diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6768 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 5570,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.968609865470853e-08,
+      "loss": 2.0239,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.484304932735426e-07,
+      "loss": 1.9842,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.968609865470852e-07,
+      "loss": 1.9791,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.345291479820628e-06,
+      "loss": 1.9583,
+      "step": 15
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.7937219730941704e-06,
+      "loss": 1.9019,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.242152466367713e-06,
+      "loss": 1.8031,
+      "step": 25
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.690582959641256e-06,
+      "loss": 1.7329,
+      "step": 30
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.1390134529147986e-06,
+      "loss": 1.6897,
+      "step": 35
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.587443946188341e-06,
+      "loss": 1.6468,
+      "step": 40
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.0358744394618836e-06,
+      "loss": 1.6347,
+      "step": 45
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.484304932735426e-06,
+      "loss": 1.5639,
+      "step": 50
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.932735426008969e-06,
+      "loss": 1.5379,
+      "step": 55
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.381165919282512e-06,
+      "loss": 1.5118,
+      "step": 60
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.8295964125560544e-06,
+      "loss": 1.484,
+      "step": 65
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 6.278026905829597e-06,
+      "loss": 1.4948,
+      "step": 70
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 6.72645739910314e-06,
+      "loss": 1.4806,
+      "step": 75
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 7.174887892376682e-06,
+      "loss": 1.4396,
+      "step": 80
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 7.6233183856502244e-06,
+      "loss": 1.4329,
+      "step": 85
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 8.071748878923767e-06,
+      "loss": 1.42,
+      "step": 90
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 8.52017937219731e-06,
+      "loss": 1.4031,
+      "step": 95
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 8.968609865470853e-06,
+      "loss": 1.4101,
+      "step": 100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.417040358744395e-06,
+      "loss": 1.3792,
+      "step": 105
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.865470852017938e-06,
+      "loss": 1.4017,
+      "step": 110
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.0313901345291482e-05,
+      "loss": 1.3627,
+      "step": 115
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.0762331838565023e-05,
+      "loss": 1.3775,
+      "step": 120
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.1210762331838564e-05,
+      "loss": 1.3622,
+      "step": 125
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.1659192825112109e-05,
+      "loss": 1.3488,
+      "step": 130
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.210762331838565e-05,
+      "loss": 1.375,
+      "step": 135
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.2556053811659194e-05,
+      "loss": 1.363,
+      "step": 140
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.3004484304932737e-05,
+      "loss": 1.3265,
+      "step": 145
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.345291479820628e-05,
+      "loss": 1.3403,
+      "step": 150
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.3901345291479822e-05,
+      "loss": 1.343,
+      "step": 155
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.4349775784753363e-05,
+      "loss": 1.3238,
+      "step": 160
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.4798206278026908e-05,
+      "loss": 1.3275,
+      "step": 165
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.5246636771300449e-05,
+      "loss": 1.2888,
+      "step": 170
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.569506726457399e-05,
+      "loss": 1.2978,
+      "step": 175
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.6143497757847534e-05,
+      "loss": 1.2675,
+      "step": 180
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.6591928251121077e-05,
+      "loss": 1.302,
+      "step": 185
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.704035874439462e-05,
+      "loss": 1.2428,
+      "step": 190
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7488789237668162e-05,
+      "loss": 1.2886,
+      "step": 195
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7937219730941705e-05,
+      "loss": 1.2717,
+      "step": 200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8385650224215248e-05,
+      "loss": 1.2753,
+      "step": 205
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.883408071748879e-05,
+      "loss": 1.2653,
+      "step": 210
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9282511210762333e-05,
+      "loss": 1.2403,
+      "step": 215
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9730941704035876e-05,
+      "loss": 1.2646,
+      "step": 220
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9999950946779145e-05,
+      "loss": 1.2439,
+      "step": 225
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9999399103571276e-05,
+      "loss": 1.2371,
+      "step": 230
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.999823413457922e-05,
+      "loss": 1.2457,
+      "step": 235
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.999645611123453e-05,
+      "loss": 1.223,
+      "step": 240
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.999406514255894e-05,
+      "loss": 1.2375,
+      "step": 245
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9991061375157722e-05,
+      "loss": 1.2303,
+      "step": 250
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9987444993210666e-05,
+      "loss": 1.2654,
+      "step": 255
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.99832162184608e-05,
+      "loss": 1.2186,
+      "step": 260
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.997837531020079e-05,
+      "loss": 1.2295,
+      "step": 265
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.997292256525703e-05,
+      "loss": 1.2621,
+      "step": 270
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.996685831797147e-05,
+      "loss": 1.2542,
+      "step": 275
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.996018294018109e-05,
+      "loss": 1.2517,
+      "step": 280
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.995289684119511e-05,
+      "loss": 1.2157,
+      "step": 285
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.994500046776989e-05,
+      "loss": 1.2732,
+      "step": 290
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9936494304081536e-05,
+      "loss": 1.2268,
+      "step": 295
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9927378871696224e-05,
+      "loss": 1.2414,
+      "step": 300
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9917654729538193e-05,
+      "loss": 1.2552,
+      "step": 305
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9907322473855508e-05,
+      "loss": 1.2284,
+      "step": 310
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.989638273818347e-05,
+      "loss": 1.205,
+      "step": 315
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9884836193305788e-05,
+      "loss": 1.2287,
+      "step": 320
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9872683547213446e-05,
+      "loss": 1.2228,
+      "step": 325
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9859925545061288e-05,
+      "loss": 1.21,
+      "step": 330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.984656296912232e-05,
+      "loss": 1.2205,
+      "step": 335
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9832596638739766e-05,
+      "loss": 1.2289,
+      "step": 340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9818027410276813e-05,
+      "loss": 1.1928,
+      "step": 345
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9802856177064105e-05,
+      "loss": 1.1985,
+      "step": 350
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.978708386934496e-05,
+      "loss": 1.2002,
+      "step": 355
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9770711454218345e-05,
+      "loss": 1.1955,
+      "step": 360
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9753739935579576e-05,
+      "loss": 1.223,
+      "step": 365
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9736170354058746e-05,
+      "loss": 1.2069,
+      "step": 370
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9718003786956937e-05,
+      "loss": 1.1913,
+      "step": 375
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9699241348180155e-05,
+      "loss": 1.2101,
+      "step": 380
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9679884188171023e-05,
+      "loss": 1.1892,
+      "step": 385
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.965993349383826e-05,
+      "loss": 1.185,
+      "step": 390
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9639390488483873e-05,
+      "loss": 1.2346,
+      "step": 395
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.961825643172819e-05,
+      "loss": 1.1946,
+      "step": 400
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.959653261943259e-05,
+      "loss": 1.2074,
+      "step": 405
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.9574220383620054e-05,
+      "loss": 1.2111,
+      "step": 410
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.9551321092393505e-05,
+      "loss": 1.1687,
+      "step": 415
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.9527836149851897e-05,
+      "loss": 1.1676,
+      "step": 420
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.9503766996004143e-05,
+      "loss": 1.2078,
+      "step": 425
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.9479115106680807e-05,
+      "loss": 1.1775,
+      "step": 430
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9453881993443624e-05,
+      "loss": 1.2089,
+      "step": 435
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9428069203492784e-05,
+      "loss": 1.1762,
+      "step": 440
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.9401678319572115e-05,
+      "loss": 1.1667,
+      "step": 445
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9374710959871986e-05,
+      "loss": 1.1833,
+      "step": 450
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9347168777930118e-05,
+      "loss": 1.1816,
+      "step": 455
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.931905346253017e-05,
+      "loss": 1.2159,
+      "step": 460
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9290366737598226e-05,
+      "loss": 1.2015,
+      "step": 465
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9261110362097046e-05,
+      "loss": 1.1925,
+      "step": 470
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.9231286129918244e-05,
+      "loss": 1.2161,
+      "step": 475
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9200895869772273e-05,
+      "loss": 1.2061,
+      "step": 480
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9169941445076325e-05,
+      "loss": 1.1669,
+      "step": 485
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9138424753840036e-05,
+      "loss": 1.1644,
+      "step": 490
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.9106347728549134e-05,
+      "loss": 1.142,
+      "step": 495
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.907371233604693e-05,
+      "loss": 1.163,
+      "step": 500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9040520577413734e-05,
+      "loss": 1.1706,
+      "step": 505
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.9006774487844145e-05,
+      "loss": 1.1712,
+      "step": 510
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.897247613652226e-05,
+      "loss": 1.1564,
+      "step": 515
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.8937627626494804e-05,
+      "loss": 1.1617,
+      "step": 520
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.890223109454218e-05,
+      "loss": 1.1242,
+      "step": 525
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.8866288711047452e-05,
+      "loss": 1.1557,
+      "step": 530
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.8829802679863246e-05,
+      "loss": 1.1654,
+      "step": 535
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.8792775238176655e-05,
+      "loss": 1.1622,
+      "step": 540
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.8755208656372016e-05,
+      "loss": 1.192,
+      "step": 545
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.871710523789174e-05,
+      "loss": 1.1524,
+      "step": 550
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.8678467319095045e-05,
+      "loss": 1.1752,
+      "step": 555
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.8639297269114725e-05,
+      "loss": 1.1476,
+      "step": 560
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.859959748971186e-05,
+      "loss": 1.1557,
+      "step": 565
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.8559370415128562e-05,
+      "loss": 1.175,
+      "step": 570
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.851861851193871e-05,
+      "loss": 1.1588,
+      "step": 575
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.8477344278896708e-05,
+      "loss": 1.1646,
+      "step": 580
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.843555024678429e-05,
+      "loss": 1.1518,
+      "step": 585
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.839323897825531e-05,
+      "loss": 1.1741,
+      "step": 590
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.8350413067678636e-05,
+      "loss": 1.1663,
+      "step": 595
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8307075140979057e-05,
+      "loss": 1.1659,
+      "step": 600
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8263227855476275e-05,
+      "loss": 1.1646,
+      "step": 605
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.821887389972198e-05,
+      "loss": 1.1876,
+      "step": 610
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.817401599333499e-05,
+      "loss": 1.1587,
+      "step": 615
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.812865688683447e-05,
+      "loss": 1.1564,
+      "step": 620
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.808279936147133e-05,
+      "loss": 1.1344,
+      "step": 625
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.8036446229057653e-05,
+      "loss": 1.1542,
+      "step": 630
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.798960033179428e-05,
+      "loss": 1.1723,
+      "step": 635
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.7942264542096574e-05,
+      "loss": 1.1705,
+      "step": 640
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.7894441762418254e-05,
+      "loss": 1.1291,
+      "step": 645
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.7846134925073452e-05,
+      "loss": 1.1621,
+      "step": 650
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.7797346992056913e-05,
+      "loss": 1.174,
+      "step": 655
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.774808095486235e-05,
+      "loss": 1.1611,
+      "step": 660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.7698339834299064e-05,
+      "loss": 1.1515,
+      "step": 665
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.7648126680306674e-05,
+      "loss": 1.1548,
+      "step": 670
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.7597444571768133e-05,
+      "loss": 1.1478,
+      "step": 675
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.754629661632093e-05,
+      "loss": 1.1534,
+      "step": 680
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.7494685950166547e-05,
+      "loss": 1.1537,
+      "step": 685
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.744261573787816e-05,
+      "loss": 1.152,
+      "step": 690
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.7390089172206594e-05,
+      "loss": 1.1377,
+      "step": 695
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.7337109473884558e-05,
+      "loss": 1.1366,
+      "step": 700
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.728367989142916e-05,
+      "loss": 1.1217,
+      "step": 705
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.722980370094273e-05,
+      "loss": 1.142,
