diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.3849855630413859,
+  "eval_steps": 500,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019249278152069297,
+      "grad_norm": 1.9932657480239868,
+      "learning_rate": 0.00016,
+      "loss": 2.962,
+      "step": 5
+    },
+    {
+      "epoch": 0.00038498556304138594,
+      "grad_norm": 1.2827370166778564,
+      "learning_rate": 0.000199999997073812,
+      "loss": 2.2983,
+      "step": 10
+    },
+    {
+      "epoch": 0.0005774783445620789,
+      "grad_norm": 1.3585917949676514,
+      "learning_rate": 0.0001999999851861734,
+      "loss": 1.9389,
+      "step": 15
+    },
+    {
+      "epoch": 0.0007699711260827719,
+      "grad_norm": 2.117544412612915,
+      "learning_rate": 0.00019999996415419864,
+      "loss": 1.6659,
+      "step": 20
+    },
+    {
+      "epoch": 0.0009624639076034649,
+      "grad_norm": 0.8802940249443054,
+      "learning_rate": 0.0001999999339778896,
+      "loss": 1.6015,
+      "step": 25
+    },
+    {
+      "epoch": 0.0011549566891241579,
+      "grad_norm": 1.256873369216919,
+      "learning_rate": 0.000199999894657249,
+      "loss": 1.7428,
+      "step": 30
+    },
+    {
+      "epoch": 0.001347449470644851,
+      "grad_norm": 1.9709804058074951,
+      "learning_rate": 0.0001999998461922805,
+      "loss": 1.4316,
+      "step": 35
+    },
+    {
+      "epoch": 0.0015399422521655437,
+      "grad_norm": 1.2085392475128174,
+      "learning_rate": 0.00019999978858298848,
+      "loss": 1.8392,
+      "step": 40
+    },
+    {
+      "epoch": 0.0017324350336862368,
+      "grad_norm": 0.9966161847114563,
+      "learning_rate": 0.00019999972182937827,
+      "loss": 1.6381,
+      "step": 45
+    },
+    {
+      "epoch": 0.0019249278152069298,
+      "grad_norm": 1.5572378635406494,
+      "learning_rate": 0.0001999996459314559,
+      "loss": 1.6214,
+      "step": 50
+    },
+    {
+      "epoch": 0.0021174205967276227,
+      "grad_norm": 0.9813450574874878,
+      "learning_rate": 0.00019999956088922837,
+      "loss": 1.5337,
+      "step": 55
+    },
+    {
+      "epoch": 0.0023099133782483157,
+      "grad_norm": 1.140754222869873,
+      "learning_rate": 0.00019999946670270341,
+      "loss": 1.5865,
+      "step": 60
+    },
+    {
+      "epoch": 0.0025024061597690088,
+      "grad_norm": 1.7033613920211792,
+      "learning_rate": 0.0001999993633718897,
+      "loss": 1.5483,
+      "step": 65
+    },
+    {
+      "epoch": 0.002694898941289702,
+      "grad_norm": 0.8782416582107544,
+      "learning_rate": 0.00019999925089679658,
+      "loss": 1.7574,
+      "step": 70
+    },
+    {
+      "epoch": 0.0028873917228103944,
+      "grad_norm": 0.94110506772995,
+      "learning_rate": 0.00019999912927743445,
+      "loss": 1.747,
+      "step": 75
+    },
+    {
+      "epoch": 0.0030798845043310875,
+      "grad_norm": 2.9130144119262695,
+      "learning_rate": 0.00019999899851381436,
+      "loss": 1.5482,
+      "step": 80
+    },
+    {
+      "epoch": 0.0032723772858517805,
+      "grad_norm": 1.444981336593628,
+      "learning_rate": 0.00019999885860594828,
+      "loss": 1.7935,
+      "step": 85
+    },
+    {
+      "epoch": 0.0034648700673724736,
+      "grad_norm": 0.8361923098564148,
+      "learning_rate": 0.00019999870955384906,
+      "loss": 1.5566,
+      "step": 90
+    },
+    {
+      "epoch": 0.0036573628488931666,
+      "grad_norm": 1.0198391675949097,
+      "learning_rate": 0.00019999855135753025,
+      "loss": 1.6608,
+      "step": 95
+    },
+    {
+      "epoch": 0.0038498556304138597,
+      "grad_norm": 0.9720978736877441,
+      "learning_rate": 0.00019999838401700632,
+      "loss": 1.4217,
+      "step": 100
+    },
+    {
+      "epoch": 0.004042348411934553,
+      "grad_norm": 0.7735599279403687,
+      "learning_rate": 0.00019999820753229263,
+      "loss": 1.4195,
+      "step": 105
+    },
+    {
+      "epoch": 0.004234841193455245,
+      "grad_norm": 1.1776920557022095,
+      "learning_rate": 0.0001999980219034053,
+      "loss": 1.7147,
+      "step": 110
+    },
+    {
+      "epoch": 0.004427333974975939,
+      "grad_norm": 1.156069278717041,
+      "learning_rate": 0.0001999978271303613,
+      "loss": 1.7,
+      "step": 115
+    },
+    {
+      "epoch": 0.0046198267564966315,
+      "grad_norm": 1.2335503101348877,
+      "learning_rate": 0.0001999976232131784,
+      "loss": 1.3309,
+      "step": 120
+    },
+    {
+      "epoch": 0.004812319538017324,
+      "grad_norm": 1.0332967042922974,
+      "learning_rate": 0.0001999974101518753,
+      "loss": 1.7515,
+      "step": 125
+    },
+    {
+      "epoch": 0.0050048123195380175,
+      "grad_norm": 1.561087727546692,
+      "learning_rate": 0.00019999718794647145,
+      "loss": 1.5517,
+      "step": 130
+    },
+    {
+      "epoch": 0.00519730510105871,
+      "grad_norm": 1.3611408472061157,
+      "learning_rate": 0.00019999695659698717,
+      "loss": 1.5771,
+      "step": 135
+    },
+    {
+      "epoch": 0.005389797882579404,
+      "grad_norm": 1.5531154870986938,
+      "learning_rate": 0.0001999967161034437,
+      "loss": 1.4217,
+      "step": 140
+    },
+    {
+      "epoch": 0.005582290664100096,
+      "grad_norm": 1.5827676057815552,
+      "learning_rate": 0.00019999646646586287,
+      "loss": 1.611,
+      "step": 145
+    },
+    {
+      "epoch": 0.005774783445620789,
+      "grad_norm": 1.1693483591079712,
+      "learning_rate": 0.00019999620768426763,
+      "loss": 1.3961,
+      "step": 150
+    },
+    {
+      "epoch": 0.005967276227141482,
+      "grad_norm": 1.4277936220169067,
+      "learning_rate": 0.00019999593975868164,
+      "loss": 1.638,
+      "step": 155
+    },
+    {
+      "epoch": 0.006159769008662175,
+      "grad_norm": 1.2951083183288574,
+      "learning_rate": 0.00019999566268912933,
+      "loss": 1.6187,
+      "step": 160
+    },
+    {
+      "epoch": 0.0063522617901828685,
+      "grad_norm": 2.4885995388031006,
+      "learning_rate": 0.0001999953764756361,
+      "loss": 1.5669,
+      "step": 165
+    },
+    {
+      "epoch": 0.006544754571703561,
+      "grad_norm": 1.3352105617523193,
+      "learning_rate": 0.00019999508111822811,
+      "loss": 1.3157,
+      "step": 170
+    },
+    {
+      "epoch": 0.006737247353224254,
+      "grad_norm": 1.2560889720916748,
+      "learning_rate": 0.00019999477661693233,
+      "loss": 1.7011,
+      "step": 175
+    },
+    {
+      "epoch": 0.006929740134744947,
+      "grad_norm": 2.4167582988739014,
+      "learning_rate": 0.00019999446297177666,
+      "loss": 1.4827,
+      "step": 180
+    },
+    {
+      "epoch": 0.00712223291626564,
+      "grad_norm": 1.0598788261413574,
+      "learning_rate": 0.00019999414018278974,
+      "loss": 1.5718,
+      "step": 185
+    },
+    {
+      "epoch": 0.007314725697786333,
+      "grad_norm": 1.5576567649841309,
+      "learning_rate": 0.00019999380825000111,
+      "loss": 1.7717,
+      "step": 190
+    },
+    {
+      "epoch": 0.007507218479307026,
+      "grad_norm": 1.005711317062378,
+      "learning_rate": 0.0001999934671734411,
+      "loss": 1.5085,
+      "step": 195
+    },
+    {
+      "epoch": 0.007699711260827719,
+      "grad_norm": 1.7211413383483887,
+      "learning_rate": 0.00019999311695314095,
+      "loss": 1.623,
+      "step": 200
+    },
+    {
+      "epoch": 0.007892204042348411,
+      "grad_norm": 1.5765767097473145,
+      "learning_rate": 0.00019999275758913261,
+      "loss": 1.5982,
+      "step": 205
+    },
+    {
+      "epoch": 0.008084696823869105,
+      "grad_norm": 1.0989298820495605,
+      "learning_rate": 0.00019999238908144896,
+      "loss": 1.3306,
+      "step": 210
+    },
+    {
+      "epoch": 0.008277189605389798,
+      "grad_norm": 1.0234464406967163,
+      "learning_rate": 0.0001999920114301238,
+      "loss": 1.5856,
+      "step": 215
+    },
+    {
+      "epoch": 0.00846968238691049,
+      "grad_norm": 1.6681355237960815,
+      "learning_rate": 0.0001999916246351915,
+      "loss": 1.4777,
+      "step": 220
+    },
+    {
+      "epoch": 0.008662175168431183,
+      "grad_norm": 0.9723508358001709,
+      "learning_rate": 0.00019999122869668754,
+      "loss": 1.5357,
+      "step": 225
+    },
+    {
+      "epoch": 0.008854667949951878,
+      "grad_norm": 0.8840959072113037,
+      "learning_rate": 0.0001999908236146481,
+      "loss": 1.5296,
+      "step": 230
+    },
+    {
+      "epoch": 0.00904716073147257,
+      "grad_norm": 0.9913238883018494,
+      "learning_rate": 0.0001999904093891102,
+      "loss": 1.5846,
+      "step": 235
+    },
+    {
+      "epoch": 0.009239653512993263,
+      "grad_norm": 1.129952073097229,
+      "learning_rate": 0.00019998998602011178,
+      "loss": 1.4455,
+      "step": 240
+    },
+    {
+      "epoch": 0.009432146294513956,
+      "grad_norm": 1.0377521514892578,
+      "learning_rate": 0.00019998955350769148,
+      "loss": 1.4212,
+      "step": 245
+    },
+    {
+      "epoch": 0.009624639076034648,
+      "grad_norm": 2.2103137969970703,
+      "learning_rate": 0.00019998911185188886,
+      "loss": 1.5812,
+      "step": 250
+    },
+    {
+      "epoch": 0.009817131857555342,
+      "grad_norm": 0.8716953992843628,
+      "learning_rate": 0.00019998866105274437,
+      "loss": 1.5326,
+      "step": 255
+    },
+    {
+      "epoch": 0.010009624639076035,
+      "grad_norm": 1.1956042051315308,
+      "learning_rate": 0.00019998820111029916,
+      "loss": 1.7183,
+      "step": 260
+    },
+    {
+      "epoch": 0.010202117420596728,
+      "grad_norm": 2.747600555419922,
+      "learning_rate": 0.00019998773202459534,
+      "loss": 1.7952,
+      "step": 265
+    },
+    {
+      "epoch": 0.01039461020211742,
+      "grad_norm": 1.3412338495254517,
+      "learning_rate": 0.00019998725379567577,
+      "loss": 1.3538,
+      "step": 270
+    },
+    {
+      "epoch": 0.010587102983638113,
+      "grad_norm": 1.651822805404663,
+      "learning_rate": 0.00019998676642358422,
+      "loss": 1.5458,
+      "step": 275
+    },
+    {
+      "epoch": 0.010779595765158807,
+      "grad_norm": 1.3036198616027832,
+      "learning_rate": 0.00019998626990836522,
+      "loss": 1.7305,
+      "step": 280
+    },
+    {
+      "epoch": 0.0109720885466795,
+      "grad_norm": 0.8263657093048096,
+      "learning_rate": 0.00019998576425006416,
+      "loss": 1.3767,
+      "step": 285
+    },
+    {
+      "epoch": 0.011164581328200193,
+      "grad_norm": 2.022136926651001,
+      "learning_rate": 0.00019998524944872737,
+      "loss": 1.5823,
+      "step": 290
+    },
+    {
+      "epoch": 0.011357074109720885,
+      "grad_norm": 1.1224019527435303,
+      "learning_rate": 0.00019998472550440178,
+      "loss": 1.5723,
+      "step": 295
+    },
+    {
+      "epoch": 0.011549566891241578,
+      "grad_norm": 1.375664234161377,
+      "learning_rate": 0.00019998419241713542,
+      "loss": 1.5224,
+      "step": 300
+    },
+    {
+      "epoch": 0.011742059672762272,
+      "grad_norm": 1.2721813917160034,
+      "learning_rate": 0.000199983650186977,
+      "loss": 1.7217,
+      "step": 305
+    },
+    {
+      "epoch": 0.011934552454282965,
+      "grad_norm": 1.4723321199417114,
+      "learning_rate": 0.0001999830988139761,
+      "loss": 1.4666,
+      "step": 310
+    },
+    {
+      "epoch": 0.012127045235803657,
+      "grad_norm": 0.695198118686676,
+      "learning_rate": 0.00019998253829818315,
+      "loss": 1.2672,
+      "step": 315
+    },
+    {
+      "epoch": 0.01231953801732435,
+      "grad_norm": 1.716638207435608,
+      "learning_rate": 0.00019998196863964937,
+      "loss": 1.3461,
+      "step": 320
+    },
+    {
+      "epoch": 0.012512030798845043,
+      "grad_norm": 1.1060154438018799,
+      "learning_rate": 0.0001999813898384269,
+      "loss": 1.3816,
+      "step": 325
+    },
+    {
+      "epoch": 0.012704523580365737,
+      "grad_norm": 1.6124354600906372,
+      "learning_rate": 0.00019998080189456862,
+      "loss": 1.5232,
+      "step": 330
+    },
+    {
+      "epoch": 0.01289701636188643,
+      "grad_norm": 1.5060306787490845,
+      "learning_rate": 0.00019998020480812832,
+      "loss": 1.5767,
+      "step": 335
+    },
+    {
+      "epoch": 0.013089509143407122,
+      "grad_norm": 1.1920175552368164,
+      "learning_rate": 0.00019997959857916063,
+      "loss": 1.6112,
+      "step": 340
+    },
+    {
+      "epoch": 0.013282001924927815,
+      "grad_norm": 1.1669896841049194,
+      "learning_rate": 0.00019997898320772096,
+      "loss": 1.3679,
+      "step": 345
+    },
+    {
+      "epoch": 0.013474494706448507,
+      "grad_norm": 1.1692086458206177,
+      "learning_rate": 0.00019997835869386553,
+      "loss": 1.4147,
+      "step": 350
+    },
+    {
+      "epoch": 0.013666987487969202,
+      "grad_norm": 2.0466034412384033,
+      "learning_rate": 0.00019997772503765153,
+      "loss": 1.5261,
+      "step": 355
+    },
+    {
+      "epoch": 0.013859480269489894,
+      "grad_norm": 1.1581529378890991,
+      "learning_rate": 0.00019997708223913686,
+      "loss": 1.5441,
+      "step": 360
+    },
+    {
+      "epoch": 0.014051973051010587,
+      "grad_norm": 1.4370143413543701,
+      "learning_rate": 0.0001999764302983803,
+      "loss": 1.651,
+      "step": 365
+    },
+    {
+      "epoch": 0.01424446583253128,
+      "grad_norm": 0.998635470867157,
+      "learning_rate": 0.00019997576921544147,
+      "loss": 1.4311,
+      "step": 370
+    },
+    {
+      "epoch": 0.014436958614051972,
+      "grad_norm": 1.2625153064727783,
+      "learning_rate": 0.00019997509899038086,
+      "loss": 1.4634,
+      "step": 375
+    },
+    {
+      "epoch": 0.014629451395572667,
+      "grad_norm": 1.171949863433838,
+      "learning_rate": 0.00019997441962325968,
+      "loss": 1.2474,
+      "step": 380
+    },
+    {
+      "epoch": 0.01482194417709336,
+      "grad_norm": 1.4312052726745605,
+      "learning_rate": 0.00019997373111414009,
+      "loss": 1.4814,
+      "step": 385
+    },
+    {
+      "epoch": 0.015014436958614052,
+      "grad_norm": 1.1508846282958984,
+      "learning_rate": 0.00019997303346308508,
+      "loss": 1.6291,
+      "step": 390
+    },
+    {
+      "epoch": 0.015206929740134744,
+      "grad_norm": 1.2096014022827148,
+      "learning_rate": 0.0001999723266701584,
+      "loss": 1.5507,
+      "step": 395
+    },
+    {
+      "epoch": 0.015399422521655439,
+      "grad_norm": 0.996391773223877,
+      "learning_rate": 0.00019997161073542473,
+      "loss": 1.6402,
+      "step": 400
+    },
+    {
+      "epoch": 0.015591915303176131,
+      "grad_norm": 1.6977828741073608,
+      "learning_rate": 0.00019997088565894947,
+      "loss": 1.5706,
+      "step": 405
+    },
+    {
+      "epoch": 0.015784408084696822,
+      "grad_norm": 1.4707343578338623,
+      "learning_rate": 0.000199970151440799,
+      "loss": 1.6348,
+      "step": 410
+    },
+    {
+      "epoch": 0.015976900866217517,
+      "grad_norm": 1.5461647510528564,
+      "learning_rate": 0.0001999694080810404,
+      "loss": 1.4836,
+      "step": 415
+    },
+    {
+      "epoch": 0.01616939364773821,
+      "grad_norm": 1.6253695487976074,
+      "learning_rate": 0.00019996865557974166,
+      "loss": 1.5834,
+      "step": 420
+    },
+    {
+      "epoch": 0.016361886429258902,
+      "grad_norm": 1.671321988105774,
+      "learning_rate": 0.00019996789393697165,
+      "loss": 1.3816,
+      "step": 425
+    },
+    {
+      "epoch": 0.016554379210779596,
+      "grad_norm": 0.9412807822227478,
+      "learning_rate": 0.00019996712315279992,
+      "loss": 1.443,
+      "step": 430
+    },
+    {
+      "epoch": 0.016746871992300287,
+      "grad_norm": 0.8705793023109436,
+      "learning_rate": 0.000199966343227297,
+      "loss": 1.4938,
+      "step": 435
+    },
+    {
+      "epoch": 0.01693936477382098,
+      "grad_norm": 1.6019854545593262,
+      "learning_rate": 0.00019996555416053422,
+      "loss": 1.3622,
+      "step": 440
+    },
+    {
+      "epoch": 0.017131857555341676,
+      "grad_norm": 1.0340136289596558,
+      "learning_rate": 0.00019996475595258372,
+      "loss": 1.5803,
+      "step": 445
+    },
+    {
+      "epoch": 0.017324350336862367,
+      "grad_norm": 1.4469108581542969,
+      "learning_rate": 0.0001999639486035185,
+      "loss": 1.6074,
+      "step": 450
+    },
+    {
+      "epoch": 0.01751684311838306,
+      "grad_norm": 1.3311457633972168,
+      "learning_rate": 0.00019996313211341238,
+      "loss": 1.5337,
+      "step": 455
+    },
+    {
+      "epoch": 0.017709335899903755,
+      "grad_norm": 0.9691542387008667,
+      "learning_rate": 0.00019996230648234003,
+      "loss": 1.3835,
+      "step": 460
+    },
+    {
+      "epoch": 0.017901828681424446,
+      "grad_norm": 1.0229564905166626,
+      "learning_rate": 0.00019996147171037691,
+      "loss": 1.4925,
+      "step": 465
+    },
+    {
+      "epoch": 0.01809432146294514,
+      "grad_norm": 1.0120052099227905,
+      "learning_rate": 0.00019996062779759942,
+      "loss": 1.4781,
+      "step": 470
+    },
+    {
+      "epoch": 0.01828681424446583,
+      "grad_norm": 0.8471246361732483,
+      "learning_rate": 0.00019995977474408468,
+      "loss": 1.4961,
+      "step": 475
+    },
+    {
+      "epoch": 0.018479307025986526,
+      "grad_norm": 2.020277261734009,
+      "learning_rate": 0.00019995891254991072,
+      "loss": 1.5299,
+      "step": 480
+    },
+    {
+      "epoch": 0.01867179980750722,
+      "grad_norm": 1.2169212102890015,
+      "learning_rate": 0.00019995804121515637,
+      "loss": 1.4626,
+      "step": 485
+    },
+    {
+      "epoch": 0.01886429258902791,
+      "grad_norm": 2.31048321723938,
+      "learning_rate": 0.00019995716073990133,
+      "loss": 1.3653,
+      "step": 490
+    },
+    {
+      "epoch": 0.019056785370548605,
+      "grad_norm": 1.8170429468154907,
+      "learning_rate": 0.0001999562711242261,
+      "loss": 1.3537,
+      "step": 495
+    },
+    {
+      "epoch": 0.019249278152069296,
+      "grad_norm": 1.1187188625335693,
+      "learning_rate": 0.00019995537236821198,
+      "loss": 1.6358,
+      "step": 500
+    },
+    {
+      "epoch": 0.01944177093358999,
+      "grad_norm": 1.2112963199615479,
+      "learning_rate": 0.0001999544644719412,
+      "loss": 1.4565,
+      "step": 505
+    },
+    {
+      "epoch": 0.019634263715110685,
+      "grad_norm": 1.3345009088516235,
+      "learning_rate": 0.0001999535474354968,
+      "loss": 1.647,
+      "step": 510
+    },
+    {
+      "epoch": 0.019826756496631376,
+      "grad_norm": 1.3109021186828613,
+      "learning_rate": 0.00019995262125896266,
+      "loss": 1.5462,
+      "step": 515
+    },
+    {
+      "epoch": 0.02001924927815207,
+      "grad_norm": 1.1681957244873047,
+      "learning_rate": 0.00019995168594242338,
+      "loss": 1.5292,
+      "step": 520
+    },
+    {
+      "epoch": 0.02021174205967276,
+      "grad_norm": 0.9509350657463074,
+      "learning_rate": 0.00019995074148596457,
+      "loss": 1.5566,
+      "step": 525
+    },
+    {
+      "epoch": 0.020404234841193455,
+      "grad_norm": 0.6594029664993286,
+      "learning_rate": 0.00019994978788967255,
+      "loss": 1.3693,
+      "step": 530
+    },
+    {
+      "epoch": 0.02059672762271415,
+      "grad_norm": 0.8029458522796631,
+      "learning_rate": 0.00019994882515363452,
+      "loss": 1.4664,
+      "step": 535
+    },
+    {
+      "epoch": 0.02078922040423484,
+      "grad_norm": 1.1551908254623413,
+      "learning_rate": 0.00019994785327793856,
+      "loss": 1.5342,
+      "step": 540
+    },
+    {
+      "epoch": 0.020981713185755535,
+      "grad_norm": 1.3600980043411255,
+      "learning_rate": 0.0001999468722626735,
+      "loss": 1.5262,
+      "step": 545
+    },
+    {
+      "epoch": 0.021174205967276226,
+      "grad_norm": 1.0333319902420044,
+      "learning_rate": 0.00019994588210792906,
+      "loss": 1.5079,
+      "step": 550
+    },
+    {
+      "epoch": 0.02136669874879692,
+      "grad_norm": 1.2757694721221924,
+      "learning_rate": 0.00019994488281379578,
+      "loss": 1.7721,
+      "step": 555
+    },
+    {
+      "epoch": 0.021559191530317615,
+      "grad_norm": 1.1292661428451538,
+      "learning_rate": 0.00019994387438036505,
+      "loss": 1.5077,
+      "step": 560
+    },
+    {
+      "epoch": 0.021751684311838305,
+      "grad_norm": 1.105522871017456,
+      "learning_rate": 0.00019994285680772906,
+      "loss": 1.6468,
+      "step": 565
+    },
+    {
+      "epoch": 0.021944177093359,
+      "grad_norm": 1.6378583908081055,
+      "learning_rate": 0.00019994183009598086,
+      "loss": 1.5432,
+      "step": 570
+    },
+    {
+      "epoch": 0.02213666987487969,
+      "grad_norm": 0.931384801864624,
+      "learning_rate": 0.0001999407942452144,
+      "loss": 1.3818,
+      "step": 575
+    },
+    {
+      "epoch": 0.022329162656400385,
+      "grad_norm": 1.0986119508743286,
+      "learning_rate": 0.0001999397492555243,
+      "loss": 1.552,
+      "step": 580
+    },
+    {
+      "epoch": 0.02252165543792108,
+      "grad_norm": 1.121957540512085,
+      "learning_rate": 0.00019993869512700623,
+      "loss": 1.5241,
+      "step": 585
+    },
+    {
+      "epoch": 0.02271414821944177,
+      "grad_norm": 1.2508270740509033,
+      "learning_rate": 0.00019993763185975646,
+      "loss": 1.6431,
+      "step": 590
+    },
+    {
+      "epoch": 0.022906641000962465,
+      "grad_norm": 1.293603777885437,
+      "learning_rate": 0.00019993655945387234,
+      "loss": 1.3788,
+      "step": 595
+    },
+    {
+      "epoch": 0.023099133782483156,
+      "grad_norm": 1.3218696117401123,
+      "learning_rate": 0.00019993547790945183,
+      "loss": 1.398,
+      "step": 600
+    },
+    {
+      "epoch": 0.02329162656400385,
+      "grad_norm": 0.8816308975219727,
+      "learning_rate": 0.0001999343872265939,
+      "loss": 1.4239,
+      "step": 605
+    },
+    {
+      "epoch": 0.023484119345524544,
+      "grad_norm": 1.9127452373504639,
+      "learning_rate": 0.00019993328740539824,
+      "loss": 1.549,
+      "step": 610
+    },
+    {
+      "epoch": 0.023676612127045235,
+      "grad_norm": 2.071992874145508,
+      "learning_rate": 0.0001999321784459655,
+      "loss": 1.6769,
+      "step": 615
+    },
+    {
+      "epoch": 0.02386910490856593,
+      "grad_norm": 1.335153579711914,
+      "learning_rate": 0.000199931060348397,
+      "loss": 1.6157,
+      "step": 620
+    },
+    {
+      "epoch": 0.02406159769008662,
+      "grad_norm": 1.1237496137619019,
+      "learning_rate": 0.000199929933112795,
+      "loss": 1.4733,
+      "step": 625
+    },
+    {
+      "epoch": 0.024254090471607315,
+      "grad_norm": 1.2557927370071411,
+      "learning_rate": 0.00019992879673926258,
+      "loss": 1.3888,
+      "step": 630
+    },
+    {
+      "epoch": 0.02444658325312801,
+      "grad_norm": 1.0877735614776611,
+      "learning_rate": 0.00019992765122790371,
+      "loss": 1.4241,
+      "step": 635
+    },
+    {
+      "epoch": 0.0246390760346487,
+      "grad_norm": 1.0029325485229492,
+      "learning_rate": 0.00019992649657882307,
+      "loss": 1.6504,
+      "step": 640
+    },
+    {
+      "epoch": 0.024831568816169394,
+      "grad_norm": 1.5832372903823853,
+      "learning_rate": 0.00019992533279212626,
+      "loss": 1.4662,
+      "step": 645
+    },
+    {
+      "epoch": 0.025024061597690085,
+      "grad_norm": 1.1658433675765991,
+      "learning_rate": 0.00019992415986791974,
+      "loss": 1.3723,
+      "step": 650
+    },
