diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4799 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.999841596705211,
+  "global_step": 3945,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.329113924050633e-05,
+      "loss": 5.9056,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00012658227848101267,
+      "loss": 4.729,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000189873417721519,
+      "loss": 4.2747,
+      "step": 15
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00025316455696202533,
+      "loss": 3.9938,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00031645569620253165,
+      "loss": 3.7136,
+      "step": 25
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.000379746835443038,
+      "loss": 3.6955,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0004430379746835443,
+      "loss": 3.4653,
+      "step": 35
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0005063291139240507,
+      "loss": 3.3515,
+      "step": 40
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000569620253164557,
+      "loss": 3.2781,
+      "step": 45
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0006329113924050633,
+      "loss": 3.2616,
+      "step": 50
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0006962025316455697,
+      "loss": 3.2314,
+      "step": 55
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.000759493670886076,
+      "loss": 3.1758,
+      "step": 60
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0008227848101265824,
+      "loss": 3.1956,
+      "step": 65
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0008860759493670886,
+      "loss": 3.1717,
+      "step": 70
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0009493670886075949,
+      "loss": 3.0096,
+      "step": 75
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0009999998349118006,
+      "loss": 3.0798,
+      "step": 80
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.000999994056836268,
+      "loss": 3.0737,
+      "step": 85
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.000999980024459781,
+      "loss": 3.1079,
+      "step": 90
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0009999577380139975,
+      "loss": 2.9692,
+      "step": 95
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0009999271978668395,
+      "loss": 3.0226,
+      "step": 100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0009998884045224886,
+      "loss": 2.8905,
+      "step": 105
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0009998413586213759,
+      "loss": 2.9354,
+      "step": 110
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0009997860609401732,
+      "loss": 2.9668,
+      "step": 115
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.000999722512391778,
+      "loss": 2.939,
+      "step": 120
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0009996507140253013,
+      "loss": 2.9213,
+      "step": 125
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.000999570667026047,
+      "loss": 2.9371,
+      "step": 130
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0009994823727154955,
+      "loss": 2.9326,
+      "step": 135
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0009993858325512794,
+      "loss": 2.9915,
+      "step": 140
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.000999281048127161,
+      "loss": 2.9352,
+      "step": 145
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0009991680211730051,
+      "loss": 2.8284,
+      "step": 150
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.000999046753554751,
+      "loss": 2.8232,
+      "step": 155
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.000998917247274381,
+      "loss": 2.8925,
+      "step": 160
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0009987795044698884,
+      "loss": 2.8381,
+      "step": 165
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0009986335274152413,
+      "loss": 2.8939,
+      "step": 170
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0009984793185203455,
+      "loss": 2.9206,
+      "step": 175
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0009983168803310042,
+      "loss": 2.733,
+      "step": 180
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0009981462155288774,
+      "loss": 2.8812,
+      "step": 185
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.000997967326931435,
+      "loss": 2.7199,
+      "step": 190
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0009977802174919133,
+      "loss": 2.8825,
+      "step": 195
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0009975848902992641,
+      "loss": 2.8283,
+      "step": 200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0009973813485781044,
+      "loss": 2.8447,
+      "step": 205
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.000997169595688664,
+      "loss": 2.7746,
+      "step": 210
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0009969496351267277,
+      "loss": 2.6779,
+      "step": 215
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.000996721470523581,
+      "loss": 2.8015,
+      "step": 220
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0009964851056459465,
+      "loss": 2.8746,
+      "step": 225
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0009962405443959244,
+      "loss": 2.7497,
+      "step": 230
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0009959877908109274,
+      "loss": 2.8197,
+      "step": 235
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0009957268490636132,
+      "loss": 2.7054,
+      "step": 240
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.000995457723461816,
+      "loss": 2.7127,
+      "step": 245
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0009951804184484766,
+      "loss": 2.7058,
+      "step": 250
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0009948949386015676,
+      "loss": 2.7129,
+      "step": 255
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0009946012886340176,
+      "loss": 2.7425,
+      "step": 260
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000994299473393635,
+      "loss": 2.7676,
+      "step": 265
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0009939894978630267,
+      "loss": 2.6728,
+      "step": 270
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0009936713671595157,
+      "loss": 2.7671,
+      "step": 275
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0009933450865350583,
+      "loss": 2.6662,
+      "step": 280
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000993010661376155,
+      "loss": 2.7334,
+      "step": 285
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.000992668097203763,
+      "loss": 2.7489,
+      "step": 290
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0009923173996732057,
+      "loss": 2.6428,
+      "step": 295
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0009919585745740776,
+      "loss": 2.7059,
+      "step": 300
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0009915916278301496,
+      "loss": 2.6961,
+      "step": 305
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0009912165654992716,
+      "loss": 2.8165,
+      "step": 310
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0009908333937732718,
+      "loss": 2.5846,
+      "step": 315
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0009904421189778551,
+      "loss": 2.6058,
+      "step": 320
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.000990042747572498,
+      "loss": 2.6858,
+      "step": 325
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0009896352861503426,
+      "loss": 2.6797,
+      "step": 330
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0009892197414380869,
+      "loss": 2.7078,
+      "step": 335
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0009887961202958747,
+      "loss": 2.5945,
+      "step": 340
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.000988364429717182,
+      "loss": 2.6061,
+      "step": 345
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0009879246768287019,
+      "loss": 2.6557,
+      "step": 350
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0009874768688902251,
+      "loss": 2.6387,
+      "step": 355
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0009870210132945235,
+      "loss": 2.6668,
+      "step": 360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0009865571175672243,
+      "loss": 2.7257,
+      "step": 365
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0009860851893666888,
+      "loss": 2.7604,
+      "step": 370
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0009856052364838846,
+      "loss": 2.6524,
+      "step": 375
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0009851172668422564,
+      "loss": 2.5279,
+      "step": 380
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0009846212884975969,
+      "loss": 2.5394,
+      "step": 385
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0009841173096379124,
+      "loss": 2.6585,
+      "step": 390
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.000983605338583288,
+      "loss": 2.6494,
+      "step": 395
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0009830853837857507,
+      "loss": 2.6424,
+      "step": 400
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0009825574538291292,
+      "loss": 2.6346,
+      "step": 405
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.000982021557428913,
+      "loss": 2.6241,
+      "step": 410
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0009814777034321067,
+      "loss": 2.5922,
+      "step": 415
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0009809259008170868,
+      "loss": 2.6258,
+      "step": 420
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0009803661586934513,
+      "loss": 2.6141,
+      "step": 425
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0009797984863018697,
+      "loss": 2.624,
+      "step": 430
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.000979222893013931,
+      "loss": 2.6261,
+      "step": 435
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0009786393883319884,
+      "loss": 2.4675,
+      "step": 440
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0009780479818890032,
+      "loss": 2.6471,
+      "step": 445
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0009774486834483848,
+      "loss": 2.6596,
+      "step": 450
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0009768415029038306,
+      "loss": 2.6316,
+      "step": 455
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0009762264502791609,
+      "loss": 2.6561,
+      "step": 460
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0009756035357281559,
+      "loss": 2.6207,
+      "step": 465
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0009749727695343859,
+      "loss": 2.4761,
+      "step": 470
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0009743341621110429,
+      "loss": 2.5767,
+      "step": 475
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0009736877240007677,
+      "loss": 2.5302,
+      "step": 480
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0009730334658754766,
+      "loss": 2.5426,
+      "step": 485
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0009723713985361853,
+      "loss": 2.5384,
+      "step": 490
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0009717015329128294,
+      "loss": 2.4947,
+      "step": 495
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0009710238800640854,
+      "loss": 2.5495,
+      "step": 500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0009703384511771874,
+      "loss": 2.5251,
+      "step": 505
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0009696452575677426,
