3.1028,
+      "step": 650
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.944976716867984e-05,
+      "loss": 3.2322,
+      "step": 655
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.9441314335720616e-05,
+      "loss": 3.1332,
+      "step": 660
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.943279780289292e-05,
+      "loss": 3.2028,
+      "step": 665
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.9424217592392854e-05,
+      "loss": 3.2408,
+      "step": 670
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.9415573726582465e-05,
+      "loss": 3.0121,
+      "step": 675
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.940686622798974e-05,
+      "loss": 3.2182,
+      "step": 680
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.93980951193085e-05,
+      "loss": 3.4794,
+      "step": 685
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.93892604233983e-05,
+      "loss": 3.3226,
+      "step": 690
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.938036216328448e-05,
+      "loss": 3.1917,
+      "step": 695
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.9371400362158016e-05,
+      "loss": 3.2176,
+      "step": 700
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.936237504337549e-05,
+      "loss": 2.933,
+      "step": 705
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.935328623045902e-05,
+      "loss": 3.134,
+      "step": 710
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.934413394709621e-05,
+      "loss": 2.983,
+      "step": 715
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.933491821714009e-05,
+      "loss": 3.4337,
+      "step": 720
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.9325639064609045e-05,
+      "loss": 3.4699,
+      "step": 725
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.931629651368673e-05,
+      "loss": 3.4081,
+      "step": 730
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.930689058872208e-05,
+      "loss": 3.2174,
+      "step": 735
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.929742131422913e-05,
+      "loss": 3.4953,
+      "step": 740
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.92878887148871e-05,
+      "loss": 3.217,
+      "step": 745
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.927829281554017e-05,
+      "loss": 3.2689,
+      "step": 750
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.926863364119756e-05,
+      "loss": 3.2994,
+      "step": 755
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.9258911217033355e-05,
+      "loss": 3.3868,
+      "step": 760
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.924912556838651e-05,
+      "loss": 3.3169,
+      "step": 765
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.923927672076075e-05,
+      "loss": 3.0466,
+      "step": 770
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.92293646998245e-05,
+      "loss": 3.0857,
+      "step": 775
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.921938953141084e-05,
+      "loss": 3.5001,
+      "step": 780
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.920935124151744e-05,
+      "loss": 3.1633,
+      "step": 785
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.919924985630647e-05,
+      "loss": 3.4816,
+      "step": 790
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.9189085402104516e-05,
+      "loss": 3.2032,
+      "step": 795
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.917885790540258e-05,
+      "loss": 3.164,
+      "step": 800
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.916856739285593e-05,
+      "loss": 3.3047,
+      "step": 805
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.91582138912841e-05,
+      "loss": 3.2366,
+      "step": 810
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.9147797427670766e-05,
+      "loss": 3.3105,
+      "step": 815
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.91373180291637e-05,
+      "loss": 3.4871,
+      "step": 820
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.912677572307473e-05,
+      "loss": 3.0442,
+      "step": 825
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.911617053687957e-05,
+      "loss": 3.1934,
+      "step": 830
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.910550249821787e-05,
+      "loss": 3.27,
+      "step": 835
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.9094771634893075e-05,
+      "loss": 3.3252,
+      "step": 840
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.908397797487237e-05,
+      "loss": 2.9852,
+      "step": 845
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.907312154628658e-05,
+      "loss": 3.1331,
+      "step": 850
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.906220237743014e-05,
+      "loss": 3.1444,
+      "step": 855
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.905122049676101e-05,
+      "loss": 3.3702,
+      "step": 860
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.904017593290056e-05,
+      "loss": 3.0531,
+      "step": 865
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.902906871463355e-05,
+      "loss": 3.2007,
+      "step": 870
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.901789887090804e-05,
+      "loss": 3.0837,
+      "step": 875
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.900666643083528e-05,
+      "loss": 3.2268,
+      "step": 880
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.8995371423689684e-05,
+      "loss": 2.9249,
+      "step": 885
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.898401387890873e-05,
+      "loss": 3.2058,
+      "step": 890
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.897259382609285e-05,
+      "loss": 3.0971,
+      "step": 895
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.8961111295005446e-05,
+      "loss": 3.2227,
+      "step": 900
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.894956631557269e-05,
+      "loss": 3.2907,
+      "step": 905
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.893795891788357e-05,
+      "loss": 3.1382,
+      "step": 910
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.892628913218969e-05,
+      "loss": 2.8996,
+      "step": 915
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.8914556988905315e-05,
+      "loss": 2.9693,
+      "step": 920
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.890276251860717e-05,
+      "loss": 3.0134,
+      "step": 925
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.889090575203445e-05,
+      "loss": 2.9696,
+      "step": 930
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.887898672008871e-05,
+      "loss": 3.3306,
+      "step": 935
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.8867005453833774e-05,
+      "loss": 3.2199,
+      "step": 940
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.885496198449567e-05,
+      "loss": 3.2795,
+      "step": 945
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.8842856343462535e-05,
+      "loss": 3.3482,
+      "step": 950
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.883068856228454e-05,
+      "loss": 3.322,
+      "step": 955
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.881845867267382e-05,
+      "loss": 3.0443,
+      "step": 960
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.8806166706504375e-05,
+      "loss": 3.1826,
+      "step": 965
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.879381269581197e-05,
+      "loss": 3.3732,
+      "step": 970
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.878139667279411e-05,
+      "loss": 3.3457,
+      "step": 975
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.876891866980988e-05,
+      "loss": 3.182,
+      "step": 980
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.875637871937994e-05,
+      "loss": 3.2416,
+      "step": 985
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.874377685418635e-05,
+      "loss": 3.2411,
+      "step": 990
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.873111310707259e-05,
+      "loss": 3.2836,
+      "step": 995
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.8718387511043385e-05,
+      "loss": 3.1773,
+      "step": 1000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.870560009926465e-05,
+      "loss": 3.1723,
+      "step": 1005
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.8692750905063424e-05,
+      "loss": 3.1535,
+      "step": 1010
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.867983996192776e-05,
+      "loss": 2.977,
+      "step": 1015
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.866686730350663e-05,
+      "loss": 3.1236,
+      "step": 1020
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.865383296360987e-05,
+      "loss": 3.4016,
+      "step": 1025
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.864073697620807e-05,
+      "loss": 3.1547,
+      "step": 1030
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.862757937543248e-05,
+      "loss": 3.143,
+      "step": 1035
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.861436019557492e-05,
+      "loss": 3.1989,
+      "step": 1040
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.860107947108772e-05,
+      "loss": 3.0308,
+      "step": 1045
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.858773723658361e-05,
+      "loss": 3.409,
+      "step": 1050
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.8574333526835605e-05,
+      "loss": 3.2011,
+      "step": 1055
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.856086837677697e-05,
+      "loss": 3.1954,
+      "step": 1060
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.854734182150107e-05,
+      "loss": 3.1366,
+      "step": 1065
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.853375389626133e-05,
+      "loss": 3.3789,
+      "step": 1070
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.85201046364711e-05,
+      "loss": 3.1337,
+      "step": 1075
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.85063940777036e-05,
+      "loss": 3.2278,
+      "step": 1080
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.849262225569181e-05,
+      "loss": 3.1339,
+      "step": 1085
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.847878920632834e-05,
+      "loss": 2.954,
+      "step": 1090
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.846489496566542e-05,
+      "loss": 3.0878,
+      "step": 1095
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.845093956991475e-05,
+      "loss": 3.2329,
+      "step": 1100
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.843692305544738e-05,
+      "loss": 3.2245,
+      "step": 1105
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.84228454587937e-05,
+      "loss": 3.2865,
+      "step": 1110
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.8408706816643254e-05,
+      "loss": 3.1569,
+      "step": 1115
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.83945071658447e-05,
+      "loss": 3.2279,
+      "step": 1120
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.838024654340571e-05,
+      "loss": 3.0554,
+      "step": 1125
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.836592498649283e-05,
+      "loss": 3.3924,
+      "step": 1130
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.835154253243147e-05,
+      "loss": 3.0695,
+      "step": 1135
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.8337099218705695e-05,
+      "loss": 3.2666,
+      "step": 1140
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.832259508295822e-05,
+      "loss": 3.135,
+      "step": 1145
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.830803016299027e-05,
+      "loss": 2.947,
+      "step": 1150
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.829340449676147e-05,
+      "loss": 3.13,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.82787181223898e-05,
+      "loss": 3.038,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.826397107815142e-05,
+      "loss": 3.0761,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.824916340248064e-05,
+      "loss": 3.1496,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.823429513396977e-05,
+      "loss": 2.8758,
+      "step": 1175
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.821936631136906e-05,
+      "loss": 3.218,
+      "step": 1180
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.820437697358656e-05,
+      "loss": 3.0933,
+      "step": 1185
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.8189327159688035e-05,
+      "loss": 3.2388,
+      "step": 1190
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.817421690889689e-05,
+      "loss": 3.1914,
+      "step": 1195
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.815904626059401e-05,
+      "loss": 3.2259,
+      "step": 1200
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.814381525431771e-05,
+      "loss": 3.1867,
+      "step": 1205
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.8128523929763626e-05,
+      "loss": 3.1134,
+      "step": 1210
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.811317232678456e-05,
+      "loss": 3.3846,
+      "step": 1215
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.8097760485390444e-05,
+      "loss": 3.2507,
+      "step": 1220
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.8082288445748194e-05,
+      "loss": 3.0992,
+      "step": 1225
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.806675624818162e-05,
+      "loss": 3.0715,
+      "step": 1230
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.8051163933171316e-05,
+      "loss": 3.2028,
+      "step": 1235
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.8035511541354547e-05,
+      "loss": 3.2714,
+      "step": 1240
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.801979911352519e-05,
+      "loss": 3.1238,
+      "step": 1245
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.800402669063353e-05,
+      "loss": 3.2508,
+      "step": 1250
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.7988194313786275e-05,
+      "loss": 3.2359,
+      "step": 1255
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.797230202424633e-05,
+      "loss": 2.8538,
+      "step": 1260
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.7956349863432794e-05,
+      "loss": 3.1991,
+      "step": 1265
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.794033787292078e-05,
+      "loss": 3.0467,
+      "step": 1270
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.7924266094441346e-05,
+      "loss": 3.1088,
+      "step": 1275
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.7908134569881344e-05,
+      "loss": 3.128,
+      "step": 1280
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.789194334128338e-05,
+      "loss": 3.3195,
+      "step": 1285
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.7875692450845624e-05,
+      "loss": 3.2049,
+      "step": 1290
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.785938194092176e-05,
+      "loss": 3.1477,
+      "step": 1295
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.7843011854020846e-05,
+      "loss": 3.2506,
+      "step": 1300
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.7826582232807213e-05,
+      "loss": 3.1822,
+      "step": 1305
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.7810093120100374e-05,
+      "loss": 2.9576,
+      "step": 1310
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.779354455887485e-05,
+      "loss": 3.2086,
+      "step": 1315
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.777693659226013e-05,
+      "loss": 3.2935,
+      "step": 1320
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.776026926354052e-05,
+      "loss": 2.9873,
+      "step": 1325
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.774354261615503e-05,
+      "loss": 3.2671,
+      "step": 1330
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.772675669369728e-05,
+      "loss": 3.0882,
+      "step": 1335
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.770991153991537e-05,
+      "loss": 3.2727,
+      "step": 1340
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.769300719871177e-05,
+      "loss": 2.9907,
+      "step": 1345
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.7676043714143195e-05,
+      "loss": 3.1303,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.765902113042053e-05,
+      "loss": 3.1412,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.7641939491908674e-05,
+      "loss": 3.3585,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.7624798843126404e-05,
+      "loss": 3.1958,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.7607599228746355e-05,
+      "loss": 3.1581,
+      "step": 1370
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.759034069359478e-05,
+      "loss": 3.224,
+      "step": 1375
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.7573023282651536e-05,
+      "loss": 3.1031,
+      "step": 1380
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.75556470410499e-05,
+      "loss": 3.0914,
+      "step": 1385
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.753821201407648e-05,
+      "loss": 3.2474,
+      "step": 1390
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.75207182471711e-05,
+      "loss": 3.2045,
+      "step": 1395
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.7503165785926676e-05,
+      "loss": 2.9288,
+      "step": 1400
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.748555467608909e-05,
+      "loss": 3.3084,