+    {
+      "epoch": 0.02521655437921078,
+      "grad_norm": 1.8895657062530518,
+      "learning_rate": 0.00019992297780631072,
+      "loss": 1.457,
+      "step": 655
+    },
+    {
+      "epoch": 0.025409047160731474,
+      "grad_norm": 1.193961501121521,
+      "learning_rate": 0.00019992178660740732,
+      "loss": 1.623,
+      "step": 660
+    },
+    {
+      "epoch": 0.025601539942252165,
+      "grad_norm": 0.9851275086402893,
+      "learning_rate": 0.00019992058627131844,
+      "loss": 1.6884,
+      "step": 665
+    },
+    {
+      "epoch": 0.02579403272377286,
+      "grad_norm": 1.5353829860687256,
+      "learning_rate": 0.00019991937679815386,
+      "loss": 1.3246,
+      "step": 670
+    },
+    {
+      "epoch": 0.02598652550529355,
+      "grad_norm": 1.2476325035095215,
+      "learning_rate": 0.0001999181581880242,
+      "loss": 1.596,
+      "step": 675
+    },
+    {
+      "epoch": 0.026179018286814244,
+      "grad_norm": 1.1163430213928223,
+      "learning_rate": 0.00019991693044104083,
+      "loss": 1.5077,
+      "step": 680
+    },
+    {
+      "epoch": 0.02637151106833494,
+      "grad_norm": 1.1388076543807983,
+      "learning_rate": 0.0001999156935573161,
+      "loss": 1.4827,
+      "step": 685
+    },
+    {
+      "epoch": 0.02656400384985563,
+      "grad_norm": 0.9100907444953918,
+      "learning_rate": 0.00019991444753696304,
+      "loss": 1.3429,
+      "step": 690
+    },
+    {
+      "epoch": 0.026756496631376324,
+      "grad_norm": 2.032510995864868,
+      "learning_rate": 0.00019991319238009565,
+      "loss": 1.5473,
+      "step": 695
+    },
+    {
+      "epoch": 0.026948989412897015,
+      "grad_norm": 1.0866800546646118,
+      "learning_rate": 0.00019991192808682868,
+      "loss": 1.5552,
+      "step": 700
+    },
+    {
+      "epoch": 0.02714148219441771,
+      "grad_norm": 1.3941971063613892,
+      "learning_rate": 0.00019991065465727774,
+      "loss": 1.4103,
+      "step": 705
+    },
+    {
+      "epoch": 0.027333974975938403,
+      "grad_norm": 1.721247911453247,
+      "learning_rate": 0.0001999093720915593,
+      "loss": 1.4965,
+      "step": 710
+    },
+    {
+      "epoch": 0.027526467757459094,
+      "grad_norm": 1.4090749025344849,
+      "learning_rate": 0.00019990808038979058,
+      "loss": 1.3159,
+      "step": 715
+    },
+    {
+      "epoch": 0.02771896053897979,
+      "grad_norm": 1.731886625289917,
+      "learning_rate": 0.00019990677955208973,
+      "loss": 1.4392,
+      "step": 720
+    },
+    {
+      "epoch": 0.02791145332050048,
+      "grad_norm": 1.9695488214492798,
+      "learning_rate": 0.00019990546957857576,
+      "loss": 1.6206,
+      "step": 725
+    },
+    {
+      "epoch": 0.028103946102021174,
+      "grad_norm": 0.7977893352508545,
+      "learning_rate": 0.0001999041504693684,
+      "loss": 1.5764,
+      "step": 730
+    },
+    {
+      "epoch": 0.02829643888354187,
+      "grad_norm": 0.9448668360710144,
+      "learning_rate": 0.00019990282222458826,
+      "loss": 1.3149,
+      "step": 735
+    },
+    {
+      "epoch": 0.02848893166506256,
+      "grad_norm": 1.0612679719924927,
+      "learning_rate": 0.00019990148484435682,
+      "loss": 1.4942,
+      "step": 740
+    },
+    {
+      "epoch": 0.028681424446583254,
+      "grad_norm": 1.4038052558898926,
+      "learning_rate": 0.0001999001383287964,
+      "loss": 1.5184,
+      "step": 745
+    },
+    {
+      "epoch": 0.028873917228103944,
+      "grad_norm": 1.0545177459716797,
+      "learning_rate": 0.0001998987826780301,
+      "loss": 1.5617,
+      "step": 750
+    },
+    {
+      "epoch": 0.02906641000962464,
+      "grad_norm": 2.392878532409668,
+      "learning_rate": 0.0001998974178921819,
+      "loss": 1.3638,
+      "step": 755
+    },
+    {
+      "epoch": 0.029258902791145333,
+      "grad_norm": 1.1004624366760254,
+      "learning_rate": 0.0001998960439713766,
+      "loss": 1.5162,
+      "step": 760
+    },
+    {
+      "epoch": 0.029451395572666024,
+      "grad_norm": 1.2530279159545898,
+      "learning_rate": 0.0001998946609157398,
+      "loss": 1.5422,
+      "step": 765
+    },
+    {
+      "epoch": 0.02964388835418672,
+      "grad_norm": 0.8240470290184021,
+      "learning_rate": 0.00019989326872539803,
+      "loss": 1.3828,
+      "step": 770
+    },
+    {
+      "epoch": 0.029836381135707413,
+      "grad_norm": 0.9734111428260803,
+      "learning_rate": 0.00019989186740047857,
+      "loss": 1.7041,
+      "step": 775
+    },
+    {
+      "epoch": 0.030028873917228104,
+      "grad_norm": 0.9785217642784119,
+      "learning_rate": 0.00019989045694110953,
+      "loss": 1.6267,
+      "step": 780
+    },
+    {
+      "epoch": 0.030221366698748798,
+      "grad_norm": 1.3278164863586426,
+      "learning_rate": 0.00019988903734741994,
+      "loss": 1.5041,
+      "step": 785
+    },
+    {
+      "epoch": 0.03041385948026949,
+      "grad_norm": 1.9143437147140503,
+      "learning_rate": 0.00019988760861953958,
+      "loss": 1.4728,
+      "step": 790
+    },
+    {
+      "epoch": 0.030606352261790183,
+      "grad_norm": 1.5717315673828125,
+      "learning_rate": 0.0001998861707575991,
+      "loss": 1.3824,
+      "step": 795
+    },
+    {
+      "epoch": 0.030798845043310877,
+      "grad_norm": 1.0486010313034058,
+      "learning_rate": 0.00019988472376173,
+      "loss": 1.6186,
+      "step": 800
+    },
+    {
+      "epoch": 0.03099133782483157,
+      "grad_norm": 1.1566083431243896,
+      "learning_rate": 0.00019988326763206458,
+      "loss": 1.3773,
+      "step": 805
+    },
+    {
+      "epoch": 0.031183830606352263,
+      "grad_norm": 1.6336543560028076,
+      "learning_rate": 0.00019988180236873602,
+      "loss": 1.2998,
+      "step": 810
+    },
+    {
+      "epoch": 0.031376323387872954,
+      "grad_norm": 1.4655206203460693,
+      "learning_rate": 0.00019988032797187824,
+      "loss": 1.3966,
+      "step": 815
+    },
+    {
+      "epoch": 0.031568816169393644,
+      "grad_norm": 2.0325050354003906,
+      "learning_rate": 0.00019987884444162618,
+      "loss": 1.3464,
+      "step": 820
+    },
+    {
+      "epoch": 0.03176130895091434,
+      "grad_norm": 1.254342794418335,
+      "learning_rate": 0.0001998773517781154,
+      "loss": 1.5236,
+      "step": 825
+    },
+    {
+      "epoch": 0.03195380173243503,
+      "grad_norm": 0.8909908533096313,
+      "learning_rate": 0.00019987584998148244,
+      "loss": 1.4838,
+      "step": 830
+    },
+    {
+      "epoch": 0.032146294513955724,
+      "grad_norm": 1.1440258026123047,
+      "learning_rate": 0.00019987433905186458,
+      "loss": 1.3952,
+      "step": 835
+    },
+    {
+      "epoch": 0.03233878729547642,
+      "grad_norm": 1.2138668298721313,
+      "learning_rate": 0.00019987281898940003,
+      "loss": 1.5982,
+      "step": 840
+    },
+    {
+      "epoch": 0.03253128007699711,
+      "grad_norm": 1.1847470998764038,
+      "learning_rate": 0.00019987128979422782,
+      "loss": 1.4313,
+      "step": 845
+    },
+    {
+      "epoch": 0.032723772858517804,
+      "grad_norm": 1.4961762428283691,
+      "learning_rate": 0.0001998697514664877,
+      "loss": 1.5187,
+      "step": 850
+    },
+    {
+      "epoch": 0.0329162656400385,
+      "grad_norm": 1.4735344648361206,
+      "learning_rate": 0.00019986820400632043,
+      "loss": 1.5443,
+      "step": 855
+    },
+    {
+      "epoch": 0.03310875842155919,
+      "grad_norm": 1.1350771188735962,
+      "learning_rate": 0.00019986664741386743,
+      "loss": 1.5219,
+      "step": 860
+    },
+    {
+      "epoch": 0.03330125120307988,
+      "grad_norm": 1.098781943321228,
+      "learning_rate": 0.0001998650816892711,
+      "loss": 1.6074,
+      "step": 865
+    },
+    {
+      "epoch": 0.033493743984600574,
+      "grad_norm": 1.9639078378677368,
+      "learning_rate": 0.0001998635068326746,
+      "loss": 1.342,
+      "step": 870
+    },
+    {
+      "epoch": 0.03368623676612127,
+      "grad_norm": 1.1193336248397827,
+      "learning_rate": 0.00019986192284422193,
+      "loss": 1.5647,
+      "step": 875
+    },
+    {
+      "epoch": 0.03387872954764196,
+      "grad_norm": 1.0558106899261475,
+      "learning_rate": 0.00019986032972405793,
+      "loss": 1.2448,
+      "step": 880
+    },
+    {
+      "epoch": 0.034071222329162654,
+      "grad_norm": 1.1178051233291626,
+      "learning_rate": 0.0001998587274723283,
+      "loss": 1.3455,
+      "step": 885
+    },
+    {
+      "epoch": 0.03426371511068335,
+      "grad_norm": 1.728400468826294,
+      "learning_rate": 0.0001998571160891795,
+      "loss": 1.44,
+      "step": 890
+    },
+    {
+      "epoch": 0.03445620789220404,
+      "grad_norm": 1.158931016921997,
+      "learning_rate": 0.000199855495574759,
+      "loss": 1.4247,
+      "step": 895
+    },
+    {
+      "epoch": 0.03464870067372473,
+      "grad_norm": 1.8745627403259277,
+      "learning_rate": 0.0001998538659292149,
+      "loss": 1.4036,
+      "step": 900
+    },
+    {
+      "epoch": 0.03484119345524543,
+      "grad_norm": 1.4273000955581665,
+      "learning_rate": 0.0001998522271526962,
+      "loss": 1.4857,
+      "step": 905
+    },
+    {
+      "epoch": 0.03503368623676612,
+      "grad_norm": 1.1671931743621826,
+      "learning_rate": 0.0001998505792453528,
+      "loss": 1.7199,
+      "step": 910
+    },
+    {
+      "epoch": 0.03522617901828681,
+      "grad_norm": 1.1703475713729858,
+      "learning_rate": 0.00019984892220733537,
+      "loss": 1.5659,
+      "step": 915
+    },
+    {
+      "epoch": 0.03541867179980751,
+      "grad_norm": 0.8550274968147278,
+      "learning_rate": 0.00019984725603879546,
+      "loss": 1.3608,
+      "step": 920
+    },
+    {
+      "epoch": 0.0356111645813282,
+      "grad_norm": 1.676072359085083,
+      "learning_rate": 0.0001998455807398854,
+      "loss": 1.4841,
+      "step": 925
+    },
+    {
+      "epoch": 0.03580365736284889,
+      "grad_norm": 1.362423062324524,
+      "learning_rate": 0.00019984389631075842,
+      "loss": 1.5501,
+      "step": 930
+    },
+    {
+      "epoch": 0.03599615014436958,
+      "grad_norm": 1.1643259525299072,
+      "learning_rate": 0.0001998422027515685,
+      "loss": 1.4954,
+      "step": 935
+    },
+    {
+      "epoch": 0.03618864292589028,
+      "grad_norm": 1.4984415769577026,
+      "learning_rate": 0.00019984050006247053,
+      "loss": 1.337,
+      "step": 940
+    },
+    {
+      "epoch": 0.03638113570741097,
+      "grad_norm": 1.399708867073059,
+      "learning_rate": 0.00019983878824362023,
+      "loss": 1.5546,
+      "step": 945
+    },
+    {
+      "epoch": 0.03657362848893166,
+      "grad_norm": 1.8458516597747803,
+      "learning_rate": 0.00019983706729517412,
+      "loss": 1.5268,
+      "step": 950
+    },
+    {
+      "epoch": 0.03676612127045236,
+      "grad_norm": 1.1428085565567017,
+      "learning_rate": 0.00019983533721728956,
+      "loss": 1.4454,
+      "step": 955
+    },
+    {
+      "epoch": 0.03695861405197305,
+      "grad_norm": 1.2200374603271484,
+      "learning_rate": 0.00019983359801012475,
+      "loss": 1.5586,
+      "step": 960
+    },
+    {
+      "epoch": 0.03715110683349374,
+      "grad_norm": 1.3679723739624023,
+      "learning_rate": 0.00019983184967383875,
+      "loss": 1.3948,
+      "step": 965
+    },
+    {
+      "epoch": 0.03734359961501444,
+      "grad_norm": 1.489397644996643,
+      "learning_rate": 0.00019983009220859142,
+      "loss": 1.5154,
+      "step": 970
+    },
+    {
+      "epoch": 0.03753609239653513,
+      "grad_norm": 1.0442456007003784,
+      "learning_rate": 0.00019982832561454345,
+      "loss": 1.5704,
+      "step": 975
+    },
+    {
+      "epoch": 0.03772858517805582,
+      "grad_norm": 1.7480882406234741,
+      "learning_rate": 0.00019982654989185642,
+      "loss": 1.5235,
+      "step": 980
+    },
+    {
+      "epoch": 0.03792107795957651,
+      "grad_norm": 1.0078760385513306,
+      "learning_rate": 0.00019982476504069272,
+      "loss": 1.3936,
+      "step": 985
+    },
+    {
+      "epoch": 0.03811357074109721,
+      "grad_norm": 1.0461446046829224,
+      "learning_rate": 0.0001998229710612155,
+      "loss": 1.6994,
+      "step": 990
+    },
+    {
+      "epoch": 0.0383060635226179,
+      "grad_norm": 2.1919922828674316,
+      "learning_rate": 0.00019982116795358885,
+      "loss": 1.5739,
+      "step": 995
+    },
+    {
+      "epoch": 0.03849855630413859,
+      "grad_norm": 1.7092692852020264,
+      "learning_rate": 0.00019981935571797768,
+      "loss": 1.2746,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03869104908565929,
+      "grad_norm": 1.3044835329055786,
+      "learning_rate": 0.00019981753435454764,
+      "loss": 1.5254,
+      "step": 1005
+    },
+    {
+      "epoch": 0.03888354186717998,
+      "grad_norm": 1.1550064086914062,
+      "learning_rate": 0.0001998157038634653,
+      "loss": 1.6154,
+      "step": 1010
+    },
+    {
+      "epoch": 0.03907603464870067,
+      "grad_norm": 2.0250370502471924,
+      "learning_rate": 0.00019981386424489808,
+      "loss": 1.4807,
+      "step": 1015
+    },
+    {
+      "epoch": 0.03926852743022137,
+      "grad_norm": 1.036095380783081,
+      "learning_rate": 0.00019981201549901419,
+      "loss": 1.4124,
+      "step": 1020
+    },
+    {
+      "epoch": 0.03946102021174206,
+      "grad_norm": 1.126434564590454,
+      "learning_rate": 0.0001998101576259827,
+      "loss": 1.4959,
+      "step": 1025
+    },
+    {
+      "epoch": 0.03965351299326275,
+      "grad_norm": 1.2912375926971436,
+      "learning_rate": 0.00019980829062597342,
+      "loss": 1.5006,
+      "step": 1030
+    },
+    {
+      "epoch": 0.03984600577478344,
+      "grad_norm": 1.5378974676132202,
+      "learning_rate": 0.00019980641449915713,
+      "loss": 1.3073,
+      "step": 1035
+    },
+    {
+      "epoch": 0.04003849855630414,
+      "grad_norm": 1.52741277217865,
+      "learning_rate": 0.0001998045292457054,
+      "loss": 1.3709,
+      "step": 1040
+    },
+    {
+      "epoch": 0.04023099133782483,
+      "grad_norm": 1.6989667415618896,
+      "learning_rate": 0.00019980263486579064,
+      "loss": 1.4784,
+      "step": 1045
+    },
+    {
+      "epoch": 0.04042348411934552,
+      "grad_norm": 1.0623974800109863,
+      "learning_rate": 0.00019980073135958607,
+      "loss": 1.5163,
+      "step": 1050
+    },
+    {
+      "epoch": 0.04061597690086622,
+      "grad_norm": 1.323283314704895,
+      "learning_rate": 0.0001997988187272657,
+      "loss": 1.4793,
+      "step": 1055
+    },
+    {
+      "epoch": 0.04080846968238691,
+      "grad_norm": 1.4508922100067139,
+      "learning_rate": 0.00019979689696900447,
+      "loss": 1.4746,
+      "step": 1060
+    },
+    {
+      "epoch": 0.0410009624639076,
+      "grad_norm": 1.159579873085022,
+      "learning_rate": 0.0001997949660849781,
+      "loss": 1.2928,
+      "step": 1065
+    },
+    {
+      "epoch": 0.0411934552454283,
+      "grad_norm": 1.5187591314315796,
+      "learning_rate": 0.0001997930260753632,
+      "loss": 1.5116,
+      "step": 1070
+    },
+    {
+      "epoch": 0.04138594802694899,
+      "grad_norm": 1.7137175798416138,
+      "learning_rate": 0.0001997910769403371,
+      "loss": 1.6406,
+      "step": 1075
+    },
+    {
+      "epoch": 0.04157844080846968,
+      "grad_norm": 1.221326470375061,
+      "learning_rate": 0.00019978911868007807,
+      "loss": 1.418,
+      "step": 1080
+    },
+    {
+      "epoch": 0.04177093358999037,
+      "grad_norm": 1.0666981935501099,
+      "learning_rate": 0.0001997871512947652,
+      "loss": 1.3768,
+      "step": 1085
+    },
+    {
+      "epoch": 0.04196342637151107,
+      "grad_norm": 0.9577809572219849,
+      "learning_rate": 0.00019978517478457834,
+      "loss": 1.4915,
+      "step": 1090
+    },
+    {
+      "epoch": 0.04215591915303176,
+      "grad_norm": 2.3966264724731445,
+      "learning_rate": 0.00019978318914969827,
+      "loss": 1.7057,
+      "step": 1095
+    },
+    {
+      "epoch": 0.04234841193455245,
+      "grad_norm": 1.0523775815963745,
+      "learning_rate": 0.0001997811943903066,
+      "loss": 1.3887,
+      "step": 1100
+    },
+    {
+      "epoch": 0.04254090471607315,
+      "grad_norm": 1.3975977897644043,
+      "learning_rate": 0.00019977919050658566,
+      "loss": 1.5335,
+      "step": 1105
+    },
+    {
+      "epoch": 0.04273339749759384,
+      "grad_norm": 1.5198701620101929,
+      "learning_rate": 0.0001997771774987187,
+      "loss": 1.3939,
+      "step": 1110
+    },
+    {
+      "epoch": 0.04292589027911453,
+      "grad_norm": 0.7943345308303833,
+      "learning_rate": 0.00019977515536688984,
+      "loss": 1.5908,
+      "step": 1115
+    },
+    {
+      "epoch": 0.04311838306063523,
+      "grad_norm": 0.9602519869804382,
+      "learning_rate": 0.00019977312411128398,
+      "loss": 1.3225,
+      "step": 1120
+    },
+    {
+      "epoch": 0.04331087584215592,
+      "grad_norm": 1.0204732418060303,
+      "learning_rate": 0.00019977108373208687,
+      "loss": 1.518,
+      "step": 1125
+    },
+    {
+      "epoch": 0.04350336862367661,
+      "grad_norm": 1.2130141258239746,
+      "learning_rate": 0.00019976903422948503,
+      "loss": 1.3693,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0436958614051973,
+      "grad_norm": 0.854958176612854,
+      "learning_rate": 0.00019976697560366598,
+      "loss": 1.4907,
+      "step": 1135
+    },
+    {
+      "epoch": 0.043888354186718,
+      "grad_norm": 1.3699367046356201,
+      "learning_rate": 0.00019976490785481789,
+      "loss": 1.4448,
+      "step": 1140
+    },
+    {
+      "epoch": 0.04408084696823869,
+      "grad_norm": 1.1766821146011353,
+      "learning_rate": 0.00019976283098312983,
+      "loss": 1.5171,
+      "step": 1145
+    },
+    {
+      "epoch": 0.04427333974975938,
+      "grad_norm": 1.6543035507202148,
+      "learning_rate": 0.00019976074498879174,
+      "loss": 1.2751,
+      "step": 1150
+    },
+    {
+      "epoch": 0.04446583253128008,
+      "grad_norm": 1.2228333950042725,
+      "learning_rate": 0.0001997586498719944,
+      "loss": 1.4522,
+      "step": 1155
+    },
+    {
+      "epoch": 0.04465832531280077,
+      "grad_norm": 1.2733262777328491,
+      "learning_rate": 0.00019975654563292937,
+      "loss": 1.6292,
+      "step": 1160
+    },
+    {
+      "epoch": 0.04485081809432146,
+      "grad_norm": 1.3934366703033447,
+      "learning_rate": 0.00019975443227178904,
+      "loss": 1.433,
+      "step": 1165
+    },
+    {
+      "epoch": 0.04504331087584216,
+      "grad_norm": 1.5495753288269043,
+      "learning_rate": 0.00019975230978876672,
+      "loss": 1.5803,
+      "step": 1170
+    },
+    {
+      "epoch": 0.04523580365736285,
+      "grad_norm": 1.0099114179611206,
+      "learning_rate": 0.00019975017818405646,
+      "loss": 1.3434,
+      "step": 1175
+    },
+    {
+      "epoch": 0.04542829643888354,
+      "grad_norm": 0.9009067416191101,
+      "learning_rate": 0.0001997480374578532,
+      "loss": 1.2312,
+      "step": 1180
+    },
+    {
+      "epoch": 0.04562078922040423,
+      "grad_norm": 1.8678425550460815,
+      "learning_rate": 0.00019974588761035266,
+      "loss": 1.6331,
+      "step": 1185
+    },
+    {
+      "epoch": 0.04581328200192493,
+      "grad_norm": 0.8258862495422363,
+      "learning_rate": 0.00019974372864175148,
+      "loss": 1.4584,
+      "step": 1190
+    },
+    {
+      "epoch": 0.04600577478344562,
+      "grad_norm": 1.44557523727417,
+      "learning_rate": 0.00019974156055224706,
+      "loss": 1.4866,
+      "step": 1195
+    },
+    {
+      "epoch": 0.04619826756496631,
+      "grad_norm": 1.7249491214752197,
+      "learning_rate": 0.00019973938334203763,
+      "loss": 1.3704,
+      "step": 1200
+    },
+    {
+      "epoch": 0.04639076034648701,
+      "grad_norm": 1.005623698234558,
+      "learning_rate": 0.0001997371970113223,
+      "loss": 1.1993,
+      "step": 1205
+    },
+    {
+      "epoch": 0.0465832531280077,
+      "grad_norm": 1.4596670866012573,
+      "learning_rate": 0.00019973500156030105,
+      "loss": 1.4996,
+      "step": 1210
+    },
+    {
+      "epoch": 0.04677574590952839,
+      "grad_norm": 1.3085503578186035,
+      "learning_rate": 0.00019973279698917454,
+      "loss": 1.441,
+      "step": 1215
+    },
+    {
+      "epoch": 0.04696823869104909,
+      "grad_norm": 0.9477142691612244,
+      "learning_rate": 0.00019973058329814445,
+      "loss": 1.5278,
+      "step": 1220
+    },
+    {
+      "epoch": 0.04716073147256978,
+      "grad_norm": 0.9040088653564453,
+      "learning_rate": 0.00019972836048741318,
+      "loss": 1.5374,
+      "step": 1225
+    },
+    {
+      "epoch": 0.04735322425409047,
+      "grad_norm": 1.7435801029205322,
+      "learning_rate": 0.00019972612855718395,
+      "loss": 1.3884,
+      "step": 1230
+    },
+    {
+      "epoch": 0.04754571703561117,
+      "grad_norm": 1.180665135383606,
+      "learning_rate": 0.00019972388750766088,
+      "loss": 1.2097,
+      "step": 1235
+    },
+    {
+      "epoch": 0.04773820981713186,
+      "grad_norm": 1.066064715385437,
+      "learning_rate": 0.00019972163733904895,
+      "loss": 1.4299,
+      "step": 1240
+    },
+    {
+      "epoch": 0.04793070259865255,
+      "grad_norm": 1.1051660776138306,
+      "learning_rate": 0.00019971937805155382,
+      "loss": 1.5055,
+      "step": 1245
+    },
+    {
+      "epoch": 0.04812319538017324,
+      "grad_norm": 1.2021822929382324,
+      "learning_rate": 0.0001997171096453822,
+      "loss": 1.5842,
+      "step": 1250
+    },
+    {
+      "epoch": 0.04831568816169394,
+      "grad_norm": 2.1715807914733887,
+      "learning_rate": 0.00019971483212074146,
+      "loss": 1.4096,
+      "step": 1255
+    },
+    {
+      "epoch": 0.04850818094321463,
+      "grad_norm": 1.1615819931030273,
+      "learning_rate": 0.00019971254547783987,
+      "loss": 1.2554,
+      "step": 1260
+    },
+    {
+      "epoch": 0.04870067372473532,
+      "grad_norm": 1.5363492965698242,
+      "learning_rate": 0.00019971024971688652,
+      "loss": 1.5773,
+      "step": 1265
+    },
+    {
+      "epoch": 0.04889316650625602,
+      "grad_norm": 1.3774447441101074,
+      "learning_rate": 0.00019970794483809137,
+      "loss": 1.3441,
+      "step": 1270
+    },
+    {
+      "epoch": 0.04908565928777671,
+      "grad_norm": 2.065901041030884,
+      "learning_rate": 0.00019970563084166515,
+      "loss": 1.6342,
+      "step": 1275
+    },
+    {
+      "epoch": 0.0492781520692974,
+      "grad_norm": 1.3221025466918945,
+      "learning_rate": 0.0001997033077278195,
+      "loss": 1.4967,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0494706448508181,
+      "grad_norm": 1.6636276245117188,
+      "learning_rate": 0.00019970097549676684,
+      "loss": 1.4936,
+      "step": 1285
+    },
+    {
+      "epoch": 0.04966313763233879,
+      "grad_norm": 1.4630615711212158,
+      "learning_rate": 0.0001996986341487204,
+      "loss": 1.4096,
+      "step": 1290
+    },
+    {
+      "epoch": 0.04985563041385948,
+      "grad_norm": 1.9586588144302368,
+      "learning_rate": 0.00019969628368389432,
+      "loss": 1.5956,
+      "step": 1295
+    },
+    {
+      "epoch": 0.05004812319538017,
+      "grad_norm": 1.0234311819076538,
+      "learning_rate": 0.00019969392410250353,
+      "loss": 1.247,
+      "step": 1300
+    },
+    {
+      "epoch": 0.05024061597690087,
+      "grad_norm": 1.7005319595336914,
+      "learning_rate": 0.0001996915554047638,