+      "loss": 2.4983,
+      "step": 510
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0009689443106795442,
+      "loss": 2.4939,
+      "step": 515
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.000968235622084383,
+      "loss": 2.3895,
+      "step": 520
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0009675192034818561,
+      "loss": 2.4947,
+      "step": 525
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0009667950666991733,
+      "loss": 2.4988,
+      "step": 530
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0009660632236909627,
+      "loss": 2.573,
+      "step": 535
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0009653236865390729,
+      "loss": 2.6162,
+      "step": 540
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0009645764674523731,
+      "loss": 2.613,
+      "step": 545
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0009638215787665529,
+      "loss": 2.508,
+      "step": 550
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0009630590329439169,
+      "loss": 2.5019,
+      "step": 555
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0009622888425731801,
+      "loss": 2.622,
+      "step": 560
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0009615110203692601,
+      "loss": 2.4998,
+      "step": 565
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0009607255791730664,
+      "loss": 2.467,
+      "step": 570
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0009599325319512893,
+      "loss": 2.6073,
+      "step": 575
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0009591318917961853,
+      "loss": 2.6101,
+      "step": 580
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.000958323671925361,
+      "loss": 2.4277,
+      "step": 585
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0009575078856815554,
+      "loss": 2.5171,
+      "step": 590
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0009566845465324185,
+      "loss": 2.5376,
+      "step": 595
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0009558536680702903,
+      "loss": 2.5084,
+      "step": 600
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0009550152640119756,
+      "loss": 2.532,
+      "step": 605
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0009541693481985178,
+      "loss": 2.5169,
+      "step": 610
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0009533159345949704,
+      "loss": 2.6213,
+      "step": 615
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.000952455037290166,
+      "loss": 2.4695,
+      "step": 620
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0009515866704964846,
+      "loss": 2.4018,
+      "step": 625
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0009507108485496185,
+      "loss": 2.5084,
+      "step": 630
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0009498275859083352,
+      "loss": 2.5441,
+      "step": 635
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0009489368971542393,
+      "loss": 2.4815,
+      "step": 640
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0009480387969915317,
+      "loss": 2.468,
+      "step": 645
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0009471333002467668,
+      "loss": 2.5972,
+      "step": 650
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0009462204218686074,
+      "loss": 2.4825,
+      "step": 655
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0009453001769275786,
+      "loss": 2.5862,
+      "step": 660
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0009443725806158182,
+      "loss": 2.5174,
+      "step": 665
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0009434376482468265,
+      "loss": 2.3676,
+      "step": 670
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0009424953952552133,
+      "loss": 2.4715,
+      "step": 675
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0009415458371964429,
+      "loss": 2.5297,
+      "step": 680
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0009405889897465778,
+      "loss": 2.5442,
+      "step": 685
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0009396248687020191,
+      "loss": 2.5398,
+      "step": 690
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.000938653489979247,
+      "loss": 2.6478,
+      "step": 695
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.000937674869614556,
+      "loss": 2.4237,
+      "step": 700
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.000936689023763793,
+      "loss": 2.5133,
+      "step": 705
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0009356959687020876,
+      "loss": 2.3837,
+      "step": 710
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0009346957208235855,
+      "loss": 2.5308,
+      "step": 715
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.000933688296641177,
+      "loss": 2.4463,
+      "step": 720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0009326737127862249,
+      "loss": 2.4815,
+      "step": 725
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0009316519860082886,
+      "loss": 2.5258,
+      "step": 730
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0009306231331748496,
+      "loss": 2.4161,
+      "step": 735
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0009295871712710317,
+      "loss": 2.5333,
+      "step": 740
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0009285441173993206,
+      "loss": 2.3897,
+      "step": 745
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0009274939887792823,
+      "loss": 2.5545,
+      "step": 750
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0009264368027472784,
+      "loss": 2.3463,
+      "step": 755
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0009253725767561793,
+      "loss": 2.4669,
+      "step": 760
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0009243013283750773,
+      "loss": 2.331,
+      "step": 765
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0009232230752889957,
+      "loss": 2.2515,
+      "step": 770
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0009221378352985967,
+      "loss": 2.462,
+      "step": 775
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0009210456263198886,
+      "loss": 2.3609,
+      "step": 780
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0009199464663839289,
+      "loss": 2.4393,
+      "step": 785
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 2.382097005844116,
+      "eval_runtime": 95.3367,
+      "eval_samples_per_second": 134.345,
+      "eval_steps_per_second": 16.793,
+      "step": 789
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.000918840373636527,
+      "loss": 2.5088,
+      "step": 790
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0009177273663379448,
+      "loss": 2.1884,
+      "step": 795
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0009166074628625952,
+      "loss": 2.4646,
+      "step": 800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0009154806816987386,
+      "loss": 2.4983,
+      "step": 805
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0009143470414481776,
+      "loss": 2.3381,
+      "step": 810
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0009132065608259505,
+      "loss": 2.3715,
+      "step": 815
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0009120592586600216,
+      "loss": 2.4095,
+      "step": 820
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0009109051538909707,
+      "loss": 2.3306,
+      "step": 825
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0009097442655716803,
+      "loss": 2.3049,
+      "step": 830
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0009085766128670217,
+      "loss": 2.3163,
+      "step": 835
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0009074022150535374,
+      "loss": 2.3609,
+      "step": 840
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.000906221091519124,
+      "loss": 2.1964,
+      "step": 845
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0009050332617627115,
+      "loss": 2.3304,
+      "step": 850
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0009038387453939415,
+      "loss": 2.3855,
+      "step": 855
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0009026375621328436,
+      "loss": 2.2852,
+      "step": 860
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0009014297318095099,
+      "loss": 2.2837,
+      "step": 865
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0009002152743637673,
+      "loss": 2.2163,
+      "step": 870
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0008989942098448485,
+      "loss": 2.2896,
+      "step": 875
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0008977665584110612,
+      "loss": 2.3854,
+      "step": 880
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0008965323403294553,
+      "loss": 2.3061,
+      "step": 885
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0008952915759754878,
+      "loss": 2.3625,
+      "step": 890
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.000894044285832687,
+      "loss": 2.4091,
+      "step": 895
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0008927904904923142,
+      "loss": 2.3056,
+      "step": 900
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0008915302106530234,
+      "loss": 2.3323,
+      "step": 905
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0008902634671205201,
+      "loss": 2.3386,
+      "step": 910
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0008889902808072176,
+      "loss": 2.3661,
+      "step": 915
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0008877106727318918,
+      "loss": 2.3923,
+      "step": 920
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0008864246640193341,
+      "loss": 2.2784,
+      "step": 925
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0008851322759000026,
+      "loss": 2.3131,
+      "step": 930
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0008838335297096721,
+      "loss": 2.4255,
+      "step": 935
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0008825284468890808,
+      "loss": 2.4198,
+      "step": 940
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0008812170489835783,
+      "loss": 2.3279,
+      "step": 945
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0008798993576427671,
+      "loss": 2.3672,
+      "step": 950
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0008785753946201483,
+      "loss": 2.3138,
+      "step": 955
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0008772451817727598,
+      "loss": 2.2753,
+      "step": 960
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0008759087410608169,
+      "loss": 2.2821,
+      "step": 965
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0008745660945473495,
+      "loss": 2.2495,
+      "step": 970
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0008732172643978382,
+      "loss": 2.47,
+      "step": 975
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0008718622728798476,
+      "loss": 2.3261,
+      "step": 980
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0008705011423626587,
+      "loss": 2.3,
+      "step": 985
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0008691338953169009,
+      "loss": 2.3209,
+      "step": 990
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0008677605543141796,
+      "loss": 2.3039,
+      "step": 995
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.000866381142026704,
+      "loss": 2.2033,
+      "step": 1000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0008649956812269133,
+      "loss": 2.3297,
+      "step": 1005
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0008636041947871001,
+      "loss": 2.273,