+      "step": 1405
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.7467884963557064e-05,
+      "loss": 3.1761,
+      "step": 1410
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.745015669438207e-05,
+      "loss": 3.0644,
+      "step": 1415
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.7432369914768185e-05,
+      "loss": 3.0874,
+      "step": 1420
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.7414524671071994e-05,
+      "loss": 3.1106,
+      "step": 1425
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.739662100980242e-05,
+      "loss": 2.8364,
+      "step": 1430
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.737865897762066e-05,
+      "loss": 3.0884,
+      "step": 1435
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.736063862134004e-05,
+      "loss": 3.1134,
+      "step": 1440
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.734255998792589e-05,
+      "loss": 3.3668,
+      "step": 1445
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.732442312449541e-05,
+      "loss": 2.9823,
+      "step": 1450
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.730622807831758e-05,
+      "loss": 3.0893,
+      "step": 1455
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.728797489681301e-05,
+      "loss": 3.1091,
+      "step": 1460
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.726966362755383e-05,
+      "loss": 2.9348,
+      "step": 1465
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.725129431826355e-05,
+      "loss": 3.2506,
+      "step": 1470
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.723286701681695e-05,
+      "loss": 2.9876,
+      "step": 1475
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.7214381771239946e-05,
+      "loss": 2.9158,
+      "step": 1480
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.719583862970949e-05,
+      "loss": 3.2035,
+      "step": 1485
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.7177237640553414e-05,
+      "loss": 2.9031,
+      "step": 1490
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.7158578852250304e-05,
+      "loss": 3.2779,
+      "step": 1495
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.7139862313429385e-05,
+      "loss": 3.1219,
+      "step": 1500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.712108807287042e-05,
+      "loss": 3.1643,
+      "step": 1505
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.7102256179503525e-05,
+      "loss": 3.1993,
+      "step": 1510
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.708336668240908e-05,
+      "loss": 3.1773,
+      "step": 1515
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.706441963081762e-05,
+      "loss": 2.9617,
+      "step": 1520
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.7045415074109646e-05,
+      "loss": 3.2621,
+      "step": 1525
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.702635306181554e-05,
+      "loss": 3.2206,
+      "step": 1530
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.700723364361545e-05,
+      "loss": 3.3413,
+      "step": 1535
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.6988056869339104e-05,
+      "loss": 3.3453,
+      "step": 1540
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.696882278896574e-05,
+      "loss": 3.0481,
+      "step": 1545
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.694953145262393e-05,
+      "loss": 3.2183,
+      "step": 1550
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.693018291059147e-05,
+      "loss": 3.4063,
+      "step": 1555
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.691077721329527e-05,
+      "loss": 3.1694,
+      "step": 1560
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.6891314411311184e-05,
+      "loss": 3.0084,
+      "step": 1565
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.6871794555363885e-05,
+      "loss": 3.318,
+      "step": 1570
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.685221769632676e-05,
+      "loss": 3.3126,
+      "step": 1575
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.683258388522175e-05,
+      "loss": 3.1085,
+      "step": 1580
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.681289317321923e-05,
+      "loss": 3.0566,
+      "step": 1585
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.679314561163787e-05,
+      "loss": 3.0642,
+      "step": 1590
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.6773341251944496e-05,
+      "loss": 2.9335,
+      "step": 1595
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.675348014575398e-05,
+      "loss": 3.1551,
+      "step": 1600
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.673356234482908e-05,
+      "loss": 3.0211,
+      "step": 1605
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.671358790108033e-05,
+      "loss": 3.28,
+      "step": 1610
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.6693556866565835e-05,
+      "loss": 3.0621,
+      "step": 1615
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.667346929349126e-05,
+      "loss": 3.1141,
+      "step": 1620
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.6653325234209575e-05,
+      "loss": 3.1941,
+      "step": 1625
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.663312474122099e-05,
+      "loss": 3.1874,
+      "step": 1630
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.661286786717278e-05,
+      "loss": 3.2085,
+      "step": 1635
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.659255466485918e-05,
+      "loss": 3.398,
+      "step": 1640
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.65721851872212e-05,
+      "loss": 3.1911,
+      "step": 1645
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.655175948734654e-05,
+      "loss": 3.2273,
+      "step": 1650
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.653127761846943e-05,
+      "loss": 3.2084,
+      "step": 1655
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.6510739633970485e-05,
+      "loss": 3.2789,
+      "step": 1660
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.649014558737656e-05,
+      "loss": 3.1193,
+      "step": 1665
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.646949553236064e-05,
+      "loss": 3.2083,
+      "step": 1670
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.6448789522741665e-05,
+      "loss": 3.0306,
+      "step": 1675
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.642802761248441e-05,
+      "loss": 3.0084,
+      "step": 1680
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.6407209855699365e-05,
+      "loss": 3.2976,
+      "step": 1685
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.6386336306642514e-05,
+      "loss": 3.3373,
+      "step": 1690
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.63654070197153e-05,
+      "loss": 3.049,
+      "step": 1695
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.634442204946442e-05,
+      "loss": 3.1669,
+      "step": 1700
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.632338145058167e-05,
+      "loss": 3.1078,
+      "step": 1705
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.630228527790386e-05,
+      "loss": 3.1724,
+      "step": 1710
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.628113358641263e-05,
+      "loss": 3.2197,
+      "step": 1715
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.625992643123429e-05,
+      "loss": 3.0047,
+      "step": 1720
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.623866386763974e-05,
+      "loss": 2.9798,
+      "step": 1725
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.621734595104427e-05,
+      "loss": 3.2065,
+      "step": 1730
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.619597273700742e-05,
+      "loss": 3.0981,
+      "step": 1735
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.6174544281232865e-05,
+      "loss": 2.7712,
+      "step": 1740
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.615306063956826e-05,
+      "loss": 3.2469,
+      "step": 1745
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.613152186800506e-05,
+      "loss": 3.3318,
+      "step": 1750
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.610992802267842e-05,
+      "loss": 3.216,
+      "step": 1755
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.608827915986704e-05,
+      "loss": 3.0298,
+      "step": 1760
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.606657533599298e-05,
+      "loss": 3.1146,
+      "step": 1765
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.604481660762158e-05,
+      "loss": 3.1621,
+      "step": 1770
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.6023003031461234e-05,
+      "loss": 2.9654,
+      "step": 1775
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.600113466436331e-05,
+      "loss": 3.3586,
+      "step": 1780
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.597921156332196e-05,
+      "loss": 3.1212,
+      "step": 1785
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.5957233785474e-05,
+      "loss": 2.8321,
+      "step": 1790
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.5935201388098716e-05,
+      "loss": 2.9001,
+      "step": 1795
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.5913114428617785e-05,
+      "loss": 3.248,
+      "step": 1800
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.589097296459508e-05,
+      "loss": 2.8819,
+      "step": 1805
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.5868777053736476e-05,
+      "loss": 3.0377,
+      "step": 1810
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.584652675388982e-05,
+      "loss": 3.3161,
+      "step": 1815
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.582422212304467e-05,
+      "loss": 3.1995,
+      "step": 1820
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.580186321933217e-05,
+      "loss": 3.4196,
+      "step": 1825
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.577945010102496e-05,
+      "loss": 3.35,
+      "step": 1830
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.575698282653693e-05,
+      "loss": 3.2269,
+      "step": 1835
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.573446145442314e-05,
+      "loss": 3.1207,
+      "step": 1840
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.571188604337963e-05,
+      "loss": 3.3682,
+      "step": 1845
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.568925665224327e-05,
+      "loss": 3.0202,
+      "step": 1850
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.5666573339991646e-05,
+      "loss": 3.0227,
+      "step": 1855
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.564383616574285e-05,
+      "loss": 3.1553,
+      "step": 1860
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.562104518875535e-05,
+      "loss": 3.1961,
+      "step": 1865
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.5598200468427854e-05,
+      "loss": 2.9751,
+      "step": 1870
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.557530206429912e-05,
+      "loss": 2.9953,
+      "step": 1875
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.555235003604782e-05,
+      "loss": 2.9119,
+      "step": 1880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.55293444434924e-05,
+      "loss": 3.1541,
+      "step": 1885
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.5506285346590894e-05,
+      "loss": 2.9233,
+      "step": 1890
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.5483172805440785e-05,
+      "loss": 3.3379,
+      "step": 1895
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.546000688027884e-05,
+      "loss": 3.1169,
+      "step": 1900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.543678763148096e-05,
+      "loss": 3.3383,
+      "step": 1905
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.5413515119562023e-05,
+      "loss": 3.1096,
+      "step": 1910
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.539018940517572e-05,
+      "loss": 3.1755,
+      "step": 1915
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.53668105491144e-05,
+      "loss": 3.1069,
+      "step": 1920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.534337861230891e-05,
+      "loss": 3.045,
+      "step": 1925
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.5319893655828446e-05,
+      "loss": 2.975,
+      "step": 1930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.5296355740880385e-05,
+      "loss": 3.4444,
+      "step": 1935
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.5272764928810115e-05,
+      "loss": 3.1052,
+      "step": 1940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.52491212811009e-05,
+      "loss": 3.0536,
+      "step": 1945
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.522542485937369e-05,
+      "loss": 3.0542,
+      "step": 1950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.520167572538699e-05,
+      "loss": 3.158,
+      "step": 1955
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5177873941036683e-05,
+      "loss": 3.0392,
+      "step": 1960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5154019568355876e-05,
+      "loss": 3.1487,
+      "step": 1965
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5130112669514726e-05,
+      "loss": 2.9061,
+      "step": 1970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.510615330682028e-05,
+      "loss": 2.9249,
+      "step": 1975
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5082141542716336e-05,
+      "loss": 3.0888,
+      "step": 1980
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505807743978324e-05,
+      "loss": 2.9417,
+      "step": 1985
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5033961060737774e-05,
+      "loss": 2.8967,
+      "step": 1990
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.5009792468432946e-05,
+      "loss": 2.9344,
+      "step": 1995
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.498557172585782e-05,
+      "loss": 3.1597,
+      "step": 2000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.496129889613744e-05,
+      "loss": 2.9188,
+      "step": 2005
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.493697404253254e-05,
+      "loss": 2.7781,
+      "step": 2010
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.4912597228439466e-05,
+      "loss": 3.0411,
+      "step": 2015
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.488816851738999e-05,
+      "loss": 3.0727,
+      "step": 2020
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.486368797305114e-05,
+      "loss": 3.0783,
+      "step": 2025
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.483915565922501e-05,
+      "loss": 2.823,
+      "step": 2030
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.481457163984864e-05,
+      "loss": 2.6414,
+      "step": 2035
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.478993597899382e-05,
+      "loss": 2.9076,
+      "step": 2040
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.476524874086695e-05,
+      "loss": 2.8118,
+      "step": 2045
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.4740509989808815e-05,
+      "loss": 3.0352,
+      "step": 2050
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.4715719790294474e-05,
+      "loss": 3.0765,
+      "step": 2055
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.4690878206933086e-05,
+      "loss": 2.6218,
+      "step": 2060
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.466598530446771e-05,
+      "loss": 3.0266,
+      "step": 2065
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.464104114777515e-05,
+      "loss": 3.0852,
+      "step": 2070
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.4616045801865814e-05,
+      "loss": 2.9536,
+      "step": 2075
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.4590999331883505e-05,
+      "loss": 2.7272,
+      "step": 2080
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.456590180310527e-05,
+      "loss": 3.0185,
+      "step": 2085
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.454075328094123e-05,
+      "loss": 3.1669,
+      "step": 2090
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.45155538309344e-05,
+      "loss": 2.7681,
+      "step": 2095
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.449030351876053e-05,
+      "loss": 3.0262,
+      "step": 2100
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.446500241022794e-05,
+      "loss": 2.7877,
+      "step": 2105
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.443965057127731e-05,
+      "loss": 2.9973,
+      "step": 2110
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.441424806798156e-05,
+      "loss": 2.9694,
+      "step": 2115
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.438879496654565e-05,
+      "loss": 2.9082,
+      "step": 2120
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.43632913333064e-05,
+      "loss": 3.1398,
+      "step": 2125
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.4337737234732334e-05,
+      "loss": 2.9788,
+      "step": 2130
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.431213273742351e-05,
+      "loss": 3.0885,
+      "step": 2135
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.42864779081113e-05,
+      "loss": 3.1436,
+      "step": 2140
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.4260772813658316e-05,
+      "loss": 2.9983,
+      "step": 2145
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.423501752105812e-05,
+      "loss": 3.0249,
+      "step": 2150