+      "loss": 1.4179,
+      "step": 1305
+    },
+    {
+      "epoch": 0.05043310875842156,
+      "grad_norm": 1.3052936792373657,
+      "learning_rate": 0.0001996891775908917,
+      "loss": 1.4002,
+      "step": 1310
+    },
+    {
+      "epoch": 0.05062560153994225,
+      "grad_norm": 1.0146903991699219,
+      "learning_rate": 0.00019968679066110473,
+      "loss": 1.5062,
+      "step": 1315
+    },
+    {
+      "epoch": 0.05081809432146295,
+      "grad_norm": 0.9611810445785522,
+      "learning_rate": 0.00019968439461562104,
+      "loss": 1.5303,
+      "step": 1320
+    },
+    {
+      "epoch": 0.05101058710298364,
+      "grad_norm": 0.8518236875534058,
+      "learning_rate": 0.0001996819894546599,
+      "loss": 1.3589,
+      "step": 1325
+    },
+    {
+      "epoch": 0.05120307988450433,
+      "grad_norm": 1.6918632984161377,
+      "learning_rate": 0.00019967957517844111,
+      "loss": 1.4589,
+      "step": 1330
+    },
+    {
+      "epoch": 0.05139557266602503,
+      "grad_norm": 1.4838560819625854,
+      "learning_rate": 0.00019967715178718551,
+      "loss": 1.2714,
+      "step": 1335
+    },
+    {
+      "epoch": 0.05158806544754572,
+      "grad_norm": 1.291231632232666,
+      "learning_rate": 0.00019967471928111465,
+      "loss": 1.6378,
+      "step": 1340
+    },
+    {
+      "epoch": 0.05178055822906641,
+      "grad_norm": 1.2091941833496094,
+      "learning_rate": 0.00019967227766045102,
+      "loss": 1.3985,
+      "step": 1345
+    },
+    {
+      "epoch": 0.0519730510105871,
+      "grad_norm": 1.2294058799743652,
+      "learning_rate": 0.00019966982692541785,
+      "loss": 1.498,
+      "step": 1350
+    },
+    {
+      "epoch": 0.0521655437921078,
+      "grad_norm": 1.1644397974014282,
+      "learning_rate": 0.00019966736707623928,
+      "loss": 1.4185,
+      "step": 1355
+    },
+    {
+      "epoch": 0.05235803657362849,
+      "grad_norm": 1.7669397592544556,
+      "learning_rate": 0.0001996648981131402,
+      "loss": 1.3564,
+      "step": 1360
+    },
+    {
+      "epoch": 0.05255052935514918,
+      "grad_norm": 0.7178487777709961,
+      "learning_rate": 0.00019966242003634644,
+      "loss": 1.2015,
+      "step": 1365
+    },
+    {
+      "epoch": 0.05274302213666988,
+      "grad_norm": 0.8149698376655579,
+      "learning_rate": 0.00019965993284608457,
+      "loss": 1.4046,
+      "step": 1370
+    },
+    {
+      "epoch": 0.05293551491819057,
+      "grad_norm": 1.3934742212295532,
+      "learning_rate": 0.00019965743654258198,
+      "loss": 1.5289,
+      "step": 1375
+    },
+    {
+      "epoch": 0.05312800769971126,
+      "grad_norm": 1.060002326965332,
+      "learning_rate": 0.00019965493112606702,
+      "loss": 1.391,
+      "step": 1380
+    },
+    {
+      "epoch": 0.05332050048123196,
+      "grad_norm": 1.1154258251190186,
+      "learning_rate": 0.00019965241659676875,
+      "loss": 1.3004,
+      "step": 1385
+    },
+    {
+      "epoch": 0.05351299326275265,
+      "grad_norm": 1.8101186752319336,
+      "learning_rate": 0.00019964989295491713,
+      "loss": 1.4968,
+      "step": 1390
+    },
+    {
+      "epoch": 0.05370548604427334,
+      "grad_norm": 1.075211524963379,
+      "learning_rate": 0.00019964736020074294,
+      "loss": 1.5198,
+      "step": 1395
+    },
+    {
+      "epoch": 0.05389797882579403,
+      "grad_norm": 2.0130980014801025,
+      "learning_rate": 0.00019964481833447775,
+      "loss": 1.5495,
+      "step": 1400
+    },
+    {
+      "epoch": 0.05409047160731473,
+      "grad_norm": 1.214570164680481,
+      "learning_rate": 0.000199642267356354,
+      "loss": 1.5886,
+      "step": 1405
+    },
+    {
+      "epoch": 0.05428296438883542,
+      "grad_norm": 1.6430037021636963,
+      "learning_rate": 0.00019963970726660497,
+      "loss": 1.5293,
+      "step": 1410
+    },
+    {
+      "epoch": 0.05447545717035611,
+      "grad_norm": 0.94575035572052,
+      "learning_rate": 0.00019963713806546478,
+      "loss": 1.276,
+      "step": 1415
+    },
+    {
+      "epoch": 0.05466794995187681,
+      "grad_norm": 1.1988322734832764,
+      "learning_rate": 0.00019963455975316832,
+      "loss": 1.3151,
+      "step": 1420
+    },
+    {
+      "epoch": 0.0548604427333975,
+      "grad_norm": 1.2768787145614624,
+      "learning_rate": 0.00019963197232995142,
+      "loss": 1.5559,
+      "step": 1425
+    },
+    {
+      "epoch": 0.05505293551491819,
+      "grad_norm": 1.5184259414672852,
+      "learning_rate": 0.0001996293757960506,
+      "loss": 1.2998,
+      "step": 1430
+    },
+    {
+      "epoch": 0.055245428296438887,
+      "grad_norm": 6.240184783935547,
+      "learning_rate": 0.0001996267701517034,
+      "loss": 1.4497,
+      "step": 1435
+    },
+    {
+      "epoch": 0.05543792107795958,
+      "grad_norm": 1.4356882572174072,
+      "learning_rate": 0.00019962415539714803,
+      "loss": 1.6364,
+      "step": 1440
+    },
+    {
+      "epoch": 0.05563041385948027,
+      "grad_norm": 0.9310120940208435,
+      "learning_rate": 0.00019962153153262358,
+      "loss": 1.417,
+      "step": 1445
+    },
+    {
+      "epoch": 0.05582290664100096,
+      "grad_norm": 1.2131333351135254,
+      "learning_rate": 0.00019961889855837,
+      "loss": 1.4059,
+      "step": 1450
+    },
+    {
+      "epoch": 0.05601539942252166,
+      "grad_norm": 1.2134804725646973,
+      "learning_rate": 0.00019961625647462808,
+      "loss": 1.458,
+      "step": 1455
+    },
+    {
+      "epoch": 0.05620789220404235,
+      "grad_norm": 1.5725634098052979,
+      "learning_rate": 0.0001996136052816394,
+      "loss": 1.352,
+      "step": 1460
+    },
+    {
+      "epoch": 0.05640038498556304,
+      "grad_norm": 0.9882212281227112,
+      "learning_rate": 0.00019961094497964642,
+      "loss": 1.1665,
+      "step": 1465
+    },
+    {
+      "epoch": 0.05659287776708374,
+      "grad_norm": 1.055966854095459,
+      "learning_rate": 0.00019960827556889235,
+      "loss": 1.388,
+      "step": 1470
+    },
+    {
+      "epoch": 0.05678537054860443,
+      "grad_norm": 1.0809309482574463,
+      "learning_rate": 0.00019960559704962133,
+      "loss": 1.4287,
+      "step": 1475
+    },
+    {
+      "epoch": 0.05697786333012512,
+      "grad_norm": 1.0014935731887817,
+      "learning_rate": 0.00019960290942207828,
+      "loss": 1.5539,
+      "step": 1480
+    },
+    {
+      "epoch": 0.057170356111645816,
+      "grad_norm": 1.1717151403427124,
+      "learning_rate": 0.000199600212686509,
+      "loss": 1.3619,
+      "step": 1485
+    },
+    {
+      "epoch": 0.05736284889316651,
+      "grad_norm": 1.3981553316116333,
+      "learning_rate": 0.00019959750684316,
+      "loss": 1.3303,
+      "step": 1490
+    },
+    {
+      "epoch": 0.0575553416746872,
+      "grad_norm": 0.7471413016319275,
+      "learning_rate": 0.00019959479189227884,
+      "loss": 1.4048,
+      "step": 1495
+    },
+    {
+      "epoch": 0.05774783445620789,
+      "grad_norm": 1.1570223569869995,
+      "learning_rate": 0.00019959206783411372,
+      "loss": 1.6713,
+      "step": 1500
+    },
+    {
+      "epoch": 0.05794032723772859,
+      "grad_norm": 1.4656585454940796,
+      "learning_rate": 0.00019958933466891366,
+      "loss": 1.3911,
+      "step": 1505
+    },
+    {
+      "epoch": 0.05813282001924928,
+      "grad_norm": 1.5338329076766968,
+      "learning_rate": 0.0001995865923969287,
+      "loss": 1.578,
+      "step": 1510
+    },
+    {
+      "epoch": 0.05832531280076997,
+      "grad_norm": 0.9481655955314636,
+      "learning_rate": 0.0001995838410184096,
+      "loss": 1.2903,
+      "step": 1515
+    },
+    {
+      "epoch": 0.058517805582290666,
+      "grad_norm": 1.4928970336914062,
+      "learning_rate": 0.00019958108053360788,
+      "loss": 1.4139,
+      "step": 1520
+    },
+    {
+      "epoch": 0.05871029836381136,
+      "grad_norm": 1.015381932258606,
+      "learning_rate": 0.00019957831094277604,
+      "loss": 1.5427,
+      "step": 1525
+    },
+    {
+      "epoch": 0.05890279114533205,
+      "grad_norm": 1.3471331596374512,
+      "learning_rate": 0.0001995755322461673,
+      "loss": 1.3763,
+      "step": 1530
+    },
+    {
+      "epoch": 0.059095283926852746,
+      "grad_norm": 2.0942165851593018,
+      "learning_rate": 0.00019957274444403576,
+      "loss": 1.4669,
+      "step": 1535
+    },
+    {
+      "epoch": 0.05928777670837344,
+      "grad_norm": 1.4853599071502686,
+      "learning_rate": 0.00019956994753663634,
+      "loss": 1.4259,
+      "step": 1540
+    },
+    {
+      "epoch": 0.05948026948989413,
+      "grad_norm": 1.3337596654891968,
+      "learning_rate": 0.0001995671415242248,
+      "loss": 1.4169,
+      "step": 1545
+    },
+    {
+      "epoch": 0.059672762271414825,
+      "grad_norm": 1.3816536664962769,
+      "learning_rate": 0.00019956432640705777,
+      "loss": 1.3679,
+      "step": 1550
+    },
+    {
+      "epoch": 0.059865255052935516,
+      "grad_norm": 1.1726235151290894,
+      "learning_rate": 0.00019956150218539262,
+      "loss": 1.4076,
+      "step": 1555
+    },
+    {
+      "epoch": 0.06005774783445621,
+      "grad_norm": 1.419520378112793,
+      "learning_rate": 0.00019955866885948764,
+      "loss": 1.3621,
+      "step": 1560
+    },
+    {
+      "epoch": 0.0602502406159769,
+      "grad_norm": 1.4154486656188965,
+      "learning_rate": 0.0001995558264296019,
+      "loss": 1.4221,
+      "step": 1565
+    },
+    {
+      "epoch": 0.060442733397497596,
+      "grad_norm": 1.4721988439559937,
+      "learning_rate": 0.00019955297489599537,
+      "loss": 1.3641,
+      "step": 1570
+    },
+    {
+      "epoch": 0.06063522617901829,
+      "grad_norm": 1.1087952852249146,
+      "learning_rate": 0.0001995501142589287,
+      "loss": 1.3734,
+      "step": 1575
+    },
+    {
+      "epoch": 0.06082771896053898,
+      "grad_norm": 1.4815518856048584,
+      "learning_rate": 0.00019954724451866357,
+      "loss": 1.4042,
+      "step": 1580
+    },
+    {
+      "epoch": 0.061020211742059675,
+      "grad_norm": 1.835754632949829,
+      "learning_rate": 0.00019954436567546236,
+      "loss": 1.2457,
+      "step": 1585
+    },
+    {
+      "epoch": 0.061212704523580366,
+      "grad_norm": 1.3139601945877075,
+      "learning_rate": 0.00019954147772958836,
+      "loss": 1.4457,
+      "step": 1590
+    },
+    {
+      "epoch": 0.06140519730510106,
+      "grad_norm": 1.155369758605957,
+      "learning_rate": 0.0001995385806813056,
+      "loss": 1.3483,
+      "step": 1595
+    },
+    {
+      "epoch": 0.061597690086621755,
+      "grad_norm": 1.1897907257080078,
+      "learning_rate": 0.00019953567453087902,
+      "loss": 1.467,
+      "step": 1600
+    },
+    {
+      "epoch": 0.061790182868142446,
+      "grad_norm": 1.0794181823730469,
+      "learning_rate": 0.00019953275927857438,
+      "loss": 1.5171,
+      "step": 1605
+    },
+    {
+      "epoch": 0.06198267564966314,
+      "grad_norm": 0.9538444876670837,
+      "learning_rate": 0.00019952983492465824,
+      "loss": 1.2643,
+      "step": 1610
+    },
+    {
+      "epoch": 0.06217516843118383,
+      "grad_norm": 1.1179461479187012,
+      "learning_rate": 0.00019952690146939804,
+      "loss": 1.408,
+      "step": 1615
+    },
+    {
+      "epoch": 0.062367661212704525,
+      "grad_norm": 1.8034144639968872,
+      "learning_rate": 0.00019952395891306197,
+      "loss": 1.3685,
+      "step": 1620
+    },
+    {
+      "epoch": 0.06256015399422522,
+      "grad_norm": 1.04547119140625,
+      "learning_rate": 0.00019952100725591912,
+      "loss": 1.4271,
+      "step": 1625
+    },
+    {
+      "epoch": 0.06275264677574591,
+      "grad_norm": 1.3097724914550781,
+      "learning_rate": 0.00019951804649823949,
+      "loss": 1.3303,
+      "step": 1630
+    },
+    {
+      "epoch": 0.0629451395572666,
+      "grad_norm": 1.8794469833374023,
+      "learning_rate": 0.00019951507664029374,
+      "loss": 1.5223,
+      "step": 1635
+    },
+    {
+      "epoch": 0.06313763233878729,
+      "grad_norm": 1.4077703952789307,
+      "learning_rate": 0.00019951209768235344,
+      "loss": 1.5582,
+      "step": 1640
+    },
+    {
+      "epoch": 0.06333012512030799,
+      "grad_norm": 1.2244471311569214,
+      "learning_rate": 0.000199509109624691,
+      "loss": 1.3437,
+      "step": 1645
+    },
+    {
+      "epoch": 0.06352261790182868,
+      "grad_norm": 1.4610791206359863,
+      "learning_rate": 0.00019950611246757972,
+      "loss": 1.6944,
+      "step": 1650
+    },
+    {
+      "epoch": 0.06371511068334937,
+      "grad_norm": 1.544989824295044,
+      "learning_rate": 0.00019950310621129358,
+      "loss": 1.3288,
+      "step": 1655
+    },
+    {
+      "epoch": 0.06390760346487007,
+      "grad_norm": 1.4837945699691772,
+      "learning_rate": 0.00019950009085610755,
+      "loss": 1.1296,
+      "step": 1660
+    },
+    {
+      "epoch": 0.06410009624639076,
+      "grad_norm": 2.2527410984039307,
+      "learning_rate": 0.0001994970664022973,
+      "loss": 1.3105,
+      "step": 1665
+    },
+    {
+      "epoch": 0.06429258902791145,
+      "grad_norm": 1.3723945617675781,
+      "learning_rate": 0.00019949403285013948,
+      "loss": 1.3976,
+      "step": 1670
+    },
+    {
+      "epoch": 0.06448508180943215,
+      "grad_norm": 1.571265459060669,
+      "learning_rate": 0.0001994909901999114,
+      "loss": 1.4603,
+      "step": 1675
+    },
+    {
+      "epoch": 0.06467757459095284,
+      "grad_norm": 1.2445194721221924,
+      "learning_rate": 0.00019948793845189137,
+      "loss": 1.3072,
+      "step": 1680
+    },
+    {
+      "epoch": 0.06487006737247353,
+      "grad_norm": 2.068112373352051,
+      "learning_rate": 0.00019948487760635842,
+      "loss": 1.4638,
+      "step": 1685
+    },
+    {
+      "epoch": 0.06506256015399423,
+      "grad_norm": 1.0896637439727783,
+      "learning_rate": 0.00019948180766359244,
+      "loss": 1.3184,
+      "step": 1690
+    },
+    {
+      "epoch": 0.06525505293551492,
+      "grad_norm": 2.0666351318359375,
+      "learning_rate": 0.00019947872862387413,
+      "loss": 1.3944,
+      "step": 1695
+    },
+    {
+      "epoch": 0.06544754571703561,
+      "grad_norm": 1.5204085111618042,
+      "learning_rate": 0.00019947564048748508,
+      "loss": 1.3795,
+      "step": 1700
+    },
+    {
+      "epoch": 0.0656400384985563,
+      "grad_norm": 0.9768043160438538,
+      "learning_rate": 0.00019947254325470768,
+      "loss": 1.3329,
+      "step": 1705
+    },
+    {
+      "epoch": 0.065832531280077,
+      "grad_norm": 1.3453469276428223,
+      "learning_rate": 0.00019946943692582516,
+      "loss": 1.304,
+      "step": 1710
+    },
+    {
+      "epoch": 0.06602502406159769,
+      "grad_norm": 1.0725489854812622,
+      "learning_rate": 0.00019946632150112152,
+      "loss": 1.5547,
+      "step": 1715
+    },
+    {
+      "epoch": 0.06621751684311838,
+      "grad_norm": 1.5973418951034546,
+      "learning_rate": 0.0001994631969808817,
+      "loss": 1.3263,
+      "step": 1720
+    },
+    {
+      "epoch": 0.06641000962463908,
+      "grad_norm": 1.2451751232147217,
+      "learning_rate": 0.0001994600633653914,
+      "loss": 1.4935,
+      "step": 1725
+    },
+    {
+      "epoch": 0.06660250240615977,
+      "grad_norm": 1.3474830389022827,
+      "learning_rate": 0.00019945692065493717,
+      "loss": 1.6282,
+      "step": 1730
+    },
+    {
+      "epoch": 0.06679499518768046,
+      "grad_norm": 1.7913939952850342,
+      "learning_rate": 0.00019945376884980643,
+      "loss": 1.2935,
+      "step": 1735
+    },
+    {
+      "epoch": 0.06698748796920115,
+      "grad_norm": 1.0764446258544922,
+      "learning_rate": 0.00019945060795028728,
+      "loss": 1.6034,
+      "step": 1740
+    },
+    {
+      "epoch": 0.06717998075072185,
+      "grad_norm": 1.0572975873947144,
+      "learning_rate": 0.00019944743795666887,
+      "loss": 1.3997,
+      "step": 1745
+    },
+    {
+      "epoch": 0.06737247353224254,
+      "grad_norm": 1.3195079565048218,
+      "learning_rate": 0.00019944425886924102,
+      "loss": 1.4838,
+      "step": 1750
+    },
+    {
+      "epoch": 0.06756496631376323,
+      "grad_norm": 1.0044989585876465,
+      "learning_rate": 0.00019944107068829448,
+      "loss": 1.388,
+      "step": 1755
+    },
+    {
+      "epoch": 0.06775745909528393,
+      "grad_norm": 1.8276032209396362,
+      "learning_rate": 0.0001994378734141207,
+      "loss": 1.447,
+      "step": 1760
+    },
+    {
+      "epoch": 0.06794995187680462,
+      "grad_norm": 1.5056366920471191,
+      "learning_rate": 0.00019943466704701218,
+      "loss": 1.5153,
+      "step": 1765
+    },
+    {
+      "epoch": 0.06814244465832531,
+      "grad_norm": 1.6947304010391235,
+      "learning_rate": 0.00019943145158726205,
+      "loss": 1.5551,
+      "step": 1770
+    },
+    {
+      "epoch": 0.068334937439846,
+      "grad_norm": 0.9702686667442322,
+      "learning_rate": 0.00019942822703516433,
+      "loss": 1.3168,
+      "step": 1775
+    },
+    {
+      "epoch": 0.0685274302213667,
+      "grad_norm": 1.6755216121673584,
+      "learning_rate": 0.0001994249933910139,
+      "loss": 1.6223,
+      "step": 1780
+    },
+    {
+      "epoch": 0.06871992300288739,
+      "grad_norm": 1.3666303157806396,
+      "learning_rate": 0.00019942175065510643,
+      "loss": 1.5748,
+      "step": 1785
+    },
+    {
+      "epoch": 0.06891241578440808,
+      "grad_norm": 1.3785196542739868,
+      "learning_rate": 0.0001994184988277385,
+      "loss": 1.4033,
+      "step": 1790
+    },
+    {
+      "epoch": 0.06910490856592878,
+      "grad_norm": 1.081828236579895,
+      "learning_rate": 0.00019941523790920743,
+      "loss": 1.4,
+      "step": 1795
+    },
+    {
+      "epoch": 0.06929740134744947,
+      "grad_norm": 1.1024401187896729,
+      "learning_rate": 0.0001994119678998114,
+      "loss": 1.4751,
+      "step": 1800
+    },
+    {
+      "epoch": 0.06948989412897016,
+      "grad_norm": 3.584055185317993,
+      "learning_rate": 0.0001994086887998495,
+      "loss": 1.3449,
+      "step": 1805
+    },
+    {
+      "epoch": 0.06968238691049086,
+      "grad_norm": 0.9418397545814514,
+      "learning_rate": 0.0001994054006096215,
+      "loss": 1.3217,
+      "step": 1810
+    },
+    {
+      "epoch": 0.06987487969201155,
+      "grad_norm": 1.6071193218231201,
+      "learning_rate": 0.00019940210332942813,
+      "loss": 1.3636,
+      "step": 1815
+    },
+    {
+      "epoch": 0.07006737247353224,
+      "grad_norm": 2.0080580711364746,
+      "learning_rate": 0.00019939879695957084,
+      "loss": 1.4779,
+      "step": 1820
+    },
+    {
+      "epoch": 0.07025986525505294,
+      "grad_norm": 1.169058918952942,
+      "learning_rate": 0.00019939548150035207,
+      "loss": 1.4031,
+      "step": 1825
+    },
+    {
+      "epoch": 0.07045235803657363,
+      "grad_norm": 0.9863006472587585,
+      "learning_rate": 0.00019939215695207496,
+      "loss": 1.3832,
+      "step": 1830
+    },
+    {
+      "epoch": 0.07064485081809432,
+      "grad_norm": 1.2257460355758667,
+      "learning_rate": 0.00019938882331504347,
+      "loss": 1.4967,
+      "step": 1835
+    },
+    {
+      "epoch": 0.07083734359961502,
+      "grad_norm": 1.0062893629074097,
+      "learning_rate": 0.00019938548058956253,
+      "loss": 1.2637,
+      "step": 1840
+    },
+    {
+      "epoch": 0.0710298363811357,
+      "grad_norm": 1.4179530143737793,
+      "learning_rate": 0.0001993821287759377,
+      "loss": 1.2961,
+      "step": 1845
+    },
+    {
+      "epoch": 0.0712223291626564,
+      "grad_norm": 1.2181779146194458,
+      "learning_rate": 0.00019937876787447557,
+      "loss": 1.4104,
+      "step": 1850
+    },
+    {
+      "epoch": 0.07141482194417709,
+      "grad_norm": 1.6110061407089233,
+      "learning_rate": 0.00019937539788548344,
+      "loss": 1.4045,
+      "step": 1855
+    },
+    {
+      "epoch": 0.07160731472569778,
+      "grad_norm": 1.2814903259277344,
+      "learning_rate": 0.0001993720188092695,
+      "loss": 1.4194,
+      "step": 1860
+    },
+    {
+      "epoch": 0.07179980750721848,
+      "grad_norm": 1.382265329360962,
+      "learning_rate": 0.00019936863064614268,
+      "loss": 1.5848,
+      "step": 1865
+    },
+    {
+      "epoch": 0.07199230028873917,
+      "grad_norm": 1.4708553552627563,
+      "learning_rate": 0.00019936523339641286,
+      "loss": 1.6196,
+      "step": 1870
+    },
+    {
+      "epoch": 0.07218479307025986,
+      "grad_norm": 1.0691862106323242,
+      "learning_rate": 0.0001993618270603907,
+      "loss": 1.4939,
+      "step": 1875
+    },
+    {
+      "epoch": 0.07237728585178056,
+      "grad_norm": 0.9476374387741089,
+      "learning_rate": 0.0001993584116383876,
+      "loss": 1.5043,
+      "step": 1880
+    },
+    {
+      "epoch": 0.07256977863330125,
+      "grad_norm": 1.37090003490448,
+      "learning_rate": 0.000199354987130716,
+      "loss": 1.4371,
+      "step": 1885
+    },
+    {
+      "epoch": 0.07276227141482194,
+      "grad_norm": 1.2001820802688599,
+      "learning_rate": 0.000199351553537689,
+      "loss": 1.3048,
+      "step": 1890
+    },
+    {
+      "epoch": 0.07295476419634264,
+      "grad_norm": 1.1123398542404175,
+      "learning_rate": 0.00019934811085962055,
+      "loss": 1.4398,
+      "step": 1895
+    },
+    {
+      "epoch": 0.07314725697786333,
+      "grad_norm": 1.638574242591858,
+      "learning_rate": 0.0001993446590968255,
+      "loss": 1.3563,
+      "step": 1900
+    },
+    {
+      "epoch": 0.07333974975938402,
+      "grad_norm": 1.9532630443572998,
+      "learning_rate": 0.00019934119824961948,
+      "loss": 1.3723,
+      "step": 1905
+    },
+    {
+      "epoch": 0.07353224254090472,
+      "grad_norm": 1.3247241973876953,
+      "learning_rate": 0.0001993377283183189,
+      "loss": 1.4474,
+      "step": 1910
+    },
+    {
+      "epoch": 0.0737247353224254,
+      "grad_norm": 1.203049659729004,
+      "learning_rate": 0.00019933424930324118,
+      "loss": 1.3347,
+      "step": 1915
+    },
+    {
+      "epoch": 0.0739172281039461,
+      "grad_norm": 1.8858312368392944,
+      "learning_rate": 0.00019933076120470436,
+      "loss": 1.4754,
+      "step": 1920
+    },
+    {
+      "epoch": 0.0741097208854668,
+      "grad_norm": 1.117814540863037,
+      "learning_rate": 0.00019932726402302744,
+      "loss": 1.4828,
+      "step": 1925
+    },
+    {
+      "epoch": 0.07430221366698748,
+      "grad_norm": 1.0317554473876953,
+      "learning_rate": 0.00019932375775853021,
+      "loss": 1.5034,
+      "step": 1930
+    },
+    {
+      "epoch": 0.07449470644850818,
+      "grad_norm": 2.315903902053833,
+      "learning_rate": 0.00019932024241153332,
+      "loss": 1.4311,
+      "step": 1935
+    },
+    {
+      "epoch": 0.07468719923002888,
+      "grad_norm": 1.5780115127563477,
+      "learning_rate": 0.00019931671798235817,
+      "loss": 1.3917,
+      "step": 1940
+    },
+    {
+      "epoch": 0.07487969201154956,
+      "grad_norm": 1.3360038995742798,
+      "learning_rate": 0.00019931318447132706,
+      "loss": 1.3634,
+      "step": 1945
+    },
+    {
+      "epoch": 0.07507218479307026,
+      "grad_norm": 2.275620937347412,
+      "learning_rate": 0.00019930964187876314,
+      "loss": 1.414,
+      "step": 1950
+    },
+    {
+      "epoch": 0.07526467757459095,
+      "grad_norm": 1.7956300973892212,
+      "learning_rate": 0.00019930609020499032,
+      "loss": 1.5117,
+      "step": 1955
+    },
+    {
+      "epoch": 0.07545717035611164,