+      "step": 1010
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0008622067056790332,
+      "loss": 2.2885,
+      "step": 1015
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0008608032369735781,
+      "loss": 2.3532,
+      "step": 1020
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0008593938118403164,
+      "loss": 2.3094,
+      "step": 1025
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0008579784535471631,
+      "loss": 2.3238,
+      "step": 1030
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0008565571854599825,
+      "loss": 2.2605,
+      "step": 1035
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0008551300310422027,
+      "loss": 2.2964,
+      "step": 1040
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0008536970138544277,
+      "loss": 2.2659,
+      "step": 1045
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0008522581575540491,
+      "loss": 2.3661,
+      "step": 1050
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.0008508134858948553,
+      "loss": 2.3099,
+      "step": 1055
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.0008493630227266387,
+      "loss": 2.291,
+      "step": 1060
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0008479067919948031,
+      "loss": 2.4099,
+      "step": 1065
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.000846444817739968,
+      "loss": 2.2842,
+      "step": 1070
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0008449771240975707,
+      "loss": 2.3229,
+      "step": 1075
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0008435037352974696,
+      "loss": 2.2471,
+      "step": 1080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.000842024675663543,
+      "loss": 2.3572,
+      "step": 1085
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0008405399696132878,
+      "loss": 2.1652,
+      "step": 1090
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0008390496416574165,
+      "loss": 2.3283,
+      "step": 1095
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0008375537163994522,
+      "loss": 2.2474,
+      "step": 1100
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0008360522185353234,
+      "loss": 2.3934,
+      "step": 1105
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0008345451728529552,
+      "loss": 2.2949,
+      "step": 1110
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0008330326042318603,
+      "loss": 2.2382,
+      "step": 1115
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0008315145376427288,
+      "loss": 2.356,
+      "step": 1120
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0008299909981470158,
+      "loss": 2.2622,
+      "step": 1125
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0008284620108965271,
+      "loss": 2.2402,
+      "step": 1130
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0008269276011330047,
+      "loss": 2.2855,
+      "step": 1135
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0008253877941877095,
+      "loss": 2.1968,
+      "step": 1140
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0008238426154810035,
+      "loss": 2.2679,
+      "step": 1145
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0008222920905219302,
+      "loss": 2.2169,
+      "step": 1150
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0008207362449077932,
+      "loss": 2.1951,
+      "step": 1155
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0008191751043237336,
+      "loss": 2.3119,
+      "step": 1160
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.000817608694542306,
+      "loss": 2.408,
+      "step": 1165
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0008160370414230535,
+      "loss": 2.1998,
+      "step": 1170
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00081446017091208,
+      "loss": 2.3845,
+      "step": 1175
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0008128781090416228,
+      "loss": 2.2619,
+      "step": 1180
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0008112908819296217,
+      "loss": 2.3258,
+      "step": 1185
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0008096985157792886,
+      "loss": 2.3633,
+      "step": 1190
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0008081010368786751,
+      "loss": 2.2268,
+      "step": 1195
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0008064984716002377,
+      "loss": 2.2499,
+      "step": 1200
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0008048908464004032,
+      "loss": 2.258,
+      "step": 1205
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0008032781878191313,
+      "loss": 2.2347,
+      "step": 1210
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0008016605224794773,
+      "loss": 2.2717,
+      "step": 1215
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.0008000378770871515,
+      "loss": 2.1741,
+      "step": 1220
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.0007984102784300793,
+      "loss": 2.1593,
+      "step": 1225
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0007967777533779582,
+      "loss": 2.2707,
+      "step": 1230
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0007951403288818149,
+      "loss": 2.2641,
+      "step": 1235
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0007934980319735599,
+      "loss": 2.2677,
+      "step": 1240
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.000791850889765541,
+      "loss": 2.2982,
+      "step": 1245
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0007901989294500968,
+      "loss": 2.1954,
+      "step": 1250
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0007885421782991064,
+      "loss": 2.2439,
+      "step": 1255
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0007868806636635405,
+      "loss": 2.2238,
+      "step": 1260
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0007852144129730086,
+      "loss": 2.234,
+      "step": 1265
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0007835434537353067,
+      "loss": 2.31,
+      "step": 1270
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0007818678135359641,
+      "loss": 2.2452,
+      "step": 1275
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0007801875200377862,
+      "loss": 2.1722,
+      "step": 1280
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0007785026009803993,
+      "loss": 2.1681,
+      "step": 1285
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0007768130841797919,
+      "loss": 2.2128,
+      "step": 1290
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0007751189975278561,
+      "loss": 2.121,
+      "step": 1295
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0007734203689919261,
+      "loss": 2.2649,
+      "step": 1300
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0007717172266143178,
+      "loss": 2.1976,
+      "step": 1305
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.000770009598511865,
+      "loss": 2.3699,
+      "step": 1310
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.0007682975128754548,
+      "loss": 2.1847,
+      "step": 1315
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.0007665809979695639,
+      "loss": 2.2154,
+      "step": 1320
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00076486008213179,
+      "loss": 2.2729,
+      "step": 1325
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.0007631347937723856,
+      "loss": 2.3637,
+      "step": 1330
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.0007614051613737878,
+      "loss": 2.244,
+      "step": 1335
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0007596712134901487,
+      "loss": 2.2387,
+      "step": 1340
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0007579329787468638,
+      "loss": 2.1745,
+      "step": 1345
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.0007561904858400999,
+      "loss": 2.1605,
+      "step": 1350
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.0007544437635363206,
+      "loss": 2.2522,
+      "step": 1355
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.0007526928406718115,
+      "loss": 2.2604,
+      "step": 1360
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0007509377461522049,
+      "loss": 2.2935,
+      "step": 1365
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.000749178508952002,
+      "loss": 2.2513,
+      "step": 1370
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.0007474151581140947,
+      "loss": 2.1089,
+      "step": 1375
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.0007456477227492858,
+      "loss": 2.2211,
+      "step": 1380
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0007438762320358088,
+      "loss": 2.2009,
+      "step": 1385
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0007421007152188468,
+      "loss": 2.1967,
+      "step": 1390
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0007403212016100484,
+      "loss": 2.3052,
+      "step": 1395
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0007385377205870442,
+      "loss": 2.2927,
+      "step": 1400
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0007367503015929626,
+      "loss": 2.1667,
+      "step": 1405
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.0007349589741359431,
+      "loss": 2.2273,
+      "step": 1410
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.000733163767788649,
+      "loss": 2.3456,
+      "step": 1415
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0007313647121877793,
+      "loss": 2.279,
+      "step": 1420
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.0007295618370335799,
+      "loss": 2.3189,
+      "step": 1425
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.0007277551720893528,
+      "loss": 2.1636,
+      "step": 1430
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.0007259447471809651,
+      "loss": 2.1341,
+      "step": 1435
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.0007241305921963557,
+      "loss": 2.2994,
+      "step": 1440
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.0007223127370850433,
+      "loss": 2.2897,
+      "step": 1445
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.000720491211857631,
+      "loss": 2.2286,
+      "step": 1450
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.000718666046585311,
+      "loss": 2.2796,
+      "step": 1455
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.0007168372713993683,
+      "loss": 2.2812,
+      "step": 1460
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.0007150049164906835,
+      "loss": 2.1562,
+      "step": 1465
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.0007131690121092336,
+      "loss": 2.305,
+      "step": 1470
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.0007113295885635935,
+      "loss": 2.2369,
+      "step": 1475
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.0007094866762204354,
+      "loss": 2.2044,
+      "step": 1480
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.0007076403055040271,
+      "loss": 2.1396,
+      "step": 1485
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.0007057905068957299,
+      "loss": 2.3368,
+      "step": 1490
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.0007039373109334957,
+      "loss": 2.2138,
+      "step": 1495
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.0007020807482113627,
+      "loss": 2.1798,
+      "step": 1500
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00070022084937895,
+      "loss": 2.1819,
+      "step": 1505
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.0006983576451409517,
+      "loss": 2.1694,
+      "step": 1510
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.0006964911662566308,
+      "loss": 2.1964,