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.420921209743511e-05,
+      "loss": 2.8603,
+      "step": 2155
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.4183356610044364e-05,
+      "loss": 2.8446,
+      "step": 2160
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.415745112627141e-05,
+      "loss": 3.0299,
+      "step": 2165
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.4131495713632104e-05,
+      "loss": 2.8172,
+      "step": 2170
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.4105490439772404e-05,
+      "loss": 2.9729,
+      "step": 2175
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.407943537246822e-05,
+      "loss": 2.9726,
+      "step": 2180
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.4053330579625265e-05,
+      "loss": 2.8896,
+      "step": 2185
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.4027176129278815e-05,
+      "loss": 3.1341,
+      "step": 2190
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.4000972089593566e-05,
+      "loss": 2.8074,
+      "step": 2195
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.397471852886349e-05,
+      "loss": 3.0711,
+      "step": 2200
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.3948415515511586e-05,
+      "loss": 3.0245,
+      "step": 2205
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.3922063118089765e-05,
+      "loss": 3.0007,
+      "step": 2210
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.389566140527861e-05,
+      "loss": 2.9818,
+      "step": 2215
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.386921044588727e-05,
+      "loss": 2.9922,
+      "step": 2220
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.384271030885322e-05,
+      "loss": 2.8606,
+      "step": 2225
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.381616106324211e-05,
+      "loss": 2.7615,
+      "step": 2230
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.3789562778247585e-05,
+      "loss": 2.7838,
+      "step": 2235
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.3762915523191084e-05,
+      "loss": 2.9275,
+      "step": 2240
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.373621936752168e-05,
+      "loss": 3.0331,
+      "step": 2245
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.37094743808159e-05,
+      "loss": 2.9586,
+      "step": 2250
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.3682680632777515e-05,
+      "loss": 2.7475,
+      "step": 2255
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.36558381932374e-05,
+      "loss": 3.0034,
+      "step": 2260
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.362894713215334e-05,
+      "loss": 2.9839,
+      "step": 2265
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.360200751960982e-05,
+      "loss": 2.8333,
+      "step": 2270
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.3575019425817856e-05,
+      "loss": 2.853,
+      "step": 2275
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.354798292111483e-05,
+      "loss": 3.266,
+      "step": 2280
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.35208980759643e-05,
+      "loss": 3.1553,
+      "step": 2285
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.34937649609558e-05,
+      "loss": 3.1441,
+      "step": 2290
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.3466583646804674e-05,
+      "loss": 2.8948,
+      "step": 2295
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.3439354204351866e-05,
+      "loss": 3.0963,
+      "step": 2300
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.341207670456379e-05,
+      "loss": 3.0771,
+      "step": 2305
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.3384751218532086e-05,
+      "loss": 3.0364,
+      "step": 2310
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.335737781747343e-05,
+      "loss": 2.9928,
+      "step": 2315
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.332995657272944e-05,
+      "loss": 3.1993,
+      "step": 2320
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.330248755576636e-05,
+      "loss": 2.9873,
+      "step": 2325
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.3274970838175e-05,
+      "loss": 2.9258,
+      "step": 2330
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.324740649167044e-05,
+      "loss": 2.7857,
+      "step": 2335
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.321979458809191e-05,
+      "loss": 3.1293,
+      "step": 2340
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.31921351994026e-05,
+      "loss": 3.0588,
+      "step": 2345
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.3164428397689436e-05,
+      "loss": 3.116,
+      "step": 2350
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.313667425516292e-05,
+      "loss": 3.0814,
+      "step": 2355
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.310887284415695e-05,
+      "loss": 2.9781,
+      "step": 2360
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.308102423712859e-05,
+      "loss": 2.8611,
+      "step": 2365
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.3053128506657934e-05,
+      "loss": 3.1201,
+      "step": 2370
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.302518572544787e-05,
+      "loss": 3.108,
+      "step": 2375
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.299719596632392e-05,
+      "loss": 2.8409,
+      "step": 2380
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.2969159302234044e-05,
+      "loss": 2.8566,
+      "step": 2385
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.294107580624845e-05,
+      "loss": 2.7636,
+      "step": 2390
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.291294555155939e-05,
+      "loss": 3.1108,
+      "step": 2395
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.2884768611481e-05,
+      "loss": 3.0492,
+      "step": 2400
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.2856545059449063e-05,
+      "loss": 3.1709,
+      "step": 2405
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.282827496902086e-05,
+      "loss": 3.0867,
+      "step": 2410
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.279995841387496e-05,
+      "loss": 2.8054,
+      "step": 2415
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.277159546781103e-05,
+      "loss": 2.8368,
+      "step": 2420
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.274318620474964e-05,
+      "loss": 3.0,
+      "step": 2425
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.2714730698732076e-05,
+      "loss": 2.9218,
+      "step": 2430
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.268622902392014e-05,
+      "loss": 2.9109,
+      "step": 2435
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.265768125459597e-05,
+      "loss": 3.074,
+      "step": 2440
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.263480990108211e-05,
+      "loss": 2.8238,
+      "step": 2445
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.260617934921023e-05,
+      "loss": 2.9209,
+      "step": 2450
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.257750291145457e-05,
+      "loss": 2.8323,
+      "step": 2455
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.254878066255274e-05,
+      "loss": 3.2091,
+      "step": 2460
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.252001267736174e-05,
+      "loss": 2.8372,
+      "step": 2465
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.2491199030857764e-05,
+      "loss": 3.0106,
+      "step": 2470
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.246233979813602e-05,
+      "loss": 2.9969,
+      "step": 2475
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.243343505441052e-05,
+      "loss": 3.1595,
+      "step": 2480
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.2404484875013895e-05,
+      "loss": 2.8051,
+      "step": 2485
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.237548933539718e-05,
+      "loss": 2.8842,
+      "step": 2490
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.234644851112964e-05,
+      "loss": 2.9729,
+      "step": 2495
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.231736247789856e-05,
+      "loss": 2.7805,
+      "step": 2500
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.228823131150904e-05,
+      "loss": 2.9725,
+      "step": 2505
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.225905508788382e-05,
+      "loss": 3.0599,
+      "step": 2510
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.222983388306308e-05,
+      "loss": 2.7567,
+      "step": 2515
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.2200567773204194e-05,
+      "loss": 2.884,
+      "step": 2520
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.217125683458161e-05,
+      "loss": 2.8594,
+      "step": 2525
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.214190114358659e-05,
+      "loss": 2.9515,
+      "step": 2530
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.211250077672704e-05,
+      "loss": 2.992,
+      "step": 2535
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.208305581062728e-05,
+      "loss": 3.1784,
+      "step": 2540
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.205356632202789e-05,
+      "loss": 3.0196,
+      "step": 2545
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.20240323877855e-05,
+      "loss": 2.7891,
+      "step": 2550
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.1994454084872525e-05,
+      "loss": 2.8974,
+      "step": 2555
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.196483149037707e-05,
+      "loss": 2.8219,
+      "step": 2560
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.1935164681502624e-05,
+      "loss": 2.8833,
+      "step": 2565
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.190545373556796e-05,
+      "loss": 3.0166,
+      "step": 2570
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.187569873000684e-05,
+      "loss": 3.1676,
+      "step": 2575
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.184589974236788e-05,
+      "loss": 2.9239,
+      "step": 2580
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.181605685031432e-05,
+      "loss": 3.0094,
+      "step": 2585
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.1786170131623825e-05,
+      "loss": 3.0235,
+      "step": 2590
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.1756239664188273e-05,
+      "loss": 3.058,
+      "step": 2595
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.172626552601358e-05,
+      "loss": 2.7391,
+      "step": 2600
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.169624779521944e-05,
+      "loss": 2.9975,
+      "step": 2605
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.1666186550039214e-05,
+      "loss": 2.7703,
+      "step": 2610
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.163608186881964e-05,
+      "loss": 2.8288,
+      "step": 2615
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.160593383002066e-05,
+      "loss": 2.9945,
+      "step": 2620
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.157574251221521e-05,
+      "loss": 2.8768,
+      "step": 2625
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.1545507994089064e-05,
+      "loss": 2.7161,
+      "step": 2630
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.1515230354440524e-05,
+      "loss": 3.0821,
+      "step": 2635
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.148490967218033e-05,
+      "loss": 3.1119,
+      "step": 2640
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.145454602633137e-05,
+      "loss": 2.8815,
+      "step": 2645
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.14241394960285e-05,
+      "loss": 2.9891,
+      "step": 2650
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.1393690160518384e-05,
+      "loss": 3.2624,
+      "step": 2655
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.1363198099159205e-05,
+      "loss": 2.7588,
+      "step": 2660
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.1332663391420515e-05,
+      "loss": 3.0405,
+      "step": 2665
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.130208611688302e-05,
+      "loss": 2.9254,
+      "step": 2670
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.1271466355238354e-05,
+      "loss": 2.8499,
+      "step": 2675
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.124080418628888e-05,
+      "loss": 3.0828,
+      "step": 2680
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.1210099689947513e-05,
+      "loss": 2.8524,
+      "step": 2685
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.117935294623743e-05,
+      "loss": 3.0711,
+      "step": 2690
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.1148564035291956e-05,
+      "loss": 3.0343,
+      "step": 2695
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.1117733037354314e-05,
+      "loss": 2.8869,
+      "step": 2700
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.1086860032777395e-05,
+      "loss": 3.0381,
+      "step": 2705
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.105594510202359e-05,
+      "loss": 2.9049,
+      "step": 2710
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.102498832566454e-05,
+      "loss": 3.0071,
+      "step": 2715
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.0993989784380947e-05,
+      "loss": 2.9955,
+      "step": 2720
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.096294955896239e-05,
+      "loss": 2.9568,
+      "step": 2725
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.093186773030705e-05,
+      "loss": 2.9727,
+      "step": 2730
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.090074437942155e-05,
+      "loss": 3.0501,
+      "step": 2735
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.086957958742075e-05,
+      "loss": 3.2328,
+      "step": 2740
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.0838373435527494e-05,
+      "loss": 2.7966,
+      "step": 2745
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.08071260050724e-05,
+      "loss": 2.8666,
+      "step": 2750
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.077583737749373e-05,
+      "loss": 2.9765,
+      "step": 2755
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.074450763433704e-05,
+      "loss": 2.6923,
+      "step": 2760
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.07131368572551e-05,
+      "loss": 2.9031,
+      "step": 2765
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.06817251280076e-05,
+      "loss": 2.8287,
+      "step": 2770
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.065027252846095e-05,
+      "loss": 3.1046,
+      "step": 2775
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.0618779140588124e-05,
+      "loss": 3.0414,
+      "step": 2780
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.058724504646834e-05,
+      "loss": 2.8644,
+      "step": 2785
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.055567032828695e-05,
+      "loss": 2.9589,
+      "step": 2790
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.052405506833516e-05,
+      "loss": 2.886,
+      "step": 2795
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.0492399349009846e-05,
+      "loss": 2.8026,
+      "step": 2800
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.0460703252813326e-05,
+      "loss": 3.0376,
+      "step": 2805
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.042896686235316e-05,
+      "loss": 2.9751,
+      "step": 2810
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.039719026034191e-05,
+      "loss": 2.8638,
+      "step": 2815
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.036537352959694e-05,
+      "loss": 3.0403,
+      "step": 2820
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.033351675304022e-05,
+      "loss": 2.8241,
+      "step": 2825
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.030162001369807e-05,
+      "loss": 3.1196,
+      "step": 2830
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.026968339470096e-05,
+      "loss": 2.8388,
+      "step": 2835
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.0237706979283304e-05,
+      "loss": 2.9974,
+      "step": 2840
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.020569085078324e-05,
+      "loss": 3.0178,
+      "step": 2845
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.0173635092642386e-05,
+      "loss": 3.0166,
+      "step": 2850
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.014153978840567e-05,
+      "loss": 3.0657,
+      "step": 2855
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.010940502172108e-05,
+      "loss": 3.0229,
+      "step": 2860
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.007723087633943e-05,
+      "loss": 2.5904,
+      "step": 2865
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.0045017436114205e-05,
+      "loss": 2.8607,
+      "step": 2870
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.001276478500127e-05,
+      "loss": 3.0797,
+      "step": 2875
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.998047300705868e-05,
+      "loss": 2.9475,
+      "step": 2880
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.9948142186446486e-05,
+      "loss": 2.7829,
+      "step": 2885
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.991577240742649e-05,
+      "loss": 3.0715,
+      "step": 2890
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.9883363754362004e-05,
+      "loss": 3.0126,
+      "step": 2895