+      "grad_norm": 1.6429657936096191,
+      "learning_rate": 0.0001993025294503334,
+      "loss": 1.4436,
+      "step": 1960
+    },
+    {
+      "epoch": 0.07564966313763234,
+      "grad_norm": 1.432246446609497,
+      "learning_rate": 0.000199298959615118,
+      "loss": 1.3952,
+      "step": 1965
+    },
+    {
+      "epoch": 0.07584215591915303,
+      "grad_norm": 1.0579869747161865,
+      "learning_rate": 0.00019929538069967051,
+      "loss": 1.4369,
+      "step": 1970
+    },
+    {
+      "epoch": 0.07603464870067372,
+      "grad_norm": 1.766543984413147,
+      "learning_rate": 0.00019929179270431824,
+      "loss": 1.5033,
+      "step": 1975
+    },
+    {
+      "epoch": 0.07622714148219442,
+      "grad_norm": 1.0774848461151123,
+      "learning_rate": 0.00019928819562938928,
+      "loss": 1.3399,
+      "step": 1980
+    },
+    {
+      "epoch": 0.0764196342637151,
+      "grad_norm": 1.0951963663101196,
+      "learning_rate": 0.00019928458947521252,
+      "loss": 1.3656,
+      "step": 1985
+    },
+    {
+      "epoch": 0.0766121270452358,
+      "grad_norm": 1.278283953666687,
+      "learning_rate": 0.0001992809742421178,
+      "loss": 1.3467,
+      "step": 1990
+    },
+    {
+      "epoch": 0.0768046198267565,
+      "grad_norm": 1.139508605003357,
+      "learning_rate": 0.00019927734993043566,
+      "loss": 1.4316,
+      "step": 1995
+    },
+    {
+      "epoch": 0.07699711260827719,
+      "grad_norm": 1.39482581615448,
+      "learning_rate": 0.00019927371654049748,
+      "loss": 1.2032,
+      "step": 2000
+    },
+    {
+      "epoch": 0.07718960538979788,
+      "grad_norm": 0.9154567718505859,
+      "learning_rate": 0.0001992700740726356,
+      "loss": 1.5053,
+      "step": 2005
+    },
+    {
+      "epoch": 0.07738209817131858,
+      "grad_norm": 1.5105671882629395,
+      "learning_rate": 0.00019926642252718303,
+      "loss": 1.5059,
+      "step": 2010
+    },
+    {
+      "epoch": 0.07757459095283926,
+      "grad_norm": 1.4019540548324585,
+      "learning_rate": 0.00019926276190447367,
+      "loss": 1.4051,
+      "step": 2015
+    },
+    {
+      "epoch": 0.07776708373435996,
+      "grad_norm": 1.619841456413269,
+      "learning_rate": 0.00019925909220484234,
+      "loss": 1.1784,
+      "step": 2020
+    },
+    {
+      "epoch": 0.07795957651588066,
+      "grad_norm": 1.6128195524215698,
+      "learning_rate": 0.0001992554134286245,
+      "loss": 1.4623,
+      "step": 2025
+    },
+    {
+      "epoch": 0.07815206929740134,
+      "grad_norm": 1.2766104936599731,
+      "learning_rate": 0.00019925172557615665,
+      "loss": 1.3162,
+      "step": 2030
+    },
+    {
+      "epoch": 0.07834456207892204,
+      "grad_norm": 1.2187426090240479,
+      "learning_rate": 0.00019924802864777598,
+      "loss": 1.2874,
+      "step": 2035
+    },
+    {
+      "epoch": 0.07853705486044274,
+      "grad_norm": 1.1050268411636353,
+      "learning_rate": 0.00019924432264382055,
+      "loss": 1.433,
+      "step": 2040
+    },
+    {
+      "epoch": 0.07872954764196342,
+      "grad_norm": 1.6128287315368652,
+      "learning_rate": 0.00019924060756462925,
+      "loss": 1.4698,
+      "step": 2045
+    },
+    {
+      "epoch": 0.07892204042348412,
+      "grad_norm": 1.6588749885559082,
+      "learning_rate": 0.00019923688341054176,
+      "loss": 1.4972,
+      "step": 2050
+    },
+    {
+      "epoch": 0.0791145332050048,
+      "grad_norm": 1.135289192199707,
+      "learning_rate": 0.0001992331501818987,
+      "loss": 1.3991,
+      "step": 2055
+    },
+    {
+      "epoch": 0.0793070259865255,
+      "grad_norm": 1.757759928703308,
+      "learning_rate": 0.00019922940787904137,
+      "loss": 1.3736,
+      "step": 2060
+    },
+    {
+      "epoch": 0.0794995187680462,
+      "grad_norm": 0.9943239092826843,
+      "learning_rate": 0.00019922565650231207,
+      "loss": 1.4476,
+      "step": 2065
+    },
+    {
+      "epoch": 0.07969201154956689,
+      "grad_norm": 0.9459586143493652,
+      "learning_rate": 0.00019922189605205379,
+      "loss": 1.3913,
+      "step": 2070
+    },
+    {
+      "epoch": 0.07988450433108758,
+      "grad_norm": 1.2325133085250854,
+      "learning_rate": 0.00019921812652861037,
+      "loss": 1.4658,
+      "step": 2075
+    },
+    {
+      "epoch": 0.08007699711260828,
+      "grad_norm": 1.2397321462631226,
+      "learning_rate": 0.00019921434793232658,
+      "loss": 1.2552,
+      "step": 2080
+    },
+    {
+      "epoch": 0.08026948989412896,
+      "grad_norm": 0.9636020660400391,
+      "learning_rate": 0.0001992105602635479,
+      "loss": 1.3296,
+      "step": 2085
+    },
+    {
+      "epoch": 0.08046198267564966,
+      "grad_norm": 0.900841474533081,
+      "learning_rate": 0.00019920676352262067,
+      "loss": 1.2329,
+      "step": 2090
+    },
+    {
+      "epoch": 0.08065447545717036,
+      "grad_norm": 1.0425807237625122,
+      "learning_rate": 0.00019920295770989213,
+      "loss": 1.1604,
+      "step": 2095
+    },
+    {
+      "epoch": 0.08084696823869104,
+      "grad_norm": 1.1449722051620483,
+      "learning_rate": 0.00019919914282571024,
+      "loss": 1.3233,
+      "step": 2100
+    },
+    {
+      "epoch": 0.08103946102021174,
+      "grad_norm": 1.2076728343963623,
+      "learning_rate": 0.00019919531887042387,
+      "loss": 1.3449,
+      "step": 2105
+    },
+    {
+      "epoch": 0.08123195380173244,
+      "grad_norm": 0.968323826789856,
+      "learning_rate": 0.00019919148584438272,
+      "loss": 1.4273,
+      "step": 2110
+    },
+    {
+      "epoch": 0.08142444658325312,
+      "grad_norm": 1.7322039604187012,
+      "learning_rate": 0.00019918764374793726,
+      "loss": 1.4994,
+      "step": 2115
+    },
+    {
+      "epoch": 0.08161693936477382,
+      "grad_norm": 1.4216794967651367,
+      "learning_rate": 0.00019918379258143884,
+      "loss": 1.4071,
+      "step": 2120
+    },
+    {
+      "epoch": 0.08180943214629452,
+      "grad_norm": 1.2262970209121704,
+      "learning_rate": 0.00019917993234523963,
+      "loss": 1.3528,
+      "step": 2125
+    },
+    {
+      "epoch": 0.0820019249278152,
+      "grad_norm": 1.3137859106063843,
+      "learning_rate": 0.0001991760630396926,
+      "loss": 1.4367,
+      "step": 2130
+    },
+    {
+      "epoch": 0.0821944177093359,
+      "grad_norm": 1.364478588104248,
+      "learning_rate": 0.00019917218466515156,
+      "loss": 1.6896,
+      "step": 2135
+    },
+    {
+      "epoch": 0.0823869104908566,
+      "grad_norm": 1.2037614583969116,
+      "learning_rate": 0.00019916829722197124,
+      "loss": 1.5371,
+      "step": 2140
+    },
+    {
+      "epoch": 0.08257940327237728,
+      "grad_norm": 1.7590453624725342,
+      "learning_rate": 0.00019916440071050706,
+      "loss": 1.6331,
+      "step": 2145
+    },
+    {
+      "epoch": 0.08277189605389798,
+      "grad_norm": 1.6112565994262695,
+      "learning_rate": 0.00019916049513111532,
+      "loss": 1.5066,
+      "step": 2150
+    },
+    {
+      "epoch": 0.08296438883541868,
+      "grad_norm": 0.937174916267395,
+      "learning_rate": 0.00019915658048415318,
+      "loss": 1.4698,
+      "step": 2155
+    },
+    {
+      "epoch": 0.08315688161693936,
+      "grad_norm": 1.8568309545516968,
+      "learning_rate": 0.00019915265676997862,
+      "loss": 1.3197,
+      "step": 2160
+    },
+    {
+      "epoch": 0.08334937439846006,
+      "grad_norm": 1.9865350723266602,
+      "learning_rate": 0.00019914872398895043,
+      "loss": 1.4883,
+      "step": 2165
+    },
+    {
+      "epoch": 0.08354186717998074,
+      "grad_norm": 1.0227729082107544,
+      "learning_rate": 0.0001991447821414282,
+      "loss": 1.3967,
+      "step": 2170
+    },
+    {
+      "epoch": 0.08373435996150144,
+      "grad_norm": 1.3028923273086548,
+      "learning_rate": 0.00019914083122777245,
+      "loss": 1.4296,
+      "step": 2175
+    },
+    {
+      "epoch": 0.08392685274302214,
+      "grad_norm": 1.6131690740585327,
+      "learning_rate": 0.00019913687124834442,
+      "loss": 1.2983,
+      "step": 2180
+    },
+    {
+      "epoch": 0.08411934552454282,
+      "grad_norm": 1.1791858673095703,
+      "learning_rate": 0.00019913290220350622,
+      "loss": 1.4632,
+      "step": 2185
+    },
+    {
+      "epoch": 0.08431183830606352,
+      "grad_norm": 1.8457857370376587,
+      "learning_rate": 0.00019912892409362085,
+      "loss": 1.3623,
+      "step": 2190
+    },
+    {
+      "epoch": 0.08450433108758422,
+      "grad_norm": 1.525680422782898,
+      "learning_rate": 0.00019912493691905198,
+      "loss": 1.2729,
+      "step": 2195
+    },
+    {
+      "epoch": 0.0846968238691049,
+      "grad_norm": 1.3267451524734497,
+      "learning_rate": 0.0001991209406801643,
+      "loss": 1.3808,
+      "step": 2200
+    },
+    {
+      "epoch": 0.0848893166506256,
+      "grad_norm": 1.37312912940979,
+      "learning_rate": 0.00019911693537732323,
+      "loss": 1.6072,
+      "step": 2205
+    },
+    {
+      "epoch": 0.0850818094321463,
+      "grad_norm": 1.3433706760406494,
+      "learning_rate": 0.000199112921010895,
+      "loss": 1.4956,
+      "step": 2210
+    },
+    {
+      "epoch": 0.08527430221366698,
+      "grad_norm": 1.220732569694519,
+      "learning_rate": 0.00019910889758124672,
+      "loss": 1.4875,
+      "step": 2215
+    },
+    {
+      "epoch": 0.08546679499518768,
+      "grad_norm": 0.9385544657707214,
+      "learning_rate": 0.00019910486508874627,
+      "loss": 1.4202,
+      "step": 2220
+    },
+    {
+      "epoch": 0.08565928777670838,
+      "grad_norm": 0.8727134466171265,
+      "learning_rate": 0.0001991008235337624,
+      "loss": 1.2268,
+      "step": 2225
+    },
+    {
+      "epoch": 0.08585178055822906,
+      "grad_norm": 2.276063919067383,
+      "learning_rate": 0.00019909677291666473,
+      "loss": 1.3911,
+      "step": 2230
+    },
+    {
+      "epoch": 0.08604427333974976,
+      "grad_norm": 1.2023353576660156,
+      "learning_rate": 0.00019909271323782364,
+      "loss": 1.4754,
+      "step": 2235
+    },
+    {
+      "epoch": 0.08623676612127046,
+      "grad_norm": 0.9018556475639343,
+      "learning_rate": 0.00019908864449761033,
+      "loss": 1.4073,
+      "step": 2240
+    },
+    {
+      "epoch": 0.08642925890279114,
+      "grad_norm": 1.2011221647262573,
+      "learning_rate": 0.00019908456669639687,
+      "loss": 1.3213,
+      "step": 2245
+    },
+    {
+      "epoch": 0.08662175168431184,
+      "grad_norm": 1.9858746528625488,
+      "learning_rate": 0.0001990804798345562,
+      "loss": 1.3403,
+      "step": 2250
+    },
+    {
+      "epoch": 0.08681424446583254,
+      "grad_norm": 1.0072557926177979,
+      "learning_rate": 0.000199076383912462,
+      "loss": 1.3387,
+      "step": 2255
+    },
+    {
+      "epoch": 0.08700673724735322,
+      "grad_norm": 1.4516913890838623,
+      "learning_rate": 0.00019907227893048877,
+      "loss": 1.3755,
+      "step": 2260
+    },
+    {
+      "epoch": 0.08719923002887392,
+      "grad_norm": 1.0636364221572876,
+      "learning_rate": 0.00019906816488901195,
+      "loss": 1.2495,
+      "step": 2265
+    },
+    {
+      "epoch": 0.0873917228103946,
+      "grad_norm": 1.8495078086853027,
+      "learning_rate": 0.0001990640417884077,
+      "loss": 1.4166,
+      "step": 2270
+    },
+    {
+      "epoch": 0.0875842155919153,
+      "grad_norm": 2.327951431274414,
+      "learning_rate": 0.00019905990962905312,
+      "loss": 1.3934,
+      "step": 2275
+    },
+    {
+      "epoch": 0.087776708373436,
+      "grad_norm": 1.5719425678253174,
+      "learning_rate": 0.00019905576841132595,
+      "loss": 1.3932,
+      "step": 2280
+    },
+    {
+      "epoch": 0.08796920115495668,
+      "grad_norm": 1.5799787044525146,
+      "learning_rate": 0.000199051618135605,
+      "loss": 1.5148,
+      "step": 2285
+    },
+    {
+      "epoch": 0.08816169393647738,
+      "grad_norm": 0.7972100377082825,
+      "learning_rate": 0.00019904745880226966,
+      "loss": 1.2456,
+      "step": 2290
+    },
+    {
+      "epoch": 0.08835418671799808,
+      "grad_norm": 1.4252464771270752,
+      "learning_rate": 0.00019904329041170042,
+      "loss": 1.4287,
+      "step": 2295
+    },
+    {
+      "epoch": 0.08854667949951876,
+      "grad_norm": 1.5532910823822021,
+      "learning_rate": 0.00019903911296427834,
+      "loss": 1.3685,
+      "step": 2300
+    },
+    {
+      "epoch": 0.08873917228103946,
+      "grad_norm": 1.3019160032272339,
+      "learning_rate": 0.00019903492646038544,
+      "loss": 1.3928,
+      "step": 2305
+    },
+    {
+      "epoch": 0.08893166506256016,
+      "grad_norm": 1.7292853593826294,
+      "learning_rate": 0.00019903073090040457,
+      "loss": 1.369,
+      "step": 2310
+    },
+    {
+      "epoch": 0.08912415784408084,
+      "grad_norm": 1.1780908107757568,
+      "learning_rate": 0.00019902652628471938,
+      "loss": 1.2541,
+      "step": 2315
+    },
+    {
+      "epoch": 0.08931665062560154,
+      "grad_norm": 1.353721261024475,
+      "learning_rate": 0.00019902231261371433,
+      "loss": 1.2658,
+      "step": 2320
+    },
+    {
+      "epoch": 0.08950914340712224,
+      "grad_norm": 1.0020657777786255,
+      "learning_rate": 0.0001990180898877748,
+      "loss": 1.3319,
+      "step": 2325
+    },
+    {
+      "epoch": 0.08970163618864292,
+      "grad_norm": 1.1655325889587402,
+      "learning_rate": 0.00019901385810728686,
+      "loss": 1.3783,
+      "step": 2330
+    },
+    {
+      "epoch": 0.08989412897016362,
+      "grad_norm": 1.2237039804458618,
+      "learning_rate": 0.00019900961727263748,
+      "loss": 1.2919,
+      "step": 2335
+    },
+    {
+      "epoch": 0.09008662175168432,
+      "grad_norm": 1.6417179107666016,
+      "learning_rate": 0.0001990053673842145,
+      "loss": 1.471,
+      "step": 2340
+    },
+    {
+      "epoch": 0.090279114533205,
+      "grad_norm": 1.2170498371124268,
+      "learning_rate": 0.00019900110844240653,
+      "loss": 1.3889,
+      "step": 2345
+    },
+    {
+      "epoch": 0.0904716073147257,
+      "grad_norm": 1.1462334394454956,
+      "learning_rate": 0.00019899684044760304,
+      "loss": 1.4191,
+      "step": 2350
+    },
+    {
+      "epoch": 0.0906641000962464,
+      "grad_norm": 0.961063802242279,
+      "learning_rate": 0.00019899256340019425,
+      "loss": 1.5019,
+      "step": 2355
+    },
+    {
+      "epoch": 0.09085659287776708,
+      "grad_norm": 0.9323278069496155,
+      "learning_rate": 0.0001989882773005713,
+      "loss": 1.3988,
+      "step": 2360
+    },
+    {
+      "epoch": 0.09104908565928778,
+      "grad_norm": 1.8326833248138428,
+      "learning_rate": 0.00019898398214912612,
+      "loss": 1.4211,
+      "step": 2365
+    },
+    {
+      "epoch": 0.09124157844080846,
+      "grad_norm": 1.2725722789764404,
+      "learning_rate": 0.00019897967794625153,
+      "loss": 1.3274,
+      "step": 2370
+    },
+    {
+      "epoch": 0.09143407122232916,
+      "grad_norm": 0.9105005860328674,
+      "learning_rate": 0.00019897536469234102,
+      "loss": 1.3309,
+      "step": 2375
+    },
+    {
+      "epoch": 0.09162656400384986,
+      "grad_norm": 1.3157737255096436,
+      "learning_rate": 0.00019897104238778907,
+      "loss": 1.4086,
+      "step": 2380
+    },
+    {
+      "epoch": 0.09181905678537054,
+      "grad_norm": 1.9295995235443115,
+      "learning_rate": 0.00019896671103299094,
+      "loss": 1.3849,
+      "step": 2385
+    },
+    {
+      "epoch": 0.09201154956689124,
+      "grad_norm": 1.0183601379394531,
+      "learning_rate": 0.00019896237062834267,
+      "loss": 1.4397,
+      "step": 2390
+    },
+    {
+      "epoch": 0.09220404234841194,
+      "grad_norm": 1.118998646736145,
+      "learning_rate": 0.00019895802117424118,
+      "loss": 1.568,
+      "step": 2395
+    },
+    {
+      "epoch": 0.09239653512993262,
+      "grad_norm": 1.6463871002197266,
+      "learning_rate": 0.00019895366267108416,
+      "loss": 1.2755,
+      "step": 2400
+    },
+    {
+      "epoch": 0.09258902791145332,
+      "grad_norm": 1.3326902389526367,
+      "learning_rate": 0.00019894929511927022,
+      "loss": 1.4369,
+      "step": 2405
+    },
+    {
+      "epoch": 0.09278152069297402,
+      "grad_norm": 1.4168566465377808,
+      "learning_rate": 0.00019894491851919871,
+      "loss": 1.4323,
+      "step": 2410
+    },
+    {
+      "epoch": 0.0929740134744947,
+      "grad_norm": 1.3266388177871704,
+      "learning_rate": 0.00019894053287126986,
+      "loss": 1.17,
+      "step": 2415
+    },
+    {
+      "epoch": 0.0931665062560154,
+      "grad_norm": 1.7362377643585205,
+      "learning_rate": 0.0001989361381758847,
+      "loss": 1.5996,
+      "step": 2420
+    },
+    {
+      "epoch": 0.0933589990375361,
+      "grad_norm": 1.1684424877166748,
+      "learning_rate": 0.00019893173443344511,
+      "loss": 1.3486,
+      "step": 2425
+    },
+    {
+      "epoch": 0.09355149181905678,
+      "grad_norm": 1.3784310817718506,
+      "learning_rate": 0.00019892732164435376,
+      "loss": 1.2775,
+      "step": 2430
+    },
+    {
+      "epoch": 0.09374398460057748,
+      "grad_norm": 1.1288561820983887,
+      "learning_rate": 0.00019892289980901414,
+      "loss": 1.2044,
+      "step": 2435
+    },
+    {
+      "epoch": 0.09393647738209818,
+      "grad_norm": 1.1601535081863403,
+      "learning_rate": 0.00019891846892783067,
+      "loss": 1.4937,
+      "step": 2440
+    },
+    {
+      "epoch": 0.09412897016361886,
+      "grad_norm": 1.3866316080093384,
+      "learning_rate": 0.0001989140290012085,
+      "loss": 1.913,
+      "step": 2445
+    },
+    {
+      "epoch": 0.09432146294513956,
+      "grad_norm": 1.4638808965682983,
+      "learning_rate": 0.00019890958002955362,
+      "loss": 1.4114,
+      "step": 2450
+    },
+    {
+      "epoch": 0.09451395572666026,
+      "grad_norm": 1.4660701751708984,
+      "learning_rate": 0.00019890512201327284,
+      "loss": 1.3607,
+      "step": 2455
+    },
+    {
+      "epoch": 0.09470644850818094,
+      "grad_norm": 0.9787619113922119,
+      "learning_rate": 0.00019890065495277388,
+      "loss": 1.3729,
+      "step": 2460
+    },
+    {
+      "epoch": 0.09489894128970164,
+      "grad_norm": 1.4845494031906128,
+      "learning_rate": 0.00019889617884846517,
+      "loss": 1.3326,
+      "step": 2465
+    },
+    {
+      "epoch": 0.09509143407122234,
+      "grad_norm": 1.2955145835876465,
+      "learning_rate": 0.000198891693700756,
+      "loss": 1.3738,
+      "step": 2470
+    },
+    {
+      "epoch": 0.09528392685274302,
+      "grad_norm": 1.7431209087371826,
+      "learning_rate": 0.00019888719951005656,
+      "loss": 1.3676,
+      "step": 2475
+    },
+    {
+      "epoch": 0.09547641963426372,
+      "grad_norm": 0.923613965511322,
+      "learning_rate": 0.00019888269627677777,
+      "loss": 1.4142,
+      "step": 2480
+    },
+    {
+      "epoch": 0.0956689124157844,
+      "grad_norm": 1.0258625745773315,
+      "learning_rate": 0.0001988781840013315,
+      "loss": 1.3868,
+      "step": 2485
+    },
+    {
+      "epoch": 0.0958614051973051,
+      "grad_norm": 1.1365761756896973,
+      "learning_rate": 0.00019887366268413025,
+      "loss": 1.2871,
+      "step": 2490
+    },
+    {
+      "epoch": 0.0960538979788258,
+      "grad_norm": 2.3250112533569336,
+      "learning_rate": 0.00019886913232558754,
+      "loss": 1.4345,
+      "step": 2495
+    },
+    {
+      "epoch": 0.09624639076034648,
+      "grad_norm": 1.1625771522521973,
+      "learning_rate": 0.00019886459292611767,
+      "loss": 1.5796,
+      "step": 2500
+    },
+    {
+      "epoch": 0.09643888354186718,
+      "grad_norm": 1.7454233169555664,
+      "learning_rate": 0.00019886004448613562,
+      "loss": 1.6151,
+      "step": 2505
+    },
+    {
+      "epoch": 0.09663137632338788,
+      "grad_norm": 1.3514907360076904,
+      "learning_rate": 0.00019885548700605745,
+      "loss": 1.4529,
+      "step": 2510
+    },
+    {
+      "epoch": 0.09682386910490856,
+      "grad_norm": 1.9735958576202393,
+      "learning_rate": 0.00019885092048629982,
+      "loss": 1.4945,
+      "step": 2515
+    },
+    {
+      "epoch": 0.09701636188642926,
+      "grad_norm": 1.190207600593567,
+      "learning_rate": 0.00019884634492728037,
+      "loss": 1.473,
+      "step": 2520
+    },
+    {
+      "epoch": 0.09720885466794996,
+      "grad_norm": 1.1596134901046753,
+      "learning_rate": 0.00019884176032941743,
+      "loss": 1.3745,
+      "step": 2525
+    },
+    {
+      "epoch": 0.09740134744947064,
+      "grad_norm": 1.0496324300765991,
+      "learning_rate": 0.0001988371666931303,
+      "loss": 1.3853,
+      "step": 2530
+    },
+    {
+      "epoch": 0.09759384023099134,
+      "grad_norm": 1.2820552587509155,
+      "learning_rate": 0.000198832564018839,
+      "loss": 1.4205,
+      "step": 2535
+    },
+    {
+      "epoch": 0.09778633301251204,
+      "grad_norm": 0.9559310674667358,
+      "learning_rate": 0.00019882795230696446,
+      "loss": 1.2517,
+      "step": 2540
+    },
+    {
+      "epoch": 0.09797882579403272,
+      "grad_norm": 1.026782751083374,
+      "learning_rate": 0.00019882333155792835,
+      "loss": 1.335,
+      "step": 2545
+    },
+    {
+      "epoch": 0.09817131857555342,
+      "grad_norm": 1.3378793001174927,
+      "learning_rate": 0.00019881870177215319,
+      "loss": 1.3419,
+      "step": 2550
+    },
+    {
+      "epoch": 0.09836381135707412,
+      "grad_norm": 1.0646761655807495,
+      "learning_rate": 0.00019881406295006238,
+      "loss": 1.3793,
+      "step": 2555
+    },
+    {
+      "epoch": 0.0985563041385948,
+      "grad_norm": 1.3302899599075317,
+      "learning_rate": 0.00019880941509208005,
+      "loss": 1.3056,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0987487969201155,
+      "grad_norm": 1.3029305934906006,
+      "learning_rate": 0.00019880475819863134,
+      "loss": 1.3028,
+      "step": 2565
+    },
+    {
+      "epoch": 0.0989412897016362,
+      "grad_norm": 1.6653764247894287,
+      "learning_rate": 0.00019880009227014197,
+      "loss": 1.4698,
+      "step": 2570
+    },
+    {
+      "epoch": 0.09913378248315688,
+      "grad_norm": 1.5575610399246216,
+      "learning_rate": 0.00019879541730703865,
+      "loss": 1.2843,
+      "step": 2575
+    },
+    {
+      "epoch": 0.09932627526467758,
+      "grad_norm": 1.1219451427459717,
+      "learning_rate": 0.0001987907333097489,
+      "loss": 1.2824,
+      "step": 2580
+    },
+    {
+      "epoch": 0.09951876804619826,
+      "grad_norm": 1.680050253868103,
+      "learning_rate": 0.000198786040278701,
+      "loss": 1.431,
+      "step": 2585
+    },
+    {
+      "epoch": 0.09971126082771896,
+      "grad_norm": 2.5341451168060303,
+      "learning_rate": 0.00019878133821432412,
+      "loss": 1.3925,
+      "step": 2590
+    },
+    {
+      "epoch": 0.09990375360923966,
+      "grad_norm": 1.132542610168457,
+      "learning_rate": 0.00019877662711704824,
+      "loss": 1.4082,
+      "step": 2595
+    },
+    {
+      "epoch": 0.10009624639076034,
+      "grad_norm": 1.0605584383010864,
+      "learning_rate": 0.0001987719069873041,
+      "loss": 1.2904,
+      "step": 2600
+    },
+    {
+      "epoch": 0.10028873917228104,
+      "grad_norm": 1.161116361618042,
+      "learning_rate": 0.0001987671778255234,
+      "loss": 1.2922,
+      "step": 2605
+    },
+    {
+      "epoch": 0.10048123195380174,
+      "grad_norm": 2.2763168811798096,
+      "learning_rate": 0.0001987624396321386,