+      "step": 1515
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.0006946214435393103,
+      "loss": 2.0662,
+      "step": 1520
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.0006927485078558652,
+      "loss": 2.1458,
+      "step": 1525
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.0006908723901262126,
+      "loss": 2.2772,
+      "step": 1530
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.0006889931213228015,
+      "loss": 2.1833,
+      "step": 1535
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.0006871107324701008,
+      "loss": 2.2241,
+      "step": 1540
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.0006852252546440885,
+      "loss": 2.1994,
+      "step": 1545
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.0006833367189717372,
+      "loss": 2.229,
+      "step": 1550
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.0006814451566305014,
+      "loss": 2.2222,
+      "step": 1555
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.0006795505988478015,
+      "loss": 2.2849,
+      "step": 1560
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.0006776530769005098,
+      "loss": 2.11,
+      "step": 1565
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.0006757526221144329,
+      "loss": 2.1437,
+      "step": 1570
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.0006738492658637956,
+      "loss": 2.1549,
+      "step": 1575
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 2.1981911659240723,
+      "eval_runtime": 95.327,
+      "eval_samples_per_second": 134.359,
+      "eval_steps_per_second": 16.795,
+      "step": 1578
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.0006719430395707215,
+      "loss": 2.3153,
+      "step": 1580
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.0006700339747047161,
+      "loss": 2.0107,
+      "step": 1585
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.0006681221027821458,
+      "loss": 2.1278,
+      "step": 1590
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.000666207455365718,
+      "loss": 2.0655,
+      "step": 1595
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.0006642900640639608,
+      "loss": 2.0664,
+      "step": 1600
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.0006623699605306999,
+      "loss": 1.9791,
+      "step": 1605
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.0006604471764645369,
+      "loss": 2.0518,
+      "step": 1610
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.000658521743608326,
+      "loss": 2.0644,
+      "step": 1615
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.0006565936937486492,
+      "loss": 2.0737,
+      "step": 1620
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.0006546630587152929,
+      "loss": 2.0784,
+      "step": 1625
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.000652729870380721,
+      "loss": 2.1277,
+      "step": 1630
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.0006507941606595491,
+      "loss": 2.1686,
+      "step": 1635
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0006488559615080188,
+      "loss": 2.1398,
+      "step": 1640
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0006469153049234683,
+      "loss": 2.1453,
+      "step": 1645
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.0006449722229438053,
+      "loss": 2.0425,
+      "step": 1650
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.0006430267476469783,
+      "loss": 2.0705,
+      "step": 1655
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.0006410789111504461,
+      "loss": 2.1157,
+      "step": 1660
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.0006391287456106483,
+      "loss": 2.1243,
+      "step": 1665
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.000637176283222474,
+      "loss": 2.0428,
+      "step": 1670
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.0006352215562187306,
+      "loss": 2.0638,
+      "step": 1675
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.000633264596869612,
+      "loss": 2.0722,
+      "step": 1680
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.0006313054374821646,
+      "loss": 2.1089,
+      "step": 1685
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.0006293441103997555,
+      "loss": 2.0444,
+      "step": 1690
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.0006273806480015374,
+      "loss": 2.0067,
+      "step": 1695
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.0006254150827019148,
+      "loss": 2.2493,
+      "step": 1700
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.0006234474469500085,
+      "loss": 2.0435,
+      "step": 1705
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.00062147777322912,
+      "loss": 2.1426,
+      "step": 1710
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.0006195060940561949,
+      "loss": 2.0512,
+      "step": 1715
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.0006175324419812869,
+      "loss": 2.1365,
+      "step": 1720
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.0006155568495870196,
+      "loss": 2.137,
+      "step": 1725
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.000613579349488049,
+      "loss": 1.9907,
+      "step": 1730
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.0006115999743305252,
+      "loss": 2.085,
+      "step": 1735
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.0006096187567915528,
+      "loss": 2.1276,
+      "step": 1740
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.0006076357295786525,
+      "loss": 1.9774,
+      "step": 1745
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.0006056509254292202,
+      "loss": 2.0054,
+      "step": 1750
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.0006036643771099871,
+      "loss": 2.0158,
+      "step": 1755
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.000601676117416478,
+      "loss": 1.9952,
+      "step": 1760
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.0005996861791724712,
+      "loss": 2.0267,
+      "step": 1765
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.0005976945952294553,
+      "loss": 1.9985,
+      "step": 1770
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.0005957013984660875,
+      "loss": 2.0334,
+      "step": 1775
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.0005937066217876511,
+      "loss": 2.0078,
+      "step": 1780
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.0005917102981255113,
+      "loss": 2.0976,
+      "step": 1785
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.0005897124604365727,
+      "loss": 2.0843,
+      "step": 1790
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.0005877131417027343,
+      "loss": 1.9916,
+      "step": 1795
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.0005857123749303454,
+      "loss": 2.0005,
+      "step": 1800
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.000583710193149661,
+      "loss": 2.134,
+      "step": 1805
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.0005817066294142957,
+      "loss": 2.1503,
+      "step": 1810
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.0005797017168006791,
+      "loss": 2.033,
+      "step": 1815
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.0005776954884075085,
+      "loss": 2.0627,
+      "step": 1820
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.0005756879773552036,
+      "loss": 1.985,
+      "step": 1825
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.000573679216785359,
+      "loss": 2.0933,
+      "step": 1830
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.0005716692398601975,
+      "loss": 2.1007,
+      "step": 1835
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.0005696580797620226,
+      "loss": 2.0164,
+      "step": 1840
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.0005676457696926703,
+      "loss": 2.0588,
+      "step": 1845
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.0005656323428729612,
+      "loss": 2.0087,
+      "step": 1850
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.0005636178325421524,
+      "loss": 2.0798,
+      "step": 1855
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.000561602271957388,
+      "loss": 2.1189,
+      "step": 1860
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.0005595856943931512,
+      "loss": 2.0921,
+      "step": 1865
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.0005575681331407137,
+      "loss": 2.1177,
+      "step": 1870
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.0005555496215075871,
+      "loss": 2.1912,
+      "step": 1875
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.0005535301928169722,
+      "loss": 2.1851,
+      "step": 1880
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.00055150988040721,
+      "loss": 2.1525,
+      "step": 1885
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.00054948871763123,
+      "loss": 2.0721,
+      "step": 1890
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.0005474667378560006,
+      "loss": 2.1164,
+      "step": 1895
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.0005454439744619776,
+      "loss": 2.1189,
+      "step": 1900
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.0005434204608425538,
+      "loss": 2.0945,
+      "step": 1905
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.0005413962304035068,
+      "loss": 2.2036,
+      "step": 1910
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.0005393713165624483,
+      "loss": 2.1643,
+      "step": 1915
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.0005373457527482725,
+      "loss": 2.0757,
+      "step": 1920
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.0005353195724006031,
+      "loss": 2.0375,
+      "step": 1925
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.0005332928089692425,
+      "loss": 2.0028,
+      "step": 1930
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.0005312654959136194,
+      "loss": 2.08,
+      "step": 1935
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.0005292376667022355,
+      "loss": 2.0521,
+      "step": 1940
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.0005272093548121141,
+      "loss": 1.9639,
+      "step": 1945
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.0005251805937282467,
+      "loss": 2.1265,
+      "step": 1950
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.0005231514169430402,
+      "loss": 2.0656,
+      "step": 1955
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.000521121857955765,
+      "loss": 2.1371,
+      "step": 1960
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.0005190919502720005,
+      "loss": 2.0298,
+      "step": 1965
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.0005170617274030827,
+      "loss": 2.166,
+      "step": 1970
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.0005150312228655515,
+      "loss": 2.0785,
+      "step": 1975
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.0005130004701805963,
+      "loss": 2.0836,
+      "step": 1980
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.0005109695028735029,
+      "loss": 2.1391,
+      "step": 1985
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.0005089383544731012,
+      "loss": 1.9789,
+      "step": 1990
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.0005069070585112097,
+      "loss": 2.0498,
+      "step": 1995
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.0005048756485220836,
+      "loss": 2.1021,
+      "step": 2000
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.0005028441580418602,
+      "loss": 2.0545,
+      "step": 2005
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.0005008126206080059,