+      "step": 710
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.7175484205911922e-05,
+      "loss": 1.1475,
+      "step": 715
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.7120724737005176e-05,
+      "loss": 1.1322,
+      "step": 720
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.706552865186848e-05,
+      "loss": 1.1392,
+      "step": 725
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.7009899334919506e-05,
+      "loss": 1.1519,
+      "step": 730
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.695384019714007e-05,
+      "loss": 1.1396,
+      "step": 735
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.689735467586701e-05,
+      "loss": 1.1561,
+      "step": 740
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.1560195684432983,
+      "eval_runtime": 298.5995,
+      "eval_samples_per_second": 13.419,
+      "eval_steps_per_second": 0.211,
+      "step": 743
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.909651722005937e-05,
+      "loss": 1.0557,
+      "step": 745
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.9074695017537226e-05,
+      "loss": 1.0677,
+      "step": 750
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9052625184621195e-05,
+      "loss": 1.0752,
+      "step": 755
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9030308323553167e-05,
+      "loss": 1.0783,
+      "step": 760
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.9007745043315927e-05,
+      "loss": 1.0809,
+      "step": 765
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.898493595961656e-05,
+      "loss": 1.065,
+      "step": 770
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.8961881694869628e-05,
+      "loss": 1.0763,
+      "step": 775
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.8938582878180208e-05,
+      "loss": 1.0592,
+      "step": 780
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.8915040145326707e-05,
+      "loss": 1.0468,
+      "step": 785
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.8891254138743516e-05,
+      "loss": 1.0455,
+      "step": 790
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.8867225507503497e-05,
+      "loss": 1.0601,
+      "step": 795
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.8842954907300236e-05,
+      "loss": 1.0993,
+      "step": 800
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.881844300043019e-05,
+      "loss": 1.0603,
+      "step": 805
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.879369045577459e-05,
+      "loss": 1.0636,
+      "step": 810
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.876869794878118e-05,
+      "loss": 1.1158,
+      "step": 815
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.8743466161445823e-05,
+      "loss": 1.0916,
+      "step": 820
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.871799578229385e-05,
+      "loss": 1.041,
+      "step": 825
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.869228750636129e-05,
+      "loss": 1.0784,
+      "step": 830
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.8666342035175912e-05,
+      "loss": 1.0706,
+      "step": 835
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.8640160076738063e-05,
+      "loss": 1.0649,
+      "step": 840
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.861374234550136e-05,
+      "loss": 1.0736,
+      "step": 845
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.8587089562353197e-05,
+      "loss": 1.0509,
+      "step": 850
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.8560202454595052e-05,
+      "loss": 1.0961,
+      "step": 855
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.8533081755922673e-05,
+      "loss": 1.0487,
+      "step": 860
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.850572820640603e-05,
+      "loss": 1.0896,
+      "step": 865
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.8478142552469135e-05,
+      "loss": 1.0771,
+      "step": 870
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.8450325546869662e-05,
+      "loss": 1.0476,
+      "step": 875
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.842227794867842e-05,
+      "loss": 1.0773,
+      "step": 880
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.839400052325863e-05,
+      "loss": 1.0997,
+      "step": 885
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.8365494042245032e-05,
+      "loss": 1.0832,
+      "step": 890
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.833675928352285e-05,
+      "loss": 1.07,
+      "step": 895
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8307797031206543e-05,
+      "loss": 1.0794,
+      "step": 900
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8278608075618433e-05,
+      "loss": 1.0934,
+      "step": 905
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.8249193213267103e-05,
+      "loss": 1.0513,
+      "step": 910
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.82195532468257e-05,
+      "loss": 1.12,
+      "step": 915
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.8189688985110007e-05,
+      "loss": 1.0765,
+      "step": 920
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.815960124305638e-05,
+      "loss": 1.0554,
+      "step": 925
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.8129290841699508e-05,
+      "loss": 1.0726,
+      "step": 930
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.8098758608150017e-05,
+      "loss": 1.0272,
+      "step": 935
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.806800537557188e-05,
+      "loss": 1.0755,
+      "step": 940
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.8037031983159702e-05,
+      "loss": 1.0735,
+      "step": 945
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.8005839276115817e-05,
+      "loss": 1.0892,
+      "step": 950
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.797442810562721e-05,
+      "loss": 1.0842,
+      "step": 955
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.7942799328842308e-05,
+      "loss": 1.0767,
+      "step": 960
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.791095380884757e-05,
+      "loss": 1.0623,
+      "step": 965
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.7878892414643963e-05,
+      "loss": 1.0484,
+      "step": 970
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.7846616021123216e-05,
+      "loss": 1.0691,
+      "step": 975
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.781412550904397e-05,
+      "loss": 1.0957,
+      "step": 980
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.778142176500774e-05,
+      "loss": 1.0709,
+      "step": 985
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.77485056814347e-05,
+      "loss": 1.0784,
+      "step": 990
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.7715378156539376e-05,
+      "loss": 1.0557,
+      "step": 995
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.7682040094306077e-05,
+      "loss": 1.0809,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.7648492404464284e-05,
+      "loss": 1.0706,
+      "step": 1005
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.761473600246378e-05,
+      "loss": 1.0717,
+      "step": 1010
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.7580771809449696e-05,
+      "loss": 1.0617,
+      "step": 1015
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.754660075223737e-05,
+      "loss": 1.0623,
+      "step": 1020
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.751222376328704e-05,
+      "loss": 1.0665,
+      "step": 1025
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.7477641780678435e-05,
+      "loss": 1.064,
+      "step": 1030
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.7442855748085125e-05,
+      "loss": 1.0715,
+      "step": 1035
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.7407866614748825e-05,
+      "loss": 1.0547,
+      "step": 1040
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.7372675335453445e-05,
+      "loss": 1.0481,
+      "step": 1045
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.733728287049907e-05,
+      "loss": 1.0567,
+      "step": 1050
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.7301690185675734e-05,
+      "loss": 1.0519,
+      "step": 1055
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.7265898252237075e-05,
+      "loss": 1.0468,
+      "step": 1060
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.722990804687384e-05,
+      "loss": 1.0498,
+      "step": 1065
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.7193720551687205e-05,
+      "loss": 1.0638,
+      "step": 1070
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.7157336754162003e-05,
+      "loss": 1.06,
+      "step": 1075
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.712075764713977e-05,
+      "loss": 1.0448,
+      "step": 1080
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.708398422879165e-05,
+      "loss": 1.069,
+      "step": 1085
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.7047017502591147e-05,
+      "loss": 1.0614,
+      "step": 1090
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.7009858477286762e-05,
+      "loss": 1.0555,
+      "step": 1095
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.6972508166874455e-05,
+      "loss": 1.0565,
+      "step": 1100
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.6934967590569974e-05,
+      "loss": 1.0523,
+      "step": 1105
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6897237772781046e-05,
+      "loss": 1.0827,
+      "step": 1110
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.1386452913284302,
+      "eval_runtime": 270.1529,
+      "eval_samples_per_second": 14.832,
+      "eval_steps_per_second": 0.233,
+      "step": 1114
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6859319743079418e-05,
+      "loss": 1.0631,
+      "step": 1115
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6821214536172776e-05,
+      "loss": 1.0932,
+      "step": 1120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.678292319187649e-05,
+      "loss": 1.0894,
+      "step": 1125
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6744446755085262e-05,
+      "loss": 1.0334,
+      "step": 1130
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6705786275744583e-05,
+      "loss": 1.0676,
+      "step": 1135
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6666942808822108e-05,
+      "loss": 1.0437,
+      "step": 1140
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.6627917414278868e-05,
+      "loss": 1.0324,
+      "step": 1145
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.658871115704032e-05,
+      "loss": 1.0557,
+      "step": 1150
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.6549325106967325e-05,
+      "loss": 1.0512,
+      "step": 1155
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.650976033882691e-05,
+      "loss": 1.0486,
+      "step": 1160
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.6470017932262984e-05,
+      "loss": 1.0355,
+      "step": 1165
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.6430098971766838e-05,
+      "loss": 1.0232,
+      "step": 1170
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.6390004546647585e-05,
+      "loss": 1.0065,
+      "step": 1175
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.6349735751002406e-05,
+      "loss": 1.0284,
+      "step": 1180
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.6309293683686717e-05,
+      "loss": 1.0459,
+      "step": 1185
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.6268679448284166e-05,
+      "loss": 1.0362,
+      "step": 1190
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.6227894153076527e-05,
+      "loss": 1.0292,
+      "step": 1195
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.6186938911013462e-05,
+      "loss": 1.0333,
+      "step": 1200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.614581483968214e-05,
+      "loss": 1.0149,
+      "step": 1205
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.610452306127675e-05,
+      "loss": 1.0859,
+      "step": 1210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.606306470256786e-05,
+      "loss": 1.0579,
+      "step": 1215
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.60214408948717e-05,
+      "loss": 1.0825,
+      "step": 1220
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.5979652774019267e-05,
+      "loss": 1.0193,
+      "step": 1225
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.593770148032534e-05,
+      "loss": 1.0099,
+      "step": 1230
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.5895588158557356e-05,
+      "loss": 1.0205,
+      "step": 1235
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.5853313957904178e-05,
+      "loss": 1.0181,
+      "step": 1240
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.5810880031944743e-05,
+      "loss": 1.0367,
+      "step": 1245
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.5768287538616553e-05,
+      "loss": 1.0345,
+      "step": 1250
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.5725537640184122e-05,
+      "loss": 1.0709,
+      "step": 1255
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.568263150320722e-05,
+      "loss": 1.0312,
+      "step": 1260
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.5639570298509067e-05,
+      "loss": 1.0226,
+      "step": 1265
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.5596355201144367e-05,
+      "loss": 1.0455,
+      "step": 1270
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.5552987390367246e-05,
+      "loss": 1.0069,
+      "step": 1275
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.550946804959909e-05,
+      "loss": 1.0713,
+      "step": 1280
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.5465798366396215e-05,
+      "loss": 1.036,
+      "step": 1285
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.5421979532417502e-05,
+      "loss": 1.0206,
+      "step": 1290
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.5378012743391856e-05,
+      "loss": 1.0739,
+      "step": 1295
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.5333899199085577e-05,
+      "loss": 1.013,
+      "step": 1300
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.5289640103269626e-05,
+      "loss": 1.0229,
+      "step": 1305
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.524523666368678e-05,
+      "loss": 1.0486,
+      "step": 1310
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.520069009201866e-05,
+      "loss": 1.0147,
+      "step": 1315
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.5156001603852693e-05,
+      "loss": 1.0681,
+      "step": 1320
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.511117241864891e-05,
+      "loss": 1.0338,
+      "step": 1325
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.5066203759706696e-05,
+      "loss": 1.0418,
+      "step": 1330
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.5021096854131387e-05,
+      "loss": 1.0372,
+      "step": 1335
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.4975852932800805e-05,
+      "loss": 1.0394,
+      "step": 1340
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.4930473230331652e-05,
+      "loss": 1.0371,
+      "step": 1345
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.4884958985045835e-05,
+      "loss": 1.0233,
+      "step": 1350
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.4839311438936653e-05,
+      "loss": 1.0207,
+      "step": 1355
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.4793531837634935e-05,
+      "loss": 1.054,
+      "step": 1360
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.4747621430375032e-05,
+      "loss": 1.0572,
+      "step": 1365
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.4701581469960718e-05,
+      "loss": 1.0538,
+      "step": 1370
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.4655413212731025e-05,
+      "loss": 1.0613,
+      "step": 1375
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.4609117918525945e-05,
+      "loss": 1.0333,
+      "step": 1380
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.4562696850652068e-05,
+      "loss": 1.0556,
+      "step": 1385
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.4516151275848078e-05,
+      "loss": 1.0721,
+      "step": 1390
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.4469482464250222e-05,
+      "loss": 0.9882,
+      "step": 1395
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.4422691689357632e-05,
+      "loss": 1.0409,
+      "step": 1400
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.4375780227997565e-05,