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.985091631171769e-05,
+      "loss": 3.0835,
+      "step": 2900
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.981843016405927e-05,
+      "loss": 2.9109,
+      "step": 2905
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.978590539605338e-05,
+      "loss": 3.0058,
+      "step": 2910
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.975334209246727e-05,
+      "loss": 2.7376,
+      "step": 2915
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.972074033816866e-05,
+      "loss": 2.7173,
+      "step": 2920
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.968810021812544e-05,
+      "loss": 2.8642,
+      "step": 2925
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.965542181740553e-05,
+      "loss": 2.8643,
+      "step": 2930
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.9622705221176596e-05,
+      "loss": 3.1752,
+      "step": 2935
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.958995051470585e-05,
+      "loss": 2.7045,
+      "step": 2940
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.955715778335984e-05,
+      "loss": 3.0828,
+      "step": 2945
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.952432711260421e-05,
+      "loss": 2.982,
+      "step": 2950
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.949145858800348e-05,
+      "loss": 2.9762,
+      "step": 2955
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.945855229522084e-05,
+      "loss": 3.1026,
+      "step": 2960
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.942560832001789e-05,
+      "loss": 3.1225,
+      "step": 2965
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.9392626748254456e-05,
+      "loss": 3.0462,
+      "step": 2970
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.9359607665888346e-05,
+      "loss": 2.8954,
+      "step": 2975
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.932655115897513e-05,
+      "loss": 2.8361,
+      "step": 2980
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.92934573136679e-05,
+      "loss": 3.1644,
+      "step": 2985
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.9260326216217105e-05,
+      "loss": 3.0226,
+      "step": 2990
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.922715795297022e-05,
+      "loss": 3.0285,
+      "step": 2995
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.919395261037163e-05,
+      "loss": 3.0582,
+      "step": 3000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.9160710274962345e-05,
+      "loss": 2.9447,
+      "step": 3005
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.912743103337977e-05,
+      "loss": 3.0433,
+      "step": 3010
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.9094114972357524e-05,
+      "loss": 3.0276,
+      "step": 3015
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.9060762178725165e-05,
+      "loss": 3.1617,
+      "step": 3020
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.9027372739407994e-05,
+      "loss": 2.962,
+      "step": 3025
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 3.899394674142682e-05,
+      "loss": 2.9722,
+      "step": 3030
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 3.8960484271897736e-05,
+      "loss": 2.9605,
+      "step": 3035
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 3.892698541803187e-05,
+      "loss": 3.1036,
+      "step": 3040
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 3.889345026713519e-05,
+      "loss": 3.1859,
+      "step": 3045
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.8859878906608275e-05,
+      "loss": 3.0349,
+      "step": 3050
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.882627142394605e-05,
+      "loss": 2.9102,
+      "step": 3055
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.87926279067376e-05,
+      "loss": 2.9039,
+      "step": 3060
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.8758948442665914e-05,
+      "loss": 3.0984,
+      "step": 3065
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.872523311950768e-05,
+      "loss": 2.9782,
+      "step": 3070
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.869148202513303e-05,
+      "loss": 2.7941,
+      "step": 3075
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.8657695247505325e-05,
+      "loss": 3.1176,
+      "step": 3080
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.862387287468094e-05,
+      "loss": 2.8274,
+      "step": 3085
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.859001499480901e-05,
+      "loss": 3.0035,
+      "step": 3090
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.8556121696131194e-05,
+      "loss": 3.0127,
+      "step": 3095
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.852219306698149e-05,
+      "loss": 3.0512,
+      "step": 3100
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.848822919578596e-05,
+      "loss": 2.8802,
+      "step": 3105
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.845423017106251e-05,
+      "loss": 2.9527,
+      "step": 3110
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.842019608142069e-05,
+      "loss": 2.9869,
+      "step": 3115
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.838612701556138e-05,
+      "loss": 2.7298,
+      "step": 3120
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.835202306227669e-05,
+      "loss": 3.0647,
+      "step": 3125
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.8317884310449614e-05,
+      "loss": 2.9665,
+      "step": 3130
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.8283710849053834e-05,
+      "loss": 2.9411,
+      "step": 3135
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.824950276715351e-05,
+      "loss": 2.8853,
+      "step": 3140
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.8215260153903024e-05,
+      "loss": 2.7117,
+      "step": 3145
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.818098309854676e-05,
+      "loss": 3.1676,
+      "step": 3150
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.8146671690418865e-05,
+      "loss": 2.9812,
+      "step": 3155
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.811232601894301e-05,
+      "loss": 2.8832,
+      "step": 3160
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.8077946173632175e-05,
+      "loss": 3.0479,
+      "step": 3165
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.80435322440884e-05,
+      "loss": 2.9631,
+      "step": 3170
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.800908432000254e-05,
+      "loss": 2.9479,
+      "step": 3175
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.7974602491154094e-05,
+      "loss": 2.752,
+      "step": 3180
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.7940086847410874e-05,
+      "loss": 3.0117,
+      "step": 3185
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.7905537478728845e-05,
+      "loss": 3.0633,
+      "step": 3190
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.787095447515187e-05,
+      "loss": 3.1068,
+      "step": 3195
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.783633792681147e-05,
+      "loss": 2.9438,
+      "step": 3200
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.7801687923926577e-05,
+      "loss": 2.9187,
+      "step": 3205
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.7767004556803346e-05,
+      "loss": 3.0672,
+      "step": 3210
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.773228791583484e-05,
+      "loss": 2.9911,
+      "step": 3215
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.76975380915009e-05,
+      "loss": 3.0206,
+      "step": 3220
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.7662755174367795e-05,
+      "loss": 3.1758,
+      "step": 3225
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.7627939255088064e-05,
+      "loss": 2.5882,
+      "step": 3230
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.759309042440028e-05,
+      "loss": 3.1835,
+      "step": 3235
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.755820877312876e-05,
+      "loss": 3.0573,
+      "step": 3240
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.752329439218337e-05,
+      "loss": 2.9557,
+      "step": 3245
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.748834737255927e-05,
+      "loss": 2.7683,
+      "step": 3250
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.74533678053367e-05,
+      "loss": 2.9666,
+      "step": 3255
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.741835578168071e-05,
+      "loss": 3.1286,
+      "step": 3260
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.738331139284096e-05,
+      "loss": 2.8983,
+      "step": 3265
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.734823473015143e-05,
+      "loss": 2.9258,
+      "step": 3270
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.7313125885030244e-05,
+      "loss": 3.2793,
+      "step": 3275
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.727798494897938e-05,
+      "loss": 2.7685,
+      "step": 3280
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.724281201358447e-05,
+      "loss": 2.7007,
+      "step": 3285
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.7207607170514534e-05,
+      "loss": 3.0827,
+      "step": 3290
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.717237051152175e-05,
+      "loss": 3.1145,
+      "step": 3295
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.7137102128441216e-05,
+      "loss": 2.9999,
+      "step": 3300
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.710180211319071e-05,
+      "loss": 2.9787,
+      "step": 3305
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.7066470557770463e-05,
+      "loss": 2.9137,
+      "step": 3310
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.703110755426289e-05,
+      "loss": 3.0994,
+      "step": 3315
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6995713194832375e-05,
+      "loss": 2.8384,
+      "step": 3320
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.696028757172503e-05,
+      "loss": 2.9642,
+      "step": 3325
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6924830777268425e-05,
+      "loss": 2.7456,
+      "step": 3330
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.68893429038714e-05,
+      "loss": 2.8045,
+      "step": 3335
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.685382404402378e-05,
+      "loss": 3.2544,
+      "step": 3340
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.681827429029613e-05,
+      "loss": 2.9062,
+      "step": 3345
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.678269373533956e-05,
+      "loss": 3.0275,
+      "step": 3350
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.674708247188545e-05,
+      "loss": 2.9737,
+      "step": 3355
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.6711440592745205e-05,
+      "loss": 3.2186,
+      "step": 3360
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.6675768190810026e-05,
+      "loss": 3.2481,
+      "step": 3365
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.6640065359050645e-05,
+      "loss": 2.9293,
+      "step": 3370
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.660433219051714e-05,
+      "loss": 2.9425,
+      "step": 3375
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.656856877833863e-05,
+      "loss": 2.8838,
+      "step": 3380
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.6532775215723044e-05,
+      "loss": 2.9395,
+      "step": 3385
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.6496951595956924e-05,
+      "loss": 2.9398,
+      "step": 3390
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.646109801240511e-05,
+      "loss": 3.0734,
+      "step": 3395
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.642521455851058e-05,
+      "loss": 3.0898,
+      "step": 3400
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.638930132779411e-05,
+      "loss": 2.8408,
+      "step": 3405
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.635335841385411e-05,
+      "loss": 2.8395,
+      "step": 3410
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.6317385910366364e-05,
+      "loss": 3.0667,
+      "step": 3415
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.6281383911083744e-05,
+      "loss": 2.7302,
+      "step": 3420
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.624535250983601e-05,
+      "loss": 2.8523,
+      "step": 3425
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.620929180052955e-05,
+      "loss": 2.9712,
+      "step": 3430
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.6173201877147134e-05,
+      "loss": 3.0531,
+      "step": 3435
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.613708283374767e-05,
+      "loss": 2.9067,
+      "step": 3440
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.610093476446597e-05,
+      "loss": 2.7005,
+      "step": 3445
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.606475776351248e-05,
+      "loss": 2.8974,
+      "step": 3450
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.6028551925173046e-05,
+      "loss": 2.6924,
+      "step": 3455
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.599231734380869e-05,
+      "loss": 2.9246,
+      "step": 3460
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.595605411385533e-05,
+      "loss": 2.9283,
+      "step": 3465
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.5919762329823554e-05,
+      "loss": 3.091,
+      "step": 3470
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.588344208629838e-05,
+      "loss": 2.8846,
+      "step": 3475
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.5847093477938956e-05,
+      "loss": 2.9335,
+      "step": 3480
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.581071659947842e-05,
+      "loss": 3.0189,
+      "step": 3485
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.5774311545723525e-05,
+      "loss": 2.7938,
+      "step": 3490
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.5737878411554514e-05,
+      "loss": 3.0325,
+      "step": 3495
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.5701417291924765e-05,
+      "loss": 3.0631,
+      "step": 3500
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.566492828186063e-05,
+      "loss": 2.6561,
+      "step": 3505
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.562841147646111e-05,
+      "loss": 3.0796,
+      "step": 3510
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.559186697089769e-05,
+      "loss": 2.6858,
+      "step": 3515
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.555529486041401e-05,
+      "loss": 3.1426,
+      "step": 3520
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.55186952403257e-05,
+      "loss": 3.1105,
+      "step": 3525
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.548206820602004e-05,
+      "loss": 3.189,
+      "step": 3530
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.544541385295578e-05,
+      "loss": 2.7971,
+      "step": 3535
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.540873227666288e-05,
+      "loss": 2.766,
+      "step": 3540
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.537202357274223e-05,
+      "loss": 2.9254,
+      "step": 3545
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.533528783686541e-05,
+      "loss": 2.6065,
+      "step": 3550
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.52985251647745e-05,
+      "loss": 2.9204,
+      "step": 3555
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.5261735652281727e-05,
+      "loss": 2.8152,
+      "step": 3560
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.5224919395269315e-05,
+      "loss": 2.9444,
+      "step": 3565
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.5188076489689165e-05,
+      "loss": 3.0586,
+      "step": 3570
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.515120703156264e-05,
+      "loss": 2.8679,
+      "step": 3575
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.5114311116980285e-05,
+      "loss": 2.8359,
+      "step": 3580
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.507738884210164e-05,
+      "loss": 2.978,
+      "step": 3585
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.5040440303154895e-05,
+      "loss": 3.0614,
+      "step": 3590
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.500346559643675e-05,
+      "loss": 2.9246,
+      "step": 3595
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.4966464818312025e-05,
+      "loss": 2.8654,
+      "step": 3600
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.492943806521356e-05,
+      "loss": 2.9768,
+      "step": 3605
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.489238543364187e-05,
+      "loss": 2.9184,
+      "step": 3610
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.4855307020164904e-05,
+      "loss": 3.0659,
+      "step": 3615
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.4818202921417805e-05,
+      "loss": 2.8185,
+      "step": 3620
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.4781073234102665e-05,
+      "loss": 3.0411,
+      "step": 3625
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.474391805498827e-05,
+      "loss": 3.0205,
+      "step": 3630
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.470673748090984e-05,
+      "loss": 3.0916,
+      "step": 3635
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.4669531608768775e-05,
+      "loss": 2.9119,