+      "loss": 2.0408,
+      "step": 2010
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.0004987810697587617,
+      "loss": 2.1183,
+      "step": 2015
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.0004967495390325911,
+      "loss": 2.1223,
+      "step": 2020
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.0004947180619676244,
+      "loss": 1.9974,
+      "step": 2025
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.0004926866721011064,
+      "loss": 2.1279,
+      "step": 2030
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.0004906554029688425,
+      "loss": 1.9737,
+      "step": 2035
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.0004886242881046447,
+      "loss": 1.9904,
+      "step": 2040
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.00048659336103977805,
+      "loss": 2.0324,
+      "step": 2045
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.00048456265530240784,
+      "loss": 2.1492,
+      "step": 2050
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.00048253220441704507,
+      "loss": 1.9715,
+      "step": 2055
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.00048050204190399335,
+      "loss": 2.0,
+      "step": 2060
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.00047847220127879606,
+      "loss": 2.1284,
+      "step": 2065
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.00047644271605168213,
+      "loss": 2.0984,
+      "step": 2070
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.0004744136197270135,
+      "loss": 2.057,
+      "step": 2075
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.00047238494580273193,
+      "loss": 2.0703,
+      "step": 2080
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.00047035672776980574,
+      "loss": 2.154,
+      "step": 2085
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.00046832899911167697,
+      "loss": 2.154,
+      "step": 2090
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00046630179330370866,
+      "loss": 2.0377,
+      "step": 2095
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00046427514381263205,
+      "loss": 2.1521,
+      "step": 2100
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.00046224908409599495,
+      "loss": 2.159,
+      "step": 2105
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.000460223647601608,
+      "loss": 2.0178,
+      "step": 2110
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.0004581988677669935,
+      "loss": 2.0769,
+      "step": 2115
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.00045617477801883295,
+      "loss": 2.1209,
+      "step": 2120
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.00045415141177241545,
+      "loss": 1.9985,
+      "step": 2125
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.0004521288024310856,
+      "loss": 1.9867,
+      "step": 2130
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.00045010698338569333,
+      "loss": 2.022,
+      "step": 2135
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.0004480859880140408,
+      "loss": 2.0363,
+      "step": 2140
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00044606584968033265,
+      "loss": 2.1197,
+      "step": 2145
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00044404660173462477,
+      "loss": 2.0195,
+      "step": 2150
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.0004420282775122739,
+      "loss": 2.0232,
+      "step": 2155
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.00044001091033338735,
+      "loss": 2.0646,
+      "step": 2160
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.0004379945335022726,
+      "loss": 1.9648,
+      "step": 2165
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.00043597918030688777,
+      "loss": 2.1262,
+      "step": 2170
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.0004339648840182919,
+      "loss": 2.0872,
+      "step": 2175
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.0004319516778900957,
+      "loss": 2.0648,
+      "step": 2180
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.0004299395951579126,
+      "loss": 1.9803,
+      "step": 2185
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.0004279286690388107,
+      "loss": 2.0624,
+      "step": 2190
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.00042591893273076313,
+      "loss": 1.989,
+      "step": 2195
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.00042391041941210086,
+      "loss": 2.0994,
+      "step": 2200
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.00042190316224096497,
+      "loss": 2.1275,
+      "step": 2205
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 0.00041989719435475865,
+      "loss": 1.8627,
+      "step": 2210
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.0004178925488696012,
+      "loss": 2.0948,
+      "step": 2215
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.0004158892588797801,
+      "loss": 2.1067,
+      "step": 2220
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.0004138873574572053,
+      "loss": 2.0034,
+      "step": 2225
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.0004118868776508633,
+      "loss": 2.0439,
+      "step": 2230
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.0004098878524862715,
+      "loss": 2.0353,
+      "step": 2235
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.00040789031496493255,
+      "loss": 2.0757,
+      "step": 2240
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.00040589429806379057,
+      "loss": 2.0462,
+      "step": 2245
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.0004038998347346854,
+      "loss": 1.9894,
+      "step": 2250
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.0004019069579038096,
+      "loss": 2.0345,
+      "step": 2255
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.00039991570047116456,
+      "loss": 2.044,
+      "step": 2260
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.0003979260953100169,
+      "loss": 1.9671,
+      "step": 2265
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.0003959381752663572,
+      "loss": 1.9423,
+      "step": 2270
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.00039395197315835573,
+      "loss": 1.9895,
+      "step": 2275
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.0003919675217758224,
+      "loss": 1.9851,
+      "step": 2280
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 0.00038998485387966425,
+      "loss": 2.1416,
+      "step": 2285
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 0.00038800400220134556,
+      "loss": 2.0666,
+      "step": 2290
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.0003860249994423467,
+      "loss": 2.0336,
+      "step": 2295
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.00038404787827362477,
+      "loss": 2.1442,
+      "step": 2300
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.0003820726713350742,
+      "loss": 2.0041,
+      "step": 2305
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.00038009941123498737,
+      "loss": 2.0138,
+      "step": 2310
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.0003781281305495171,
+      "loss": 2.1307,
+      "step": 2315
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.00037615886182213826,
+      "loss": 1.9921,
+      "step": 2320
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.0003741916375631105,
+      "loss": 2.1046,
+      "step": 2325
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.0003722264902489425,
+      "loss": 2.0436,
+      "step": 2330
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.00037026345232185414,
+      "loss": 1.9607,
+      "step": 2335
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.00036830255618924265,
+      "loss": 1.9253,
+      "step": 2340
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.0003663438342231462,
+      "loss": 2.127,
+      "step": 2345
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.00036438731875971034,
+      "loss": 2.1264,
+      "step": 2350
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.00036243304209865405,
+      "loss": 2.038,
+      "step": 2355
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.0003604810365027362,
+      "loss": 2.1015,
+      "step": 2360
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.00035853133419722316,
+      "loss": 2.1424,
+      "step": 2365
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 2.1064605712890625,
+      "eval_runtime": 95.2849,
+      "eval_samples_per_second": 134.418,
+      "eval_steps_per_second": 16.802,
+      "step": 2367
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0003565839673693566,
+      "loss": 2.0409,
+      "step": 2370
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.0003546389681678224,
+      "loss": 2.023,
+      "step": 2375
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.0003526963687022198,
+      "loss": 1.8723,
+      "step": 2380
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.0003507562010425312,
+      "loss": 1.852,
+      "step": 2385
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.000348818497218593,
+      "loss": 2.0228,
+      "step": 2390
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.00034688328921956635,
+      "loss": 1.8949,
+      "step": 2395
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.0003449506089934095,
+      "loss": 1.911,
+      "step": 2400
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.0003430204884463499,
+      "loss": 1.9915,
+      "step": 2405
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.0003410929594423585,
+      "loss": 1.8748,
+      "step": 2410
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.0003391680538026224,
+      "loss": 2.0819,
+      "step": 2415
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00033724580330502007,
+      "loss": 1.9166,
+      "step": 2420
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00033532623968359697,
+      "loss": 1.9435,
+      "step": 2425
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.0003334093946280417,
+      "loss": 1.9159,
+      "step": 2430
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.0003314952997831618,
+      "loss": 2.0227,
+      "step": 2435
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00032958398674836324,
+      "loss": 1.9393,
+      "step": 2440
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.00032767548707712693,
+      "loss": 1.9019,
+      "step": 2445
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.0003257698322764886,
+      "loss": 1.9114,
+      "step": 2450
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00032386705380651874,
+      "loss": 1.8678,
+      "step": 2455
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.00032196718307980277,
+      "loss": 1.9498,
+      "step": 2460
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.00032007025146092347,
+      "loss": 1.9621,
+      "step": 2465
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.00031817629026594164,
+      "loss": 1.9599,
+      "step": 2470
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.0003162853307618805,
+      "loss": 1.8853,
+      "step": 2475
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.0003143974041662087,
+      "loss": 1.9151,
+      "step": 2480
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.00031251254164632537,
+      "loss": 1.8902,
+      "step": 2485
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.0003106307743190452,
+      "loss": 1.9937,
+      "step": 2490
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.0003087521332500854,
+      "loss": 2.0179,
+      "step": 2495
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.0003068766494535522,
+      "loss": 2.0089,