+      "loss": 1.0527,
+      "step": 1405
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.4328749360290587e-05,
+      "loss": 1.0248,
+      "step": 1410
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.428160036961561e-05,
+      "loss": 1.0534,
+      "step": 1415
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.4234334542574906e-05,
+      "loss": 0.9941,
+      "step": 1420
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.418695316895896e-05,
+      "loss": 1.053,
+      "step": 1425
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.4139457541711303e-05,
+      "loss": 1.0324,
+      "step": 1430
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.4091848956893223e-05,
+      "loss": 1.0279,
+      "step": 1435
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.4044128713648392e-05,
+      "loss": 1.0278,
+      "step": 1440
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.3996298114167422e-05,
+      "loss": 1.0633,
+      "step": 1445
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.3948358463652317e-05,
+      "loss": 1.0154,
+      "step": 1450
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.3900311070280878e-05,
+      "loss": 1.0135,
+      "step": 1455
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.3852157245170991e-05,
+      "loss": 1.034,
+      "step": 1460
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3803898302344854e-05,
+      "loss": 1.0241,
+      "step": 1465
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3755535558693111e-05,
+      "loss": 1.0702,
+      "step": 1470
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3707070333938939e-05,
+      "loss": 1.038,
+      "step": 1475
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3658503950602007e-05,
+      "loss": 1.0253,
+      "step": 1480
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3609837733962405e-05,
+      "loss": 1.045,
+      "step": 1485
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3561073012024477e-05,
+      "loss": 1.0279,
+      "step": 1490
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3512211115480578e-05,
+      "loss": 1.0166,
+      "step": 1495
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.3463253377674774e-05,
+      "loss": 1.0184,
+      "step": 1500
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.3414201134566434e-05,
+      "loss": 1.0514,
+      "step": 1505
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.3365055724693795e-05,
+      "loss": 1.0346,
+      "step": 1510
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.3315818489137442e-05,
+      "loss": 1.0531,
+      "step": 1515
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.3266490771483675e-05,
+      "loss": 1.0278,
+      "step": 1520
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3217073917787897e-05,
+      "loss": 1.0558,
+      "step": 1525
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3167569276537839e-05,
+      "loss": 1.0481,
+      "step": 1530
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3117978198616787e-05,
+      "loss": 1.0687,
+      "step": 1535
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.306830203726671e-05,
+      "loss": 1.0741,
+      "step": 1540
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3018542148051336e-05,
+      "loss": 1.0073,
+      "step": 1545
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.2968699888819158e-05,
+      "loss": 1.0441,
+      "step": 1550
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2918776619666385e-05,
+      "loss": 1.004,
+      "step": 1555
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2868773702899828e-05,
+      "loss": 0.9965,
+      "step": 1560
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2818692502999717e-05,
+      "loss": 1.041,
+      "step": 1565
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2768534386582476e-05,
+      "loss": 1.0049,
+      "step": 1570
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2718300722363431e-05,
+      "loss": 1.0336,
+      "step": 1575
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.2667992881119452e-05,
+      "loss": 1.0232,
+      "step": 1580
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.2617612235651554e-05,
+      "loss": 0.9965,
+      "step": 1585
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.2567160160747442e-05,
+      "loss": 1.0322,
+      "step": 1590
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.251663803314398e-05,
+      "loss": 1.0388,
+      "step": 1595
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.2466047231489626e-05,
+      "loss": 0.9872,
+      "step": 1600
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.2415389136306835e-05,
+      "loss": 1.0115,
+      "step": 1605
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.2364665129954347e-05,
+      "loss": 1.0314,
+      "step": 1610
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.2313876596589505e-05,
+      "loss": 1.0515,
+      "step": 1615
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.2263024922130455e-05,
+      "loss": 1.0165,
+      "step": 1620
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.2212111494218339e-05,
+      "loss": 1.0365,
+      "step": 1625
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.2161137702179435e-05,
+      "loss": 1.0008,
+      "step": 1630
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.2110104936987228e-05,
+      "loss": 1.0264,
+      "step": 1635
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.205901459122447e-05,
+      "loss": 1.0287,
+      "step": 1640
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.2007868059045169e-05,
+      "loss": 0.9973,
+      "step": 1645
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.1956666736136557e-05,
+      "loss": 0.9963,
+      "step": 1650
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.1905412019680982e-05,
+      "loss": 1.0434,
+      "step": 1655
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.1854105308317805e-05,
+      "loss": 1.0148,
+      "step": 1660
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.1802748002105229e-05,
+      "loss": 1.0251,
+      "step": 1665
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.1751341502482071e-05,
+      "loss": 1.0049,
+      "step": 1670
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.1699887212229564e-05,
+      "loss": 1.0155,
+      "step": 1675
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.1648386535433035e-05,
+      "loss": 0.9966,
+      "step": 1680
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.1596840877443609e-05,
+      "loss": 1.0214,
+      "step": 1685
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.1545251644839866e-05,
+      "loss": 1.0431,
+      "step": 1690
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.149362024538944e-05,
+      "loss": 1.0451,
+      "step": 1695
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.1441948088010627e-05,
+      "loss": 1.0332,
+      "step": 1700
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.1390236582733917e-05,
+      "loss": 1.033,
+      "step": 1705
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.1338487140663532e-05,
+      "loss": 1.0366,
+      "step": 1710
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.1286701173938906e-05,
+      "loss": 1.0033,
+      "step": 1715
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.1234880095696161e-05,
+      "loss": 1.0273,
+      "step": 1720
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.1183025320029545e-05,
+      "loss": 1.0492,
+      "step": 1725
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.1131138261952845e-05,
+      "loss": 1.0282,
+      "step": 1730
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.1079220337360762e-05,
+      "loss": 1.0362,
+      "step": 1735
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.1027272962990286e-05,
+      "loss": 1.0302,
+      "step": 1740
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.0975297556382041e-05,
+      "loss": 1.0027,
+      "step": 1745
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.0923295535841581e-05,
+      "loss": 1.0089,
+      "step": 1750
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.087126832040071e-05,
+      "loss": 1.0065,
+      "step": 1755
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.0819217329778752e-05,
+      "loss": 1.0203,
+      "step": 1760
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.0767143984343808e-05,
+      "loss": 1.0367,
+      "step": 1765
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.0715049705073992e-05,
+      "loss": 1.0242,
+      "step": 1770
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.0662935913518662e-05,
+      "loss": 0.9876,
+      "step": 1775
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.0610804031759634e-05,
+      "loss": 0.9922,
+      "step": 1780
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.0558655482372358e-05,
+      "loss": 0.9913,
+      "step": 1785
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.0506491688387128e-05,
+      "loss": 0.9943,
+      "step": 1790
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.0454314073250216e-05,
+      "loss": 1.0171,
+      "step": 1795
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.0402124060785055e-05,
+      "loss": 1.0161,
+      "step": 1800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.034992307515337e-05,
+      "loss": 0.9986,
+      "step": 1805
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.0297712540816326e-05,
+      "loss": 1.0172,
+      "step": 1810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.0245493882495657e-05,
+      "loss": 1.0079,
+      "step": 1815
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.0193268525134776e-05,
+      "loss": 1.0202,
+      "step": 1820
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.0141037893859908e-05,
+      "loss": 1.007,
+      "step": 1825
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.0088803413941183e-05,
+      "loss": 1.0105,
+      "step": 1830
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.0036566510753764e-05,
+      "loss": 0.9913,
+      "step": 1835
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.984328609738927e-06,
+      "loss": 1.0487,
+      "step": 1840
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.932091136365193e-06,
+      "loss": 0.995,
+      "step": 1845
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.879855516089396e-06,
+      "loss": 1.055,
+      "step": 1850
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.827623174317815e-06,
+      "loss": 1.0318,
+      "step": 1855
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.775395536367265e-06,
+      "loss": 0.995,
+      "step": 1860
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.723174027426193e-06,
+      "loss": 1.0236,
+      "step": 1865
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 9.670960072515809e-06,
+      "loss": 1.0893,
+      "step": 1870
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 9.618755096451178e-06,
+      "loss": 1.0244,
+      "step": 1875
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 9.566560523802359e-06,
+      "loss": 1.0099,
+      "step": 1880
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 9.514377778855521e-06,
+      "loss": 1.0482,
+      "step": 1885
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 9.46220828557407e-06,
+      "loss": 1.0085,
+      "step": 1890
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 9.410053467559815e-06,
+      "loss": 1.0098,
+      "step": 1895
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 9.357914748014092e-06,
+      "loss": 1.0058,
+      "step": 1900
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 9.305793549698954e-06,
+      "loss": 1.008,
+      "step": 1905
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 9.253691294898322e-06,
+      "loss": 0.9987,
+      "step": 1910
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 9.201609405379196e-06,
+      "loss": 0.9889,
+      "step": 1915
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 9.14954930235284e-06,
+      "loss": 1.0021,
+      "step": 1920
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 9.09751240643602e-06,
+      "loss": 1.0085,
+      "step": 1925
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 9.045500137612213e-06,
+      "loss": 0.9855,
+      "step": 1930
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.993513915192875e-06,
+      "loss": 1.0406,
+      "step": 1935
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.941555157778717e-06,
+      "loss": 0.9954,
+      "step": 1940
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.889625283220964e-06,
+      "loss": 1.0048,
+      "step": 1945
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.837725708582712e-06,
+      "loss": 0.9826,
+      "step": 1950
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.785857850100212e-06,
+      "loss": 1.0078,
+      "step": 1955
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.734023123144253e-06,
+      "loss": 1.0104,
+      "step": 1960
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.682222942181523e-06,
+      "loss": 0.9944,
+      "step": 1965
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.630458720736028e-06,
+      "loss": 1.0247,
+      "step": 1970
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.57873187135051e-06,
+      "loss": 1.0013,
+      "step": 1975
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.527043805547888e-06,
+      "loss": 1.0306,
+      "step": 1980
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.47539593379277e-06,
+      "loss": 1.0086,
+      "step": 1985
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.423789665452941e-06,
+      "loss": 0.9972,
+      "step": 1990
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.372226408760911e-06,
+      "loss": 1.0358,
+      "step": 1995
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 8.320707570775492e-06,
+      "loss": 0.9843,
+      "step": 2000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 8.269234557343388e-06,
+      "loss": 1.0006,
+      "step": 2005
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 8.217808773060858e-06,
+      "loss": 1.0084,
+      "step": 2010
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 8.166431621235356e-06,
+      "loss": 1.0065,
+      "step": 2015
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 8.115104503847266e-06,
+      "loss": 0.9779,
+      "step": 2020
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 8.063828821511618e-06,
+      "loss": 0.9894,
+      "step": 2025
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 8.012605973439891e-06,
+      "loss": 0.9662,
+      "step": 2030
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.961437357401818e-06,
+      "loss": 0.9949,
+      "step": 2035
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.91032436968725e-06,
+      "loss": 0.9958,
+      "step": 2040
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.859268405068052e-06,
+      "loss": 0.9732,
+      "step": 2045
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.808270856760037e-06,
+      "loss": 1.0233,
+      "step": 2050
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.757333116384961e-06,
+      "loss": 0.983,
+      "step": 2055
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.70645657393253e-06,
+      "loss": 0.9974,
+      "step": 2060
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.655642617722489e-06,
+      "loss": 0.9945,
+      "step": 2065
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.604892634366727e-06,
+      "loss": 1.0259,
+      "step": 2070
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.554208008731442e-06,
+      "loss": 0.9975,
+      "step": 2075
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.503590123899343e-06,
+      "loss": 0.9827,
+      "step": 2080
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.4530403611319235e-06,
+      "loss": 1.0423,
+      "step": 2085
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.402560099831754e-06,
+      "loss": 1.0176,
+      "step": 2090
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.352150717504852e-06,
+      "loss": 1.0119,
+      "step": 2095
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.301813589723093e-06,
+      "loss": 0.9805,
+      "step": 2100
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.251550090086657e-06,
+      "loss": 1.0459,
+      "step": 2105
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.201361590186573e-06,
+      "loss": 1.0295,
+      "step": 2110