+      "step": 3640
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.463230053553241e-05,
+      "loss": 2.9236,
+      "step": 3645
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.459504435823378e-05,
+      "loss": 2.7746,
+      "step": 3650
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.455776317397129e-05,
+      "loss": 3.0066,
+      "step": 3655
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.4520457079908605e-05,
+      "loss": 2.667,
+      "step": 3660
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.448312617327423e-05,
+      "loss": 2.8988,
+      "step": 3665
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.444577055136139e-05,
+      "loss": 2.8698,
+      "step": 3670
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.4408390311527686e-05,
+      "loss": 2.8258,
+      "step": 3675
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.4370985551194926e-05,
+      "loss": 2.9826,
+      "step": 3680
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.4333556367848776e-05,
+      "loss": 2.7939,
+      "step": 3685
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.429610285903858e-05,
+      "loss": 2.959,
+      "step": 3690
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.425862512237708e-05,
+      "loss": 2.8027,
+      "step": 3695
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.4221123255540157e-05,
+      "loss": 2.8562,
+      "step": 3700
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.418359735626656e-05,
+      "loss": 2.8358,
+      "step": 3705
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.414604752235772e-05,
+      "loss": 2.8507,
+      "step": 3710
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.4108473851677405e-05,
+      "loss": 2.954,
+      "step": 3715
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.407087644215153e-05,
+      "loss": 3.0872,
+      "step": 3720
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.4033255391767865e-05,
+      "loss": 2.9302,
+      "step": 3725
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.39956107985758e-05,
+      "loss": 2.8434,
+      "step": 3730
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.395794276068609e-05,
+      "loss": 2.984,
+      "step": 3735
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.3920251376270575e-05,
+      "loss": 2.8695,
+      "step": 3740
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.388253674356196e-05,
+      "loss": 2.9782,
+      "step": 3745
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.384479896085353e-05,
+      "loss": 3.0113,
+      "step": 3750
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.380703812649891e-05,
+      "loss": 3.076,
+      "step": 3755
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.376925433891179e-05,
+      "loss": 2.7471,
+      "step": 3760
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.373144769656571e-05,
+      "loss": 2.89,
+      "step": 3765
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.369361829799375e-05,
+      "loss": 2.9204,
+      "step": 3770
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.365576624178828e-05,
+      "loss": 2.8335,
+      "step": 3775
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.3617891626600786e-05,
+      "loss": 3.0041,
+      "step": 3780
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.357999455114148e-05,
+      "loss": 2.9935,
+      "step": 3785
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.3542075114179164e-05,
+      "loss": 2.9154,
+      "step": 3790
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.350413341454086e-05,
+      "loss": 2.8998,
+      "step": 3795
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.346616955111168e-05,
+      "loss": 2.6922,
+      "step": 3800
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.342818362283444e-05,
+      "loss": 2.7723,
+      "step": 3805
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.33901757287095e-05,
+      "loss": 2.8529,
+      "step": 3810
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.335214596779446e-05,
+      "loss": 3.0667,
+      "step": 3815
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.33140944392039e-05,
+      "loss": 3.0768,
+      "step": 3820
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.327602124210915e-05,
+      "loss": 2.8217,
+      "step": 3825
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.3237926475737986e-05,
+      "loss": 3.0252,
+      "step": 3830
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.319981023937442e-05,
+      "loss": 3.0058,
+      "step": 3835
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.316167263235842e-05,
+      "loss": 2.8522,
+      "step": 3840
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.3123513754085634e-05,
+      "loss": 3.023,
+      "step": 3845
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.308533370400717e-05,
+      "loss": 3.0671,
+      "step": 3850
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.30471325816293e-05,
+      "loss": 2.7645,
+      "step": 3855
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.300891048651321e-05,
+      "loss": 2.9496,
+      "step": 3860
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.297066751827478e-05,
+      "loss": 3.0305,
+      "step": 3865
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.293240377658424e-05,
+      "loss": 2.9011,
+      "step": 3870
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.2894119361166e-05,
+      "loss": 2.7675,
+      "step": 3875
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.2855814371798325e-05,
+      "loss": 2.9937,
+      "step": 3880
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2817488908313134e-05,
+      "loss": 3.1056,
+      "step": 3885
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.277914307059566e-05,
+      "loss": 2.6727,
+      "step": 3890
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.274077695858428e-05,
+      "loss": 2.7927,
+      "step": 3895
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.2702390672270175e-05,
+      "loss": 2.7103,
+      "step": 3900
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.266398431169714e-05,
+      "loss": 2.7382,
+      "step": 3905
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.262555797696125e-05,
+      "loss": 2.5843,
+      "step": 3910
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.258711176821068e-05,
+      "loss": 2.7592,
+      "step": 3915
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.254864578564535e-05,
+      "loss": 2.8419,
+      "step": 3920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.2510160129516775e-05,
+      "loss": 2.6239,
+      "step": 3925
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.247165490012769e-05,
+      "loss": 2.6813,
+      "step": 3930
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.243313019783187e-05,
+      "loss": 2.8535,
+      "step": 3935
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.2394586123033844e-05,
+      "loss": 2.7436,
+      "step": 3940
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.235602277618862e-05,
+      "loss": 2.8294,
+      "step": 3945
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.231744025780143e-05,
+      "loss": 2.8097,
+      "step": 3950
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.227883866842748e-05,
+      "loss": 2.535,
+      "step": 3955
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.2240218108671684e-05,
+      "loss": 2.5515,
+      "step": 3960
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.2201578679188395e-05,
+      "loss": 2.6408,
+      "step": 3965
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.2162920480681126e-05,
+      "loss": 2.7759,
+      "step": 3970
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.212424361390232e-05,
+      "loss": 2.6673,
+      "step": 3975
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.2085548179653094e-05,
+      "loss": 2.6833,
+      "step": 3980
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.204683427878291e-05,
+      "loss": 2.9401,
+      "step": 3985
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.200810201218942e-05,
+      "loss": 2.5877,
+      "step": 3990
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.1969351480818075e-05,
+      "loss": 2.8027,
+      "step": 3995
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.193058278566199e-05,
+      "loss": 2.7086,
+      "step": 4000
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.189179602776157e-05,
+      "loss": 2.7855,
+      "step": 4005
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.185299130820432e-05,
+      "loss": 2.7144,
+      "step": 4010
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.181416872812455e-05,
+      "loss": 2.8703,
+      "step": 4015
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 3.177532838870313e-05,
+      "loss": 2.8389,
+      "step": 4020
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 3.1736470391167196e-05,
+      "loss": 2.733,
+      "step": 4025
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 3.169759483678992e-05,
+      "loss": 2.7172,
+      "step": 4030
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 3.165870182689024e-05,
+      "loss": 2.7626,
+      "step": 4035
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 3.161979146283255e-05,
+      "loss": 2.4813,
+      "step": 4040
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 3.158086384602652e-05,
+      "loss": 2.7431,
+      "step": 4045
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 3.1541919077926756e-05,
+      "loss": 2.6488,
+      "step": 4050
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 3.1502957260032564e-05,
+      "loss": 2.9131,
+      "step": 4055
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 3.1463978493887695e-05,
+      "loss": 2.834,
+      "step": 4060
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 3.142498288108007e-05,
+      "loss": 2.9725,
+      "step": 4065
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 3.1385970523241524e-05,
+      "loss": 2.7835,
+      "step": 4070
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 3.13469415220475e-05,
+      "loss": 2.5171,
+      "step": 4075
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 3.1307895979216876e-05,
+      "loss": 2.4543,
+      "step": 4080
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 3.126883399651158e-05,
+      "loss": 2.8861,
+      "step": 4085
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 3.1229755675736425e-05,
+      "loss": 2.8512,
+      "step": 4090
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 3.119066111873879e-05,
+      "loss": 2.6645,
+      "step": 4095
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 3.1151550427408385e-05,
+      "loss": 2.9194,
+      "step": 4100
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 3.111242370367694e-05,
+      "loss": 3.0082,
+      "step": 4105
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 3.1073281049517994e-05,
+      "loss": 2.8563,
+      "step": 4110
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 3.10341225669466e-05,
+      "loss": 2.8531,
+      "step": 4115
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 3.099494835801906e-05,
+      "loss": 2.7171,
+      "step": 4120
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 3.0955758524832664e-05,
+      "loss": 2.7625,
+      "step": 4125
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 3.0916553169525424e-05,
+      "loss": 2.9918,
+      "step": 4130
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 3.08773323942758e-05,
+      "loss": 2.9665,
+      "step": 4135
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 3.083809630130246e-05,
+      "loss": 2.7593,
+      "step": 4140
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 3.0798844992863954e-05,
+      "loss": 2.6305,
+      "step": 4145
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 3.075957857125854e-05,
+      "loss": 2.7587,
+      "step": 4150
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.072029713882381e-05,
+      "loss": 2.7514,
+      "step": 4155
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.068100079793652e-05,
+      "loss": 2.831,
+      "step": 4160
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.064168965101225e-05,
+      "loss": 2.7585,
+      "step": 4165
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.060236380050519e-05,
+      "loss": 2.7944,
+      "step": 4170
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.056302334890786e-05,
+      "loss": 2.8257,
+      "step": 4175
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.05236683987508e-05,
+      "loss": 2.7217,
+      "step": 4180
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.0484299052602355e-05,
+      "loss": 2.8344,
+      "step": 4185
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.04449154130684e-05,
+      "loss": 2.7256,
+      "step": 4190
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.0405517582792036e-05,
+      "loss": 2.5206,
+      "step": 4195
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.036610566445337e-05,
+      "loss": 2.6443,
+      "step": 4200
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.0326679760769226e-05,
+      "loss": 2.7363,
+      "step": 4205
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.028723997449286e-05,
+      "loss": 2.5985,
+      "step": 4210
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.0247786408413724e-05,
+      "loss": 2.9206,
+      "step": 4215
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.0208319165357174e-05,
+      "loss": 2.8855,
+      "step": 4220
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.0168838348184225e-05,
+      "loss": 2.7868,
+      "step": 4225
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.0129344059791253e-05,
+      "loss": 2.7129,
+      "step": 4230
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.0089836403109762e-05,
+      "loss": 2.4369,
+      "step": 4235
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.005031548110607e-05,
+      "loss": 2.629,
+      "step": 4240
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.00107813967811e-05,
+      "loss": 2.6645,
+      "step": 4245
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.9971234253170055e-05,
+      "loss": 2.8677,
+      "step": 4250
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.993167415334218e-05,
+      "loss": 2.7891,
+      "step": 4255
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.9892101200400485e-05,
+      "loss": 2.6506,
+      "step": 4260
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.985251549748149e-05,
+      "loss": 2.7657,
+      "step": 4265
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.9812917147754937e-05,
+      "loss": 2.5059,
+      "step": 4270
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.9773306254423515e-05,
+      "loss": 2.6195,
+      "step": 4275
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.973368292072263e-05,
+      "loss": 2.8329,
+      "step": 4280
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.9694047249920092e-05,
+      "loss": 2.8922,
+      "step": 4285
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.9654399345315864e-05,
+      "loss": 2.9282,
+      "step": 4290
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.961473931024182e-05,
+      "loss": 2.7919,
+      "step": 4295
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.95750672480614e-05,
+      "loss": 2.8352,
+      "step": 4300
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.9535383262169435e-05,
+      "loss": 2.5651,
+      "step": 4305
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.9495687455991816e-05,
+      "loss": 2.8988,
+      "step": 4310
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.9455979932985232e-05,
+      "loss": 2.7066,
+      "step": 4315
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.9416260796636925e-05,
+      "loss": 2.7755,
+      "step": 4320
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.9376530150464386e-05,
+      "loss": 2.6079,
+      "step": 4325
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.9336788098015116e-05,
+      "loss": 2.609,
+      "step": 4330
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.9297034742866337e-05,
+      "loss": 2.7253,
+      "step": 4335
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.925727018862474e-05,
+      "loss": 2.9559,
+      "step": 4340
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.9217494538926183e-05,
+      "loss": 2.9942,
+      "step": 4345
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.9177707897435463e-05,
+      "loss": 2.9635,
+      "step": 4350
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.913791036784601e-05,
+      "loss": 2.7588,
+      "step": 4355
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.9098102053879644e-05,
+      "loss": 2.9231,
+      "step": 4360
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.905828305928627e-05,
+      "loss": 2.9618,
+      "step": 4365
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.901845348784365e-05,
+      "loss": 2.6759,
+      "step": 4370
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.8978613443357105e-05,
+      "loss": 2.8548,
+      "step": 4375
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.893876302965925e-05,
+      "loss": 2.8764,
+      "step": 4380