+      "step": 2500
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.0003050043538914287,
+      "loss": 1.9892,
+      "step": 2505
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.00030313527747306444,
+      "loss": 1.8901,
+      "step": 2510
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00030126945105466487,
+      "loss": 1.9044,
+      "step": 2515
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.0002994069054387817,
+      "loss": 1.949,
+      "step": 2520
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.00029754767137380426,
+      "loss": 1.8189,
+      "step": 2525
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.0002956917795534528,
+      "loss": 1.988,
+      "step": 2530
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.00029383926061627054,
+      "loss": 2.0487,
+      "step": 2535
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.0002919901451451187,
+      "loss": 1.9252,
+      "step": 2540
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00029014446366667113,
+      "loss": 1.9607,
+      "step": 2545
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00028830224665091106,
+      "loss": 2.0455,
+      "step": 2550
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.0002864635245106272,
+      "loss": 1.9129,
+      "step": 2555
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.0002846283276009124,
+      "loss": 2.0086,
+      "step": 2560
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.0002827966862186616,
+      "loss": 1.9384,
+      "step": 2565
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.000280968630602073,
+      "loss": 1.991,
+      "step": 2570
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.00027914419093014734,
+      "loss": 1.8516,
+      "step": 2575
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.00027732339732219137,
+      "loss": 1.966,
+      "step": 2580
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.0002755062798373189,
+      "loss": 1.9472,
+      "step": 2585
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.0002736928684739558,
+      "loss": 1.8879,
+      "step": 2590
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.0002718831931693443,
+      "loss": 1.9771,
+      "step": 2595
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.00027007728379904814,
+      "loss": 1.8985,
+      "step": 2600
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.0002682751701764615,
+      "loss": 2.0464,
+      "step": 2605
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.0002664768820523137,
+      "loss": 1.8301,
+      "step": 2610
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.0002646824491141807,
+      "loss": 1.8646,
+      "step": 2615
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.0002628919009859934,
+      "loss": 1.9277,
+      "step": 2620
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.0002611052672275495,
+      "loss": 1.8923,
+      "step": 2625
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.0002593225773340251,
+      "loss": 1.8747,
+      "step": 2630
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.00025754386073548776,
+      "loss": 1.9835,
+      "step": 2635
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 0.000255769146796411,
+      "loss": 1.9034,
+      "step": 2640
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 0.0002539984648151886,
+      "loss": 1.9571,
+      "step": 2645
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 0.00025223184402365224,
+      "loss": 1.9356,
+      "step": 2650
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 0.0002504693135865875,
+      "loss": 1.9416,
+      "step": 2655
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 0.00024871090260125423,
+      "loss": 2.0262,
+      "step": 2660
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 0.000246956640096904,
+      "loss": 2.0165,
+      "step": 2665
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 0.0002452065550343026,
+      "loss": 2.0138,
+      "step": 2670
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 0.00024346067630525082,
+      "loss": 2.0022,
+      "step": 2675
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 0.00024171903273210828,
+      "loss": 1.8902,
+      "step": 2680
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 0.0002399816530673171,
+      "loss": 2.0161,
+      "step": 2685
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 0.00023824856599292762,
+      "loss": 1.973,
+      "step": 2690
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 0.00023651980012012454,
+      "loss": 1.8932,
+      "step": 2695
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 0.00023479538398875433,
+      "loss": 1.9342,
+      "step": 2700
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 0.00023307534606685525,
+      "loss": 1.8307,
+      "step": 2705
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 0.00023135971475018557,
+      "loss": 1.9087,
+      "step": 2710
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 0.00022964851836175705,
+      "loss": 1.8213,
+      "step": 2715
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 0.00022794178515136494,
+      "loss": 1.9313,
+      "step": 2720
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 0.00022623954329512348,
+      "loss": 2.0212,
+      "step": 2725
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 0.00022454182089500004,
+      "loss": 1.8266,
+      "step": 2730
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 0.00022284864597835058,
+      "loss": 1.9772,
+      "step": 2735
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 0.00022116004649745826,
+      "loss": 1.9601,
+      "step": 2740
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 0.00021947605032907099,
+      "loss": 1.8989,
+      "step": 2745
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 0.00021779668527394187,
+      "loss": 1.9805,
+      "step": 2750
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 0.0002161219790563691,
+      "loss": 1.8144,
+      "step": 2755
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 0.00021445195932373986,
+      "loss": 1.9025,
+      "step": 2760
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 0.0002127866536460727,
+      "loss": 1.8811,
+      "step": 2765
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 0.00021112608951556316,
+      "loss": 1.998,
+      "step": 2770
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 0.00020947029434612886,
+      "loss": 1.8758,
+      "step": 2775
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 0.00020781929547295815,
+      "loss": 1.8745,
+      "step": 2780
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 0.00020617312015205842,
+      "loss": 1.8758,
+      "step": 2785
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 0.00020453179555980533,
+      "loss": 1.9658,
+      "step": 2790
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 0.00020289534879249543,
+      "loss": 1.7781,
+      "step": 2795
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 0.00020126380686589807,
+      "loss": 1.9071,
+      "step": 2800
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 0.0001996371967148098,
+      "loss": 1.813,
+      "step": 2805
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 0.00019801554519260884,
+      "loss": 1.9647,
+      "step": 2810
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 0.00019639887907081294,
+      "loss": 1.9145,
+      "step": 2815
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 0.00019478722503863706,
+      "loss": 1.9741,
+      "step": 2820
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 0.0001931806097025517,
+      "loss": 1.9005,
+      "step": 2825
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 0.00019157905958584582,
+      "loss": 1.9434,
+      "step": 2830
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 0.0001899826011281865,
+      "loss": 1.8557,
+      "step": 2835
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 0.00018839126068518448,
+      "loss": 1.9641,
+      "step": 2840
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 0.0001868050645279576,
+      "loss": 1.9206,
+      "step": 2845
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 0.00018522403884269813,
+      "loss": 1.8437,
+      "step": 2850
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 0.00018364820973024,
+      "loss": 1.9311,
+      "step": 2855
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 0.00018207760320562794,
+      "loss": 1.9442,
+      "step": 2860
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 0.00018051224519768816,
+      "loss": 1.8921,
+      "step": 2865
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 0.00017895216154859966,
+      "loss": 1.8873,
+      "step": 2870
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 0.00017739737801346895,
+      "loss": 1.9562,
+      "step": 2875
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 0.00017584792025990287,
+      "loss": 1.9437,
+      "step": 2880
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 0.00017430381386758747,
+      "loss": 1.8848,
+      "step": 2885
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 0.00017276508432786304,
+      "loss": 1.9044,
+      "step": 2890
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 0.00017123175704330512,
+      "loss": 1.8581,
+      "step": 2895
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 0.00016970385732730414,
+      "loss": 1.9017,
+      "step": 2900
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 0.00016818141040364815,
+      "loss": 1.7544,
+      "step": 2905
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 0.00016666444140610615,
+      "loss": 1.92,
+      "step": 2910
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 0.00016515297537801304,
+      "loss": 1.8981,
+      "step": 2915
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 0.00016364703727185664,
+      "loss": 1.97,
+      "step": 2920
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 0.00016214665194886474,
+      "loss": 1.8704,
+      "step": 2925
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 0.00016065184417859613,
+      "loss": 1.9395,
+      "step": 2930
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 0.00015916263863853,
+      "loss": 1.9092,
+      "step": 2935
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 0.0001576790599136607,
+      "loss": 1.8171,
+      "step": 2940
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 0.00015620113249608942,
+      "loss": 1.8367,
+      "step": 2945
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 0.00015472888078462172,
+      "loss": 1.9402,
+      "step": 2950
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 0.00015326232908436405,
+      "loss": 1.9839,
+      "step": 2955
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 0.00015180150160632195,
+      "loss": 1.8956,
+      "step": 2960
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 0.00015034642246700204,
+      "loss": 1.8861,
+      "step": 2965
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 0.00014889711568801152,
+      "loss": 2.0093,
+      "step": 2970
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 0.00014745360519566382,
+      "loss": 1.9242,
+      "step": 2975
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 0.0001460159148205819,
+      "loss": 1.9005,
+      "step": 2980
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 0.0001445840682973062,
+      "loss": 1.9462,
+      "step": 2985
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 0.00014315808926390188,
+      "loss": 2.0249,
+      "step": 2990
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 0.00014173800126156917,