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.151249459567264e-06,
+      "loss": 0.9875,
+      "step": 2115
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.101215065689192e-06,
+      "loss": 0.9925,
+      "step": 2120
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.051259773891543e-06,
+      "loss": 1.0416,
+      "step": 2125
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.001384947354959e-06,
+      "loss": 1.0123,
+      "step": 2130
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.95159194706434e-06,
+      "loss": 0.9807,
+      "step": 2135
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.901882131771721e-06,
+      "loss": 1.0175,
+      "step": 2140
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.852256857959178e-06,
+      "loss": 0.9718,
+      "step": 2145
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.802717479801815e-06,
+      "loss": 1.0055,
+      "step": 2150
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.753265349130822e-06,
+      "loss": 1.0187,
+      "step": 2155
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.70390181539657e-06,
+      "loss": 1.0281,
+      "step": 2160
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.654628225631801e-06,
+      "loss": 1.0635,
+      "step": 2165
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.605445924414867e-06,
+      "loss": 0.9939,
+      "step": 2170
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.55635625383303e-06,
+      "loss": 0.9986,
+      "step": 2175
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.507360553445849e-06,
+      "loss": 0.9905,
+      "step": 2180
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.458460160248627e-06,
+      "loss": 1.0186,
+      "step": 2185
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.40965640863592e-06,
+      "loss": 0.9853,
+      "step": 2190
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.360950630365126e-06,
+      "loss": 1.0336,
+      "step": 2195
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.312344154520147e-06,
+      "loss": 1.0262,
+      "step": 2200
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.263838307475115e-06,
+      "loss": 0.9972,
+      "step": 2205
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.21543441285821e-06,
+      "loss": 0.993,
+      "step": 2210
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.167133791515527e-06,
+      "loss": 1.0201,
+      "step": 2215
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.118937761475038e-06,
+      "loss": 1.0386,
+      "step": 2220
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.070847637910632e-06,
+      "loss": 1.0138,
+      "step": 2225
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.0825759172439575,
+      "eval_runtime": 269.9437,
+      "eval_samples_per_second": 14.844,
+      "eval_steps_per_second": 0.233,
+      "step": 2228
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.022864733106213e-06,
+      "loss": 0.9617,
+      "step": 2230
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.974990356419902e-06,
+      "loss": 0.9457,
+      "step": 2235
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.927225814248303e-06,
+      "loss": 0.9495,
+      "step": 2240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.8795724099908614e-06,
+      "loss": 0.9459,
+      "step": 2245
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.832031444014274e-06,
+      "loss": 0.9169,
+      "step": 2250
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.784604213617038e-06,
+      "loss": 0.8853,
+      "step": 2255
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.7372920129940155e-06,
+      "loss": 0.9089,
+      "step": 2260
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.6900961332011485e-06,
+      "loss": 0.888,
+      "step": 2265
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.643017862120208e-06,
+      "loss": 0.9162,
+      "step": 2270
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.5960584844236565e-06,
+      "loss": 0.8906,
+      "step": 2275
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.549219281539601e-06,
+      "loss": 0.9347,
+      "step": 2280
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.502501531616799e-06,
+      "loss": 0.9007,
+      "step": 2285
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.455906509489817e-06,
+      "loss": 0.9515,
+      "step": 2290
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.409435486644208e-06,
+      "loss": 0.9317,
+      "step": 2295
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.363089731181852e-06,
+      "loss": 0.8994,
+      "step": 2300
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.316870507786305e-06,
+      "loss": 0.9224,
+      "step": 2305
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.270779077688332e-06,
+      "loss": 0.9042,
+      "step": 2310
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.224816698631468e-06,
+      "loss": 0.9196,
+      "step": 2315
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.1789846248376995e-06,
+      "loss": 0.8837,
+      "step": 2320
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.133284106973251e-06,
+      "loss": 0.916,
+      "step": 2325
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.087716392114428e-06,
+      "loss": 0.9439,
+      "step": 2330
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.042282723713623e-06,
+      "loss": 0.9258,
+      "step": 2335
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.996984341565361e-06,
+      "loss": 0.9729,
+      "step": 2340
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.951822481772473e-06,
+      "loss": 0.9285,
+      "step": 2345
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.9067983767123736e-06,
+      "loss": 0.9289,
+      "step": 2350
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.861913255003405e-06,
+      "loss": 0.8967,
+      "step": 2355
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.817168341471357e-06,
+      "loss": 0.9569,
+      "step": 2360
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.772564857116002e-06,
+      "loss": 0.9102,
+      "step": 2365
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.728104019077784e-06,
+      "loss": 0.9083,
+      "step": 2370
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.683787040604626e-06,
+      "loss": 0.9257,
+      "step": 2375
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.6396151310188045e-06,
+      "loss": 0.9114,
+      "step": 2380
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.595589495683952e-06,
+      "loss": 0.9103,
+      "step": 2385
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.551711335972172e-06,
+      "loss": 0.9248,
+      "step": 2390
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.507981849231251e-06,
+      "loss": 0.9305,
+      "step": 2395
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.464402228751976e-06,
+      "loss": 0.9022,
+      "step": 2400
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 4.420973663735593e-06,
+      "loss": 0.9529,
+      "step": 2405
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 4.377697339261341e-06,
+      "loss": 0.9035,
+      "step": 2410
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 4.334574436254115e-06,
+      "loss": 0.8879,
+      "step": 2415
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 4.291606131452247e-06,
+      "loss": 0.9072,
+      "step": 2420
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 4.248793597375387e-06,
+      "loss": 0.8995,
+      "step": 2425
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 4.206138002292513e-06,
+      "loss": 0.8912,
+      "step": 2430
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 4.163640510190056e-06,
+      "loss": 0.9425,
+      "step": 2435
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 4.121302280740115e-06,
+      "loss": 0.8785,
+      "step": 2440
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 4.0791244692688425e-06,
+      "loss": 0.9219,
+      "step": 2445
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 4.037108226724898e-06,
+      "loss": 0.9573,
+      "step": 2450
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.995254699648047e-06,
+      "loss": 0.9308,
+      "step": 2455
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.953565030137876e-06,
+      "loss": 0.9299,
+      "step": 2460
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.912040355822611e-06,
+      "loss": 0.9243,
+      "step": 2465
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.870681809828103e-06,
+      "loss": 0.9235,
+      "step": 2470
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.8294905207468805e-06,
+      "loss": 0.8999,
+      "step": 2475
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.788467612607378e-06,
+      "loss": 0.9218,
+      "step": 2480
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.7476142048432283e-06,
+      "loss": 0.9287,
+      "step": 2485
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.706931412262749e-06,
+      "loss": 0.905,
+      "step": 2490
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.6664203450185045e-06,
+      "loss": 0.9664,
+      "step": 2495
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.626082108577017e-06,
+      "loss": 0.9128,
+      "step": 2500
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.585917803688603e-06,
+      "loss": 0.9228,
+      "step": 2505
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.5459285263573227e-06,
+      "loss": 0.9113,
+      "step": 2510
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.5061153678110905e-06,
+      "loss": 0.9084,
+      "step": 2515
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.466479414471887e-06,
+      "loss": 0.8885,
+      "step": 2520
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 3.4270217479261147e-06,
+      "loss": 0.9083,
+      "step": 2525
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 3.387743444895084e-06,
+      "loss": 0.9217,
+      "step": 2530
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.3486455772056304e-06,
+      "loss": 0.9464,
+      "step": 2535
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.3097292117608694e-06,
+      "loss": 0.9189,
+      "step": 2540
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.2709954105110787e-06,
+      "loss": 0.9169,
+      "step": 2545
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 3.2324452304247266e-06,
+      "loss": 0.9018,
+      "step": 2550
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 3.1940797234596156e-06,
+      "loss": 0.8962,
+      "step": 2555
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 3.1558999365341927e-06,
+      "loss": 0.8915,
+      "step": 2560
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 3.1179069114989736e-06,
+      "loss": 0.9441,
+      "step": 2565
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 3.0801016851081124e-06,
+      "loss": 0.9303,
+      "step": 2570
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 3.0424852889911138e-06,
+      "loss": 0.9336,
+      "step": 2575
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 3.0050587496246707e-06,
+      "loss": 0.9121,
+      "step": 2580
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.967823088304671e-06,
+      "loss": 0.9122,
+      "step": 2585
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.9307793211183168e-06,
+      "loss": 0.9178,
+      "step": 2590
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.8939284589164096e-06,
+      "loss": 0.907,
+      "step": 2595
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.857271507285739e-06,
+      "loss": 0.8861,
+      "step": 2600
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.8208094665216757e-06,
+      "loss": 0.9148,
+      "step": 2605
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.784543331600854e-06,
+      "loss": 0.9317,
+      "step": 2610
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.7484740921540288e-06,
+      "loss": 0.8886,
+      "step": 2615
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.7126027324390712e-06,
+      "loss": 0.9177,
+      "step": 2620
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.6769302313140966e-06,
+      "loss": 0.8915,
+      "step": 2625
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.6414575622107787e-06,
+      "loss": 0.92,
+      "step": 2630
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.6061856931077644e-06,
+      "loss": 0.8871,
+      "step": 2635
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.5711155865042703e-06,
+      "loss": 0.8972,
+      "step": 2640
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.536248199393817e-06,
+      "loss": 0.9284,
+      "step": 2645
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.501584483238102e-06,
+      "loss": 0.8995,
+      "step": 2650
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.467125383941066e-06,
+      "loss": 0.9184,
+      "step": 2655
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.432871841823047e-06,
+      "loss": 0.9183,
+      "step": 2660
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.3988247915951445e-06,
+      "loss": 0.9288,
+      "step": 2665
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.364985162333697e-06,
+      "loss": 0.8974,
+      "step": 2670
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.3313538774549428e-06,
+      "loss": 0.914,
+      "step": 2675
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.2979318546898145e-06,
+      "loss": 0.9057,
+      "step": 2680
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.264720006058898e-06,
+      "loss": 0.9529,
+      "step": 2685
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.231719237847546e-06,
+      "loss": 0.9345,
+      "step": 2690
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.198930450581139e-06,
+      "loss": 0.9018,
+      "step": 2695
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.166354539000528e-06,
+      "loss": 0.8782,
+      "step": 2700
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.1339923920376017e-06,
+      "loss": 0.9086,
+      "step": 2705
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.1018448927910497e-06,
+      "loss": 0.9109,
+      "step": 2710
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.0699129185022324e-06,
+      "loss": 0.9217,
+      "step": 2715
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.0381973405312806e-06,
+      "loss": 0.9157,
+      "step": 2720
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.0066990243332885e-06,
+      "loss": 0.8888,
+      "step": 2725
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.975418829434713e-06,
+      "loss": 0.963,
+      "step": 2730
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.944357609409917e-06,
+      "loss": 0.8953,
+      "step": 2735
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.9135162118578644e-06,
+      "loss": 0.9401,
+      "step": 2740
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.8828954783790088e-06,
+      "loss": 0.9246,
+      "step": 2745
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.8524962445523176e-06,
+      "loss": 0.9401,
+      "step": 2750
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.822319339912474e-06,
+      "loss": 0.8813,
+      "step": 2755
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.7923655879272395e-06,
+      "loss": 0.8964,
+      "step": 2760
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.7626358059749748e-06,
+      "loss": 0.9069,
+      "step": 2765
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.7331308053223594e-06,
+      "loss": 0.9092,
+      "step": 2770
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.703851391102226e-06,
+      "loss": 0.9131,
+      "step": 2775
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.6747983622916053e-06,
+      "loss": 0.9142,
+      "step": 2780
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.6459725116899184e-06,
+      "loss": 0.9299,
+      "step": 2785
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.6173746258973443e-06,
+      "loss": 0.8959,
+      "step": 2790
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.5890054852933602e-06,
+      "loss": 0.9397,
+      "step": 2795
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.5608658640154361e-06,
+      "loss": 0.9297,
+      "step": 2800
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.5329565299379223e-06,
+      "loss": 0.8527,
+      "step": 2805
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.5052782446510772e-06,