+      "loss": 1.4692,
+      "step": 2610
+    },
+    {
+      "epoch": 0.10067372473532242,
+      "grad_norm": 1.547316312789917,
+      "learning_rate": 0.00019875769240758286,
+      "loss": 1.458,
+      "step": 2615
+    },
+    {
+      "epoch": 0.10086621751684312,
+      "grad_norm": 1.0679529905319214,
+      "learning_rate": 0.0001987529361522904,
+      "loss": 1.3075,
+      "step": 2620
+    },
+    {
+      "epoch": 0.10105871029836382,
+      "grad_norm": 1.9426227807998657,
+      "learning_rate": 0.0001987481708666961,
+      "loss": 1.4985,
+      "step": 2625
+    },
+    {
+      "epoch": 0.1012512030798845,
+      "grad_norm": 1.1619765758514404,
+      "learning_rate": 0.00019874339655123575,
+      "loss": 1.329,
+      "step": 2630
+    },
+    {
+      "epoch": 0.1014436958614052,
+      "grad_norm": 0.8115332722663879,
+      "learning_rate": 0.00019873861320634587,
+      "loss": 1.218,
+      "step": 2635
+    },
+    {
+      "epoch": 0.1016361886429259,
+      "grad_norm": 1.2575538158416748,
+      "learning_rate": 0.0001987338208324639,
+      "loss": 1.3133,
+      "step": 2640
+    },
+    {
+      "epoch": 0.10182868142444658,
+      "grad_norm": 0.9605635404586792,
+      "learning_rate": 0.00019872901943002806,
+      "loss": 1.4462,
+      "step": 2645
+    },
+    {
+      "epoch": 0.10202117420596728,
+      "grad_norm": 1.7909116744995117,
+      "learning_rate": 0.00019872420899947742,
+      "loss": 1.257,
+      "step": 2650
+    },
+    {
+      "epoch": 0.10221366698748797,
+      "grad_norm": 1.5501129627227783,
+      "learning_rate": 0.00019871938954125185,
+      "loss": 1.2825,
+      "step": 2655
+    },
+    {
+      "epoch": 0.10240615976900866,
+      "grad_norm": 1.4636069536209106,
+      "learning_rate": 0.00019871456105579208,
+      "loss": 1.3909,
+      "step": 2660
+    },
+    {
+      "epoch": 0.10259865255052936,
+      "grad_norm": 1.4283297061920166,
+      "learning_rate": 0.0001987097235435396,
+      "loss": 1.2148,
+      "step": 2665
+    },
+    {
+      "epoch": 0.10279114533205005,
+      "grad_norm": 1.316149115562439,
+      "learning_rate": 0.00019870487700493684,
+      "loss": 1.393,
+      "step": 2670
+    },
+    {
+      "epoch": 0.10298363811357074,
+      "grad_norm": 0.8449459671974182,
+      "learning_rate": 0.00019870002144042689,
+      "loss": 1.4969,
+      "step": 2675
+    },
+    {
+      "epoch": 0.10317613089509144,
+      "grad_norm": 1.3309835195541382,
+      "learning_rate": 0.00019869515685045383,
+      "loss": 1.4927,
+      "step": 2680
+    },
+    {
+      "epoch": 0.10336862367661212,
+      "grad_norm": 0.9159907102584839,
+      "learning_rate": 0.00019869028323546246,
+      "loss": 1.3526,
+      "step": 2685
+    },
+    {
+      "epoch": 0.10356111645813282,
+      "grad_norm": 2.2842464447021484,
+      "learning_rate": 0.00019868540059589845,
+      "loss": 1.3646,
+      "step": 2690
+    },
+    {
+      "epoch": 0.10375360923965352,
+      "grad_norm": 0.9444146156311035,
+      "learning_rate": 0.00019868050893220832,
+      "loss": 1.349,
+      "step": 2695
+    },
+    {
+      "epoch": 0.1039461020211742,
+      "grad_norm": 1.8546898365020752,
+      "learning_rate": 0.0001986756082448393,
+      "loss": 1.3195,
+      "step": 2700
+    },
+    {
+      "epoch": 0.1041385948026949,
+      "grad_norm": 1.310783863067627,
+      "learning_rate": 0.00019867069853423961,
+      "loss": 1.6065,
+      "step": 2705
+    },
+    {
+      "epoch": 0.1043310875842156,
+      "grad_norm": 1.248542308807373,
+      "learning_rate": 0.00019866577980085813,
+      "loss": 1.1987,
+      "step": 2710
+    },
+    {
+      "epoch": 0.10452358036573628,
+      "grad_norm": 1.421844482421875,
+      "learning_rate": 0.00019866085204514472,
+      "loss": 1.3576,
+      "step": 2715
+    },
+    {
+      "epoch": 0.10471607314725698,
+      "grad_norm": 1.1641993522644043,
+      "learning_rate": 0.00019865591526754996,
+      "loss": 1.436,
+      "step": 2720
+    },
+    {
+      "epoch": 0.10490856592877768,
+      "grad_norm": 1.1122993230819702,
+      "learning_rate": 0.0001986509694685253,
+      "loss": 1.4218,
+      "step": 2725
+    },
+    {
+      "epoch": 0.10510105871029836,
+      "grad_norm": 1.222016453742981,
+      "learning_rate": 0.00019864601464852295,
+      "loss": 1.2965,
+      "step": 2730
+    },
+    {
+      "epoch": 0.10529355149181906,
+      "grad_norm": 1.6765378713607788,
+      "learning_rate": 0.00019864105080799602,
+      "loss": 1.3908,
+      "step": 2735
+    },
+    {
+      "epoch": 0.10548604427333975,
+      "grad_norm": 1.8405592441558838,
+      "learning_rate": 0.00019863607794739845,
+      "loss": 1.2583,
+      "step": 2740
+    },
+    {
+      "epoch": 0.10567853705486044,
+      "grad_norm": 1.3908604383468628,
+      "learning_rate": 0.00019863109606718497,
+      "loss": 1.2726,
+      "step": 2745
+    },
+    {
+      "epoch": 0.10587102983638114,
+      "grad_norm": 1.3825894594192505,
+      "learning_rate": 0.0001986261051678111,
+      "loss": 1.3234,
+      "step": 2750
+    },
+    {
+      "epoch": 0.10606352261790183,
+      "grad_norm": 1.5409029722213745,
+      "learning_rate": 0.00019862110524973328,
+      "loss": 1.4151,
+      "step": 2755
+    },
+    {
+      "epoch": 0.10625601539942252,
+      "grad_norm": 2.1902191638946533,
+      "learning_rate": 0.00019861609631340868,
+      "loss": 1.3865,
+      "step": 2760
+    },
+    {
+      "epoch": 0.10644850818094322,
+      "grad_norm": 0.9851712584495544,
+      "learning_rate": 0.00019861107835929533,
+      "loss": 1.4799,
+      "step": 2765
+    },
+    {
+      "epoch": 0.10664100096246391,
+      "grad_norm": 1.2206732034683228,
+      "learning_rate": 0.0001986060513878521,
+      "loss": 1.3456,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1068334937439846,
+      "grad_norm": 1.3443645238876343,
+      "learning_rate": 0.0001986010153995387,
+      "loss": 1.2586,
+      "step": 2775
+    },
+    {
+      "epoch": 0.1070259865255053,
+      "grad_norm": 1.1602864265441895,
+      "learning_rate": 0.00019859597039481561,
+      "loss": 1.1789,
+      "step": 2780
+    },
+    {
+      "epoch": 0.107218479307026,
+      "grad_norm": 0.8068190813064575,
+      "learning_rate": 0.00019859091637414414,
+      "loss": 1.4228,
+      "step": 2785
+    },
+    {
+      "epoch": 0.10741097208854668,
+      "grad_norm": 1.4439321756362915,
+      "learning_rate": 0.0001985858533379865,
+      "loss": 1.4365,
+      "step": 2790
+    },
+    {
+      "epoch": 0.10760346487006738,
+      "grad_norm": 1.0814299583435059,
+      "learning_rate": 0.00019858078128680564,
+      "loss": 1.2755,
+      "step": 2795
+    },
+    {
+      "epoch": 0.10779595765158806,
+      "grad_norm": 1.7848068475723267,
+      "learning_rate": 0.00019857570022106536,
+      "loss": 1.4061,
+      "step": 2800
+    },
+    {
+      "epoch": 0.10798845043310876,
+      "grad_norm": 1.3163549900054932,
+      "learning_rate": 0.0001985706101412303,
+      "loss": 1.3599,
+      "step": 2805
+    },
+    {
+      "epoch": 0.10818094321462945,
+      "grad_norm": 1.439104437828064,
+      "learning_rate": 0.0001985655110477659,
+      "loss": 1.3054,
+      "step": 2810
+    },
+    {
+      "epoch": 0.10837343599615014,
+      "grad_norm": 0.892706036567688,
+      "learning_rate": 0.0001985604029411385,
+      "loss": 1.3504,
+      "step": 2815
+    },
+    {
+      "epoch": 0.10856592877767084,
+      "grad_norm": 1.102704405784607,
+      "learning_rate": 0.0001985552858218151,
+      "loss": 1.3902,
+      "step": 2820
+    },
+    {
+      "epoch": 0.10875842155919153,
+      "grad_norm": 1.21804678440094,
+      "learning_rate": 0.0001985501596902637,
+      "loss": 1.36,
+      "step": 2825
+    },
+    {
+      "epoch": 0.10895091434071222,
+      "grad_norm": 1.6015477180480957,
+      "learning_rate": 0.00019854502454695302,
+      "loss": 1.6163,
+      "step": 2830
+    },
+    {
+      "epoch": 0.10914340712223292,
+      "grad_norm": 1.3947224617004395,
+      "learning_rate": 0.00019853988039235265,
+      "loss": 1.2207,
+      "step": 2835
+    },
+    {
+      "epoch": 0.10933589990375361,
+      "grad_norm": 1.616458535194397,
+      "learning_rate": 0.00019853472722693302,
+      "loss": 1.2081,
+      "step": 2840
+    },
+    {
+      "epoch": 0.1095283926852743,
+      "grad_norm": 2.1588330268859863,
+      "learning_rate": 0.00019852956505116528,
+      "loss": 1.4428,
+      "step": 2845
+    },
+    {
+      "epoch": 0.109720885466795,
+      "grad_norm": 1.2287509441375732,
+      "learning_rate": 0.00019852439386552152,
+      "loss": 1.4548,
+      "step": 2850
+    },
+    {
+      "epoch": 0.1099133782483157,
+      "grad_norm": 1.7198657989501953,
+      "learning_rate": 0.00019851921367047463,
+      "loss": 1.2034,
+      "step": 2855
+    },
+    {
+      "epoch": 0.11010587102983638,
+      "grad_norm": 1.4924067258834839,
+      "learning_rate": 0.00019851402446649825,
+      "loss": 1.3635,
+      "step": 2860
+    },
+    {
+      "epoch": 0.11029836381135708,
+      "grad_norm": 1.3675332069396973,
+      "learning_rate": 0.00019850882625406695,
+      "loss": 1.29,
+      "step": 2865
+    },
+    {
+      "epoch": 0.11049085659287777,
+      "grad_norm": 1.2170599699020386,
+      "learning_rate": 0.00019850361903365603,
+      "loss": 1.3495,
+      "step": 2870
+    },
+    {
+      "epoch": 0.11068334937439846,
+      "grad_norm": 1.6067026853561401,
+      "learning_rate": 0.00019849840280574167,
+      "loss": 1.4679,
+      "step": 2875
+    },
+    {
+      "epoch": 0.11087584215591915,
+      "grad_norm": 1.0457261800765991,
+      "learning_rate": 0.00019849317757080092,
+      "loss": 1.3289,
+      "step": 2880
+    },
+    {
+      "epoch": 0.11106833493743985,
+      "grad_norm": 0.6958736181259155,
+      "learning_rate": 0.00019848794332931146,
+      "loss": 0.9412,
+      "step": 2885
+    },
+    {
+      "epoch": 0.11126082771896054,
+      "grad_norm": 0.9687005281448364,
+      "learning_rate": 0.00019848270008175205,
+      "loss": 1.2777,
+      "step": 2890
+    },
+    {
+      "epoch": 0.11145332050048123,
+      "grad_norm": 0.8073298931121826,
+      "learning_rate": 0.00019847744782860213,
+      "loss": 1.4295,
+      "step": 2895
+    },
+    {
+      "epoch": 0.11164581328200192,
+      "grad_norm": 0.8794350624084473,
+      "learning_rate": 0.00019847218657034193,
+      "loss": 1.2199,
+      "step": 2900
+    },
+    {
+      "epoch": 0.11183830606352262,
+      "grad_norm": 1.644554853439331,
+      "learning_rate": 0.00019846691630745258,
+      "loss": 1.3076,
+      "step": 2905
+    },
+    {
+      "epoch": 0.11203079884504331,
+      "grad_norm": 1.0819231271743774,
+      "learning_rate": 0.00019846163704041603,
+      "loss": 1.385,
+      "step": 2910
+    },
+    {
+      "epoch": 0.112223291626564,
+      "grad_norm": 1.4424269199371338,
+      "learning_rate": 0.000198456348769715,
+      "loss": 1.4287,
+      "step": 2915
+    },
+    {
+      "epoch": 0.1124157844080847,
+      "grad_norm": 1.289413332939148,
+      "learning_rate": 0.00019845105149583308,
+      "loss": 1.25,
+      "step": 2920
+    },
+    {
+      "epoch": 0.1126082771896054,
+      "grad_norm": 1.4669229984283447,
+      "learning_rate": 0.00019844574521925474,
+      "loss": 1.5371,
+      "step": 2925
+    },
+    {
+      "epoch": 0.11280076997112608,
+      "grad_norm": 2.102736473083496,
+      "learning_rate": 0.0001984404299404651,
+      "loss": 1.5017,
+      "step": 2930
+    },
+    {
+      "epoch": 0.11299326275264678,
+      "grad_norm": 1.1487330198287964,
+      "learning_rate": 0.00019843510565995025,
+      "loss": 1.3164,
+      "step": 2935
+    },
+    {
+      "epoch": 0.11318575553416747,
+      "grad_norm": 1.259538173675537,
+      "learning_rate": 0.00019842977237819707,
+      "loss": 1.2946,
+      "step": 2940
+    },
+    {
+      "epoch": 0.11337824831568816,
+      "grad_norm": 2.3158466815948486,
+      "learning_rate": 0.00019842443009569324,
+      "loss": 1.4614,
+      "step": 2945
+    },
+    {
+      "epoch": 0.11357074109720885,
+      "grad_norm": 1.5077046155929565,
+      "learning_rate": 0.0001984190788129273,
+      "loss": 1.3478,
+      "step": 2950
+    },
+    {
+      "epoch": 0.11376323387872955,
+      "grad_norm": 1.2548809051513672,
+      "learning_rate": 0.00019841371853038852,
+      "loss": 1.3351,
+      "step": 2955
+    },
+    {
+      "epoch": 0.11395572666025024,
+      "grad_norm": 1.4622430801391602,
+      "learning_rate": 0.00019840834924856715,
+      "loss": 1.2788,
+      "step": 2960
+    },
+    {
+      "epoch": 0.11414821944177093,
+      "grad_norm": 0.9759154319763184,
+      "learning_rate": 0.00019840297096795415,
+      "loss": 1.2793,
+      "step": 2965
+    },
+    {
+      "epoch": 0.11434071222329163,
+      "grad_norm": 1.2217987775802612,
+      "learning_rate": 0.00019839758368904128,
+      "loss": 1.284,
+      "step": 2970
+    },
+    {
+      "epoch": 0.11453320500481232,
+      "grad_norm": 2.180697441101074,
+      "learning_rate": 0.00019839326738746614,
+      "loss": 1.4163,
+      "step": 2975
+    },
+    {
+      "epoch": 0.11472569778633301,
+      "grad_norm": 1.156293511390686,
+      "learning_rate": 0.00019838786391285554,
+      "loss": 1.3045,
+      "step": 2980
+    },
+    {
+      "epoch": 0.11491819056785371,
+      "grad_norm": 1.1444417238235474,
+      "learning_rate": 0.00019838245144132658,
+      "loss": 1.4522,
+      "step": 2985
+    },
+    {
+      "epoch": 0.1151106833493744,
+      "grad_norm": 1.3959949016571045,
+      "learning_rate": 0.00019837702997337414,
+      "loss": 1.3959,
+      "step": 2990
+    },
+    {
+      "epoch": 0.1153031761308951,
+      "grad_norm": 1.2789435386657715,
+      "learning_rate": 0.00019837159950949402,
+      "loss": 1.2951,
+      "step": 2995
+    },
+    {
+      "epoch": 0.11549566891241578,
+      "grad_norm": 1.0902299880981445,
+      "learning_rate": 0.00019836616005018275,
+      "loss": 1.4573,
+      "step": 3000
+    },
+    {
+      "epoch": 0.11568816169393648,
+      "grad_norm": 1.452920913696289,
+      "learning_rate": 0.0001983607115959378,
+      "loss": 1.4688,
+      "step": 3005
+    },
+    {
+      "epoch": 0.11588065447545717,
+      "grad_norm": 2.192514419555664,
+      "learning_rate": 0.0001983552541472573,
+      "loss": 1.4282,
+      "step": 3010
+    },
+    {
+      "epoch": 0.11607314725697786,
+      "grad_norm": 1.938883900642395,
+      "learning_rate": 0.0001983497877046404,
+      "loss": 1.6123,
+      "step": 3015
+    },
+    {
+      "epoch": 0.11626564003849855,
+      "grad_norm": 2.4365732669830322,
+      "learning_rate": 0.0001983443122685869,
+      "loss": 1.4987,
+      "step": 3020
+    },
+    {
+      "epoch": 0.11645813282001925,
+      "grad_norm": 1.827972173690796,
+      "learning_rate": 0.0001983388278395975,
+      "loss": 1.2196,
+      "step": 3025
+    },
+    {
+      "epoch": 0.11665062560153994,
+      "grad_norm": 1.6184618473052979,
+      "learning_rate": 0.00019833333441817374,
+      "loss": 1.5257,
+      "step": 3030
+    },
+    {
+      "epoch": 0.11684311838306063,
+      "grad_norm": 1.0191036462783813,
+      "learning_rate": 0.00019832783200481797,
+      "loss": 1.4799,
+      "step": 3035
+    },
+    {
+      "epoch": 0.11703561116458133,
+      "grad_norm": 1.1552925109863281,
+      "learning_rate": 0.0001983223206000333,
+      "loss": 1.2014,
+      "step": 3040
+    },
+    {
+      "epoch": 0.11722810394610202,
+      "grad_norm": 0.9793531894683838,
+      "learning_rate": 0.00019831680020432376,
+      "loss": 1.2092,
+      "step": 3045
+    },
+    {
+      "epoch": 0.11742059672762271,
+      "grad_norm": 1.480634331703186,
+      "learning_rate": 0.0001983112708181941,
+      "loss": 1.3238,
+      "step": 3050
+    },
+    {
+      "epoch": 0.11761308950914341,
+      "grad_norm": 1.5112073421478271,
+      "learning_rate": 0.00019830573244215,
+      "loss": 1.5513,
+      "step": 3055
+    },
+    {
+      "epoch": 0.1178055822906641,
+      "grad_norm": 1.4130852222442627,
+      "learning_rate": 0.00019830018507669786,
+      "loss": 1.4368,
+      "step": 3060
+    },
+    {
+      "epoch": 0.1179980750721848,
+      "grad_norm": 1.401934027671814,
+      "learning_rate": 0.000198294628722345,
+      "loss": 1.243,
+      "step": 3065
+    },
+    {
+      "epoch": 0.11819056785370549,
+      "grad_norm": 1.8309379816055298,
+      "learning_rate": 0.00019828906337959946,
+      "loss": 1.1656,
+      "step": 3070
+    },
+    {
+      "epoch": 0.11838306063522618,
+      "grad_norm": 0.8511875867843628,
+      "learning_rate": 0.0001982834890489702,
+      "loss": 1.406,
+      "step": 3075
+    },
+    {
+      "epoch": 0.11857555341674687,
+      "grad_norm": 1.4291598796844482,
+      "learning_rate": 0.00019827790573096694,
+      "loss": 1.3963,
+      "step": 3080
+    },
+    {
+      "epoch": 0.11876804619826757,
+      "grad_norm": 0.6835631132125854,
+      "learning_rate": 0.0001982723134261002,
+      "loss": 1.1238,
+      "step": 3085
+    },
+    {
+      "epoch": 0.11896053897978826,
+      "grad_norm": 1.6569236516952515,
+      "learning_rate": 0.00019826671213488145,
+      "loss": 1.3335,
+      "step": 3090
+    },
+    {
+      "epoch": 0.11915303176130895,
+      "grad_norm": 1.0488132238388062,
+      "learning_rate": 0.00019826110185782277,
+      "loss": 1.3009,
+      "step": 3095
+    },
+    {
+      "epoch": 0.11934552454282965,
+      "grad_norm": 1.3253639936447144,
+      "learning_rate": 0.00019825548259543726,
+      "loss": 1.3863,
+      "step": 3100
+    },
+    {
+      "epoch": 0.11953801732435033,
+      "grad_norm": 0.9408076405525208,
+      "learning_rate": 0.00019824985434823878,
+      "loss": 1.3184,
+      "step": 3105
+    },
+    {
+      "epoch": 0.11973051010587103,
+      "grad_norm": 0.9649772644042969,
+      "learning_rate": 0.00019824421711674194,
+      "loss": 1.2427,
+      "step": 3110
+    },
+    {
+      "epoch": 0.11992300288739172,
+      "grad_norm": 1.7673052549362183,
+      "learning_rate": 0.00019823857090146225,
+      "loss": 1.2804,
+      "step": 3115
+    },
+    {
+      "epoch": 0.12011549566891241,
+      "grad_norm": 1.230724811553955,
+      "learning_rate": 0.00019823291570291604,
+      "loss": 1.3527,
+      "step": 3120
+    },
+    {
+      "epoch": 0.12030798845043311,
+      "grad_norm": 2.382617473602295,
+      "learning_rate": 0.0001982272515216204,
+      "loss": 1.4123,
+      "step": 3125
+    },
+    {
+      "epoch": 0.1205004812319538,
+      "grad_norm": 1.2811720371246338,
+      "learning_rate": 0.00019822157835809332,
+      "loss": 1.3935,
+      "step": 3130
+    },
+    {
+      "epoch": 0.1206929740134745,
+      "grad_norm": 1.9592630863189697,
+      "learning_rate": 0.00019821589621285356,
+      "loss": 1.2387,
+      "step": 3135
+    },
+    {
+      "epoch": 0.12088546679499519,
+      "grad_norm": 1.659197449684143,
+      "learning_rate": 0.0001982102050864207,
+      "loss": 1.4228,
+      "step": 3140
+    },
+    {
+      "epoch": 0.12107795957651588,
+      "grad_norm": 1.2591451406478882,
+      "learning_rate": 0.00019820450497931517,
+      "loss": 1.3192,
+      "step": 3145
+    },
+    {
+      "epoch": 0.12127045235803657,
+      "grad_norm": 1.1670453548431396,
+      "learning_rate": 0.00019819879589205822,
+      "loss": 1.2593,
+      "step": 3150
+    },
+    {
+      "epoch": 0.12146294513955727,
+      "grad_norm": 1.680776834487915,
+      "learning_rate": 0.0001981930778251719,
+      "loss": 1.5809,
+      "step": 3155
+    },
+    {
+      "epoch": 0.12165543792107796,
+      "grad_norm": 1.388492226600647,
+      "learning_rate": 0.00019818735077917904,
+      "loss": 1.5646,
+      "step": 3160
+    },
+    {
+      "epoch": 0.12184793070259865,
+      "grad_norm": 1.3851470947265625,
+      "learning_rate": 0.00019818161475460342,
+      "loss": 1.3282,
+      "step": 3165
+    },
+    {
+      "epoch": 0.12204042348411935,
+      "grad_norm": 1.252103567123413,
+      "learning_rate": 0.0001981758697519695,
+      "loss": 1.3326,
+      "step": 3170
+    },
+    {
+      "epoch": 0.12223291626564003,
+      "grad_norm": 2.6637227535247803,
+      "learning_rate": 0.0001981701157718027,
+      "loss": 1.4247,
+      "step": 3175
+    },
+    {
+      "epoch": 0.12242540904716073,
+      "grad_norm": 1.4228829145431519,
+      "learning_rate": 0.00019816435281462907,
+      "loss": 1.3287,
+      "step": 3180
+    },
+    {
+      "epoch": 0.12261790182868143,
+      "grad_norm": 1.0654631853103638,
+      "learning_rate": 0.00019815858088097565,
+      "loss": 1.3651,
+      "step": 3185
+    },
+    {
+      "epoch": 0.12281039461020211,
+      "grad_norm": 1.1779879331588745,
+      "learning_rate": 0.00019815279997137028,
+      "loss": 1.2699,
+      "step": 3190
+    },
+    {
+      "epoch": 0.12300288739172281,
+      "grad_norm": 0.966482937335968,
+      "learning_rate": 0.0001981470100863416,
+      "loss": 1.3029,
+      "step": 3195
+    },
+    {
+      "epoch": 0.12319538017324351,
+      "grad_norm": 1.13119375705719,
+      "learning_rate": 0.00019814121122641894,
+      "loss": 1.3431,
+      "step": 3200
+    },
+    {
+      "epoch": 0.1233878729547642,
+      "grad_norm": 1.0690468549728394,
+      "learning_rate": 0.00019813540339213263,
+      "loss": 1.237,
+      "step": 3205
+    },
+    {
+      "epoch": 0.12358036573628489,
+      "grad_norm": 1.169592022895813,
+      "learning_rate": 0.00019812958658401382,
+      "loss": 1.3341,
+      "step": 3210
+    },
+    {
+      "epoch": 0.12377285851780558,
+      "grad_norm": 0.9310591816902161,
+      "learning_rate": 0.00019812376080259435,
+      "loss": 1.3168,
+      "step": 3215
+    },
+    {
+      "epoch": 0.12396535129932627,
+      "grad_norm": 1.1262513399124146,
+      "learning_rate": 0.00019811792604840694,
+      "loss": 1.322,
+      "step": 3220
+    },
+    {
+      "epoch": 0.12415784408084697,
+      "grad_norm": 1.0723376274108887,
+      "learning_rate": 0.00019811208232198518,
+      "loss": 1.2814,
+      "step": 3225
+    },
+    {
+      "epoch": 0.12435033686236766,
+      "grad_norm": 1.5084266662597656,
+      "learning_rate": 0.00019810622962386344,
+      "loss": 1.3136,
+      "step": 3230
+    },
+    {
+      "epoch": 0.12454282964388835,
+      "grad_norm": 1.5219266414642334,
+      "learning_rate": 0.0001981003679545769,
+      "loss": 1.2971,
+      "step": 3235
+    },
+    {
+      "epoch": 0.12473532242540905,
+      "grad_norm": 1.8135708570480347,
+      "learning_rate": 0.00019809449731466154,
+      "loss": 1.3987,
+      "step": 3240
+    },
+    {
+      "epoch": 0.12492781520692973,
+      "grad_norm": 1.9838290214538574,
+      "learning_rate": 0.00019808861770465424,
+      "loss": 1.4063,
+      "step": 3245
+    },
+    {
+      "epoch": 0.12512030798845045,
+      "grad_norm": 0.9821895956993103,
+      "learning_rate": 0.00019808272912509258,
+      "loss": 1.4336,
+      "step": 3250
+    },
+    {
+      "epoch": 0.12531280076997112,
+      "grad_norm": 1.0371532440185547,
+      "learning_rate": 0.00019807683157651513,
+      "loss": 1.4659,
+      "step": 3255
+    },
+    {
+      "epoch": 0.12550529355149181,
+      "grad_norm": 1.2441003322601318,
+      "learning_rate": 0.0001980709250594611,
+      "loss": 1.3807,
+      "step": 3260
+    },
+    {
+      "epoch": 0.1256977863330125,