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.889890235060972e-05,
+      "loss": 2.5567,
+      "step": 4385
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.8859031510094914e-05,
+      "loss": 2.7917,
+      "step": 4390
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.88191506120277e-05,
+      "loss": 2.6198,
+      "step": 4395
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.8779259760347187e-05,
+      "loss": 2.7678,
+      "step": 4400
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.873935905901839e-05,
+      "loss": 2.5505,
+      "step": 4405
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.869944861203202e-05,
+      "loss": 2.7345,
+      "step": 4410
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.8659528523404168e-05,
+      "loss": 2.8262,
+      "step": 4415
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.861959889717607e-05,
+      "loss": 2.7795,
+      "step": 4420
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.8579659837413812e-05,
+      "loss": 2.591,
+      "step": 4425
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8539711448208074e-05,
+      "loss": 2.6929,
+      "step": 4430
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8499753833673824e-05,
+      "loss": 3.0178,
+      "step": 4435
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.845978709795012e-05,
+      "loss": 3.0475,
+      "step": 4440
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8419811345199738e-05,
+      "loss": 2.6604,
+      "step": 4445
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8379826679609e-05,
+      "loss": 2.6987,
+      "step": 4450
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8339833205387434e-05,
+      "loss": 2.7681,
+      "step": 4455
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8299831026767526e-05,
+      "loss": 2.7053,
+      "step": 4460
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.825982024800445e-05,
+      "loss": 2.8991,
+      "step": 4465
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.8219800973375803e-05,
+      "loss": 2.7856,
+      "step": 4470
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.8179773307181306e-05,
+      "loss": 2.6422,
+      "step": 4475
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.813973735374258e-05,
+      "loss": 2.6979,
+      "step": 4480
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.8099693217402806e-05,
+      "loss": 2.9674,
+      "step": 4485
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.805964100252652e-05,
+      "loss": 2.6377,
+      "step": 4490
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.8019580813499303e-05,
+      "loss": 2.798,
+      "step": 4495
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.7979512754727527e-05,
+      "loss": 2.8589,
+      "step": 4500
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.7939436930638048e-05,
+      "loss": 2.9183,
+      "step": 4505
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.7899353445678e-05,
+      "loss": 3.087,
+      "step": 4510
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.7859262404314452e-05,
+      "loss": 2.8895,
+      "step": 4515
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.7819163911034172e-05,
+      "loss": 2.7229,
+      "step": 4520
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.7779058070343367e-05,
+      "loss": 2.7135,
+      "step": 4525
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.773894498676737e-05,
+      "loss": 2.6625,
+      "step": 4530
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.7698824764850408e-05,
+      "loss": 2.8783,
+      "step": 4535
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.7658697509155294e-05,
+      "loss": 2.528,
+      "step": 4540
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.761856332426321e-05,
+      "loss": 2.5445,
+      "step": 4545
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.7578422314773344e-05,
+      "loss": 2.7371,
+      "step": 4550
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.7538274585302703e-05,
+      "loss": 2.7187,
+      "step": 4555
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.7498120240485815e-05,
+      "loss": 2.6928,
+      "step": 4560
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.745795938497443e-05,
+      "loss": 2.6319,
+      "step": 4565
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.7417792123437263e-05,
+      "loss": 2.6975,
+      "step": 4570
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.7377618560559746e-05,
+      "loss": 2.8003,
+      "step": 4575
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.7337438801043724e-05,
+      "loss": 2.9561,
+      "step": 4580
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.7297252949607193e-05,
+      "loss": 2.5395,
+      "step": 4585
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.7257061110984007e-05,
+      "loss": 2.842,
+      "step": 4590
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.7216863389923658e-05,
+      "loss": 2.8684,
+      "step": 4595
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.717665989119094e-05,
+      "loss": 2.714,
+      "step": 4600
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.713645071956573e-05,
+      "loss": 2.8792,
+      "step": 4605
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.709623597984266e-05,
+      "loss": 2.9185,
+      "step": 4610
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.7056015776830905e-05,
+      "loss": 2.596,
+      "step": 4615
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.7015790215353853e-05,
+      "loss": 2.883,
+      "step": 4620
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.6975559400248872e-05,
+      "loss": 2.7508,
+      "step": 4625
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.6935323436367023e-05,
+      "loss": 2.7681,
+      "step": 4630
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.6895082428572776e-05,
+      "loss": 2.9875,
+      "step": 4635
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.6854836481743756e-05,
+      "loss": 2.6518,
+      "step": 4640
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.681458570077045e-05,
+      "loss": 2.5739,
+      "step": 4645
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.6774330190555953e-05,
+      "loss": 3.0008,
+      "step": 4650
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.673407005601569e-05,
+      "loss": 2.6662,
+      "step": 4655
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.669380540207712e-05,
+      "loss": 2.7791,
+      "step": 4660
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.6653536333679502e-05,
+      "loss": 2.9536,
+      "step": 4665
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.6613262955773578e-05,
+      "loss": 2.6102,
+      "step": 4670
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.6572985373321342e-05,
+      "loss": 2.8118,
+      "step": 4675
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.6532703691295734e-05,
+      "loss": 2.7839,
+      "step": 4680
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.6492418014680385e-05,
+      "loss": 2.7449,
+      "step": 4685
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.6452128448469333e-05,
+      "loss": 2.7359,
+      "step": 4690
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.6411835097666747e-05,
+      "loss": 2.784,
+      "step": 4695
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.6371538067286678e-05,
+      "loss": 2.4048,
+      "step": 4700
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.6331237462352742e-05,
+      "loss": 2.8536,
+      "step": 4705
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.6290933387897898e-05,
+      "loss": 2.8984,
+      "step": 4710
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.6250625948964124e-05,
+      "loss": 2.6593,
+      "step": 4715
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.6210315250602185e-05,
+      "loss": 2.6529,
+      "step": 4720
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.617000139787132e-05,
+      "loss": 2.8256,
+      "step": 4725
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.6129684495839014e-05,
+      "loss": 2.7692,
+      "step": 4730
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.6089364649580676e-05,
+      "loss": 2.5697,
+      "step": 4735
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.604904196417941e-05,
+      "loss": 2.7452,
+      "step": 4740
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.6008716544725693e-05,
+      "loss": 2.5728,
+      "step": 4745
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.5968388496317157e-05,
+      "loss": 2.7245,
+      "step": 4750
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.5928057924058263e-05,
+      "loss": 2.6465,
+      "step": 4755
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.5887724933060065e-05,
+      "loss": 2.7824,
+      "step": 4760
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.5847389628439906e-05,
+      "loss": 3.0099,
+      "step": 4765
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.580705211532118e-05,
+      "loss": 2.7486,
+      "step": 4770
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.5766712498833008e-05,
+      "loss": 2.7712,
+      "step": 4775
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.5726370884110018e-05,
+      "loss": 2.5944,
+      "step": 4780
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.5686027376292037e-05,
+      "loss": 2.844,
+      "step": 4785
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.5645682080523826e-05,
+      "loss": 2.4769,
+      "step": 4790
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.5605335101954797e-05,
+      "loss": 2.5758,
+      "step": 4795
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.5564986545738766e-05,
+      "loss": 2.6202,
+      "step": 4800
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.5524636517033646e-05,
+      "loss": 2.4537,
+      "step": 4805
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.54842851210012e-05,
+      "loss": 2.5937,
+      "step": 4810
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.544393246280673e-05,
+      "loss": 2.4738,
+      "step": 4815
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.5403578647618865e-05,
+      "loss": 2.6147,
+      "step": 4820
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.5363223780609214e-05,
+      "loss": 2.7137,
+      "step": 4825
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.5322867966952146e-05,
+      "loss": 2.7939,
+      "step": 4830
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.5282511311824497e-05,
+      "loss": 2.821,
+      "step": 4835
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.524215392040529e-05,
+      "loss": 2.9476,
+      "step": 4840
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.520179589787547e-05,
+      "loss": 2.5596,
+      "step": 4845
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.5161437349417622e-05,
+      "loss": 2.7911,
+      "step": 4850
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.512107838021571e-05,
+      "loss": 2.5827,
+      "step": 4855
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.5080719095454796e-05,
+      "loss": 2.6661,
+      "step": 4860
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.5040359600320747e-05,
+      "loss": 2.9065,
+      "step": 4865
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.5e-05,
+      "loss": 2.6174,
+      "step": 4870
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.4959640399679252e-05,
+      "loss": 2.7372,
+      "step": 4875
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.4919280904545213e-05,
+      "loss": 2.615,
+      "step": 4880
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.4878921619784295e-05,
+      "loss": 2.8296,
+      "step": 4885
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.483856265058238e-05,
+      "loss": 2.5063,
+      "step": 4890
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.4798204102124533e-05,
+      "loss": 2.7276,
+      "step": 4895
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.475784607959472e-05,
+      "loss": 2.7801,
+      "step": 4900
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.4717488688175512e-05,
+      "loss": 2.8598,
+      "step": 4905
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.4685203300121436e-05,
+      "loss": 2.6644,
+      "step": 4910
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.4644847309756042e-05,
+      "loss": 2.4651,
+      "step": 4915
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.4604492245002873e-05,
+      "loss": 2.8326,
+      "step": 4920
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.456413821103682e-05,
+      "loss": 2.6974,
+      "step": 4925
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.4523785313030085e-05,
+      "loss": 2.5726,
+      "step": 4930
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.44834336561519e-05,
+      "loss": 2.5724,
+      "step": 4935
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.4443083345568253e-05,
+      "loss": 2.9383,
+      "step": 4940
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.4402734486441646e-05,
+      "loss": 2.5757,
+      "step": 4945
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.4362387183930775e-05,
+      "loss": 2.9013,
+      "step": 4950
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.432204154319031e-05,
+      "loss": 2.7228,
+      "step": 4955
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.4281697669370545e-05,
+      "loss": 2.7228,
+      "step": 4960
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.424135566761721e-05,
+      "loss": 2.6018,
+      "step": 4965
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.4201015643071136e-05,
+      "loss": 2.7756,
+      "step": 4970
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.416067770086801e-05,
+      "loss": 2.8861,
+      "step": 4975
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.4120341946138085e-05,
+      "loss": 2.9409,
+      "step": 4980
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.4080008484005907e-05,
+      "loss": 2.5961,
+      "step": 4985
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.4039677419590066e-05,
+      "loss": 2.7308,
+      "step": 4990
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.3999348858002888e-05,
+      "loss": 3.0326,
+      "step": 4995
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.395902290435017e-05,
+      "loss": 2.7954,
+      "step": 5000
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.391869966373094e-05,
+      "loss": 2.8,
+      "step": 5005
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.3878379241237136e-05,
+      "loss": 2.4567,
+      "step": 5010
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.3838061741953334e-05,
+      "loss": 2.7514,
+      "step": 5015
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.3797747270956522e-05,
+      "loss": 2.7957,
+      "step": 5020
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.3757435933315785e-05,
+      "loss": 2.8452,
+      "step": 5025
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.371712783409205e-05,
+      "loss": 2.8488,
+      "step": 5030
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.367682307833777e-05,
+      "loss": 2.7709,
+      "step": 5035
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.3636521771096725e-05,
+      "loss": 2.8874,
+      "step": 5040
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.3596224017403683e-05,
+      "loss": 3.0164,
+      "step": 5045
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.355592992228418e-05,
+      "loss": 2.8052,
+      "step": 5050
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.3515639590754163e-05,
+      "loss": 2.8099,
+      "step": 5055
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.3475353127819823e-05,
+      "loss": 2.734,
+      "step": 5060
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.3435070638477237e-05,
+      "loss": 2.7923,
+      "step": 5065
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.339479222771215e-05,
+      "loss": 2.8617,
+      "step": 5070
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.3354518000499652e-05,
+      "loss": 2.4441,
+      "step": 5075
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.3314248061803948e-05,
+      "loss": 2.7458,
+      "step": 5080
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.327398251657805e-05,
+      "loss": 2.5911,
+      "step": 5085
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.3233721469763556e-05,
+      "loss": 2.7941,
+      "step": 5090
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.319346502629028e-05,
+      "loss": 2.5942,
+      "step": 5095
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.315321329107609e-05,
+      "loss": 3.008,
+      "step": 5100
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.311296636902656e-05,
+      "loss": 2.5184,
+      "step": 5105
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.307272436503474e-05,
+      "loss": 2.7171,
+      "step": 5110
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.3032487383980822e-05,
+      "loss": 2.8098,
+      "step": 5115
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.299225553073195e-05,
+      "loss": 2.7077,
+      "step": 5120
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.2952028910141875e-05,
+      "loss": 2.4607,
+      "step": 5125