+      "loss": 0.9119,
+      "step": 2810
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.4778317634403082e-06,
+      "loss": 0.9469,
+      "step": 2815
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.4506178352655475e-06,
+      "loss": 0.873,
+      "step": 2820
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.423637202740813e-06,
+      "loss": 0.9326,
+      "step": 2825
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.3968906021139517e-06,
+      "loss": 0.9292,
+      "step": 2830
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.3703787632465459e-06,
+      "loss": 0.8877,
+      "step": 2835
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.3441024095939913e-06,
+      "loss": 0.9064,
+      "step": 2840
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.318062258185765e-06,
+      "loss": 0.8927,
+      "step": 2845
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.2922590196058539e-06,
+      "loss": 0.903,
+      "step": 2850
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.2666933979733575e-06,
+      "loss": 0.911,
+      "step": 2855
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.2413660909232893e-06,
+      "loss": 0.9241,
+      "step": 2860
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.21627778958753e-06,
+      "loss": 0.9215,
+      "step": 2865
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.1914291785759658e-06,
+      "loss": 0.9333,
+      "step": 2870
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.1668209359578153e-06,
+      "loss": 0.9004,
+      "step": 2875
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.1424537332431119e-06,
+      "loss": 0.8789,
+      "step": 2880
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.118328235364401e-06,
+      "loss": 0.9247,
+      "step": 2885
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.0944451006585766e-06,
+      "loss": 0.9213,
+      "step": 2890
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.0708049808489263e-06,
+      "loss": 0.9284,
+      "step": 2895
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.0474085210273378e-06,
+      "loss": 0.8989,
+      "step": 2900
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.0242563596367094e-06,
+      "loss": 0.9243,
+      "step": 2905
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.001349128453516e-06,
+      "loss": 0.8905,
+      "step": 2910
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 9.78687452570577e-07,
+      "loss": 0.9108,
+      "step": 2915
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 9.562719503799967e-07,
+      "loss": 0.8905,
+      "step": 2920
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.341032335562805e-07,
+      "loss": 0.9233,
+      "step": 2925
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.121819070396598e-07,
+      "loss": 0.9552,
+      "step": 2930
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.905085690195725e-07,
+      "loss": 0.8892,
+      "step": 2935
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 8.690838109183463e-07,
+      "loss": 0.9288,
+      "step": 2940
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 8.47908217375053e-07,
+      "loss": 0.9249,
+      "step": 2945
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 8.269823662295596e-07,
+      "loss": 0.9242,
+      "step": 2950
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 8.063068285067632e-07,
+      "loss": 0.8873,
+      "step": 2955
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 7.858821684009987e-07,
+      "loss": 0.9275,
+      "step": 2960
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 7.657089432606568e-07,
+      "loss": 0.9259,
+      "step": 2965
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 7.457877035729588e-07,
+      "loss": 0.9342,
+      "step": 2970
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 7.261189929489509e-07,
+      "loss": 0.9211,
+      "step": 2975
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 7.067033481086605e-07,
+      "loss": 0.9178,
+      "step": 2980
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.875412988664532e-07,
+      "loss": 0.8959,
+      "step": 2985
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.686333681165735e-07,
+      "loss": 0.9205,
+      "step": 2990
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 6.499800718188776e-07,
+      "loss": 0.9074,
+      "step": 2995
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 6.315819189847517e-07,
+      "loss": 0.918,
+      "step": 3000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 6.134394116632314e-07,
+      "loss": 0.889,
+      "step": 3005
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.955530449272861e-07,
+      "loss": 0.9108,
+      "step": 3010
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.779233068603196e-07,
+      "loss": 0.9774,
+      "step": 3015
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.605506785428505e-07,
+      "loss": 0.9115,
+      "step": 3020
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.434356340393832e-07,
+      "loss": 0.8917,
+      "step": 3025
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.265786403854711e-07,
+      "loss": 0.8878,
+      "step": 3030
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.099801575749752e-07,
+      "loss": 0.9124,
+      "step": 3035
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.936406385475024e-07,
+      "loss": 0.9115,
+      "step": 3040
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.775605291760609e-07,
+      "loss": 0.922,
+      "step": 3045
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.6174026825487997e-07,
+      "loss": 0.9283,
+      "step": 3050
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.4618028748744213e-07,
+      "loss": 0.8847,
+      "step": 3055
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.3088101147470306e-07,
+      "loss": 0.9202,
+      "step": 3060
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.158428577035012e-07,
+      "loss": 0.9252,
+      "step": 3065
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.0106623653516916e-07,
+      "loss": 0.8904,
+      "step": 3070
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.865515511943363e-07,
+      "loss": 0.9257,
+      "step": 3075
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.7229919775792155e-07,
+      "loss": 0.9058,
+      "step": 3080
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.58309565144328e-07,
+      "loss": 0.8647,
+      "step": 3085
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.4458303510282787e-07,
+      "loss": 0.9449,
+      "step": 3090
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.311199822031508e-07,
+      "loss": 0.8925,
+      "step": 3095
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.1792077382525544e-07,
+      "loss": 0.895,
+      "step": 3100
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.04985770149312e-07,
+      "loss": 0.9363,
+      "step": 3105
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.9231532414586314e-07,
+      "loss": 0.9493,
+      "step": 3110
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.799097815662011e-07,
+      "loss": 0.9185,
+      "step": 3115
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.677694809329334e-07,
+      "loss": 0.9278,
+      "step": 3120
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.5589475353073987e-07,
+      "loss": 0.9356,
+      "step": 3125
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.4428592339733157e-07,
+      "loss": 0.885,
+      "step": 3130
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.329433073146148e-07,
+      "loss": 0.9391,
+      "step": 3135
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.2186721480004114e-07,
+      "loss": 0.9084,
+      "step": 3140
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.1105794809816782e-07,
+      "loss": 0.9021,
+      "step": 3145
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.0051580217240185e-07,
+      "loss": 0.9618,
+      "step": 3150
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.9024106469695547e-07,
+      "loss": 0.9278,
+      "step": 3155
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.802340160489968e-07,
+      "loss": 0.922,
+      "step": 3160
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.704949293009972e-07,
+      "loss": 0.8844,
+      "step": 3165
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.610240702132826e-07,
+      "loss": 0.9047,
+      "step": 3170
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.5182169722677497e-07,
+      "loss": 0.8983,
+      "step": 3175
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.4288806145594692e-07,
+      "loss": 0.9385,
+      "step": 3180
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.3422340668196588e-07,
+      "loss": 0.9109,
+      "step": 3185
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.2582796934604292e-07,
+      "loss": 0.9284,
+      "step": 3190
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1770197854297893e-07,
+      "loss": 0.877,
+      "step": 3195
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.0984565601491415e-07,
+      "loss": 0.9205,
+      "step": 3200
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0225921614527956e-07,
+      "loss": 0.9207,
+      "step": 3205
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 9.494286595294389e-08,
+      "loss": 0.9282,
+      "step": 3210
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 8.789680508656473e-08,
+      "loss": 0.922,
+      "step": 3215
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 8.112122581913962e-08,
+      "loss": 0.9086,
+      "step": 3220
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.461631304276462e-08,
+      "loss": 0.9127,
+      "step": 3225
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.838224426358065e-08,
+      "loss": 0.9173,
+      "step": 3230
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.241918959693838e-08,
+      "loss": 0.9251,
+      "step": 3235
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.672731176274981e-08,
+      "loss": 0.8894,
+      "step": 3240
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.1306766081048456e-08,
+      "loss": 0.8684,
+      "step": 3245
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 4.615770046775603e-08,
+      "loss": 0.8916,
+      "step": 3250
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 4.128025543064018e-08,
+      "loss": 0.9037,
+      "step": 3255
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.66745640654842e-08,
+      "loss": 0.9356,
+      "step": 3260
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.234075205245213e-08,
+      "loss": 0.9158,
+      "step": 3265
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.8278937652662653e-08,
+      "loss": 0.9056,
+      "step": 3270
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.4489231704958317e-08,
+      "loss": 0.8952,
+      "step": 3275
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.0971737622883515e-08,
+      "loss": 0.9064,
+      "step": 3280
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.77265513918623e-08,
+      "loss": 0.8911,
+      "step": 3285
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.4753761566578261e-08,
+      "loss": 0.9326,
+      "step": 3290
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.205344926855867e-08,
+      "loss": 0.9235,
+      "step": 3295
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.625688183958481e-09,
+      "loss": 0.9424,
+      "step": 3300
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 7.47054456155416e-09,
+      "loss": 0.8567,
+      "step": 3305
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.588077210932907e-09,
+      "loss": 0.8862,
+      "step": 3310
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.978337500887275e-09,
+      "loss": 0.9048,
+      "step": 3315
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.641369358016288e-09,
+      "loss": 0.9303,
+      "step": 3320
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.5772092655219618e-09,
+      "loss": 0.9105,
+      "step": 3325
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 7.858862622189823e-10,
+      "loss": 0.9435,
+      "step": 3330
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.674219417408974e-10,
+      "loss": 0.9219,
+      "step": 3335
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.183045194725786e-11,
+      "loss": 0.928,
+      "step": 3340
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.0884039402008057,
+      "eval_runtime": 162.651,
+      "eval_samples_per_second": 24.636,
+      "eval_steps_per_second": 0.387,
+      "step": 3342
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.245006253484668e-06,
+      "loss": 0.8916,
+      "step": 3345
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 8.214166787806253e-06,
+      "loss": 0.8846,
+      "step": 3350
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 8.183344856284602e-06,
+      "loss": 0.9028,
+      "step": 3355
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 8.152540761544041e-06,
+      "loss": 0.9094,
+      "step": 3360
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 8.121754806033773e-06,
+      "loss": 0.8841,
+      "step": 3365
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 8.0909872920249e-06,
+      "loss": 0.8714,
+      "step": 3370
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 8.060238521607455e-06,
+      "loss": 0.8917,
+      "step": 3375
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 8.029508796687432e-06,
+      "loss": 0.874,
+      "step": 3380
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 7.998798418983841e-06,
+      "loss": 0.8712,
+      "step": 3385
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 7.968107690025716e-06,
+      "loss": 0.8787,
+      "step": 3390
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 7.937436911149184e-06,
+      "loss": 0.8686,
+      "step": 3395
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 7.906786383494479e-06,
+      "loss": 0.8882,
+      "step": 3400
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 7.87615640800301e-06,
+      "loss": 0.9286,
+      "step": 3405
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 7.845547285414399e-06,
+      "loss": 0.8802,
+      "step": 3410
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 7.814959316263507e-06,
+      "loss": 0.8919,
+      "step": 3415
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 7.78439280087751e-06,
+      "loss": 0.8981,
+      "step": 3420
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 7.75384803937295e-06,
+      "loss": 0.9084,
+      "step": 3425
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 7.723325331652765e-06,
+      "loss": 0.9348,
+      "step": 3430
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 7.692824977403362e-06,
+      "loss": 0.8967,
+      "step": 3435
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 7.662347276091678e-06,
+      "loss": 0.9366,
+      "step": 3440
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 7.63189252696222e-06,
+      "loss": 0.889,
+      "step": 3445
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 7.601461029034155e-06,
+      "loss": 0.8943,
+      "step": 3450
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 7.571053081098346e-06,
+      "loss": 0.9559,
+      "step": 3455
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 7.540668981714434e-06,
+      "loss": 0.9159,
+      "step": 3460
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 7.510309029207914e-06,
+      "loss": 0.8611,
+      "step": 3465
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 7.4799735216671786e-06,
+      "loss": 0.8762,
+      "step": 3470
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 7.449662756940617e-06,
+      "loss": 0.8756,
+      "step": 3475
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 7.419377032633689e-06,
+      "loss": 0.9262,
+      "step": 3480
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 7.389116646105977e-06,
+      "loss": 0.8761,
+      "step": 3485
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 7.358881894468304e-06,
+      "loss": 0.8921,
+      "step": 3490
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 7.328673074579788e-06,
+      "loss": 0.9248,
+      "step": 3495
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 7.298490483044935e-06,
+      "loss": 0.8906,
+      "step": 3500
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 7.268334416210737e-06,
+      "loss": 0.9038,
+      "step": 3505