+      "grad_norm": 1.6097456216812134,
+      "learning_rate": 0.00019806500957447067,
+      "loss": 1.4115,
+      "step": 3265
+    },
+    {
+      "epoch": 0.1258902791145332,
+      "grad_norm": 1.4005634784698486,
+      "learning_rate": 0.0001980590851220847,
+      "loss": 1.6008,
+      "step": 3270
+    },
+    {
+      "epoch": 0.1260827718960539,
+      "grad_norm": 1.1883544921875,
+      "learning_rate": 0.00019805315170284498,
+      "loss": 1.3768,
+      "step": 3275
+    },
+    {
+      "epoch": 0.12627526467757458,
+      "grad_norm": 1.2404242753982544,
+      "learning_rate": 0.00019804720931729413,
+      "loss": 1.463,
+      "step": 3280
+    },
+    {
+      "epoch": 0.12646775745909528,
+      "grad_norm": 0.625027596950531,
+      "learning_rate": 0.00019804125796597544,
+      "loss": 1.3286,
+      "step": 3285
+    },
+    {
+      "epoch": 0.12666025024061597,
+      "grad_norm": 1.5616633892059326,
+      "learning_rate": 0.0001980352976494332,
+      "loss": 1.4161,
+      "step": 3290
+    },
+    {
+      "epoch": 0.12685274302213667,
+      "grad_norm": 0.8003360629081726,
+      "learning_rate": 0.0001980293283682124,
+      "loss": 1.4117,
+      "step": 3295
+    },
+    {
+      "epoch": 0.12704523580365737,
+      "grad_norm": 1.0671011209487915,
+      "learning_rate": 0.0001980233501228589,
+      "loss": 1.4192,
+      "step": 3300
+    },
+    {
+      "epoch": 0.12723772858517807,
+      "grad_norm": 1.4135669469833374,
+      "learning_rate": 0.0001980173629139194,
+      "loss": 1.3046,
+      "step": 3305
+    },
+    {
+      "epoch": 0.12743022136669874,
+      "grad_norm": 1.0450470447540283,
+      "learning_rate": 0.00019801136674194134,
+      "loss": 1.4156,
+      "step": 3310
+    },
+    {
+      "epoch": 0.12762271414821943,
+      "grad_norm": 1.1435261964797974,
+      "learning_rate": 0.00019800536160747306,
+      "loss": 1.2311,
+      "step": 3315
+    },
+    {
+      "epoch": 0.12781520692974013,
+      "grad_norm": 1.5508229732513428,
+      "learning_rate": 0.0001979993475110637,
+      "loss": 1.4224,
+      "step": 3320
+    },
+    {
+      "epoch": 0.12800769971126083,
+      "grad_norm": 0.9542085528373718,
+      "learning_rate": 0.0001979933244532632,
+      "loss": 1.2423,
+      "step": 3325
+    },
+    {
+      "epoch": 0.12820019249278153,
+      "grad_norm": 1.5797593593597412,
+      "learning_rate": 0.0001979872924346223,
+      "loss": 1.3357,
+      "step": 3330
+    },
+    {
+      "epoch": 0.12839268527430223,
+      "grad_norm": 1.0982688665390015,
+      "learning_rate": 0.00019798125145569263,
+      "loss": 1.2404,
+      "step": 3335
+    },
+    {
+      "epoch": 0.1285851780558229,
+      "grad_norm": 1.5471248626708984,
+      "learning_rate": 0.0001979752015170266,
+      "loss": 1.3556,
+      "step": 3340
+    },
+    {
+      "epoch": 0.1287776708373436,
+      "grad_norm": 1.64442777633667,
+      "learning_rate": 0.0001979691426191774,
+      "loss": 1.3407,
+      "step": 3345
+    },
+    {
+      "epoch": 0.1289701636188643,
+      "grad_norm": 1.494186520576477,
+      "learning_rate": 0.0001979630747626991,
+      "loss": 1.4509,
+      "step": 3350
+    },
+    {
+      "epoch": 0.129162656400385,
+      "grad_norm": 0.9598186612129211,
+      "learning_rate": 0.00019795699794814654,
+      "loss": 1.3221,
+      "step": 3355
+    },
+    {
+      "epoch": 0.1293551491819057,
+      "grad_norm": 1.1328315734863281,
+      "learning_rate": 0.00019795091217607544,
+      "loss": 1.5129,
+      "step": 3360
+    },
+    {
+      "epoch": 0.12954764196342639,
+      "grad_norm": 1.0476043224334717,
+      "learning_rate": 0.00019794481744704227,
+      "loss": 1.3448,
+      "step": 3365
+    },
+    {
+      "epoch": 0.12974013474494706,
+      "grad_norm": 1.2570463418960571,
+      "learning_rate": 0.0001979387137616044,
+      "loss": 1.2726,
+      "step": 3370
+    },
+    {
+      "epoch": 0.12993262752646775,
+      "grad_norm": 1.395627498626709,
+      "learning_rate": 0.00019793260112031992,
+      "loss": 1.1469,
+      "step": 3375
+    },
+    {
+      "epoch": 0.13012512030798845,
+      "grad_norm": 2.2382960319519043,
+      "learning_rate": 0.00019792647952374782,
+      "loss": 1.3375,
+      "step": 3380
+    },
+    {
+      "epoch": 0.13031761308950915,
+      "grad_norm": 1.4930087327957153,
+      "learning_rate": 0.00019792034897244784,
+      "loss": 1.3684,
+      "step": 3385
+    },
+    {
+      "epoch": 0.13051010587102985,
+      "grad_norm": 0.9732452034950256,
+      "learning_rate": 0.00019791420946698064,
+      "loss": 1.0792,
+      "step": 3390
+    },
+    {
+      "epoch": 0.13070259865255052,
+      "grad_norm": 1.9484987258911133,
+      "learning_rate": 0.0001979080610079076,
+      "loss": 1.4284,
+      "step": 3395
+    },
+    {
+      "epoch": 0.13089509143407121,
+      "grad_norm": 1.3746837377548218,
+      "learning_rate": 0.00019790190359579097,
+      "loss": 1.4393,
+      "step": 3400
+    },
+    {
+      "epoch": 0.1310875842155919,
+      "grad_norm": 1.2191319465637207,
+      "learning_rate": 0.0001978957372311938,
+      "loss": 1.2184,
+      "step": 3405
+    },
+    {
+      "epoch": 0.1312800769971126,
+      "grad_norm": 1.0825196504592896,
+      "learning_rate": 0.00019788956191467994,
+      "loss": 1.3891,
+      "step": 3410
+    },
+    {
+      "epoch": 0.1314725697786333,
+      "grad_norm": 1.9972898960113525,
+      "learning_rate": 0.00019788337764681412,
+      "loss": 1.3207,
+      "step": 3415
+    },
+    {
+      "epoch": 0.131665062560154,
+      "grad_norm": 1.3864003419876099,
+      "learning_rate": 0.00019787718442816182,
+      "loss": 1.3791,
+      "step": 3420
+    },
+    {
+      "epoch": 0.13185755534167468,
+      "grad_norm": 1.3315006494522095,
+      "learning_rate": 0.0001978709822592894,
+      "loss": 1.4253,
+      "step": 3425
+    },
+    {
+      "epoch": 0.13205004812319537,
+      "grad_norm": 1.0171843767166138,
+      "learning_rate": 0.00019786477114076397,
+      "loss": 1.2974,
+      "step": 3430
+    },
+    {
+      "epoch": 0.13224254090471607,
+      "grad_norm": 1.293380618095398,
+      "learning_rate": 0.00019785855107315353,
+      "loss": 1.3616,
+      "step": 3435
+    },
+    {
+      "epoch": 0.13243503368623677,
+      "grad_norm": 2.0498528480529785,
+      "learning_rate": 0.00019785232205702681,
+      "loss": 1.3431,
+      "step": 3440
+    },
+    {
+      "epoch": 0.13262752646775747,
+      "grad_norm": 0.8635803461074829,
+      "learning_rate": 0.0001978460840929535,
+      "loss": 1.3672,
+      "step": 3445
+    },
+    {
+      "epoch": 0.13282001924927817,
+      "grad_norm": 0.9983857274055481,
+      "learning_rate": 0.00019783983718150392,
+      "loss": 1.4856,
+      "step": 3450
+    },
+    {
+      "epoch": 0.13301251203079884,
+      "grad_norm": 4.542407989501953,
+      "learning_rate": 0.00019783358132324937,
+      "loss": 1.4599,
+      "step": 3455
+    },
+    {
+      "epoch": 0.13320500481231953,
+      "grad_norm": 1.5495860576629639,
+      "learning_rate": 0.00019782731651876194,
+      "loss": 1.3641,
+      "step": 3460
+    },
+    {
+      "epoch": 0.13339749759384023,
+      "grad_norm": 1.2070780992507935,
+      "learning_rate": 0.00019782104276861443,
+      "loss": 1.3596,
+      "step": 3465
+    },
+    {
+      "epoch": 0.13358999037536093,
+      "grad_norm": 1.1749752759933472,
+      "learning_rate": 0.00019781476007338058,
+      "loss": 1.2387,
+      "step": 3470
+    },
+    {
+      "epoch": 0.13378248315688163,
+      "grad_norm": 1.8580079078674316,
+      "learning_rate": 0.00019780846843363485,
+      "loss": 1.3966,
+      "step": 3475
+    },
+    {
+      "epoch": 0.1339749759384023,
+      "grad_norm": 1.9713795185089111,
+      "learning_rate": 0.00019780216784995265,
+      "loss": 1.2541,
+      "step": 3480
+    },
+    {
+      "epoch": 0.134167468719923,
+      "grad_norm": 1.4017597436904907,
+      "learning_rate": 0.00019779585832291002,
+      "loss": 1.4827,
+      "step": 3485
+    },
+    {
+      "epoch": 0.1343599615014437,
+      "grad_norm": 1.188761591911316,
+      "learning_rate": 0.00019778953985308406,
+      "loss": 1.3972,
+      "step": 3490
+    },
+    {
+      "epoch": 0.1345524542829644,
+      "grad_norm": 1.0930372476577759,
+      "learning_rate": 0.00019778321244105242,
+      "loss": 1.4706,
+      "step": 3495
+    },
+    {
+      "epoch": 0.1347449470644851,
+      "grad_norm": 1.3041532039642334,
+      "learning_rate": 0.0001977768760873938,
+      "loss": 1.1929,
+      "step": 3500
+    },
+    {
+      "epoch": 0.13493743984600579,
+      "grad_norm": 2.6741833686828613,
+      "learning_rate": 0.00019777053079268753,
+      "loss": 1.268,
+      "step": 3505
+    },
+    {
+      "epoch": 0.13512993262752646,
+      "grad_norm": 1.091823935508728,
+      "learning_rate": 0.0001977641765575139,
+      "loss": 1.2776,
+      "step": 3510
+    },
+    {
+      "epoch": 0.13532242540904715,
+      "grad_norm": 0.9205764532089233,
+      "learning_rate": 0.00019775781338245398,
+      "loss": 1.3007,
+      "step": 3515
+    },
+    {
+      "epoch": 0.13551491819056785,
+      "grad_norm": 1.6321576833724976,
+      "learning_rate": 0.00019775144126808958,
+      "loss": 1.4214,
+      "step": 3520
+    },
+    {
+      "epoch": 0.13570741097208855,
+      "grad_norm": 1.7947146892547607,
+      "learning_rate": 0.00019774506021500343,
+      "loss": 1.3895,
+      "step": 3525
+    },
+    {
+      "epoch": 0.13589990375360925,
+      "grad_norm": 1.6696717739105225,
+      "learning_rate": 0.00019773867022377902,
+      "loss": 1.3968,
+      "step": 3530
+    },
+    {
+      "epoch": 0.13609239653512994,
+      "grad_norm": 1.1003444194793701,
+      "learning_rate": 0.0001977322712950007,
+      "loss": 1.4084,
+      "step": 3535
+    },
+    {
+      "epoch": 0.13628488931665061,
+      "grad_norm": 1.0268352031707764,
+      "learning_rate": 0.00019772586342925357,
+      "loss": 1.254,
+      "step": 3540
+    },
+    {
+      "epoch": 0.1364773820981713,
+      "grad_norm": 1.3906810283660889,
+      "learning_rate": 0.0001977194466271236,
+      "loss": 1.3266,
+      "step": 3545
+    },
+    {
+      "epoch": 0.136669874879692,
+      "grad_norm": 1.1786664724349976,
+      "learning_rate": 0.00019771302088919757,
+      "loss": 1.3114,
+      "step": 3550
+    },
+    {
+      "epoch": 0.1368623676612127,
+      "grad_norm": 1.0252714157104492,
+      "learning_rate": 0.00019770658621606307,
+      "loss": 1.2089,
+      "step": 3555
+    },
+    {
+      "epoch": 0.1370548604427334,
+      "grad_norm": 0.8099033236503601,
+      "learning_rate": 0.00019770014260830853,
+      "loss": 1.2607,
+      "step": 3560
+    },
+    {
+      "epoch": 0.1372473532242541,
+      "grad_norm": 1.3679542541503906,
+      "learning_rate": 0.0001976936900665231,
+      "loss": 1.376,
+      "step": 3565
+    },
+    {
+      "epoch": 0.13743984600577477,
+      "grad_norm": 1.7685283422470093,
+      "learning_rate": 0.00019768722859129693,
+      "loss": 1.4522,
+      "step": 3570
+    },
+    {
+      "epoch": 0.13763233878729547,
+      "grad_norm": 1.0158277750015259,
+      "learning_rate": 0.00019768075818322081,
+      "loss": 1.2714,
+      "step": 3575
+    },
+    {
+      "epoch": 0.13782483156881617,
+      "grad_norm": 1.7043020725250244,
+      "learning_rate": 0.00019767427884288642,
+      "loss": 1.5669,
+      "step": 3580
+    },
+    {
+      "epoch": 0.13801732435033687,
+      "grad_norm": 1.8171344995498657,
+      "learning_rate": 0.00019766779057088627,
+      "loss": 1.4186,
+      "step": 3585
+    },
+    {
+      "epoch": 0.13820981713185757,
+      "grad_norm": 1.0524088144302368,
+      "learning_rate": 0.00019766129336781365,
+      "loss": 1.167,
+      "step": 3590
+    },
+    {
+      "epoch": 0.13840230991337824,
+      "grad_norm": 1.558383584022522,
+      "learning_rate": 0.0001976547872342627,
+      "loss": 1.5015,
+      "step": 3595
+    },
+    {
+      "epoch": 0.13859480269489893,
+      "grad_norm": 1.9925919771194458,
+      "learning_rate": 0.00019764827217082838,
+      "loss": 1.3661,
+      "step": 3600
+    },
+    {
+      "epoch": 0.13878729547641963,
+      "grad_norm": 1.5693559646606445,
+      "learning_rate": 0.0001976417481781064,
+      "loss": 1.3389,
+      "step": 3605
+    },
+    {
+      "epoch": 0.13897978825794033,
+      "grad_norm": 1.2609871625900269,
+      "learning_rate": 0.00019763521525669343,
+      "loss": 1.2883,
+      "step": 3610
+    },
+    {
+      "epoch": 0.13917228103946103,
+      "grad_norm": 1.4910306930541992,
+      "learning_rate": 0.00019762867340718674,
+      "loss": 1.4237,
+      "step": 3615
+    },
+    {
+      "epoch": 0.13936477382098172,
+      "grad_norm": 0.9409481287002563,
+      "learning_rate": 0.0001976221226301846,
+      "loss": 1.4289,
+      "step": 3620
+    },
+    {
+      "epoch": 0.1395572666025024,
+      "grad_norm": 0.9263445138931274,
+      "learning_rate": 0.00019761556292628604,
+      "loss": 1.2987,
+      "step": 3625
+    },
+    {
+      "epoch": 0.1397497593840231,
+      "grad_norm": 0.9329832792282104,
+      "learning_rate": 0.0001976089942960909,
+      "loss": 1.3709,
+      "step": 3630
+    },
+    {
+      "epoch": 0.1399422521655438,
+      "grad_norm": 1.7852829694747925,
+      "learning_rate": 0.00019760241674019984,
+      "loss": 1.2282,
+      "step": 3635
+    },
+    {
+      "epoch": 0.1401347449470645,
+      "grad_norm": 1.0068609714508057,
+      "learning_rate": 0.0001975958302592143,
+      "loss": 1.3143,
+      "step": 3640
+    },
+    {
+      "epoch": 0.14032723772858519,
+      "grad_norm": 2.1680188179016113,
+      "learning_rate": 0.0001975892348537366,
+      "loss": 1.4447,
+      "step": 3645
+    },
+    {
+      "epoch": 0.14051973051010588,
+      "grad_norm": 1.633169412612915,
+      "learning_rate": 0.00019758263052436988,
+      "loss": 1.2633,
+      "step": 3650
+    },
+    {
+      "epoch": 0.14071222329162655,
+      "grad_norm": 1.3609623908996582,
+      "learning_rate": 0.000197576017271718,
+      "loss": 1.3352,
+      "step": 3655
+    },
+    {
+      "epoch": 0.14090471607314725,
+      "grad_norm": 1.50294828414917,
+      "learning_rate": 0.00019756939509638573,
+      "loss": 1.3557,
+      "step": 3660
+    },
+    {
+      "epoch": 0.14109720885466795,
+      "grad_norm": 0.9931232333183289,
+      "learning_rate": 0.0001975627639989786,
+      "loss": 1.4719,
+      "step": 3665
+    },
+    {
+      "epoch": 0.14128970163618865,
+      "grad_norm": 1.3870011568069458,
+      "learning_rate": 0.000197556123980103,
+      "loss": 1.5173,
+      "step": 3670
+    },
+    {
+      "epoch": 0.14148219441770934,
+      "grad_norm": 1.274064540863037,
+      "learning_rate": 0.00019754947504036608,
+      "loss": 1.3951,
+      "step": 3675
+    },
+    {
+      "epoch": 0.14167468719923004,
+      "grad_norm": 1.6096014976501465,
+      "learning_rate": 0.00019754281718037593,
+      "loss": 1.4478,
+      "step": 3680
+    },
+    {
+      "epoch": 0.1418671799807507,
+      "grad_norm": 1.155772089958191,
+      "learning_rate": 0.00019753615040074131,
+      "loss": 1.229,
+      "step": 3685
+    },
+    {
+      "epoch": 0.1420596727622714,
+      "grad_norm": 1.123856544494629,
+      "learning_rate": 0.0001975294747020718,
+      "loss": 1.5036,
+      "step": 3690
+    },
+    {
+      "epoch": 0.1422521655437921,
+      "grad_norm": 1.541308879852295,
+      "learning_rate": 0.00019752279008497796,
+      "loss": 1.1174,
+      "step": 3695
+    },
+    {
+      "epoch": 0.1424446583253128,
+      "grad_norm": 1.8912441730499268,
+      "learning_rate": 0.00019751609655007098,
+      "loss": 1.3753,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1426371511068335,
+      "grad_norm": 1.7746648788452148,
+      "learning_rate": 0.00019750939409796293,
+      "loss": 1.3115,
+      "step": 3705
+    },
+    {
+      "epoch": 0.14282964388835417,
+      "grad_norm": 1.2228045463562012,
+      "learning_rate": 0.00019750268272926676,
+      "loss": 1.3477,
+      "step": 3710
+    },
+    {
+      "epoch": 0.14302213666987487,
+      "grad_norm": 1.5031695365905762,
+      "learning_rate": 0.00019749596244459614,
+      "loss": 1.1905,
+      "step": 3715
+    },
+    {
+      "epoch": 0.14321462945139557,
+      "grad_norm": 2.871879816055298,
+      "learning_rate": 0.0001974892332445656,
+      "loss": 1.3334,
+      "step": 3720
+    },
+    {
+      "epoch": 0.14340712223291627,
+      "grad_norm": 1.1911511421203613,
+      "learning_rate": 0.00019748249512979048,
+      "loss": 1.2528,
+      "step": 3725
+    },
+    {
+      "epoch": 0.14359961501443697,
+      "grad_norm": 1.2722115516662598,
+      "learning_rate": 0.00019747574810088697,
+      "loss": 1.3314,
+      "step": 3730
+    },
+    {
+      "epoch": 0.14379210779595766,
+      "grad_norm": 1.0464539527893066,
+      "learning_rate": 0.00019746899215847198,
+      "loss": 1.1621,
+      "step": 3735
+    },
+    {
+      "epoch": 0.14398460057747833,
+      "grad_norm": 1.8877158164978027,
+      "learning_rate": 0.00019746222730316338,
+      "loss": 1.2534,
+      "step": 3740
+    },
+    {
+      "epoch": 0.14417709335899903,
+      "grad_norm": 1.5137780904769897,
+      "learning_rate": 0.00019745545353557967,
+      "loss": 1.1738,
+      "step": 3745
+    },
+    {
+      "epoch": 0.14436958614051973,
+      "grad_norm": 1.7104227542877197,
+      "learning_rate": 0.00019744867085634034,
+      "loss": 1.2868,
+      "step": 3750
+    },
+    {
+      "epoch": 0.14456207892204043,
+      "grad_norm": 1.2920212745666504,
+      "learning_rate": 0.00019744187926606558,
+      "loss": 1.3054,
+      "step": 3755
+    },
+    {
+      "epoch": 0.14475457170356112,
+      "grad_norm": 2.3661959171295166,
+      "learning_rate": 0.00019743507876537647,
+      "loss": 1.3187,
+      "step": 3760
+    },
+    {
+      "epoch": 0.14494706448508182,
+      "grad_norm": 1.4622807502746582,
+      "learning_rate": 0.00019742826935489487,
+      "loss": 1.1548,
+      "step": 3765
+    },
+    {
+      "epoch": 0.1451395572666025,
+      "grad_norm": 1.7818437814712524,
+      "learning_rate": 0.00019742145103524342,
+      "loss": 1.4081,
+      "step": 3770
+    },
+    {
+      "epoch": 0.1453320500481232,
+      "grad_norm": 1.023716926574707,
+      "learning_rate": 0.00019741462380704566,
+      "loss": 1.3367,
+      "step": 3775
+    },
+    {
+      "epoch": 0.1455245428296439,
+      "grad_norm": 1.4382961988449097,
+      "learning_rate": 0.00019740778767092585,
+      "loss": 1.3498,
+      "step": 3780
+    },
+    {
+      "epoch": 0.14571703561116459,
+      "grad_norm": 1.5282870531082153,
+      "learning_rate": 0.0001974009426275091,
+      "loss": 1.2685,
+      "step": 3785
+    },
+    {
+      "epoch": 0.14590952839268528,
+      "grad_norm": 1.2222365140914917,
+      "learning_rate": 0.0001973940886774214,
+      "loss": 1.2273,
+      "step": 3790
+    },
+    {
+      "epoch": 0.14610202117420595,
+      "grad_norm": 1.3231360912322998,
+      "learning_rate": 0.00019738722582128944,
+      "loss": 1.5449,
+      "step": 3795
+    },
+    {
+      "epoch": 0.14629451395572665,
+      "grad_norm": 1.2198995351791382,
+      "learning_rate": 0.00019738035405974085,
+      "loss": 1.4927,
+      "step": 3800
+    },
+    {
+      "epoch": 0.14648700673724735,
+      "grad_norm": 1.1108288764953613,
+      "learning_rate": 0.00019737347339340394,
+      "loss": 1.3894,
+      "step": 3805
+    },
+    {
+      "epoch": 0.14667949951876805,
+      "grad_norm": 1.1478091478347778,
+      "learning_rate": 0.0001973665838229079,
+      "loss": 1.342,
+      "step": 3810
+    },
+    {
+      "epoch": 0.14687199230028875,
+      "grad_norm": 1.555680751800537,
+      "learning_rate": 0.0001973596853488828,
+      "loss": 1.269,
+      "step": 3815
+    },
+    {
+      "epoch": 0.14706448508180944,
+      "grad_norm": 1.2819339036941528,
+      "learning_rate": 0.0001973527779719594,
+      "loss": 1.3462,
+      "step": 3820
+    },
+    {
+      "epoch": 0.1472569778633301,
+      "grad_norm": 1.6733057498931885,
+      "learning_rate": 0.00019734586169276939,
+      "loss": 1.3179,
+      "step": 3825
+    },
+    {
+      "epoch": 0.1474494706448508,
+      "grad_norm": 1.8622225522994995,
+      "learning_rate": 0.00019733893651194517,
+      "loss": 1.452,
+      "step": 3830
+    },
+    {
+      "epoch": 0.1476419634263715,
+      "grad_norm": 1.2225052118301392,
+      "learning_rate": 0.00019733200243012006,
+      "loss": 1.2925,
+      "step": 3835
+    },
+    {
+      "epoch": 0.1478344562078922,
+      "grad_norm": 0.7980884313583374,
+      "learning_rate": 0.00019732505944792804,
+      "loss": 1.1505,
+      "step": 3840
+    },
+    {
+      "epoch": 0.1480269489894129,
+      "grad_norm": 1.3874131441116333,
+      "learning_rate": 0.00019731810756600405,
+      "loss": 1.2989,
+      "step": 3845
+    },
+    {
+      "epoch": 0.1482194417709336,
+      "grad_norm": 1.4387590885162354,
+      "learning_rate": 0.00019731114678498378,
+      "loss": 1.3295,
+      "step": 3850
+    },
+    {
+      "epoch": 0.14841193455245427,
+      "grad_norm": 1.8189646005630493,
+      "learning_rate": 0.00019730417710550383,
+      "loss": 1.2926,
+      "step": 3855
+    },
+    {
+      "epoch": 0.14860442733397497,
+      "grad_norm": 0.9577664732933044,
+      "learning_rate": 0.0001972971985282014,
+      "loss": 1.2375,
+      "step": 3860
+    },
+    {
+      "epoch": 0.14879692011549567,
+      "grad_norm": 1.7154825925827026,
+      "learning_rate": 0.00019729021105371474,
+      "loss": 1.2853,
+      "step": 3865
+    },
+    {
+      "epoch": 0.14898941289701637,
+      "grad_norm": 2.1061089038848877,
+      "learning_rate": 0.00019728321468268277,
+      "loss": 1.3391,
+      "step": 3870
+    },
+    {
+      "epoch": 0.14918190567853706,
+      "grad_norm": 1.0177017450332642,
+      "learning_rate": 0.00019727620941574524,
+      "loss": 1.2801,
+      "step": 3875
+    },
+    {
+      "epoch": 0.14937439846005776,
+      "grad_norm": 1.0773547887802124,
+      "learning_rate": 0.00019726919525354277,
+      "loss": 1.3063,
+      "step": 3880
+    },
+    {
+      "epoch": 0.14956689124157843,
+      "grad_norm": 0.9082854986190796,
+      "learning_rate": 0.00019726217219671673,
+      "loss": 1.3601,
+      "step": 3885
+    },
+    {
+      "epoch": 0.14975938402309913,
+      "grad_norm": 1.341280221939087,
+      "learning_rate": 0.00019725514024590934,
+      "loss": 1.4052,
+      "step": 3890
+    },
+    {
+      "epoch": 0.14995187680461983,
+      "grad_norm": 2.240399122238159,
+      "learning_rate": 0.00019724809940176364,
+      "loss": 1.1955,
+      "step": 3895
+    },
+    {
+      "epoch": 0.15014436958614052,
+      "grad_norm": 1.549137830734253,
+      "learning_rate": 0.00019724104966492348,
+      "loss": 1.3089,
+      "step": 3900
+    },
+    {
+      "epoch": 0.15033686236766122,
+      "grad_norm": 1.6887294054031372,
+      "learning_rate": 0.00019723399103603346,
+      "loss": 1.4147,
+      "step": 3905
+    },
+    {
+      "epoch": 0.1505293551491819,
+      "grad_norm": 1.793087363243103,
+      "learning_rate": 0.0001972269235157391,
+      "loss": 1.2674,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1507218479307026,
+      "grad_norm": 1.718336820602417,
+      "learning_rate": 0.00019721984710468663,
+      "loss": 1.2716,