+      "loss": 1.9317,
+      "step": 2995
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 0.00014032382773425384,
+      "loss": 2.0039,
+      "step": 3000
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 0.00013891559202826133,
+      "loss": 1.9658,
+      "step": 3005
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 0.0001375133173918709,
+      "loss": 1.9448,
+      "step": 3010
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 0.00013611702697495086,
+      "loss": 1.8653,
+      "step": 3015
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 0.0001347267438285786,
+      "loss": 1.9222,
+      "step": 3020
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 0.0001333424909046574,
+      "loss": 1.8796,
+      "step": 3025
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 0.0001319642910555396,
+      "loss": 1.8942,
+      "step": 3030
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 0.00013059216703364813,
+      "loss": 1.9026,
+      "step": 3035
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 0.00012922614149110152,
+      "loss": 1.9212,
+      "step": 3040
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 0.0001278662369793398,
+      "loss": 1.9068,
+      "step": 3045
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 0.000126512475948752,
+      "loss": 1.9956,
+      "step": 3050
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 0.00012516488074830584,
+      "loss": 1.8775,
+      "step": 3055
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 0.00012382347362517815,
+      "loss": 1.9005,
+      "step": 3060
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 0.00012248827672438868,
+      "loss": 1.9538,
+      "step": 3065
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 0.00012115931208843318,
+      "loss": 1.8309,
+      "step": 3070
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 0.00011983660165692079,
+      "loss": 2.0001,
+      "step": 3075
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 0.00011852016726621096,
+      "loss": 1.8293,
+      "step": 3080
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 0.00011721003064905328,
+      "loss": 1.9337,
+      "step": 3085
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 0.0001159062134342289,
+      "loss": 1.912,
+      "step": 3090
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 0.00011460873714619274,
+      "loss": 1.8585,
+      "step": 3095
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 0.00011331762320471906,
+      "loss": 1.9535,
+      "step": 3100
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 0.00011203289292454726,
+      "loss": 1.9018,
+      "step": 3105
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 0.00011075456751503043,
+      "loss": 1.9418,
+      "step": 3110
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 0.0001094826680797843,
+      "loss": 1.9581,
+      "step": 3115
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 0.00010821721561634024,
+      "loss": 1.9045,
+      "step": 3120
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 0.00010695823101579727,
+      "loss": 1.8314,
+      "step": 3125
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 0.00010570573506247811,
+      "loss": 1.9717,
+      "step": 3130
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 0.00010445974843358564,
+      "loss": 1.9037,
+      "step": 3135
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 0.0001032202916988616,
+      "loss": 1.8119,
+      "step": 3140
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.00010198738532024716,
+      "loss": 1.8549,
+      "step": 3145
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.00010076104965154459,
+      "loss": 1.8318,
+      "step": 3150
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 9.9541304938082e-05,
+      "loss": 1.9885,
+      "step": 3155
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 2.0513949394226074,
+      "eval_runtime": 95.3176,
+      "eval_samples_per_second": 134.372,
+      "eval_steps_per_second": 16.796,
+      "step": 3156
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 9.832817131637867e-05,
+      "loss": 1.9155,
+      "step": 3160
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 9.71216688138128e-05,
+      "loss": 1.8082,
+      "step": 3165
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.592181734829036e-05,
+      "loss": 1.8647,
+      "step": 3170
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.472863672791721e-05,
+      "loss": 1.8277,
+      "step": 3175
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.35421466506714e-05,
+      "loss": 1.9529,
+      "step": 3180
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 9.236236670407771e-05,
+      "loss": 1.8167,
+      "step": 3185
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 9.11893163648857e-05,
+      "loss": 1.9159,
+      "step": 3190
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 9.002301499874621e-05,
+      "loss": 1.7958,
+      "step": 3195
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8.886348185989334e-05,
+      "loss": 1.865,
+      "step": 3200
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8.771073609082531e-05,
+      "loss": 1.859,
+      "step": 3205
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 8.656479672198947e-05,
+      "loss": 1.8602,
+      "step": 3210
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 8.54256826714676e-05,
+      "loss": 1.8442,
+      "step": 3215
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 8.429341274466368e-05,
+      "loss": 1.8399,
+      "step": 3220
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 8.316800563399307e-05,
+      "loss": 1.895,
+      "step": 3225
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 8.204947991857486e-05,
+      "loss": 1.8071,
+      "step": 3230
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.093785406392429e-05,
+      "loss": 1.9761,
+      "step": 3235
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 7.983314642164786e-05,
+      "loss": 1.8828,
+      "step": 3240
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 7.873537522914154e-05,
+      "loss": 1.9077,
+      "step": 3245
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 7.764455860928793e-05,
+      "loss": 1.867,
+      "step": 3250
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 7.656071457015878e-05,
+      "loss": 1.8226,
+      "step": 3255
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 7.548386100471621e-05,
+      "loss": 1.821,
+      "step": 3260
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 7.441401569051847e-05,
+      "loss": 1.8415,
+      "step": 3265
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 7.335119628942594e-05,
+      "loss": 1.8206,
+      "step": 3270
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 7.229542034730952e-05,
+      "loss": 1.8202,
+      "step": 3275
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 7.124670529376126e-05,
+      "loss": 1.9373,
+      "step": 3280
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 7.020506844180608e-05,
+      "loss": 1.8372,
+      "step": 3285
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 6.917052698761666e-05,
+      "loss": 1.8923,
+      "step": 3290
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 6.814309801022872e-05,
+      "loss": 1.8606,
+      "step": 3295
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 6.712279847126035e-05,
+      "loss": 1.8864,
+      "step": 3300
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 6.610964521463031e-05,
+      "loss": 1.8178,
+      "step": 3305
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 6.510365496628129e-05,
+      "loss": 1.8438,
+      "step": 3310
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 6.410484433390334e-05,
+      "loss": 1.79,
+      "step": 3315
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 6.311322980665934e-05,
+      "loss": 1.8641,
+      "step": 3320
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 6.212882775491352e-05,
+      "loss": 1.8712,
+      "step": 3325
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 6.115165442996056e-05,
+      "loss": 1.863,
+      "step": 3330
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 6.018172596375776e-05,
+      "loss": 1.8744,
+      "step": 3335
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 5.921905836865821e-05,
+      "loss": 1.8524,
+      "step": 3340
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 5.826366753714707e-05,
+      "loss": 1.9031,
+      "step": 3345
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 5.731556924157844e-05,
+      "loss": 1.784,
+      "step": 3350
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 5.6374779133916035e-05,
+      "loss": 1.8172,
+      "step": 3355
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 5.544131274547365e-05,
+      "loss": 1.8779,
+      "step": 3360
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 5.4515185486659454e-05,
+      "loss": 1.7402,
+      "step": 3365
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 5.359641264672155e-05,
+      "loss": 1.9493,
+      "step": 3370
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 5.268500939349513e-05,
+      "loss": 1.8279,
+      "step": 3375
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 5.178099077315257e-05,
+      "loss": 1.7599,
+      "step": 3380
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 5.0884371709954804e-05,
+      "loss": 1.7905,
+      "step": 3385
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 4.999516700600504e-05,
+      "loss": 1.8888,
+      "step": 3390
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 4.911339134100401e-05,
+      "loss": 1.8539,
+      "step": 3395
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 4.823905927200833e-05,
+      "loss": 1.8406,
+      "step": 3400
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 4.737218523318965e-05,
+      "loss": 1.8983,
+      "step": 3405
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 4.65127835355964e-05,
+      "loss": 1.7739,
+      "step": 3410
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 4.566086836691791e-05,
+      "loss": 1.8877,
+      "step": 3415
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 4.4816453791249436e-05,
+      "loss": 1.8403,
+      "step": 3420
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 4.3979553748861035e-05,
+      "loss": 1.9217,
+      "step": 3425
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 4.3150182055966384e-05,
+      "loss": 1.8122,
+      "step": 3430
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 4.232835240449534e-05,
+      "loss": 1.8977,
+      "step": 3435
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 4.1514078361867826e-05,
+      "loss": 1.9035,
+      "step": 3440
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 4.070737337076963e-05,
+      "loss": 1.8811,
+      "step": 3445
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 3.990825074893056e-05,
+      "loss": 1.9786,
+      "step": 3450
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 3.911672368890462e-05,
+      "loss": 1.9312,
+      "step": 3455
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 3.8332805257852374e-05,
+      "loss": 1.8757,
+      "step": 3460
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 3.755650839732489e-05,
+      "loss": 1.8269,
+      "step": 3465
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 3.67878459230504e-05,
+      "loss": 1.7784,
+      "step": 3470
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 3.6026830524722444e-05,
+      "loss": 1.8365,
+      "step": 3475