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 7.238205170163744e-06,
+      "loss": 0.9138,
+      "step": 3510
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 7.2081030407271715e-06,
+      "loss": 0.9013,
+      "step": 3515
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 7.1780283234579976e-06,
+      "loss": 0.9174,
+      "step": 3520
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 7.147981313644041e-06,
+      "loss": 0.8708,
+      "step": 3525
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.117962306301085e-06,
+      "loss": 0.9359,
+      "step": 3530
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.0879715961699746e-06,
+      "loss": 0.8862,
+      "step": 3535
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.058009477713705e-06,
+      "loss": 0.9141,
+      "step": 3540
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.028076245114565e-06,
+      "loss": 0.9017,
+      "step": 3545
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 6.998172192271209e-06,
+      "loss": 0.8628,
+      "step": 3550
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 6.968297612795798e-06,
+      "loss": 0.9472,
+      "step": 3555
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 6.938452800011119e-06,
+      "loss": 0.8825,
+      "step": 3560
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 6.908638046947679e-06,
+      "loss": 0.8804,
+      "step": 3565
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 6.878853646340853e-06,
+      "loss": 0.9433,
+      "step": 3570
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 6.8490998906280085e-06,
+      "loss": 0.9129,
+      "step": 3575
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 6.819377071945608e-06,
+      "loss": 0.8972,
+      "step": 3580
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 6.789685482126378e-06,
+      "loss": 0.914,
+      "step": 3585
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 6.760025412696419e-06,
+      "loss": 0.9269,
+      "step": 3590
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 6.730397154872341e-06,
+      "loss": 0.8669,
+      "step": 3595
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 6.700800999558428e-06,
+      "loss": 0.875,
+      "step": 3600
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 6.6712372373437514e-06,
+      "loss": 0.907,
+      "step": 3605
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 6.6417061584993394e-06,
+      "loss": 0.8929,
+      "step": 3610
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 6.612208052975326e-06,
+      "loss": 0.8968,
+      "step": 3615
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 6.58274321039808e-06,
+      "loss": 0.9157,
+      "step": 3620
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 6.553311920067393e-06,
+      "loss": 0.8825,
+      "step": 3625
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 6.523914470953625e-06,
+      "loss": 0.8943,
+      "step": 3630
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 6.494551151694854e-06,
+      "loss": 0.8973,
+      "step": 3635
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 6.465222250594071e-06,
+      "loss": 0.8858,
+      "step": 3640
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 6.435928055616324e-06,
+      "loss": 0.905,
+      "step": 3645
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 6.406668854385895e-06,
+      "loss": 0.926,
+      "step": 3650
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 6.3774449341834985e-06,
+      "loss": 0.9108,
+      "step": 3655
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 6.348256581943419e-06,
+      "loss": 0.8875,
+      "step": 3660
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 6.319104084250742e-06,
+      "loss": 0.8755,
+      "step": 3665
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 6.289987727338502e-06,
+      "loss": 0.9258,
+      "step": 3670
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 6.260907797084883e-06,
+      "loss": 0.8721,
+      "step": 3675
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 6.231864579010427e-06,
+      "loss": 0.8946,
+      "step": 3680
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 6.202858358275215e-06,
+      "loss": 0.894,
+      "step": 3685
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 6.173889419676057e-06,
+      "loss": 0.8819,
+      "step": 3690
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 6.144958047643732e-06,
+      "loss": 0.887,
+      "step": 3695
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 6.116064526240151e-06,
+      "loss": 0.8457,
+      "step": 3700
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 6.087209139155607e-06,
+      "loss": 0.8673,
+      "step": 3705
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 6.058392169705962e-06,
+      "loss": 0.886,
+      "step": 3710
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 6.029613900829876e-06,
+      "loss": 0.8601,
+      "step": 3715
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 6.000874615086034e-06,
+      "loss": 0.8498,
+      "step": 3720
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5.972174594650363e-06,
+      "loss": 0.8934,
+      "step": 3725
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5.943514121313257e-06,
+      "loss": 0.906,
+      "step": 3730
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5.914893476476834e-06,
+      "loss": 0.9065,
+      "step": 3735
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5.886312941152146e-06,
+      "loss": 0.872,
+      "step": 3740
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5.857772795956427e-06,
+      "loss": 0.8978,
+      "step": 3745
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5.829273321110356e-06,
+      "loss": 0.8698,
+      "step": 3750
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5.800814796435275e-06,
+      "loss": 0.8934,
+      "step": 3755
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5.772397501350466e-06,
+      "loss": 0.911,
+      "step": 3760
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5.7440217148704095e-06,
+      "loss": 0.8939,
+      "step": 3765
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5.715687715602011e-06,
+      "loss": 0.8998,
+      "step": 3770
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 5.687395781741905e-06,
+      "loss": 0.8859,
+      "step": 3775
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 5.659146191073715e-06,
+      "loss": 0.8924,
+      "step": 3780
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5.6309392209652924e-06,
+      "loss": 0.8709,
+      "step": 3785
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5.602775148366053e-06,
+      "loss": 0.899,
+      "step": 3790
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5.574654249804204e-06,
+      "loss": 0.905,
+      "step": 3795
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5.546576801384045e-06,
+      "loss": 0.9167,
+      "step": 3800
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5.518543078783283e-06,
+      "loss": 0.8507,
+      "step": 3805
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5.490553357250283e-06,
+      "loss": 0.8897,
+      "step": 3810
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5.4626079116013904e-06,
+      "loss": 0.8695,
+      "step": 3815
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5.434707016218251e-06,
+      "loss": 0.8604,
+      "step": 3820
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5.4068509450450555e-06,
+      "loss": 0.8791,
+      "step": 3825
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5.379039971585929e-06,
+      "loss": 0.8768,
+      "step": 3830
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5.351274368902184e-06,
+      "loss": 0.9302,
+      "step": 3835
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5.323554409609657e-06,
+      "loss": 0.8816,
+      "step": 3840
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5.2958803658760584e-06,
+      "loss": 0.8981,
+      "step": 3845
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5.26825250941825e-06,
+      "loss": 0.9107,
+      "step": 3850
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5.2406711114996245e-06,
+      "loss": 0.8965,
+      "step": 3855
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5.213136442927425e-06,
+      "loss": 0.9214,
+      "step": 3860
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5.185648774050065e-06,
+      "loss": 0.9129,
+      "step": 3865
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5.158208374754503e-06,
+      "loss": 0.825,
+      "step": 3870
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 5.130815514463595e-06,
+      "loss": 0.9019,
+      "step": 3875
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 5.103470462133411e-06,
+      "loss": 0.9306,
+      "step": 3880
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 5.076173486250642e-06,
+      "loss": 0.9126,
+      "step": 3885
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 5.048924854829934e-06,
+      "loss": 0.9109,
+      "step": 3890
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 5.021724835411252e-06,
+      "loss": 0.8636,
+      "step": 3895
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 4.994573695057292e-06,
+      "loss": 0.8564,
+      "step": 3900
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 4.967471700350804e-06,
+      "loss": 0.9006,
+      "step": 3905
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 4.940419117392017e-06,
+      "loss": 0.8914,
+      "step": 3910
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 4.913416211796019e-06,
+      "loss": 0.898,
+      "step": 3915
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 4.886463248690122e-06,
+      "loss": 0.8824,
+      "step": 3920
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 4.8595604927113e-06,
+      "loss": 0.887,
+      "step": 3925
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 4.832708208003558e-06,
+      "loss": 0.8436,
+      "step": 3930
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 4.805906658215345e-06,
+      "loss": 0.9058,
+      "step": 3935
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 4.779156106496986e-06,
+      "loss": 0.8642,
+      "step": 3940
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 4.752456815498075e-06,
+      "loss": 0.8989,
+      "step": 3945
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 4.725809047364894e-06,
+      "loss": 0.9035,
+      "step": 3950
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 4.699213063737873e-06,
+      "loss": 0.848,
+      "step": 3955
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 4.672669125748967e-06,
+      "loss": 0.8869,
+      "step": 3960
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 4.646177494019148e-06,
+      "loss": 0.8993,
+      "step": 3965
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 4.619738428655814e-06,
+      "loss": 0.8863,
+      "step": 3970
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 4.593352189250223e-06,
+      "loss": 0.9149,
+      "step": 3975
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 4.567019034874992e-06,
+      "loss": 0.8988,
+      "step": 3980
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 4.540739224081507e-06,
+      "loss": 0.8707,
+      "step": 3985
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 4.514513014897395e-06,
+      "loss": 0.8805,
+      "step": 3990
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 4.488340664824019e-06,
+      "loss": 0.8947,
+      "step": 3995
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 4.462222430833901e-06,
+      "loss": 0.8609,
+      "step": 4000
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 4.436158569368251e-06,
+      "loss": 0.9086,
+      "step": 4005
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 4.410149336334412e-06,
+      "loss": 0.9107,
+      "step": 4010
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 4.384194987103347e-06,
+      "loss": 0.9069,
+      "step": 4015
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 4.358295776507169e-06,
+      "loss": 0.8991,
+      "step": 4020
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 4.3324519588366e-06,
+      "loss": 0.9011,
+      "step": 4025
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 4.3066637878384786e-06,
+      "loss": 0.888,
+      "step": 4030
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 4.2809315167132995e-06,
+      "loss": 0.8755,
+      "step": 4035
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 4.25525539811269e-06,
+      "loss": 0.8731,
+      "step": 4040
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 4.22963568413695e-06,
+      "loss": 0.8573,
+      "step": 4045
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 4.204072626332578e-06,
+      "loss": 0.92,
+      "step": 4050
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 4.178566475689777e-06,
+      "loss": 0.8779,
+      "step": 4055
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 4.153117482640032e-06,
+      "loss": 0.8831,
+      "step": 4060
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 4.127725897053619e-06,
+      "loss": 0.8932,
+      "step": 4065
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 4.1023919682371425e-06,
+      "loss": 0.8764,
+      "step": 4070
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 4.077115944931137e-06,
+      "loss": 0.8703,
+      "step": 4075
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 4.051898075307573e-06,
+      "loss": 0.8988,
+      "step": 4080
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 4.026738606967438e-06,
+      "loss": 0.882,
+      "step": 4085
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 4.001637786938322e-06,
+      "loss": 0.88,
+      "step": 4090
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.976595861671971e-06,
+      "loss": 0.879,
+      "step": 4095
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.951613077041873e-06,
+      "loss": 0.8769,
+      "step": 4100
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.9266896783408495e-06,
+      "loss": 0.9086,
+      "step": 4105
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.90182591027863e-06,
+      "loss": 0.8886,
+      "step": 4110
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.877022016979481e-06,
+      "loss": 0.8914,
+      "step": 4115
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.852278241979782e-06,
+      "loss": 0.8912,
+      "step": 4120
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.8275948282256325e-06,
+      "loss": 0.8714,
+      "step": 4125
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 3.8029720180704954e-06,
+      "loss": 0.8594,
+      "step": 4130
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 3.7784100532727853e-06,
+      "loss": 0.9012,
+      "step": 4135
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 3.753909174993513e-06,
+      "loss": 0.8871,
+      "step": 4140
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 3.7294696237939156e-06,
+      "loss": 0.8889,
+      "step": 4145
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.705091639633078e-06,
+      "loss": 0.856,
+      "step": 4150
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.680775461865611e-06,
+      "loss": 0.9155,
+      "step": 4155
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.656521329239271e-06,
+      "loss": 0.8735,
+      "step": 4160
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 3.6323294798926167e-06,
+      "loss": 0.8754,
+      "step": 4165
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 3.608200151352699e-06,
+      "loss": 0.8701,
+      "step": 4170
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.584133580532696e-06,
+      "loss": 0.8998,
+      "step": 4175
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.560130003729605e-06,
+      "loss": 0.9001,
+      "step": 4180
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.536189656621919e-06,
+      "loss": 0.9005,
+      "step": 4185
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.512312774267309e-06,
+      "loss": 0.9006,
+      "step": 4190
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 3.488499591100324e-06,
+      "loss": 0.8993,
+      "step": 4195
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 3.464750340930082e-06,
+      "loss": 0.8992,
+      "step": 4200
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 3.4410652569379667e-06,
+      "loss": 0.8436,
+      "step": 4205
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 3.417444571675367e-06,