+      "step": 3915
+    },
+    {
+      "epoch": 0.1509143407122233,
+      "grad_norm": 2.2342288494110107,
+      "learning_rate": 0.0001972127618035232,
+      "loss": 0.965,
+      "step": 3920
+    },
+    {
+      "epoch": 0.15110683349374399,
+      "grad_norm": 1.5450822114944458,
+      "learning_rate": 0.00019720566761289665,
+      "loss": 1.3461,
+      "step": 3925
+    },
+    {
+      "epoch": 0.15129932627526468,
+      "grad_norm": 1.4395346641540527,
+      "learning_rate": 0.0001971985645334557,
+      "loss": 1.3462,
+      "step": 3930
+    },
+    {
+      "epoch": 0.15149181905678538,
+      "grad_norm": 1.1160500049591064,
+      "learning_rate": 0.00019719145256584994,
+      "loss": 1.3334,
+      "step": 3935
+    },
+    {
+      "epoch": 0.15168431183830605,
+      "grad_norm": 1.0270999670028687,
+      "learning_rate": 0.00019718433171072967,
+      "loss": 1.2737,
+      "step": 3940
+    },
+    {
+      "epoch": 0.15187680461982675,
+      "grad_norm": 1.4266023635864258,
+      "learning_rate": 0.00019717720196874608,
+      "loss": 1.3639,
+      "step": 3945
+    },
+    {
+      "epoch": 0.15206929740134745,
+      "grad_norm": 1.552283525466919,
+      "learning_rate": 0.00019717006334055108,
+      "loss": 1.301,
+      "step": 3950
+    },
+    {
+      "epoch": 0.15226179018286815,
+      "grad_norm": 1.5459437370300293,
+      "learning_rate": 0.0001971629158267975,
+      "loss": 1.265,
+      "step": 3955
+    },
+    {
+      "epoch": 0.15245428296438884,
+      "grad_norm": 1.4866915941238403,
+      "learning_rate": 0.00019715575942813888,
+      "loss": 1.5694,
+      "step": 3960
+    },
+    {
+      "epoch": 0.15264677574590954,
+      "grad_norm": 1.1116254329681396,
+      "learning_rate": 0.00019714859414522967,
+      "loss": 1.4858,
+      "step": 3965
+    },
+    {
+      "epoch": 0.1528392685274302,
+      "grad_norm": 1.1708245277404785,
+      "learning_rate": 0.0001971414199787251,
+      "loss": 1.3582,
+      "step": 3970
+    },
+    {
+      "epoch": 0.1530317613089509,
+      "grad_norm": 1.1672711372375488,
+      "learning_rate": 0.00019713423692928114,
+      "loss": 1.3393,
+      "step": 3975
+    },
+    {
+      "epoch": 0.1532242540904716,
+      "grad_norm": 1.4800153970718384,
+      "learning_rate": 0.0001971270449975547,
+      "loss": 1.22,
+      "step": 3980
+    },
+    {
+      "epoch": 0.1534167468719923,
+      "grad_norm": 1.92826509475708,
+      "learning_rate": 0.00019711984418420338,
+      "loss": 1.3902,
+      "step": 3985
+    },
+    {
+      "epoch": 0.153609239653513,
+      "grad_norm": 1.2292252779006958,
+      "learning_rate": 0.00019711263448988567,
+      "loss": 1.2327,
+      "step": 3990
+    },
+    {
+      "epoch": 0.1538017324350337,
+      "grad_norm": 1.1007169485092163,
+      "learning_rate": 0.00019710541591526085,
+      "loss": 1.4284,
+      "step": 3995
+    },
+    {
+      "epoch": 0.15399422521655437,
+      "grad_norm": 0.9456301927566528,
+      "learning_rate": 0.00019709818846098905,
+      "loss": 1.1589,
+      "step": 4000
+    },
+    {
+      "epoch": 0.15418671799807507,
+      "grad_norm": 1.518704891204834,
+      "learning_rate": 0.0001970909521277311,
+      "loss": 1.3976,
+      "step": 4005
+    },
+    {
+      "epoch": 0.15437921077959577,
+      "grad_norm": 1.3318589925765991,
+      "learning_rate": 0.00019708370691614872,
+      "loss": 1.3635,
+      "step": 4010
+    },
+    {
+      "epoch": 0.15457170356111646,
+      "grad_norm": 1.752626657485962,
+      "learning_rate": 0.0001970764528269045,
+      "loss": 1.3175,
+      "step": 4015
+    },
+    {
+      "epoch": 0.15476419634263716,
+      "grad_norm": 2.055469512939453,
+      "learning_rate": 0.00019706918986066172,
+      "loss": 1.2873,
+      "step": 4020
+    },
+    {
+      "epoch": 0.15495668912415783,
+      "grad_norm": 2.1063289642333984,
+      "learning_rate": 0.00019706191801808457,
+      "loss": 1.3208,
+      "step": 4025
+    },
+    {
+      "epoch": 0.15514918190567853,
+      "grad_norm": 1.2449209690093994,
+      "learning_rate": 0.00019705463729983798,
+      "loss": 1.2863,
+      "step": 4030
+    },
+    {
+      "epoch": 0.15534167468719923,
+      "grad_norm": 1.4950852394104004,
+      "learning_rate": 0.00019704734770658778,
+      "loss": 1.2338,
+      "step": 4035
+    },
+    {
+      "epoch": 0.15553416746871992,
+      "grad_norm": 0.9372254014015198,
+      "learning_rate": 0.00019704004923900046,
+      "loss": 1.2105,
+      "step": 4040
+    },
+    {
+      "epoch": 0.15572666025024062,
+      "grad_norm": 1.2273038625717163,
+      "learning_rate": 0.00019703274189774347,
+      "loss": 1.3584,
+      "step": 4045
+    },
+    {
+      "epoch": 0.15591915303176132,
+      "grad_norm": 1.1560612916946411,
+      "learning_rate": 0.00019702542568348502,
+      "loss": 1.432,
+      "step": 4050
+    },
+    {
+      "epoch": 0.156111645813282,
+      "grad_norm": 1.2214939594268799,
+      "learning_rate": 0.00019701810059689415,
+      "loss": 1.3237,
+      "step": 4055
+    },
+    {
+      "epoch": 0.1563041385948027,
+      "grad_norm": 1.255182147026062,
+      "learning_rate": 0.00019701076663864066,
+      "loss": 1.5111,
+      "step": 4060
+    },
+    {
+      "epoch": 0.1564966313763234,
+      "grad_norm": 1.2496423721313477,
+      "learning_rate": 0.0001970034238093952,
+      "loss": 1.3917,
+      "step": 4065
+    },
+    {
+      "epoch": 0.15668912415784408,
+      "grad_norm": 2.773935556411743,
+      "learning_rate": 0.00019699607210982918,
+      "loss": 1.3072,
+      "step": 4070
+    },
+    {
+      "epoch": 0.15688161693936478,
+      "grad_norm": 2.5853006839752197,
+      "learning_rate": 0.00019698871154061497,
+      "loss": 1.2737,
+      "step": 4075
+    },
+    {
+      "epoch": 0.15707410972088548,
+      "grad_norm": 0.9573465585708618,
+      "learning_rate": 0.00019698134210242553,
+      "loss": 1.411,
+      "step": 4080
+    },
+    {
+      "epoch": 0.15726660250240615,
+      "grad_norm": 2.204242467880249,
+      "learning_rate": 0.00019697396379593482,
+      "loss": 1.2493,
+      "step": 4085
+    },
+    {
+      "epoch": 0.15745909528392685,
+      "grad_norm": 1.4688855409622192,
+      "learning_rate": 0.0001969665766218175,
+      "loss": 1.273,
+      "step": 4090
+    },
+    {
+      "epoch": 0.15765158806544755,
+      "grad_norm": 2.1439919471740723,
+      "learning_rate": 0.0001969591805807491,
+      "loss": 1.4691,
+      "step": 4095
+    },
+    {
+      "epoch": 0.15784408084696824,
+      "grad_norm": 1.4877434968948364,
+      "learning_rate": 0.00019695177567340594,
+      "loss": 1.4427,
+      "step": 4100
+    },
+    {
+      "epoch": 0.15803657362848894,
+      "grad_norm": 1.3709458112716675,
+      "learning_rate": 0.00019694436190046514,
+      "loss": 1.2713,
+      "step": 4105
+    },
+    {
+      "epoch": 0.1582290664100096,
+      "grad_norm": 2.1676931381225586,
+      "learning_rate": 0.00019693693926260464,
+      "loss": 1.1888,
+      "step": 4110
+    },
+    {
+      "epoch": 0.1584215591915303,
+      "grad_norm": 1.1726205348968506,
+      "learning_rate": 0.0001969295077605032,
+      "loss": 1.3544,
+      "step": 4115
+    },
+    {
+      "epoch": 0.158614051973051,
+      "grad_norm": 1.2441811561584473,
+      "learning_rate": 0.00019692206739484037,
+      "loss": 1.4796,
+      "step": 4120
+    },
+    {
+      "epoch": 0.1588065447545717,
+      "grad_norm": 1.4889960289001465,
+      "learning_rate": 0.00019691461816629652,
+      "loss": 1.418,
+      "step": 4125
+    },
+    {
+      "epoch": 0.1589990375360924,
+      "grad_norm": 1.3810794353485107,
+      "learning_rate": 0.00019690716007555282,
+      "loss": 1.6398,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1591915303176131,
+      "grad_norm": 1.589390754699707,
+      "learning_rate": 0.00019689969312329132,
+      "loss": 1.3203,
+      "step": 4135
+    },
+    {
+      "epoch": 0.15938402309913377,
+      "grad_norm": 0.8731974959373474,
+      "learning_rate": 0.00019689221731019477,
+      "loss": 1.2408,
+      "step": 4140
+    },
+    {
+      "epoch": 0.15957651588065447,
+      "grad_norm": 1.046852707862854,
+      "learning_rate": 0.00019688473263694678,
+      "loss": 1.1249,
+      "step": 4145
+    },
+    {
+      "epoch": 0.15976900866217517,
+      "grad_norm": 0.8767102360725403,
+      "learning_rate": 0.0001968772391042318,
+      "loss": 1.2611,
+      "step": 4150
+    },
+    {
+      "epoch": 0.15996150144369586,
+      "grad_norm": 1.1452685594558716,
+      "learning_rate": 0.0001968697367127351,
+      "loss": 1.2992,
+      "step": 4155
+    },
+    {
+      "epoch": 0.16015399422521656,
+      "grad_norm": 0.9254185557365417,
+      "learning_rate": 0.00019686222546314266,
+      "loss": 1.3894,
+      "step": 4160
+    },
+    {
+      "epoch": 0.16034648700673726,
+      "grad_norm": 0.9607768654823303,
+      "learning_rate": 0.00019685470535614133,
+      "loss": 1.3076,
+      "step": 4165
+    },
+    {
+      "epoch": 0.16053897978825793,
+      "grad_norm": 1.2880384922027588,
+      "learning_rate": 0.0001968471763924188,
+      "loss": 1.3868,
+      "step": 4170
+    },
+    {
+      "epoch": 0.16073147256977863,
+      "grad_norm": 1.1116464138031006,
+      "learning_rate": 0.00019683963857266356,
+      "loss": 1.2489,
+      "step": 4175
+    },
+    {
+      "epoch": 0.16092396535129933,
+      "grad_norm": 0.9132522940635681,
+      "learning_rate": 0.0001968320918975649,
+      "loss": 1.3788,
+      "step": 4180
+    },
+    {
+      "epoch": 0.16111645813282002,
+      "grad_norm": 1.1793001890182495,
+      "learning_rate": 0.00019682453636781283,
+      "loss": 1.4742,
+      "step": 4185
+    },
+    {
+      "epoch": 0.16130895091434072,
+      "grad_norm": 1.1624877452850342,
+      "learning_rate": 0.00019681697198409835,
+      "loss": 1.3547,
+      "step": 4190
+    },
+    {
+      "epoch": 0.16150144369586142,
+      "grad_norm": 1.1367181539535522,
+      "learning_rate": 0.00019680939874711312,
+      "loss": 1.3692,
+      "step": 4195
+    },
+    {
+      "epoch": 0.1616939364773821,
+      "grad_norm": 1.0168886184692383,
+      "learning_rate": 0.00019680181665754972,
+      "loss": 1.4148,
+      "step": 4200
+    },
+    {
+      "epoch": 0.1618864292589028,
+      "grad_norm": 1.3179705142974854,
+      "learning_rate": 0.0001967942257161014,
+      "loss": 1.2674,
+      "step": 4205
+    },
+    {
+      "epoch": 0.16207892204042348,
+      "grad_norm": 0.8679062724113464,
+      "learning_rate": 0.00019678662592346235,
+      "loss": 1.4001,
+      "step": 4210
+    },
+    {
+      "epoch": 0.16227141482194418,
+      "grad_norm": 0.8477693200111389,
+      "learning_rate": 0.00019677901728032754,
+      "loss": 1.3527,
+      "step": 4215
+    },
+    {
+      "epoch": 0.16246390760346488,
+      "grad_norm": 1.280357003211975,
+      "learning_rate": 0.00019677139978739266,
+      "loss": 1.2576,
+      "step": 4220
+    },
+    {
+      "epoch": 0.16265640038498555,
+      "grad_norm": 3.5572381019592285,
+      "learning_rate": 0.00019676377344535434,
+      "loss": 1.3059,
+      "step": 4225
+    },
+    {
+      "epoch": 0.16284889316650625,
+      "grad_norm": 0.9162838459014893,
+      "learning_rate": 0.0001967561382549099,
+      "loss": 1.3655,
+      "step": 4230
+    },
+    {
+      "epoch": 0.16304138594802695,
+      "grad_norm": 1.0635076761245728,
+      "learning_rate": 0.00019674849421675764,
+      "loss": 1.2356,
+      "step": 4235
+    },
+    {
+      "epoch": 0.16323387872954764,
+      "grad_norm": 2.3638720512390137,
+      "learning_rate": 0.00019674084133159642,
+      "loss": 1.3598,
+      "step": 4240
+    },
+    {
+      "epoch": 0.16342637151106834,
+      "grad_norm": 1.013108730316162,
+      "learning_rate": 0.00019673317960012615,
+      "loss": 1.6119,
+      "step": 4245
+    },
+    {
+      "epoch": 0.16361886429258904,
+      "grad_norm": 1.391450047492981,
+      "learning_rate": 0.00019672550902304737,
+      "loss": 1.2481,
+      "step": 4250
+    },
+    {
+      "epoch": 0.1638113570741097,
+      "grad_norm": 1.5574865341186523,
+      "learning_rate": 0.00019671782960106157,
+      "loss": 1.345,
+      "step": 4255
+    },
+    {
+      "epoch": 0.1640038498556304,
+      "grad_norm": 1.8456825017929077,
+      "learning_rate": 0.00019671014133487095,
+      "loss": 1.3582,
+      "step": 4260
+    },
+    {
+      "epoch": 0.1641963426371511,
+      "grad_norm": 1.4087297916412354,
+      "learning_rate": 0.00019670244422517855,
+      "loss": 1.3162,
+      "step": 4265
+    },
+    {
+      "epoch": 0.1643888354186718,
+      "grad_norm": 1.167403221130371,
+      "learning_rate": 0.0001966947382726882,
+      "loss": 1.3841,
+      "step": 4270
+    },
+    {
+      "epoch": 0.1645813282001925,
+      "grad_norm": 1.3395906686782837,
+      "learning_rate": 0.0001966870234781046,
+      "loss": 1.1306,
+      "step": 4275
+    },
+    {
+      "epoch": 0.1647738209817132,
+      "grad_norm": 0.8549813628196716,
+      "learning_rate": 0.00019667929984213317,
+      "loss": 1.3017,
+      "step": 4280
+    },
+    {
+      "epoch": 0.16496631376323387,
+      "grad_norm": 0.8681890368461609,
+      "learning_rate": 0.00019667156736548021,
+      "loss": 1.2152,
+      "step": 4285
+    },
+    {
+      "epoch": 0.16515880654475457,
+      "grad_norm": 1.8476097583770752,
+      "learning_rate": 0.00019666382604885283,
+      "loss": 1.2571,
+      "step": 4290
+    },
+    {
+      "epoch": 0.16535129932627526,
+      "grad_norm": 1.6583194732666016,
+      "learning_rate": 0.00019665607589295888,
+      "loss": 1.3866,
+      "step": 4295
+    },
+    {
+      "epoch": 0.16554379210779596,
+      "grad_norm": 1.6784121990203857,
+      "learning_rate": 0.00019664831689850712,
+      "loss": 1.2966,
+      "step": 4300
+    },
+    {
+      "epoch": 0.16573628488931666,
+      "grad_norm": 1.5268521308898926,
+      "learning_rate": 0.00019664054906620696,
+      "loss": 1.3086,
+      "step": 4305
+    },
+    {
+      "epoch": 0.16592877767083736,
+      "grad_norm": 2.0114951133728027,
+      "learning_rate": 0.00019663277239676877,
+      "loss": 1.2137,
+      "step": 4310
+    },
+    {
+      "epoch": 0.16612127045235803,
+      "grad_norm": 1.4572757482528687,
+      "learning_rate": 0.00019662498689090372,
+      "loss": 1.2505,
+      "step": 4315
+    },
+    {
+      "epoch": 0.16631376323387873,
+      "grad_norm": 1.4267566204071045,
+      "learning_rate": 0.00019661719254932369,
+      "loss": 1.1485,
+      "step": 4320
+    },
+    {
+      "epoch": 0.16650625601539942,
+      "grad_norm": 0.9921162128448486,
+      "learning_rate": 0.00019660938937274142,
+      "loss": 1.304,
+      "step": 4325
+    },
+    {
+      "epoch": 0.16669874879692012,
+      "grad_norm": 1.3901869058609009,
+      "learning_rate": 0.00019660157736187047,
+      "loss": 1.4347,
+      "step": 4330
+    },
+    {
+      "epoch": 0.16689124157844082,
+      "grad_norm": 1.5446443557739258,
+      "learning_rate": 0.0001965937565174252,
+      "loss": 1.3157,
+      "step": 4335
+    },
+    {
+      "epoch": 0.1670837343599615,
+      "grad_norm": 1.2553350925445557,
+      "learning_rate": 0.0001965859268401208,
+      "loss": 1.1882,
+      "step": 4340
+    },
+    {
+      "epoch": 0.1672762271414822,
+      "grad_norm": 1.9385195970535278,
+      "learning_rate": 0.0001965780883306732,
+      "loss": 1.4522,
+      "step": 4345
+    },
+    {
+      "epoch": 0.16746871992300288,
+      "grad_norm": 1.426032543182373,
+      "learning_rate": 0.00019657024098979916,
+      "loss": 1.1029,
+      "step": 4350
+    },
+    {
+      "epoch": 0.16766121270452358,
+      "grad_norm": 1.5562461614608765,
+      "learning_rate": 0.0001965623848182163,
+      "loss": 1.4837,
+      "step": 4355
+    },
+    {
+      "epoch": 0.16785370548604428,
+      "grad_norm": 1.0057613849639893,
+      "learning_rate": 0.00019655451981664306,
+      "loss": 1.3095,
+      "step": 4360
+    },
+    {
+      "epoch": 0.16804619826756498,
+      "grad_norm": 1.447845697402954,
+      "learning_rate": 0.00019654664598579857,
+      "loss": 1.4002,
+      "step": 4365
+    },
+    {
+      "epoch": 0.16823869104908565,
+      "grad_norm": 0.9452415108680725,
+      "learning_rate": 0.00019653876332640288,
+      "loss": 1.3324,
+      "step": 4370
+    },
+    {
+      "epoch": 0.16843118383060635,
+      "grad_norm": 1.7831186056137085,
+      "learning_rate": 0.00019653087183917677,
+      "loss": 1.3004,
+      "step": 4375
+    },
+    {
+      "epoch": 0.16862367661212704,
+      "grad_norm": 1.0656229257583618,
+      "learning_rate": 0.0001965229715248419,
+      "loss": 1.5165,
+      "step": 4380
+    },
+    {
+      "epoch": 0.16881616939364774,
+      "grad_norm": 1.0360915660858154,
+      "learning_rate": 0.0001965150623841207,
+      "loss": 1.2842,
+      "step": 4385
+    },
+    {
+      "epoch": 0.16900866217516844,
+      "grad_norm": 1.286447525024414,
+      "learning_rate": 0.00019650714441773643,
+      "loss": 1.2902,
+      "step": 4390
+    },
+    {
+      "epoch": 0.16920115495668914,
+      "grad_norm": 1.2435790300369263,
+      "learning_rate": 0.00019649921762641306,
+      "loss": 1.3049,
+      "step": 4395
+    },
+    {
+      "epoch": 0.1693936477382098,
+      "grad_norm": 1.9299678802490234,
+      "learning_rate": 0.0001964912820108755,
+      "loss": 1.3057,
+      "step": 4400
+    },
+    {
+      "epoch": 0.1695861405197305,
+      "grad_norm": 1.7493208646774292,
+      "learning_rate": 0.0001964833375718494,
+      "loss": 1.3225,
+      "step": 4405
+    },
+    {
+      "epoch": 0.1697786333012512,
+      "grad_norm": 1.3697878122329712,
+      "learning_rate": 0.0001964753843100612,
+      "loss": 1.3518,
+      "step": 4410
+    },
+    {
+      "epoch": 0.1699711260827719,
+      "grad_norm": 1.343985676765442,
+      "learning_rate": 0.0001964674222262382,
+      "loss": 1.3195,
+      "step": 4415
+    },
+    {
+      "epoch": 0.1701636188642926,
+      "grad_norm": 1.0094975233078003,
+      "learning_rate": 0.00019645945132110853,
+      "loss": 1.3184,
+      "step": 4420
+    },
+    {
+      "epoch": 0.17035611164581327,
+      "grad_norm": 1.6048771142959595,
+      "learning_rate": 0.00019645147159540096,
+      "loss": 1.3307,
+      "step": 4425
+    },
+    {
+      "epoch": 0.17054860442733397,
+      "grad_norm": 2.14099383354187,
+      "learning_rate": 0.00019644348304984524,
+      "loss": 1.3221,
+      "step": 4430
+    },
+    {
+      "epoch": 0.17074109720885466,
+      "grad_norm": 2.5571303367614746,
+      "learning_rate": 0.00019643548568517192,
+      "loss": 1.3092,
+      "step": 4435
+    },
+    {
+      "epoch": 0.17093358999037536,
+      "grad_norm": 1.1076972484588623,
+      "learning_rate": 0.00019642747950211225,
+      "loss": 1.1981,
+      "step": 4440
+    },
+    {
+      "epoch": 0.17112608277189606,
+      "grad_norm": 1.1315946578979492,
+      "learning_rate": 0.00019641946450139831,
+      "loss": 1.335,
+      "step": 4445
+    },
+    {
+      "epoch": 0.17131857555341676,
+      "grad_norm": 1.33171808719635,
+      "learning_rate": 0.00019641144068376312,
+      "loss": 1.4677,
+      "step": 4450
+    },
+    {
+      "epoch": 0.17151106833493743,
+      "grad_norm": 0.87531977891922,
+      "learning_rate": 0.0001964034080499403,
+      "loss": 1.1795,
+      "step": 4455
+    },
+    {
+      "epoch": 0.17170356111645813,
+      "grad_norm": 1.6923136711120605,
+      "learning_rate": 0.00019639536660066446,
+      "loss": 1.2491,
+      "step": 4460
+    },
+    {
+      "epoch": 0.17189605389797882,
+      "grad_norm": 1.481703519821167,
+      "learning_rate": 0.0001963873163366709,
+      "loss": 1.2894,
+      "step": 4465
+    },
+    {
+      "epoch": 0.17208854667949952,
+      "grad_norm": 3.3689515590667725,
+      "learning_rate": 0.00019637925725869576,
+      "loss": 1.3785,
+      "step": 4470
+    },
+    {
+      "epoch": 0.17228103946102022,
+      "grad_norm": 2.498059034347534,
+      "learning_rate": 0.000196371189367476,
+      "loss": 1.2854,
+      "step": 4475
+    },
+    {
+      "epoch": 0.17247353224254092,
+      "grad_norm": 1.2852959632873535,
+      "learning_rate": 0.00019636311266374939,
+      "loss": 1.2272,
+      "step": 4480
+    },
+    {
+      "epoch": 0.1726660250240616,
+      "grad_norm": 0.9257192015647888,
+      "learning_rate": 0.00019635502714825446,
+      "loss": 1.1707,
+      "step": 4485
+    },
+    {
+      "epoch": 0.17285851780558228,
+      "grad_norm": 0.989142656326294,
+      "learning_rate": 0.00019634693282173058,
+      "loss": 1.3174,
+      "step": 4490
+    },
+    {
+      "epoch": 0.17305101058710298,
+      "grad_norm": 1.4923882484436035,
+      "learning_rate": 0.00019633882968491794,
+      "loss": 1.2334,
+      "step": 4495
+    },
+    {
+      "epoch": 0.17324350336862368,
+      "grad_norm": 1.2684218883514404,
+      "learning_rate": 0.0001963307177385575,
+      "loss": 1.2468,
+      "step": 4500
+    },
+    {
+      "epoch": 0.17343599615014438,
+      "grad_norm": 0.9474775791168213,
+      "learning_rate": 0.0001963225969833911,
+      "loss": 1.2767,
+      "step": 4505
+    },
+    {
+      "epoch": 0.17362848893166508,
+      "grad_norm": 2.477541446685791,
+      "learning_rate": 0.00019631446742016126,
+      "loss": 1.4144,
+      "step": 4510
+    },
+    {
+      "epoch": 0.17382098171318575,
+      "grad_norm": 1.040477991104126,
+      "learning_rate": 0.00019630632904961138,
+      "loss": 1.5665,
+      "step": 4515
+    },
+    {
+      "epoch": 0.17401347449470644,
+      "grad_norm": 1.3127304315567017,
+      "learning_rate": 0.0001962981818724857,
+      "loss": 1.3511,
+      "step": 4520
+    },
+    {
+      "epoch": 0.17420596727622714,
+      "grad_norm": 1.6968106031417847,
+      "learning_rate": 0.0001962900258895292,
+      "loss": 1.3202,
+      "step": 4525
+    },
+    {
+      "epoch": 0.17439846005774784,
+      "grad_norm": 2.2431318759918213,
+      "learning_rate": 0.0001962818611014877,
+      "loss": 1.351,
+      "step": 4530
+    },
+    {
+      "epoch": 0.17459095283926854,
+      "grad_norm": 1.2938642501831055,
+      "learning_rate": 0.00019627368750910779,
+      "loss": 1.276,
+      "step": 4535
+    },
+    {
+      "epoch": 0.1747834456207892,
+      "grad_norm": 1.1331931352615356,
+      "learning_rate": 0.00019626550511313694,
+      "loss": 1.4734,
+      "step": 4540
+    },
+    {
+      "epoch": 0.1749759384023099,
+      "grad_norm": 1.4755507707595825,
+      "learning_rate": 0.00019625731391432333,
+      "loss": 1.24,
+      "step": 4545
+    },
+    {
+      "epoch": 0.1751684311838306,
+      "grad_norm": 1.5442554950714111,
+      "learning_rate": 0.00019624911391341604,
+      "loss": 1.0894,
+      "step": 4550
+    },
+    {
+      "epoch": 0.1753609239653513,
+      "grad_norm": 1.2970473766326904,
+      "learning_rate": 0.00019624090511116481,
+      "loss": 1.3262,
+      "step": 4555
+    },
+    {
+      "epoch": 0.175553416746872,
+      "grad_norm": 2.1946523189544678,
+      "learning_rate": 0.0001962326875083204,
+      "loss": 1.4652,
+      "step": 4560
+    },
+    {
+      "epoch": 0.1757459095283927,
+      "grad_norm": 1.1216411590576172,
+      "learning_rate": 0.00019622446110563417,
+      "loss": 1.1608,
+      "step": 4565
+    },
+    {
+      "epoch": 0.17593840230991337,
+      "grad_norm": 1.996535301208496,