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.2911807627050748e-05,
+      "loss": 2.5207,
+      "step": 5130
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.2871591786284746e-05,
+      "loss": 2.8489,
+      "step": 5135
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.283138149265592e-05,
+      "loss": 2.5236,
+      "step": 5140
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.279117685096184e-05,
+      "loss": 2.9405,
+      "step": 5145
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.2750977965985348e-05,
+      "loss": 2.896,
+      "step": 5150
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.2710784942494278e-05,
+      "loss": 2.5815,
+      "step": 5155
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.2670597885241202e-05,
+      "loss": 2.6745,
+      "step": 5160
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.2630416898963128e-05,
+      "loss": 2.521,
+      "step": 5165
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2590242088381248e-05,
+      "loss": 2.6961,
+      "step": 5170
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.255007355820065e-05,
+      "loss": 2.5516,
+      "step": 5175
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2509911413110067e-05,
+      "loss": 2.9821,
+      "step": 5180
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2469755757781597e-05,
+      "loss": 2.8287,
+      "step": 5185
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.2429606696870387e-05,
+      "loss": 2.8595,
+      "step": 5190
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.238946433501444e-05,
+      "loss": 2.8129,
+      "step": 5195
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.2349328776834278e-05,
+      "loss": 2.8261,
+      "step": 5200
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.2309200126932705e-05,
+      "loss": 2.5692,
+      "step": 5205
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.2269078489894495e-05,
+      "loss": 2.9282,
+      "step": 5210
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.2228963970286164e-05,
+      "loss": 2.6939,
+      "step": 5215
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.218885667265568e-05,
+      "loss": 2.7504,
+      "step": 5220
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.2148756701532196e-05,
+      "loss": 2.7624,
+      "step": 5225
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.210866416142573e-05,
+      "loss": 2.6252,
+      "step": 5230
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.2068579156826974e-05,
+      "loss": 2.7068,
+      "step": 5235
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.2028501792206975e-05,
+      "loss": 2.6107,
+      "step": 5240
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.198843217201686e-05,
+      "loss": 2.6384,
+      "step": 5245
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.1948370400687556e-05,
+      "loss": 2.7685,
+      "step": 5250
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.1908316582629572e-05,
+      "loss": 2.9635,
+      "step": 5255
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.1868270822232656e-05,
+      "loss": 2.6668,
+      "step": 5260
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.1828233223865583e-05,
+      "loss": 2.7072,
+      "step": 5265
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.1788203891875818e-05,
+      "loss": 2.8598,
+      "step": 5270
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.1748182930589316e-05,
+      "loss": 2.7215,
+      "step": 5275
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.1708170444310205e-05,
+      "loss": 2.5789,
+      "step": 5280
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.1668166537320535e-05,
+      "loss": 2.7569,
+      "step": 5285
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.1628171313879963e-05,
+      "loss": 2.6224,
+      "step": 5290
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.1588184878225553e-05,
+      "loss": 2.3747,
+      "step": 5295
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.1548207334571452e-05,
+      "loss": 2.7526,
+      "step": 5300
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.1508238787108635e-05,
+      "loss": 2.7644,
+      "step": 5305
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.1468279340004607e-05,
+      "loss": 2.5787,
+      "step": 5310
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.1428329097403187e-05,
+      "loss": 2.7147,
+      "step": 5315
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.1388388163424193e-05,
+      "loss": 2.8698,
+      "step": 5320
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.1348456642163196e-05,
+      "loss": 2.6175,
+      "step": 5325
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.1308534637691198e-05,
+      "loss": 2.4902,
+      "step": 5330
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.1268622254054436e-05,
+      "loss": 2.6952,
+      "step": 5335
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.1228719595274055e-05,
+      "loss": 2.9045,
+      "step": 5340
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.118882676534585e-05,
+      "loss": 2.8485,
+      "step": 5345
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.1148943868240024e-05,
+      "loss": 2.8643,
+      "step": 5350
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.110907100790086e-05,
+      "loss": 2.7929,
+      "step": 5355
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.1069208288246518e-05,
+      "loss": 2.6293,
+      "step": 5360
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.1029355813168695e-05,
+      "loss": 2.6163,
+      "step": 5365
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.0989513686532403e-05,
+      "loss": 2.9208,
+      "step": 5370
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.0949682012175693e-05,
+      "loss": 2.7804,
+      "step": 5375
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.0909860893909372e-05,
+      "loss": 2.8074,
+      "step": 5380
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.087005043551671e-05,
+      "loss": 3.0755,
+      "step": 5385
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.0830250740753222e-05,
+      "loss": 2.8733,
+      "step": 5390
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.0790461913346364e-05,
+      "loss": 2.6596,
+      "step": 5395
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.0750684056995286e-05,
+      "loss": 2.7885,
+      "step": 5400
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.071091727537049e-05,
+      "loss": 2.9107,
+      "step": 5405
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.0671161672113677e-05,
+      "loss": 2.8962,
+      "step": 5410
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.0631417350837378e-05,
+      "loss": 2.7318,
+      "step": 5415
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.0591684415124743e-05,
+      "loss": 2.8219,
+      "step": 5420
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.055196296852922e-05,
+      "loss": 2.8166,
+      "step": 5425
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.0512253114574343e-05,
+      "loss": 2.7209,
+      "step": 5430
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.047255495675341e-05,
+      "loss": 2.7705,
+      "step": 5435
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.043286859852926e-05,
+      "loss": 2.6631,
+      "step": 5440
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.0393194143333953e-05,
+      "loss": 2.8472,
+      "step": 5445
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.035353169456854e-05,
+      "loss": 2.7624,
+      "step": 5450
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.0313881355602785e-05,
+      "loss": 2.5701,
+      "step": 5455
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0274243229774896e-05,
+      "loss": 2.6064,
+      "step": 5460
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0234617420391218e-05,
+      "loss": 2.5272,
+      "step": 5465
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.019500403072603e-05,
+      "loss": 2.4602,
+      "step": 5470
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0155403164021227e-05,
+      "loss": 2.4315,
+      "step": 5475
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.011581492348609e-05,
+      "loss": 2.7391,
+      "step": 5480
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.0076239412296946e-05,
+      "loss": 2.6853,
+      "step": 5485
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.003667673359699e-05,
+      "loss": 2.5167,
+      "step": 5490
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.9997126990495947e-05,
+      "loss": 2.6499,
+      "step": 5495
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.9957590286069856e-05,
+      "loss": 2.545,
+      "step": 5500
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.9918066723360738e-05,
+      "loss": 2.8519,
+      "step": 5505
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.9878556405376385e-05,
+      "loss": 2.7581,
+      "step": 5510
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.9839059435090073e-05,
+      "loss": 2.7482,
+      "step": 5515
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.9799575915440284e-05,
+      "loss": 2.8268,
+      "step": 5520
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.9760105949330446e-05,
+      "loss": 2.595,
+      "step": 5525
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.972064963962865e-05,
+      "loss": 2.7057,
+      "step": 5530
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.9681207089167422e-05,
+      "loss": 2.8896,
+      "step": 5535
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.964177840074341e-05,
+      "loss": 2.6064,
+      "step": 5540
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.960236367711713e-05,
+      "loss": 2.6605,
+      "step": 5545
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.956296302101272e-05,
+      "loss": 2.981,
+      "step": 5550
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.952357653511765e-05,
+      "loss": 2.7744,
+      "step": 5555
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.9484204322082435e-05,
+      "loss": 2.4719,
+      "step": 5560
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.944484648452042e-05,
+      "loss": 2.4476,
+      "step": 5565
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.9405503125007478e-05,
+      "loss": 2.8826,
+      "step": 5570
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.9366174346081755e-05,
+      "loss": 2.4737,
+      "step": 5575
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.932686025024337e-05,
+      "loss": 2.9233,
+      "step": 5580
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.928756093995419e-05,
+      "loss": 2.7084,
+      "step": 5585
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.924827651763756e-05,
+      "loss": 2.9017,
+      "step": 5590
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.9209007085678017e-05,
+      "loss": 2.5049,
+      "step": 5595
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.9169752746421008e-05,
+      "loss": 2.7287,
+      "step": 5600
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.9130513602172668e-05,
+      "loss": 2.6271,
+      "step": 5605
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.909128975519953e-05,
+      "loss": 2.6457,
+      "step": 5610
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.9052081307728254e-05,
+      "loss": 2.7968,
+      "step": 5615
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.9012888361945353e-05,
+      "loss": 2.9797,
+      "step": 5620
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.8973711019996948e-05,
+      "loss": 2.8581,
+      "step": 5625
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.8934549383988503e-05,
+      "loss": 2.949,
+      "step": 5630
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.889540355598455e-05,
+      "loss": 2.9881,
+      "step": 5635
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.8856273638008388e-05,
+      "loss": 2.6324,
+      "step": 5640
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.8817159732041883e-05,
+      "loss": 2.6003,
+      "step": 5645
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.877806194002517e-05,
+      "loss": 3.043,
+      "step": 5650
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.8738980363856375e-05,
+      "loss": 2.749,
+      "step": 5655
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.8699915105391358e-05,
+      "loss": 2.593,
+      "step": 5660
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.8660866266443456e-05,
+      "loss": 2.7712,
+      "step": 5665
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.8621833948783223e-05,
+      "loss": 2.5154,
+      "step": 5670
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.858281825413816e-05,
+      "loss": 2.6295,
+      "step": 5675
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.85438192841924e-05,
+      "loss": 2.6558,
+      "step": 5680
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.8504837140586538e-05,
+      "loss": 2.4923,
+      "step": 5685
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.8465871924917294e-05,
+      "loss": 2.4985,
+      "step": 5690
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.8426923738737272e-05,
+      "loss": 2.8206,
+      "step": 5695
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.838799268355469e-05,
+      "loss": 2.6971,
+      "step": 5700
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.8349078860833123e-05,
+      "loss": 2.6192,
+      "step": 5705
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.8310182371991236e-05,
+      "loss": 2.5193,
+      "step": 5710
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.8271303318402504e-05,
+      "loss": 2.8105,
+      "step": 5715
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.823244180139499e-05,
+      "loss": 2.5983,
+      "step": 5720
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.819359792225101e-05,
+      "loss": 2.4457,
+      "step": 5725
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.8154771782206966e-05,
+      "loss": 2.6443,
+      "step": 5730
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.8115963482452968e-05,
+      "loss": 2.6344,
+      "step": 5735
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.8077173124132678e-05,
+      "loss": 2.8258,
+      "step": 5740
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.803840080834297e-05,
+      "loss": 2.7479,
+      "step": 5745
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.7999646636133726e-05,
+      "loss": 2.7194,
+      "step": 5750
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.7960910708507495e-05,
+      "loss": 2.7996,
+      "step": 5755
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.7922193126419305e-05,
+      "loss": 2.6871,
+      "step": 5760
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.7883493990776383e-05,
+      "loss": 3.0033,
+      "step": 5765
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.7844813402437863e-05,
+      "loss": 2.7075,
+      "step": 5770
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.7806151462214527e-05,
+      "loss": 2.6911,
+      "step": 5775
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.776750827086858e-05,
+      "loss": 2.7382,
+      "step": 5780
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.772888392911336e-05,
+      "loss": 2.6003,
+      "step": 5785
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.769027853761307e-05,
+      "loss": 2.3981,
+      "step": 5790
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.7651692196982517e-05,
+      "loss": 2.7203,
+      "step": 5795
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.7613125007786876e-05,
+      "loss": 3.2133,
+      "step": 5800
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.7574577070541395e-05,
+      "loss": 2.6019,
+      "step": 5805
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.753604848571116e-05,
+      "loss": 2.7039,
+      "step": 5810
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.74975393537108e-05,
+      "loss": 2.9425,
+      "step": 5815
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.7459049774904257e-05,
+      "loss": 2.6642,
+      "step": 5820
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.742057984960452e-05,
+      "loss": 2.7018,
+      "step": 5825
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.738212967807335e-05,
+      "loss": 2.851,
+      "step": 5830
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.734369936052101e-05,
+      "loss": 2.6783,
+      "step": 5835
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.730528899710604e-05,
+      "loss": 2.827,
+      "step": 5840
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.726689868793497e-05,
+      "loss": 2.8378,
+      "step": 5845
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.722852853306207e-05,
+      "loss": 2.4727,
+      "step": 5850
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.7190178632489053e-05,
+      "loss": 2.1913,
+      "step": 5855
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.7151849086164872e-05,
+      "loss": 2.7207,
+      "step": 5860
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.711353999398543e-05,
+      "loss": 2.4743,
+      "step": 5865
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.707525145579333e-05,