+      "loss": 0.8307,
+      "step": 4210
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 3.39388851706136e-06,
+      "loss": 0.9374,
+      "step": 4215
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 3.3703973243804533e-06,
+      "loss": 0.8965,
+      "step": 4220
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 3.34697122428031e-06,
+      "loss": 0.8821,
+      "step": 4225
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 3.3236104467694805e-06,
+      "loss": 0.8909,
+      "step": 4230
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 3.300315221215149e-06,
+      "loss": 0.8928,
+      "step": 4235
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 3.2770857763408816e-06,
+      "loss": 0.8993,
+      "step": 4240
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 3.2539223402243735e-06,
+      "loss": 0.8958,
+      "step": 4245
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 3.2308251402952183e-06,
+      "loss": 0.8612,
+      "step": 4250
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 3.2077944033326702e-06,
+      "loss": 0.8681,
+      "step": 4255
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 3.18483035546342e-06,
+      "loss": 0.8814,
+      "step": 4260
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 3.161933222159371e-06,
+      "loss": 0.9353,
+      "step": 4265
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 3.139103228235428e-06,
+      "loss": 0.89,
+      "step": 4270
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 3.116340597847293e-06,
+      "loss": 0.8637,
+      "step": 4275
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 3.093645554489254e-06,
+      "loss": 0.8833,
+      "step": 4280
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 3.071018320992004e-06,
+      "loss": 0.8744,
+      "step": 4285
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 3.0484591195204447e-06,
+      "loss": 0.8983,
+      "step": 4290
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 3.0259681715715094e-06,
+      "loss": 0.8347,
+      "step": 4295
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 3.003545697971972e-06,
+      "loss": 0.8931,
+      "step": 4300
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 2.981191918876317e-06,
+      "loss": 0.8751,
+      "step": 4305
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 2.9589070537645347e-06,
+      "loss": 0.8511,
+      "step": 4310
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 2.936691321439993e-06,
+      "loss": 0.8524,
+      "step": 4315
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 2.914544940027277e-06,
+      "loss": 0.8768,
+      "step": 4320
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 2.892468126970058e-06,
+      "loss": 0.9014,
+      "step": 4325
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 2.8704610990289473e-06,
+      "loss": 0.8941,
+      "step": 4330
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 2.848524072279372e-06,
+      "loss": 0.8533,
+      "step": 4335
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 2.826657262109459e-06,
+      "loss": 0.913,
+      "step": 4340
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 2.804860883217911e-06,
+      "loss": 0.8897,
+      "step": 4345
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 2.7831351496119076e-06,
+      "loss": 0.8815,
+      "step": 4350
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 2.761480274604994e-06,
+      "loss": 0.9072,
+      "step": 4355
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 2.7398964708149977e-06,
+      "loss": 0.8978,
+      "step": 4360
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 2.7183839501619304e-06,
+      "loss": 0.8789,
+      "step": 4365
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 2.696942923865915e-06,
+      "loss": 0.8813,
+      "step": 4370
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 2.6755736024451095e-06,
+      "loss": 0.859,
+      "step": 4375
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 2.6542761957136374e-06,
+      "loss": 0.8847,
+      "step": 4380
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 2.63305091277953e-06,
+      "loss": 0.8953,
+      "step": 4385
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 2.6118979620426743e-06,
+      "loss": 0.8731,
+      "step": 4390
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 2.590817551192765e-06,
+      "loss": 0.8978,
+      "step": 4395
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 2.569809887207265e-06,
+      "loss": 0.892,
+      "step": 4400
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 2.5488751763493757e-06,
+      "loss": 0.8848,
+      "step": 4405
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 2.528013624166007e-06,
+      "loss": 0.9003,
+      "step": 4410
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 2.507225435485766e-06,
+      "loss": 0.8844,
+      "step": 4415
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 2.4865108144169404e-06,
+      "loss": 0.8861,
+      "step": 4420
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 2.4658699643454953e-06,
+      "loss": 0.896,
+      "step": 4425
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 2.4453030879330788e-06,
+      "loss": 0.8928,
+      "step": 4430
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 2.4248103871150298e-06,
+      "loss": 0.9331,
+      "step": 4435
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 2.404392063098395e-06,
+      "loss": 0.8947,
+      "step": 4440
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 2.384048316359958e-06,
+      "loss": 0.8848,
+      "step": 4445
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 2.363779346644264e-06,
+      "loss": 0.8967,
+      "step": 4450
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 2.3435853529616616e-06,
+      "loss": 0.8938,
+      "step": 4455
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.0775328874588013,
+      "eval_runtime": 165.7744,
+      "eval_samples_per_second": 24.171,
+      "eval_steps_per_second": 0.507,
+      "step": 4456
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 2.3234665335863525e-06,
+      "loss": 0.862,
+      "step": 4460
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 2.3034230860544383e-06,
+      "loss": 0.8098,
+      "step": 4465
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 2.2834552071619854e-06,
+      "loss": 0.8289,
+      "step": 4470
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 2.26356309296309e-06,
+      "loss": 0.864,
+      "step": 4475
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 2.243746938767959e-06,
+      "loss": 0.8438,
+      "step": 4480
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 2.2240069391409824e-06,
+      "loss": 0.7762,
+      "step": 4485
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 2.2043432878988314e-06,
+      "loss": 0.8177,
+      "step": 4490
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 2.1847561781085535e-06,
+      "loss": 0.7772,
+      "step": 4495
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 2.1652458020856735e-06,
+      "loss": 0.8474,
+      "step": 4500
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 2.145812351392309e-06,
+      "loss": 0.8485,
+      "step": 4505
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 2.126456016835289e-06,
+      "loss": 0.8396,
+      "step": 4510
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 2.107176988464279e-06,
+      "loss": 0.811,
+      "step": 4515
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 2.087975455569915e-06,
+      "loss": 0.7833,
+      "step": 4520
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 2.0688516066819464e-06,
+      "loss": 0.8312,
+      "step": 4525
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 2.049805629567384e-06,
+      "loss": 0.8378,
+      "step": 4530
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 2.030837711228657e-06,
+      "loss": 0.8215,
+      "step": 4535
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 2.0119480379017765e-06,
+      "loss": 0.8479,
+      "step": 4540
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 1.9931367950545068e-06,
+      "loss": 0.8621,
+      "step": 4545
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 1.974404167384545e-06,
+      "loss": 0.8427,
+      "step": 4550
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 1.9557503388177057e-06,
+      "loss": 0.8022,
+      "step": 4555
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 1.937175492506118e-06,
+      "loss": 0.8118,
+      "step": 4560
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1.918679810826427e-06,
+      "loss": 0.7988,
+      "step": 4565
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1.900263475377997e-06,
+      "loss": 0.8303,
+      "step": 4570
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 1.8819266669811397e-06,
+      "loss": 0.8205,
+      "step": 4575
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 1.8636695656753278e-06,
+      "loss": 0.8022,
+      "step": 4580
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 1.845492350717435e-06,
+      "loss": 0.8677,
+      "step": 4585
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 1.827395200579971e-06,
+      "loss": 0.8406,
+      "step": 4590
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 1.8093782929493332e-06,
+      "loss": 0.8164,
+      "step": 4595
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 1.7914418047240556e-06,
+      "loss": 0.8669,
+      "step": 4600
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 1.7735859120130826e-06,
+      "loss": 0.8369,
+      "step": 4605
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.7558107901340294e-06,
+      "loss": 0.838,
+      "step": 4610
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.7381166136114646e-06,
+      "loss": 0.8319,
+      "step": 4615
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.7205035561752014e-06,
+      "loss": 0.8419,
+      "step": 4620
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.702971790758582e-06,
+      "loss": 0.7837,
+      "step": 4625
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.6855214894967886e-06,
+      "loss": 0.8214,
+      "step": 4630
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.6681528237251454e-06,
+      "loss": 0.8362,
+      "step": 4635
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.6508659639774504e-06,
+      "loss": 0.8347,
+      "step": 4640
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.6336610799842788e-06,
+      "loss": 0.831,
+      "step": 4645
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.6165383406713376e-06,
+      "loss": 0.8763,
+      "step": 4650
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.5994979141577937e-06,
+      "loss": 0.8536,
+      "step": 4655
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.5825399677546305e-06,
+      "loss": 0.8596,
+      "step": 4660
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1.5656646679630027e-06,
+      "loss": 0.7901,
+      "step": 4665
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1.5488721804726003e-06,
+      "loss": 0.8195,
+      "step": 4670
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.5321626701600213e-06,
+      "loss": 0.7907,
+      "step": 4675
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.5155363010871627e-06,
+      "loss": 0.8539,
+      "step": 4680
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 1.4989932364995874e-06,
+      "loss": 0.8054,
+      "step": 4685
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 1.4825336388249455e-06,
+      "loss": 0.8162,
+      "step": 4690
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 1.466157669671372e-06,
+      "loss": 0.8393,
+      "step": 4695
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 1.4498654898258857e-06,
+      "loss": 0.8421,
+      "step": 4700
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 1.433657259252833e-06,
+      "loss": 0.8299,
+      "step": 4705
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 1.4175331370922995e-06,
+      "loss": 0.8525,
+      "step": 4710
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 1.4014932816585602e-06,
+      "loss": 0.817,
+      "step": 4715
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 1.3855378504385164e-06,
+      "loss": 0.8278,
+      "step": 4720
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 1.3696670000901513e-06,
+      "loss": 0.8666,
+      "step": 4725
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 1.3538808864409948e-06,
+      "loss": 0.8466,
+      "step": 4730
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 1.3381796644865964e-06,
+      "loss": 0.8353,
+      "step": 4735
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 1.3225634883889882e-06,
+      "loss": 0.8093,
+      "step": 4740
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 1.307032511475188e-06,
+      "loss": 0.8146,
+      "step": 4745
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 1.2915868862356896e-06,
+      "loss": 0.859,
+      "step": 4750
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 1.2762267643229597e-06,
+      "loss": 0.8337,
+      "step": 4755
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 1.2609522965499555e-06,
+      "loss": 0.8354,
+      "step": 4760
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 1.245763632888638e-06,
+      "loss": 0.8514,
+      "step": 4765
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 1.2306609224685029e-06,
+      "loss": 0.8396,
+      "step": 4770
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1.2156443135751262e-06,
+      "loss": 0.8408,
+      "step": 4775
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1.2007139536486823e-06,
+      "loss": 0.838,
+      "step": 4780
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.1858699892825209e-06,
+      "loss": 0.8165,
+      "step": 4785
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.1711125662217248e-06,
+      "loss": 0.8046,
+      "step": 4790
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.156441829361661e-06,
+      "loss": 0.8506,
+      "step": 4795
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 1.1418579227465777e-06,
+      "loss": 0.8341,
+      "step": 4800
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 1.1273609895681815e-06,
+      "loss": 0.8449,
+      "step": 4805
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 1.1129511721642294e-06,
+      "loss": 0.8144,
+      "step": 4810
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 1.0986286120171452e-06,
+      "loss": 0.8219,
+      "step": 4815
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 1.0843934497526042e-06,
+      "loss": 0.8334,
+      "step": 4820
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 1.0702458251381765e-06,
+      "loss": 0.8281,
+      "step": 4825
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 1.05618587708195e-06,
+      "loss": 0.8388,
+      "step": 4830
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 1.042213743631153e-06,
+      "loss": 0.8378,
+      "step": 4835
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 1.02832956197081e-06,
+      "loss": 0.8276,
+      "step": 4840
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 1.014533468422405e-06,
+      "loss": 0.8324,
+      "step": 4845
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 1.000825598442514e-06,
+      "loss": 0.7807,
+      "step": 4850
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 9.872060866215071e-07,
+      "loss": 0.8854,
+      "step": 4855
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 9.736750666822004e-07,
+      "loss": 0.7875,
+      "step": 4860
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 9.602326714785592e-07,
+      "loss": 0.8137,
+      "step": 4865
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 9.468790329943956e-07,
+      "loss": 0.8352,
+      "step": 4870
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 9.336142823420558e-07,
+      "loss": 0.8216,
+      "step": 4875
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 9.204385497611468e-07,
+      "loss": 0.8317,
+      "step": 4880
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 9.073519646172613e-07,
+      "loss": 0.8239,
+      "step": 4885
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 8.943546554006855e-07,
+      "loss": 0.8441,
+      "step": 4890
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 8.814467497251678e-07,
+      "loss": 0.8267,
+      "step": 4895