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 3.527347476579068e-05,
+      "loss": 1.9105,
+      "step": 3480
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.4527791083253154e-05,
+      "loss": 1.8909,
+      "step": 3485
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.3789791787451344e-05,
+      "loss": 1.9213,
+      "step": 3490
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 3.305948906186662e-05,
+      "loss": 1.7467,
+      "step": 3495
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.233689496291931e-05,
+      "loss": 1.8005,
+      "step": 3500
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.162202141976955e-05,
+      "loss": 1.9142,
+      "step": 3505
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.09148802341202e-05,
+      "loss": 1.8589,
+      "step": 3510
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.0215483080022478e-05,
+      "loss": 1.9262,
+      "step": 3515
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 2.9523841503682757e-05,
+      "loss": 1.8654,
+      "step": 3520
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 2.8839966923272288e-05,
+      "loss": 1.8914,
+      "step": 3525
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 2.8163870628738375e-05,
+      "loss": 1.8989,
+      "step": 3530
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 2.7495563781618328e-05,
+      "loss": 1.8157,
+      "step": 3535
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 2.6835057414855036e-05,
+      "loss": 1.8452,
+      "step": 3540
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 2.6182362432614513e-05,
+      "loss": 1.8583,
+      "step": 3545
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 2.5537489610106824e-05,
+      "loss": 1.7927,
+      "step": 3550
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 2.4900449593406984e-05,
+      "loss": 1.928,
+      "step": 3555
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 2.4271252899280173e-05,
+      "loss": 1.8615,
+      "step": 3560
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 2.364990991500743e-05,
+      "loss": 1.8917,
+      "step": 3565
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 2.3036430898214832e-05,
+      "loss": 1.7866,
+      "step": 3570
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 2.2430825976703484e-05,
+      "loss": 1.768,
+      "step": 3575
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.1833105148282905e-05,
+      "loss": 1.7868,
+      "step": 3580
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.1243278280605515e-05,
+      "loss": 1.8383,
+      "step": 3585
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.0661355111004032e-05,
+      "loss": 1.8518,
+      "step": 3590
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.0087345246330713e-05,
+      "loss": 1.939,
+      "step": 3595
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 1.9521258162798263e-05,
+      "loss": 1.8259,
+      "step": 3600
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 1.8963103205824395e-05,
+      "loss": 1.9559,
+      "step": 3605
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 1.8412889589876358e-05,
+      "loss": 1.8916,
+      "step": 3610
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 1.787062639831988e-05,
+      "loss": 1.8576,
+      "step": 3615
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 1.7336322583268336e-05,
+      "loss": 1.8634,
+      "step": 3620
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 1.6809986965435677e-05,
+      "loss": 1.9129,
+      "step": 3625
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.6291628233990387e-05,
+      "loss": 1.7839,
+      "step": 3630
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 1.5781254946412028e-05,
+      "loss": 1.8129,
+      "step": 3635
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 1.5278875528350277e-05,
+      "loss": 1.8645,
+      "step": 3640
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1.4784498273485436e-05,
+      "loss": 1.7878,
+      "step": 3645
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.4298131343391918e-05,
+      "loss": 1.8806,
+      "step": 3650
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.3819782767403033e-05,
+      "loss": 1.8544,
+      "step": 3655
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.3349460442479089e-05,
+      "loss": 1.9462,
+      "step": 3660
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1.2887172133076219e-05,
+      "loss": 1.8129,
+      "step": 3665
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1.243292547101893e-05,
+      "loss": 1.8055,
+      "step": 3670
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1.1986727955373588e-05,
+      "loss": 1.8297,
+      "step": 3675
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1.1548586952324912e-05,
+      "loss": 1.7961,
+      "step": 3680
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1.1118509695054235e-05,
+      "loss": 1.9098,
+      "step": 3685
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.0696503283620319e-05,
+      "loss": 1.7803,
+      "step": 3690
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.0282574684841784e-05,
+      "loss": 1.9406,
+      "step": 3695
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 9.876730732182203e-06,
+      "loss": 1.8219,
+      "step": 3700
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 9.478978125637583e-06,
+      "loss": 1.9694,
+      "step": 3705
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 9.089323431625273e-06,
+      "loss": 1.8647,
+      "step": 3710
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 8.70777308287618e-06,
+      "loss": 1.8981,
+      "step": 3715
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 8.334333378327785e-06,
+      "loss": 1.8218,
+      "step": 3720
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 7.969010483020845e-06,
+      "loss": 1.8954,
+      "step": 3725
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 7.611810427997412e-06,
+      "loss": 1.7638,
+      "step": 3730
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 7.262739110200922e-06,
+      "loss": 1.7664,
+      "step": 3735
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 6.921802292379376e-06,
+      "loss": 1.8825,
+      "step": 3740
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 6.589005602989862e-06,
+      "loss": 1.7941,
+      "step": 3745
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 6.264354536105799e-06,
+      "loss": 1.8809,
+      "step": 3750
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 5.947854451326007e-06,
+      "loss": 1.7973,
+      "step": 3755
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 5.639510573686613e-06,
+      "loss": 1.9447,
+      "step": 3760
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 5.339327993574339e-06,
+      "loss": 1.8651,
+      "step": 3765
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 5.047311666642851e-06,
+      "loss": 1.8057,
+      "step": 3770
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 4.763466413730821e-06,
+      "loss": 1.8872,
+      "step": 3775
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 4.4877969207821615e-06,
+      "loss": 1.7565,
+      "step": 3780
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 4.2203077387688586e-06,
+      "loss": 1.9462,
+      "step": 3785
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 3.961003283615816e-06,
+      "loss": 1.8327,
+      "step": 3790
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 3.709887836128023e-06,
+      "loss": 1.8064,
+      "step": 3795
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 3.4669655419196087e-06,
+      "loss": 1.8423,
+      "step": 3800
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 3.232240411345788e-06,
+      "loss": 1.887,
+      "step": 3805
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 3.0057163194363023e-06,
+      "loss": 1.7267,
+      "step": 3810
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 2.7873970058316934e-06,
+      "loss": 1.8499,
+      "step": 3815
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 2.5772860747214078e-06,
+      "loss": 1.8719,
+      "step": 3820
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.3753869947843455e-06,
+      "loss": 1.8829,
+      "step": 3825
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.1817030991317934e-06,
+      "loss": 1.8648,
+      "step": 3830
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 1.9962375852519698e-06,
+      "loss": 1.8077,
+      "step": 3835
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1.8189935149576786e-06,
+      "loss": 1.7893,
+      "step": 3840
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1.6499738143354593e-06,
+      "loss": 1.8117,
+      "step": 3845
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.4891812736974596e-06,
+      "loss": 1.8163,
+      "step": 3850
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.3366185475351955e-06,
+      "loss": 1.8515,
+      "step": 3855
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.192288154476029e-06,
+      "loss": 1.8055,
+      "step": 3860
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.0561924772412024e-06,
+      "loss": 1.8637,
+      "step": 3865
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 9.283337626068144e-07,
+      "loss": 1.8176,
+      "step": 3870
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 8.087141213665717e-07,
+      "loss": 1.9019,
+      "step": 3875
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 6.973355282969274e-07,
+      "loss": 1.8888,
+      "step": 3880
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 5.941998221247191e-07,
+      "loss": 1.8405,
+      "step": 3885
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 4.993087054963597e-07,
+      "loss": 1.8748,
+      "step": 3890
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 4.1266374495024795e-07,
+      "loss": 1.7451,
+      "step": 3895
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 3.342663708905125e-07,
+      "loss": 1.8469,
+      "step": 3900
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 2.6411787756353e-07,
+      "loss": 1.9005,
+      "step": 3905
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.0221942303666474e-07,
+      "loss": 1.8063,
+      "step": 3910
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1.4857202917900603e-07,
+      "loss": 1.8288,
+      "step": 3915
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.0317658164454846e-07,
+      "loss": 1.8788,
+      "step": 3920
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 6.603382985759243e-08,
+      "loss": 1.937,
+      "step": 3925
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 3.714438700025413e-08,
+      "loss": 1.7552,
+      "step": 3930
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 1.650873000258457e-08,
+      "loss": 1.803,
+      "step": 3935
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 4.1271995344094316e-09,
+      "loss": 1.8769,
+      "step": 3940
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.0,
+      "loss": 1.806,
+      "step": 3945
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 2.0400655269622803,
+      "eval_runtime": 95.5556,
+      "eval_samples_per_second": 134.037,
+      "eval_steps_per_second": 16.755,
+      "step": 3945
+    },
+    {
+      "epoch": 5.0,
+      "step": 3945,
+      "total_flos": 1.3195381386195763e+17,
+      "train_loss": 2.171095475799986,
+      "train_runtime": 13808.1422,
+      "train_samples_per_second": 36.573,
+      "train_steps_per_second": 0.286
+    }
+  ],
+  "max_steps": 3945,
+  "num_train_epochs": 5,
+  "total_flos": 1.3195381386195763e+17,
+  "trial_name": null,
+  "trial_params": null
+}