+      "loss": 2.4683,
+      "step": 5870
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.7036983571377567e-05,
+      "loss": 2.7584,
+      "step": 5875
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.699873644047335e-05,
+      "loss": 2.4785,
+      "step": 5880
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.6960510162761794e-05,
+      "loss": 2.6711,
+      "step": 5885
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.692230483786963e-05,
+      "loss": 2.3343,
+      "step": 5890
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.688412056536902e-05,
+      "loss": 2.6087,
+      "step": 5895
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.6845957444777245e-05,
+      "loss": 2.5869,
+      "step": 5900
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.6807815575556458e-05,
+      "loss": 2.3216,
+      "step": 5905
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.6769695057113422e-05,
+      "loss": 2.6727,
+      "step": 5910
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.673159598879926e-05,
+      "loss": 2.7402,
+      "step": 5915
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.669351846990919e-05,
+      "loss": 2.5676,
+      "step": 5920
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.6655462599682287e-05,
+      "loss": 2.4796,
+      "step": 5925
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.6617428477301153e-05,
+      "loss": 2.5886,
+      "step": 5930
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.6579416201891756e-05,
+      "loss": 2.6075,
+      "step": 5935
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.6541425872523115e-05,
+      "loss": 2.4519,
+      "step": 5940
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.6503457588207052e-05,
+      "loss": 2.6462,
+      "step": 5945
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.6465511447897917e-05,
+      "loss": 2.611,
+      "step": 5950
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.6427587550492367e-05,
+      "loss": 2.8779,
+      "step": 5955
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.638968599482909e-05,
+      "loss": 2.8401,
+      "step": 5960
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.6351806879688542e-05,
+      "loss": 2.3667,
+      "step": 5965
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.6313950303792673e-05,
+      "loss": 2.8041,
+      "step": 5970
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.6276116365804716e-05,
+      "loss": 2.908,
+      "step": 5975
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.623830516432891e-05,
+      "loss": 2.2857,
+      "step": 5980
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.6200516797910216e-05,
+      "loss": 2.3759,
+      "step": 5985
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.6162751365034083e-05,
+      "loss": 2.4894,
+      "step": 5990
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.6125008964126197e-05,
+      "loss": 2.8261,
+      "step": 5995
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.6087289693552215e-05,
+      "loss": 2.4668,
+      "step": 6000
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.6049593651617534e-05,
+      "loss": 2.6883,
+      "step": 6005
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.601192093656696e-05,
+      "loss": 2.4301,
+      "step": 6010
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.597427164658454e-05,
+      "loss": 2.3079,
+      "step": 6015
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.593664587979327e-05,
+      "loss": 2.3398,
+      "step": 6020
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.589904373425485e-05,
+      "loss": 2.7114,
+      "step": 6025
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.5861465307969363e-05,
+      "loss": 2.5915,
+      "step": 6030
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.5823910698875128e-05,
+      "loss": 2.5822,
+      "step": 6035
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.5786380004848376e-05,
+      "loss": 2.6907,
+      "step": 6040
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.5748873323703013e-05,
+      "loss": 2.9532,
+      "step": 6045
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.5711390753190336e-05,
+      "loss": 2.5533,
+      "step": 6050
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.5673932390998835e-05,
+      "loss": 2.5146,
+      "step": 6055
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.5636498334753902e-05,
+      "loss": 2.8292,
+      "step": 6060
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.5599088682017565e-05,
+      "loss": 2.759,
+      "step": 6065
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.5561703530288262e-05,
+      "loss": 2.435,
+      "step": 6070
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.5524342977000586e-05,
+      "loss": 2.6814,
+      "step": 6075
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.5487007119525e-05,
+      "loss": 2.2766,
+      "step": 6080
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.5449696055167616e-05,
+      "loss": 2.4739,
+      "step": 6085
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.5412409881169926e-05,
+      "loss": 2.8195,
+      "step": 6090
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.537514869470855e-05,
+      "loss": 2.5721,
+      "step": 6095
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.5337912592895e-05,
+      "loss": 2.4579,
+      "step": 6100
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.530070167277538e-05,
+      "loss": 2.713,
+      "step": 6105
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.5263516031330194e-05,
+      "loss": 2.6144,
+      "step": 6110
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.5226355765474048e-05,
+      "loss": 2.6305,
+      "step": 6115
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.5189220972055435e-05,
+      "loss": 2.8604,
+      "step": 6120
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.5152111747856418e-05,
+      "loss": 2.3486,
+      "step": 6125
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.5115028189592462e-05,
+      "loss": 2.6622,
+      "step": 6130
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.5077970393912116e-05,
+      "loss": 2.6616,
+      "step": 6135
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.5040938457396817e-05,
+      "loss": 2.5339,
+      "step": 6140
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.5003932476560553e-05,
+      "loss": 2.7147,
+      "step": 6145
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.4966952547849714e-05,
+      "loss": 2.5724,
+      "step": 6150
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.4929998767642776e-05,
+      "loss": 2.6453,
+      "step": 6155
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.4893071232250062e-05,
+      "loss": 2.5868,
+      "step": 6160
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.4856170037913486e-05,
+      "loss": 2.4508,
+      "step": 6165
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.4819295280806338e-05,
+      "loss": 2.8447,
+      "step": 6170
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.4782447057032983e-05,
+      "loss": 2.6028,
+      "step": 6175
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.4745625462628654e-05,
+      "loss": 2.7932,
+      "step": 6180
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.4708830593559142e-05,
+      "loss": 2.5669,
+      "step": 6185
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.4672062545720633e-05,
+      "loss": 2.7314,
+      "step": 6190
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.4635321414939385e-05,
+      "loss": 2.5953,
+      "step": 6195
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.4598607296971517e-05,
+      "loss": 2.7438,
+      "step": 6200
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.4561920287502711e-05,
+      "loss": 2.5747,
+      "step": 6205
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.4525260482148059e-05,
+      "loss": 2.3323,
+      "step": 6210
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.4488627976451698e-05,
+      "loss": 2.5923,
+      "step": 6215
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.4452022865886644e-05,
+      "loss": 2.6783,
+      "step": 6220
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.44154452458545e-05,
+      "loss": 2.6587,
+      "step": 6225
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.4378895211685223e-05,
+      "loss": 2.6062,
+      "step": 6230
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.4342372858636896e-05,
+      "loss": 2.9328,
+      "step": 6235
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.4305878281895437e-05,
+      "loss": 2.5786,
+      "step": 6240
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.4269411576574376e-05,
+      "loss": 2.4903,
+      "step": 6245
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.4232972837714598e-05,
+      "loss": 2.4484,
+      "step": 6250
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.4196562160284138e-05,
+      "loss": 2.7365,
+      "step": 6255
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.4160179639177823e-05,
+      "loss": 2.3422,
+      "step": 6260
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.4123825369217178e-05,
+      "loss": 3.0981,
+      "step": 6265
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.4087499445150038e-05,
+      "loss": 2.6238,
+      "step": 6270
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.4051201961650415e-05,
+      "loss": 2.6983,
+      "step": 6275
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.4014933013318132e-05,
+      "loss": 2.3451,
+      "step": 6280
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.397869269467871e-05,
+      "loss": 2.3463,
+      "step": 6285
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.3942481100183002e-05,
+      "loss": 2.8322,
+      "step": 6290
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.3906298324207057e-05,
+      "loss": 2.5479,
+      "step": 6295
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.3870144461051737e-05,
+      "loss": 2.5753,
+      "step": 6300
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.3834019604942618e-05,
+      "loss": 2.9168,
+      "step": 6305
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.3797923850029653e-05,
+      "loss": 2.4357,
+      "step": 6310
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.3761857290386942e-05,
+      "loss": 2.6056,
+      "step": 6315
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.3725820020012507e-05,
+      "loss": 2.5245,
+      "step": 6320
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.3689812132828017e-05,
+      "loss": 2.5831,
+      "step": 6325
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.36538337226786e-05,
+      "loss": 2.622,
+      "step": 6330
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.3617884883332527e-05,
+      "loss": 2.6699,
+      "step": 6335
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.3581965708481e-05,
+      "loss": 2.7383,
+      "step": 6340
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.3546076291737918e-05,
+      "loss": 2.4363,
+      "step": 6345
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.3510216726639635e-05,
+      "loss": 2.4355,
+      "step": 6350
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.3474387106644688e-05,
+      "loss": 2.3816,
+      "step": 6355
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.343858752513357e-05,
+      "loss": 2.8382,
+      "step": 6360
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.340281807540848e-05,
+      "loss": 2.4771,
+      "step": 6365
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.3367078850693122e-05,
+      "loss": 2.4233,
+      "step": 6370
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.3331369944132387e-05,
+      "loss": 2.4934,
+      "step": 6375
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.3295691448792163e-05,
+      "loss": 2.5107,
+      "step": 6380
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.3260043457659077e-05,
+      "loss": 2.8295,
+      "step": 6385
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.3224426063640272e-05,
+      "loss": 2.631,
+      "step": 6390
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.3188839359563124e-05,
+      "loss": 2.459,
+      "step": 6395
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.3153283438175034e-05,
+      "loss": 2.7249,
+      "step": 6400
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.3117758392143181e-05,
+      "loss": 2.6904,
+      "step": 6405
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.3082264314054255e-05,
+      "loss": 2.7219,
+      "step": 6410
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.304680129641428e-05,
+      "loss": 2.5716,
+      "step": 6415
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.3011369431648263e-05,
+      "loss": 2.5634,
+      "step": 6420
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.2975968812100082e-05,
+      "loss": 2.8551,
+      "step": 6425
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.2940599530032154e-05,
+      "loss": 2.6722,
+      "step": 6430
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.2905261677625224e-05,
+      "loss": 2.7193,
+      "step": 6435
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.2869955346978113e-05,
+      "loss": 2.6377,
+      "step": 6440
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.2834680630107521e-05,
+      "loss": 2.5561,
+      "step": 6445
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.2799437618947733e-05,
+      "loss": 2.5548,
+      "step": 6450
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.2764226405350397e-05,
+      "loss": 2.6126,
+      "step": 6455
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.2729047081084289e-05,
+      "loss": 2.6932,
+      "step": 6460
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.2693899737835097e-05,
+      "loss": 2.6047,
+      "step": 6465
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.2658784467205143e-05,
+      "loss": 2.7439,
+      "step": 6470
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.2623701360713146e-05,
+      "loss": 2.7272,
+      "step": 6475
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.2588650509794015e-05,
+      "loss": 2.5525,
+      "step": 6480
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.2553632005798598e-05,
+      "loss": 2.7012,
+      "step": 6485
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.2518645939993435e-05,
+      "loss": 2.7472,
+      "step": 6490
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.2483692403560506e-05,
+      "loss": 2.7044,
+      "step": 6495
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.2448771487597036e-05,
+      "loss": 2.5916,
+      "step": 6500
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.2413883283115208e-05,
+      "loss": 2.2904,
+      "step": 6505
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.2379027881042e-05,
+      "loss": 2.466,
+      "step": 6510
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.2344205372218825e-05,
+      "loss": 2.7404,
+      "step": 6515
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.230941584740144e-05,
+      "loss": 2.3594,
+      "step": 6520
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.2274659397259588e-05,
+      "loss": 2.5332,
+      "step": 6525
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.2239936112376857e-05,
+      "loss": 2.6614,
+      "step": 6530
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.2205246083250332e-05,
+      "loss": 2.7086,
+      "step": 6535
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.21705894002905e-05,
+      "loss": 2.6566,
+      "step": 6540
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.2135966153820882e-05,
+      "loss": 2.6669,
+      "step": 6545
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.2101376434077912e-05,
+      "loss": 2.7785,
+      "step": 6550
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.2066820331210571e-05,
+      "loss": 2.1548,
+      "step": 6555
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.2032297935280296e-05,
+      "loss": 2.7001,
+      "step": 6560
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.1997809336260644e-05,
+      "loss": 2.8728,
+      "step": 6565
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.1963354624037084e-05,
+      "loss": 2.3047,
+      "step": 6570
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.1928933888406774e-05,
+      "loss": 2.6097,
+      "step": 6575
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.1894547219078343e-05,
+      "loss": 2.4706,
+      "step": 6580
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.18601947056716e-05,
+      "loss": 2.7075,
+      "step": 6585
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.1825876437717361e-05,
+      "loss": 2.6215,
+      "step": 6590
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.1791592504657174e-05,
+      "loss": 2.5929,
+      "step": 6595
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.1757342995843102e-05,
+      "loss": 2.4798,
+      "step": 6600
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 9730,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 300,
+  "total_flos": 7.787528111424307e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}