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 8.686283743266378e-07,
+      "loss": 0.8074,
+      "step": 4900
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 8.558996550619802e-07,
+      "loss": 0.8387,
+      "step": 4905
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 8.432607169077978e-07,
+      "loss": 0.8371,
+      "step": 4910
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 8.3071168395917e-07,
+      "loss": 0.792,
+      "step": 4915
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 8.182526794284485e-07,
+      "loss": 0.8194,
+      "step": 4920
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 8.058838256440493e-07,
+      "loss": 0.8323,
+      "step": 4925
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 7.936052440492348e-07,
+      "loss": 0.8843,
+      "step": 4930
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 7.814170552009381e-07,
+      "loss": 0.8377,
+      "step": 4935
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 7.693193787685782e-07,
+      "loss": 0.8585,
+      "step": 4940
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 7.573123335328692e-07,
+      "loss": 0.8503,
+      "step": 4945
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 7.45396037384677e-07,
+      "loss": 0.8537,
+      "step": 4950
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 7.335706073238391e-07,
+      "loss": 0.8476,
+      "step": 4955
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 7.218361594580325e-07,
+      "loss": 0.8318,
+      "step": 4960
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 7.101928090016319e-07,
+      "loss": 0.82,
+      "step": 4965
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 6.98640670274564e-07,
+      "loss": 0.8347,
+      "step": 4970
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 6.871798567012034e-07,
+      "loss": 0.8196,
+      "step": 4975
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 6.758104808092514e-07,
+      "loss": 0.8426,
+      "step": 4980
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 6.645326542286224e-07,
+      "loss": 0.8168,
+      "step": 4985
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 6.533464876903683e-07,
+      "loss": 0.8371,
+      "step": 4990
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 6.422520910255714e-07,
+      "loss": 0.8151,
+      "step": 4995
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 6.312495731642732e-07,
+      "loss": 0.8499,
+      "step": 5000
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 6.203390421344113e-07,
+      "loss": 0.8308,
+      "step": 5005
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 6.095206050607483e-07,
+      "loss": 0.8247,
+      "step": 5010
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 5.98794368163823e-07,
+      "loss": 0.8559,
+      "step": 5015
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 5.881604367589188e-07,
+      "loss": 0.8474,
+      "step": 5020
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 5.776189152550083e-07,
+      "loss": 0.8243,
+      "step": 5025
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 5.671699071537473e-07,
+      "loss": 0.829,
+      "step": 5030
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 5.5681351504845e-07,
+      "loss": 0.8279,
+      "step": 5035
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 5.465498406230774e-07,
+      "loss": 0.8062,
+      "step": 5040
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 5.363789846512546e-07,
+      "loss": 0.8327,
+      "step": 5045
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 5.263010469952623e-07,
+      "loss": 0.8103,
+      "step": 5050
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 5.163161266050664e-07,
+      "loss": 0.8473,
+      "step": 5055
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 5.064243215173525e-07,
+      "loss": 0.843,
+      "step": 5060
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 4.966257288545451e-07,
+      "loss": 0.8536,
+      "step": 5065
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 4.869204448238729e-07,
+      "loss": 0.7981,
+      "step": 5070
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 4.773085647164155e-07,
+      "loss": 0.8115,
+      "step": 5075
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 4.677901829061615e-07,
+      "loss": 0.8428,
+      "step": 5080
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 4.583653928490994e-07,
+      "loss": 0.814,
+      "step": 5085
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 4.490342870822828e-07,
+      "loss": 0.856,
+      "step": 5090
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 4.397969572229277e-07,
+      "loss": 0.8272,
+      "step": 5095
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 4.3065349396751997e-07,
+      "loss": 0.8357,
+      "step": 5100
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 4.216039870909094e-07,
+      "loss": 0.8058,
+      "step": 5105
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 4.126485254454493e-07,
+      "loss": 0.8338,
+      "step": 5110
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 4.0378719696010373e-07,
+      "loss": 0.8632,
+      "step": 5115
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 3.950200886395916e-07,
+      "loss": 0.833,
+      "step": 5120
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 3.8634728656354094e-07,
+      "loss": 0.8156,
+      "step": 5125
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 3.7776887588563237e-07,
+      "loss": 0.8148,
+      "step": 5130
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 3.692849408327637e-07,
+      "loss": 0.8747,
+      "step": 5135
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 3.608955647042345e-07,
+      "loss": 0.852,
+      "step": 5140
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 3.526008298709127e-07,
+      "loss": 0.8285,
+      "step": 5145
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 3.44400817774434e-07,
+      "loss": 0.8357,
+      "step": 5150
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 3.362956089264069e-07,
+      "loss": 0.81,
+      "step": 5155
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 3.2828528290760885e-07,
+      "loss": 0.8322,
+      "step": 5160
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 3.203699183672193e-07,
+      "loss": 0.8424,
+      "step": 5165
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 3.1254959302203904e-07,
+      "loss": 0.8368,
+      "step": 5170
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 3.0482438365572517e-07,
+      "loss": 0.8664,
+      "step": 5175
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.971943661180465e-07,
+      "loss": 0.8288,
+      "step": 5180
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.8965961532413155e-07,
+      "loss": 0.861,
+      "step": 5185
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.822202052537304e-07,
+      "loss": 0.8271,
+      "step": 5190
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.7487620895050193e-07,
+      "loss": 0.8762,
+      "step": 5195
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.6762769852127666e-07,
+      "loss": 0.8659,
+      "step": 5200
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.6047474513536953e-07,
+      "loss": 0.8202,
+      "step": 5205
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.5341741902386585e-07,
+      "loss": 0.8481,
+      "step": 5210
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.464557894789343e-07,
+      "loss": 0.8283,
+      "step": 5215
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.3958992485315523e-07,
+      "loss": 0.8172,
+      "step": 5220
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.328198925588432e-07,
+      "loss": 0.803,
+      "step": 5225
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.261457590673799e-07,
+      "loss": 0.8493,
+      "step": 5230
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.1956758990857253e-07,
+      "loss": 0.8473,
+      "step": 5235
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.13085449670003e-07,
+      "loss": 0.8326,
+      "step": 5240
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.0669940199639526e-07,
+      "loss": 0.857,
+      "step": 5245
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.0040950958899463e-07,
+      "loss": 0.8794,
+      "step": 5250
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.942158342049405e-07,
+      "loss": 0.8361,
+      "step": 5255
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.8811843665667683e-07,
+      "loss": 0.8777,
+      "step": 5260
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.8211737681134155e-07,
+      "loss": 0.8682,
+      "step": 5265
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.7621271359018144e-07,
+      "loss": 0.8455,
+      "step": 5270
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.7040450496798144e-07,
+      "loss": 0.8415,
+      "step": 5275
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.6469280797248522e-07,
+      "loss": 0.8413,
+      "step": 5280
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.5907767868383994e-07,
+      "loss": 0.8099,
+      "step": 5285
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.5355917223404682e-07,
+      "loss": 0.801,
+      "step": 5290
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.48137342806417e-07,
+      "loss": 0.8407,
+      "step": 5295
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.4281224363504099e-07,
+      "loss": 0.7909,
+      "step": 5300
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.375839270042656e-07,
+      "loss": 0.8323,
+      "step": 5305
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.3245244424818115e-07,
+      "loss": 0.8212,
+      "step": 5310
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.274178457501174e-07,
+      "loss": 0.8201,
+      "step": 5315
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.2248018094214941e-07,
+      "loss": 0.8028,
+      "step": 5320
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.1763949830460807e-07,
+      "loss": 0.799,
+      "step": 5325
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.1289584536561149e-07,
+      "loss": 0.8216,
+      "step": 5330
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.0824926870058983e-07,
+      "loss": 0.8352,
+      "step": 5335
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.036998139318368e-07,
+      "loss": 0.8685,
+      "step": 5340
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 9.924752572805563e-08,
+      "loss": 0.8498,
+      "step": 5345
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 9.489244780391927e-08,
+      "loss": 0.8421,
+      "step": 5350
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 9.063462291964753e-08,
+      "loss": 0.8048,
+      "step": 5355
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 8.647409288058405e-08,
+      "loss": 0.8427,
+      "step": 5360
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 8.241089853678442e-08,
+      "loss": 0.8407,
+      "step": 5365
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 7.844507978261418e-08,
+      "loss": 0.8291,
+      "step": 5370
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 7.457667555636372e-08,
+      "loss": 0.8161,
+      "step": 5375
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 7.08057238398574e-08,
+      "loss": 0.8289,
+      "step": 5380
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 6.713226165808606e-08,
+      "loss": 0.8805,
+      "step": 5385
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 6.35563250788429e-08,
+      "loss": 0.8414,
+      "step": 5390
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 6.00779492123671e-08,
+      "loss": 0.8315,
+      "step": 5395
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 5.6697168211002904e-08,
+      "loss": 0.865,
+      "step": 5400
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 5.341401526885781e-08,
+      "loss": 0.8091,
+      "step": 5405
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 5.022852262148492e-08,
+      "loss": 0.8233,
+      "step": 5410
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 4.714072154556104e-08,
+      "loss": 0.8586,
+      "step": 5415
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 4.415064235857913e-08,
+      "loss": 0.8414,
+      "step": 5420
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 4.1258314418554103e-08,
+      "loss": 0.8488,
+      "step": 5425
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 3.8463766123733035e-08,
+      "loss": 0.8216,
+      "step": 5430
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 3.576702491231432e-08,
+      "loss": 0.8235,
+      "step": 5435
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 3.3168117262181166e-08,
+      "loss": 0.8525,
+      "step": 5440
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 3.0667068690640736e-08,
+      "loss": 0.8211,
+      "step": 5445
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 2.8263903754174316e-08,
+      "loss": 0.7973,
+      "step": 5450
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 2.5958646048191983e-08,
+      "loss": 0.82,
+      "step": 5455
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 2.3751318206808316e-08,
+      "loss": 0.8062,
+      "step": 5460
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 2.1641941902611484e-08,
+      "loss": 0.8365,
+      "step": 5465
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 1.9630537846461184e-08,
+      "loss": 0.8252,
+      "step": 5470
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 1.7717125787272138e-08,
+      "loss": 0.8588,
+      "step": 5475
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1.590172451183314e-08,
+      "loss": 0.8541,
+      "step": 5480
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1.4184351844612753e-08,
+      "loss": 0.8257,
+      "step": 5485
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.2565024647590574e-08,
+      "loss": 0.8112,
+      "step": 5490
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.1043758820088457e-08,
+      "loss": 0.8213,
+      "step": 5495
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 9.620569298615102e-09,
+      "loss": 0.8088,
+      "step": 5500
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 8.29547005671838e-09,
+      "loss": 0.785,
+      "step": 5505
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 7.068474104852119e-09,
+      "loss": 0.85,
+      "step": 5510
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 5.939593490243978e-09,
+      "loss": 0.7999,
+      "step": 5515
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 4.908839296777768e-09,
+      "loss": 0.8472,
+      "step": 5520
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 3.976221644886868e-09,
+      "loss": 0.8288,
+      "step": 5525
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 3.14174969145431e-09,
+      "loss": 0.8157,
+      "step": 5530
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 2.4054316297195122e-09,
+      "loss": 0.7948,
+      "step": 5535
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.7672746892039016e-09,
+      "loss": 0.8029,
+      "step": 5540
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1.2272851356343042e-09,
+      "loss": 0.8308,
+      "step": 5545
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 7.854682708841044e-10,
+      "loss": 0.8331,
+      "step": 5550
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 4.4182843291884493e-10,
+      "loss": 0.8337,
+      "step": 5555
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 1.9636899575958823e-10,
+      "loss": 0.8564,
+      "step": 5560
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.909236944294904e-11,
+      "loss": 0.8432,
+      "step": 5565
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.0,
+      "loss": 0.8523,
+      "step": 5570
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 1.0952651500701904,
+      "eval_runtime": 163.428,
+      "eval_samples_per_second": 24.518,
+      "eval_steps_per_second": 0.514,
+      "step": 5570
+    },
+    {
+      "epoch": 5.0,
+      "step": 5570,
+      "total_flos": 1167473579655168.0,
+      "train_loss": 0.34447823324152127,
+      "train_runtime": 12106.1256,
+      "train_samples_per_second": 14.719,
+      "train_steps_per_second": 0.46
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 5570,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "total_flos": 1167473579655168.0,
+  "train_batch_size": 6,
+  "trial_name": null,
+  "trial_params": null
+}