+      "learning_rate": 0.00019621622590385842,
+      "loss": 1.2568,
+      "step": 4570
+    },
+    {
+      "epoch": 0.17613089509143406,
+      "grad_norm": 1.9742660522460938,
+      "learning_rate": 0.0001962079819037462,
+      "loss": 1.3335,
+      "step": 4575
+    },
+    {
+      "epoch": 0.17632338787295476,
+      "grad_norm": 1.985192060470581,
+      "learning_rate": 0.00019619972910605134,
+      "loss": 1.3529,
+      "step": 4580
+    },
+    {
+      "epoch": 0.17651588065447546,
+      "grad_norm": 0.8765020966529846,
+      "learning_rate": 0.00019619146751152848,
+      "loss": 1.3956,
+      "step": 4585
+    },
+    {
+      "epoch": 0.17670837343599616,
+      "grad_norm": 1.483407974243164,
+      "learning_rate": 0.00019618319712093319,
+      "loss": 1.4396,
+      "step": 4590
+    },
+    {
+      "epoch": 0.17690086621751686,
+      "grad_norm": 1.5663124322891235,
+      "learning_rate": 0.00019617491793502164,
+      "loss": 1.3896,
+      "step": 4595
+    },
+    {
+      "epoch": 0.17709335899903753,
+      "grad_norm": 1.3831099271774292,
+      "learning_rate": 0.00019616662995455096,
+      "loss": 1.2669,
+      "step": 4600
+    },
+    {
+      "epoch": 0.17728585178055822,
+      "grad_norm": 0.8688403964042664,
+      "learning_rate": 0.00019615833318027898,
+      "loss": 1.2098,
+      "step": 4605
+    },
+    {
+      "epoch": 0.17747834456207892,
+      "grad_norm": 1.9218660593032837,
+      "learning_rate": 0.00019615002761296446,
+      "loss": 1.1568,
+      "step": 4610
+    },
+    {
+      "epoch": 0.17767083734359962,
+      "grad_norm": 1.5095698833465576,
+      "learning_rate": 0.00019614171325336684,
+      "loss": 1.0516,
+      "step": 4615
+    },
+    {
+      "epoch": 0.17786333012512032,
+      "grad_norm": 0.9288404583930969,
+      "learning_rate": 0.00019613339010224646,
+      "loss": 1.075,
+      "step": 4620
+    },
+    {
+      "epoch": 0.17805582290664101,
+      "grad_norm": 1.414787769317627,
+      "learning_rate": 0.00019612505816036434,
+      "loss": 1.2158,
+      "step": 4625
+    },
+    {
+      "epoch": 0.17824831568816168,
+      "grad_norm": 1.3182802200317383,
+      "learning_rate": 0.0001961167174284824,
+      "loss": 1.3719,
+      "step": 4630
+    },
+    {
+      "epoch": 0.17844080846968238,
+      "grad_norm": 1.1671231985092163,
+      "learning_rate": 0.0001961083679073634,
+      "loss": 1.3067,
+      "step": 4635
+    },
+    {
+      "epoch": 0.17863330125120308,
+      "grad_norm": 1.11225163936615,
+      "learning_rate": 0.0001961000095977708,
+      "loss": 1.1593,
+      "step": 4640
+    },
+    {
+      "epoch": 0.17882579403272378,
+      "grad_norm": 1.235335111618042,
+      "learning_rate": 0.00019609164250046894,
+      "loss": 1.2232,
+      "step": 4645
+    },
+    {
+      "epoch": 0.17901828681424448,
+      "grad_norm": 1.0023348331451416,
+      "learning_rate": 0.00019608326661622291,
+      "loss": 1.2926,
+      "step": 4650
+    },
+    {
+      "epoch": 0.17921077959576515,
+      "grad_norm": 1.7143383026123047,
+      "learning_rate": 0.00019607488194579867,
+      "loss": 1.3149,
+      "step": 4655
+    },
+    {
+      "epoch": 0.17940327237728584,
+      "grad_norm": 1.135324478149414,
+      "learning_rate": 0.00019606648848996287,
+      "loss": 1.4155,
+      "step": 4660
+    },
+    {
+      "epoch": 0.17959576515880654,
+      "grad_norm": 0.7830592393875122,
+      "learning_rate": 0.0001960580862494831,
+      "loss": 1.2632,
+      "step": 4665
+    },
+    {
+      "epoch": 0.17978825794032724,
+      "grad_norm": 1.546481966972351,
+      "learning_rate": 0.0001960496752251277,
+      "loss": 1.4674,
+      "step": 4670
+    },
+    {
+      "epoch": 0.17998075072184794,
+      "grad_norm": 1.5377360582351685,
+      "learning_rate": 0.00019604125541766574,
+      "loss": 1.0782,
+      "step": 4675
+    },
+    {
+      "epoch": 0.18017324350336864,
+      "grad_norm": 2.1382510662078857,
+      "learning_rate": 0.0001960328268278672,
+      "loss": 1.3008,
+      "step": 4680
+    },
+    {
+      "epoch": 0.1803657362848893,
+      "grad_norm": 1.4963937997817993,
+      "learning_rate": 0.00019602438945650277,
+      "loss": 1.2601,
+      "step": 4685
+    },
+    {
+      "epoch": 0.18055822906641,
+      "grad_norm": 1.4736862182617188,
+      "learning_rate": 0.00019601594330434405,
+      "loss": 1.163,
+      "step": 4690
+    },
+    {
+      "epoch": 0.1807507218479307,
+      "grad_norm": 0.9905889630317688,
+      "learning_rate": 0.00019600748837216337,
+      "loss": 1.3675,
+      "step": 4695
+    },
+    {
+      "epoch": 0.1809432146294514,
+      "grad_norm": 1.1800122261047363,
+      "learning_rate": 0.00019599902466073385,
+      "loss": 1.3252,
+      "step": 4700
+    },
+    {
+      "epoch": 0.1811357074109721,
+      "grad_norm": 1.1933966875076294,
+      "learning_rate": 0.00019599055217082949,
+      "loss": 1.2163,
+      "step": 4705
+    },
+    {
+      "epoch": 0.1813282001924928,
+      "grad_norm": 1.3980772495269775,
+      "learning_rate": 0.000195982070903225,
+      "loss": 1.2807,
+      "step": 4710
+    },
+    {
+      "epoch": 0.18152069297401346,
+      "grad_norm": 2.541808605194092,
+      "learning_rate": 0.00019597358085869594,
+      "loss": 1.1333,
+      "step": 4715
+    },
+    {
+      "epoch": 0.18171318575553416,
+      "grad_norm": 1.616479516029358,
+      "learning_rate": 0.0001959650820380187,
+      "loss": 1.2991,
+      "step": 4720
+    },
+    {
+      "epoch": 0.18190567853705486,
+      "grad_norm": 0.9473749399185181,
+      "learning_rate": 0.00019595657444197037,
+      "loss": 1.2273,
+      "step": 4725
+    },
+    {
+      "epoch": 0.18209817131857556,
+      "grad_norm": 1.3119609355926514,
+      "learning_rate": 0.000195948058071329,
+      "loss": 1.2754,
+      "step": 4730
+    },
+    {
+      "epoch": 0.18229066410009626,
+      "grad_norm": 1.0062682628631592,
+      "learning_rate": 0.00019593953292687332,
+      "loss": 1.2494,
+      "step": 4735
+    },
+    {
+      "epoch": 0.18248315688161693,
+      "grad_norm": 1.2124086618423462,
+      "learning_rate": 0.0001959309990093829,
+      "loss": 1.3725,
+      "step": 4740
+    },
+    {
+      "epoch": 0.18267564966313762,
+      "grad_norm": 1.2050824165344238,
+      "learning_rate": 0.0001959224563196381,
+      "loss": 1.5103,
+      "step": 4745
+    },
+    {
+      "epoch": 0.18286814244465832,
+      "grad_norm": 0.9262427091598511,
+      "learning_rate": 0.00019591390485842008,
+      "loss": 1.4155,
+      "step": 4750
+    },
+    {
+      "epoch": 0.18306063522617902,
+      "grad_norm": 1.5612881183624268,
+      "learning_rate": 0.00019590534462651086,
+      "loss": 1.2289,
+      "step": 4755
+    },
+    {
+      "epoch": 0.18325312800769972,
+      "grad_norm": 1.5384646654129028,
+      "learning_rate": 0.00019589677562469312,
+      "loss": 1.2474,
+      "step": 4760
+    },
+    {
+      "epoch": 0.18344562078922041,
+      "grad_norm": 1.397716999053955,
+      "learning_rate": 0.00019588819785375057,
+      "loss": 1.4273,
+      "step": 4765
+    },
+    {
+      "epoch": 0.18363811357074108,
+      "grad_norm": 1.169207215309143,
+      "learning_rate": 0.00019587961131446754,
+      "loss": 1.3963,
+      "step": 4770
+    },
+    {
+      "epoch": 0.18383060635226178,
+      "grad_norm": 1.5064833164215088,
+      "learning_rate": 0.00019587101600762916,
+      "loss": 1.5192,
+      "step": 4775
+    },
+    {
+      "epoch": 0.18402309913378248,
+      "grad_norm": 0.9700071811676025,
+      "learning_rate": 0.00019586241193402147,
+      "loss": 1.2697,
+      "step": 4780
+    },
+    {
+      "epoch": 0.18421559191530318,
+      "grad_norm": 1.2304507493972778,
+      "learning_rate": 0.00019585379909443123,
+      "loss": 1.3025,
+      "step": 4785
+    },
+    {
+      "epoch": 0.18440808469682388,
+      "grad_norm": 1.3768020868301392,
+      "learning_rate": 0.00019584517748964605,
+      "loss": 1.3785,
+      "step": 4790
+    },
+    {
+      "epoch": 0.18460057747834457,
+      "grad_norm": 1.062251091003418,
+      "learning_rate": 0.0001958365471204543,
+      "loss": 1.5416,
+      "step": 4795
+    },
+    {
+      "epoch": 0.18479307025986524,
+      "grad_norm": 0.9126803874969482,
+      "learning_rate": 0.00019582790798764518,
+      "loss": 1.1479,
+      "step": 4800
+    },
+    {
+      "epoch": 0.18498556304138594,
+      "grad_norm": 1.579830288887024,
+      "learning_rate": 0.00019581926009200866,
+      "loss": 1.3315,
+      "step": 4805
+    },
+    {
+      "epoch": 0.18517805582290664,
+      "grad_norm": 2.351717710494995,
+      "learning_rate": 0.00019581060343433555,
+      "loss": 1.2503,
+      "step": 4810
+    },
+    {
+      "epoch": 0.18537054860442734,
+      "grad_norm": 1.1480222940444946,
+      "learning_rate": 0.00019580193801541746,
+      "loss": 1.2048,
+      "step": 4815
+    },
+    {
+      "epoch": 0.18556304138594804,
+      "grad_norm": 1.606439471244812,
+      "learning_rate": 0.00019579326383604675,
+      "loss": 1.5204,
+      "step": 4820
+    },
+    {
+      "epoch": 0.18575553416746873,
+      "grad_norm": 1.520969271659851,
+      "learning_rate": 0.00019578458089701664,
+      "loss": 1.2584,
+      "step": 4825
+    },
+    {
+      "epoch": 0.1859480269489894,
+      "grad_norm": 1.9096931219100952,
+      "learning_rate": 0.00019577588919912113,
+      "loss": 1.5508,
+      "step": 4830
+    },
+    {
+      "epoch": 0.1861405197305101,
+      "grad_norm": 1.004654884338379,
+      "learning_rate": 0.00019576718874315501,
+      "loss": 1.2249,
+      "step": 4835
+    },
+    {
+      "epoch": 0.1863330125120308,
+      "grad_norm": 1.0160667896270752,
+      "learning_rate": 0.00019575847952991388,
+      "loss": 1.0782,
+      "step": 4840
+    },
+    {
+      "epoch": 0.1865255052935515,
+      "grad_norm": 1.4719328880310059,
+      "learning_rate": 0.0001957497615601941,
+      "loss": 1.4679,
+      "step": 4845
+    },
+    {
+      "epoch": 0.1867179980750722,
+      "grad_norm": 1.229625940322876,
+      "learning_rate": 0.00019574103483479296,
+      "loss": 1.347,
+      "step": 4850
+    },
+    {
+      "epoch": 0.18691049085659286,
+      "grad_norm": 3.0996217727661133,
+      "learning_rate": 0.00019573229935450842,
+      "loss": 1.3325,
+      "step": 4855
+    },
+    {
+      "epoch": 0.18710298363811356,
+      "grad_norm": 1.59645676612854,
+      "learning_rate": 0.00019572355512013922,
+      "loss": 1.2983,
+      "step": 4860
+    },
+    {
+      "epoch": 0.18729547641963426,
+      "grad_norm": 1.373542070388794,
+      "learning_rate": 0.00019571480213248504,
+      "loss": 1.3285,
+      "step": 4865
+    },
+    {
+      "epoch": 0.18748796920115496,
+      "grad_norm": 0.9625198245048523,
+      "learning_rate": 0.00019570604039234626,
+      "loss": 1.2823,
+      "step": 4870
+    },
+    {
+      "epoch": 0.18768046198267566,
+      "grad_norm": 1.1096363067626953,
+      "learning_rate": 0.00019569726990052407,
+      "loss": 1.2508,
+      "step": 4875
+    },
+    {
+      "epoch": 0.18787295476419635,
+      "grad_norm": 1.2040042877197266,
+      "learning_rate": 0.0001956884906578205,
+      "loss": 1.3767,
+      "step": 4880
+    },
+    {
+      "epoch": 0.18806544754571702,
+      "grad_norm": 1.103530764579773,
+      "learning_rate": 0.00019567970266503833,
+      "loss": 1.4559,
+      "step": 4885
+    },
+    {
+      "epoch": 0.18825794032723772,
+      "grad_norm": 1.1266409158706665,
+      "learning_rate": 0.0001956709059229812,
+      "loss": 1.0687,
+      "step": 4890
+    },
+    {
+      "epoch": 0.18845043310875842,
+      "grad_norm": 1.2266972064971924,
+      "learning_rate": 0.00019566210043245344,
+      "loss": 1.1801,
+      "step": 4895
+    },
+    {
+      "epoch": 0.18864292589027912,
+      "grad_norm": 1.416676640510559,
+      "learning_rate": 0.0001956532861942603,
+      "loss": 1.346,
+      "step": 4900
+    },
+    {
+      "epoch": 0.18883541867179982,
+      "grad_norm": 1.5538910627365112,
+      "learning_rate": 0.0001956444632092078,
+      "loss": 1.3498,
+      "step": 4905
+    },
+    {
+      "epoch": 0.1890279114533205,
+      "grad_norm": 1.1525146961212158,
+      "learning_rate": 0.00019563563147810274,
+      "loss": 1.39,
+      "step": 4910
+    },
+    {
+      "epoch": 0.18922040423484118,
+      "grad_norm": 1.6796061992645264,
+      "learning_rate": 0.00019562679100175266,
+      "loss": 1.3377,
+      "step": 4915
+    },
+    {
+      "epoch": 0.18941289701636188,
+      "grad_norm": 1.6094450950622559,
+      "learning_rate": 0.00019561794178096607,
+      "loss": 1.3057,
+      "step": 4920
+    },
+    {
+      "epoch": 0.18960538979788258,
+      "grad_norm": 1.8123548030853271,
+      "learning_rate": 0.00019560908381655208,
+      "loss": 1.1257,
+      "step": 4925
+    },
+    {
+      "epoch": 0.18979788257940328,
+      "grad_norm": 1.5495673418045044,
+      "learning_rate": 0.00019560021710932074,
+      "loss": 1.303,
+      "step": 4930
+    },
+    {
+      "epoch": 0.18999037536092397,
+      "grad_norm": 1.623429298400879,
+      "learning_rate": 0.00019559134166008283,
+      "loss": 1.1491,
+      "step": 4935
+    },
+    {
+      "epoch": 0.19018286814244467,
+      "grad_norm": 1.2682925462722778,
+      "learning_rate": 0.00019558245746964997,
+      "loss": 1.3774,
+      "step": 4940
+    },
+    {
+      "epoch": 0.19037536092396534,
+      "grad_norm": 0.9362719058990479,
+      "learning_rate": 0.00019557356453883456,
+      "loss": 1.2936,
+      "step": 4945
+    },
+    {
+      "epoch": 0.19056785370548604,
+      "grad_norm": 1.4271594285964966,
+      "learning_rate": 0.00019556466286844976,
+      "loss": 1.3865,
+      "step": 4950
+    },
+    {
+      "epoch": 0.19076034648700674,
+      "grad_norm": 1.4094691276550293,
+      "learning_rate": 0.00019555575245930963,
+      "loss": 1.2941,
+      "step": 4955
+    },
+    {
+      "epoch": 0.19095283926852744,
+      "grad_norm": 0.9695935249328613,
+      "learning_rate": 0.00019554683331222893,
+      "loss": 1.1724,
+      "step": 4960
+    },
+    {
+      "epoch": 0.19114533205004813,
+      "grad_norm": 1.110616683959961,
+      "learning_rate": 0.00019553790542802327,
+      "loss": 1.3999,
+      "step": 4965
+    },
+    {
+      "epoch": 0.1913378248315688,
+      "grad_norm": 1.5389796495437622,
+      "learning_rate": 0.000195528968807509,
+      "loss": 1.2693,
+      "step": 4970
+    },
+    {
+      "epoch": 0.1915303176130895,
+      "grad_norm": 1.921168565750122,
+      "learning_rate": 0.00019552002345150338,
+      "loss": 1.2392,
+      "step": 4975
+    },
+    {
+      "epoch": 0.1917228103946102,
+      "grad_norm": 1.3342314958572388,
+      "learning_rate": 0.00019551106936082437,
+      "loss": 1.2477,
+      "step": 4980
+    },
+    {
+      "epoch": 0.1919153031761309,
+      "grad_norm": 1.745754361152649,
+      "learning_rate": 0.0001955021065362908,
+      "loss": 1.7169,
+      "step": 4985
+    },
+    {
+      "epoch": 0.1921077959576516,
+      "grad_norm": 1.090145468711853,
+      "learning_rate": 0.0001954931349787222,
+      "loss": 1.1156,
+      "step": 4990
+    },
+    {
+      "epoch": 0.1923002887391723,
+      "grad_norm": 1.5357612371444702,
+      "learning_rate": 0.00019548415468893899,
+      "loss": 1.5436,
+      "step": 4995
+    },
+    {
+      "epoch": 0.19249278152069296,
+      "grad_norm": 1.0309633016586304,
+      "learning_rate": 0.00019547516566776238,
+      "loss": 1.3212,
+      "step": 5000
+    },
+    {
+      "epoch": 0.19268527430221366,
+      "grad_norm": 1.000688076019287,
+      "learning_rate": 0.0001954661679160143,
+      "loss": 1.2821,
+      "step": 5005
+    },
+    {
+      "epoch": 0.19287776708373436,
+      "grad_norm": 1.268754243850708,
+      "learning_rate": 0.0001954571614345176,
+      "loss": 1.2168,
+      "step": 5010
+    },
+    {
+      "epoch": 0.19307025986525506,
+      "grad_norm": 1.3859111070632935,
+      "learning_rate": 0.00019544814622409582,
+      "loss": 1.0701,
+      "step": 5015
+    },
+    {
+      "epoch": 0.19326275264677575,
+      "grad_norm": 2.248309850692749,
+      "learning_rate": 0.00019543912228557337,
+      "loss": 1.3548,
+      "step": 5020
+    },
+    {
+      "epoch": 0.19345524542829645,
+      "grad_norm": 1.0269944667816162,
+      "learning_rate": 0.00019543008961977538,
+      "loss": 1.213,
+      "step": 5025
+    },
+    {
+      "epoch": 0.19364773820981712,
+      "grad_norm": 1.0082924365997314,
+      "learning_rate": 0.00019542104822752789,
+      "loss": 1.2395,
+      "step": 5030
+    },
+    {
+      "epoch": 0.19384023099133782,
+      "grad_norm": 2.1287014484405518,
+      "learning_rate": 0.00019541199810965766,
+      "loss": 1.3794,
+      "step": 5035
+    },
+    {
+      "epoch": 0.19403272377285852,
+      "grad_norm": 1.230859637260437,
+      "learning_rate": 0.0001954029392669922,
+      "loss": 1.3985,
+      "step": 5040
+    },
+    {
+      "epoch": 0.19422521655437922,
+      "grad_norm": 1.0987460613250732,
+      "learning_rate": 0.00019539387170035996,
+      "loss": 1.2637,
+      "step": 5045
+    },
+    {
+      "epoch": 0.1944177093358999,
+      "grad_norm": 1.2570157051086426,
+      "learning_rate": 0.00019538479541059007,
+      "loss": 1.2752,
+      "step": 5050
+    },
+    {
+      "epoch": 0.19461020211742058,
+      "grad_norm": 0.5122241377830505,
+      "learning_rate": 0.00019537571039851252,
+      "loss": 1.1927,
+      "step": 5055
+    },
+    {
+      "epoch": 0.19480269489894128,
+      "grad_norm": 1.7925124168395996,
+      "learning_rate": 0.00019536661666495807,
+      "loss": 1.1414,
+      "step": 5060
+    },
+    {
+      "epoch": 0.19499518768046198,
+      "grad_norm": 0.8517950773239136,
+      "learning_rate": 0.00019535751421075826,
+      "loss": 1.2359,
+      "step": 5065
+    },
+    {
+      "epoch": 0.19518768046198268,
+      "grad_norm": 0.582260012626648,
+      "learning_rate": 0.00019534840303674544,
+      "loss": 1.3528,
+      "step": 5070
+    },
+    {
+      "epoch": 0.19538017324350337,
+      "grad_norm": 1.3547414541244507,
+      "learning_rate": 0.0001953392831437528,
+      "loss": 1.296,
+      "step": 5075
+    },
+    {
+      "epoch": 0.19557266602502407,
+      "grad_norm": Infinity,
+      "learning_rate": 0.0001953319809522536,
+      "loss": 1.4074,
+      "step": 5080
+    },
+    {
+      "epoch": 0.19576515880654474,
+      "grad_norm": 2.2984917163848877,
+      "learning_rate": 0.00019532284536719936,
+      "loss": 1.2002,
+      "step": 5085
+    },
+    {
+      "epoch": 0.19595765158806544,
+      "grad_norm": 1.4113095998764038,
+      "learning_rate": 0.0001953137010655024,
+      "loss": 1.2755,
+      "step": 5090
+    },
+    {
+      "epoch": 0.19615014436958614,
+      "grad_norm": 1.921242594718933,
+      "learning_rate": 0.00019530454804799881,
+      "loss": 1.2431,
+      "step": 5095
+    },
+    {
+      "epoch": 0.19634263715110684,
+      "grad_norm": 1.3097113370895386,
+      "learning_rate": 0.0001952953863155257,
+      "loss": 1.415,
+      "step": 5100
+    },
+    {
+      "epoch": 0.19653512993262753,
+      "grad_norm": 2.1493217945098877,
+      "learning_rate": 0.00019528621586892072,
+      "loss": 1.4282,
+      "step": 5105
+    },
+    {
+      "epoch": 0.19672762271414823,
+      "grad_norm": 1.2487257719039917,
+      "learning_rate": 0.0001952770367090226,
+      "loss": 1.3512,
+      "step": 5110
+    },
+    {
+      "epoch": 0.1969201154956689,
+      "grad_norm": 0.9984391331672668,
+      "learning_rate": 0.00019526784883667055,
+      "loss": 1.5437,
+      "step": 5115
+    },
+    {
+      "epoch": 0.1971126082771896,
+      "grad_norm": 1.241417646408081,
+      "learning_rate": 0.00019525865225270486,
+      "loss": 1.2399,
+      "step": 5120
+    },
+    {
+      "epoch": 0.1973051010587103,
+      "grad_norm": 1.5192227363586426,
+      "learning_rate": 0.00019524944695796642,
+      "loss": 1.3236,
+      "step": 5125
+    },
+    {
+      "epoch": 0.197497593840231,
+      "grad_norm": 1.7465555667877197,
+      "learning_rate": 0.00019524023295329704,
+      "loss": 1.4247,
+      "step": 5130
+    },
+    {
+      "epoch": 0.1976900866217517,
+      "grad_norm": 1.455175757408142,
+      "learning_rate": 0.00019523101023953925,
+      "loss": 1.5053,
+      "step": 5135
+    },
+    {
+      "epoch": 0.1978825794032724,
+      "grad_norm": 2.164982318878174,
+      "learning_rate": 0.00019522177881753643,
+      "loss": 1.2796,
+      "step": 5140
+    },
+    {
+      "epoch": 0.19807507218479306,
+      "grad_norm": 1.58863365650177,
+      "learning_rate": 0.00019521253868813273,
+      "loss": 1.349,
+      "step": 5145
+    },
+    {
+      "epoch": 0.19826756496631376,
+      "grad_norm": 1.5380641222000122,
+      "learning_rate": 0.0001952032898521731,
+      "loss": 1.3107,
+      "step": 5150
+    },
+    {
+      "epoch": 0.19846005774783446,
+      "grad_norm": 1.1790603399276733,
+      "learning_rate": 0.00019519403231050327,
+      "loss": 1.2178,
+      "step": 5155
+    },
+    {
+      "epoch": 0.19865255052935515,
+      "grad_norm": 1.7905482053756714,
+      "learning_rate": 0.0001951847660639698,
+      "loss": 1.3579,
+      "step": 5160
+    },
+    {
+      "epoch": 0.19884504331087585,
+      "grad_norm": 1.1262041330337524,
+      "learning_rate": 0.00019517549111342,
+      "loss": 1.2988,
+      "step": 5165
+    },
+    {
+      "epoch": 0.19903753609239652,
+      "grad_norm": 1.6370010375976562,
+      "learning_rate": 0.00019516620745970199,
+      "loss": 1.2326,
+      "step": 5170
+    },
+    {
+      "epoch": 0.19923002887391722,
+      "grad_norm": 1.1789335012435913,
+      "learning_rate": 0.00019515691510366476,
+      "loss": 1.1357,
+      "step": 5175
+    },
+    {
+      "epoch": 0.19942252165543792,
+      "grad_norm": 1.167226791381836,
+      "learning_rate": 0.000195147614046158,
+      "loss": 1.4007,
+      "step": 5180
+    },
+    {
+      "epoch": 0.19961501443695862,
+      "grad_norm": 1.3708933591842651,
+      "learning_rate": 0.00019513830428803225,
+      "loss": 1.3029,
+      "step": 5185
+    },
+    {
+      "epoch": 0.1998075072184793,
+      "grad_norm": 1.6595165729522705,
+      "learning_rate": 0.00019512898583013875,
+      "loss": 1.3159,
+      "step": 5190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1252923011779785,
+      "learning_rate": 0.00019511965867332972,
+      "loss": 1.1894,
+      "step": 5195
+    },
+    {
+      "epoch": 0.20019249278152068,
+      "grad_norm": 0.8440331816673279,
+      "learning_rate": 0.00019511032281845797,
+      "loss": 1.2108,
+      "step": 5200
+    },
+    {
+      "epoch": 0.20038498556304138,
+      "grad_norm": 1.427147626876831,
+      "learning_rate": 0.0001951009782663773,
+      "loss": 1.197,
+      "step": 5205
+    },
+    {
+      "epoch": 0.20057747834456208,
+      "grad_norm": 1.3509503602981567,
+      "learning_rate": 0.00019509162501794213,
+      "loss": 1.3348,
+      "step": 5210
+    },
+    {
+      "epoch": 0.20076997112608277,
+      "grad_norm": 1.533103108406067,
+      "learning_rate": 0.00019508226307400777,
+      "loss": 1.1919,
+      "step": 5215
+    },
+    {
+      "epoch": 0.20096246390760347,
+      "grad_norm": 1.1347332000732422,
+      "learning_rate": 0.0001950728924354303,
+      "loss": 1.2954,
+      "step": 5220
+    },
+    {
+      "epoch": 0.20115495668912417,
+      "grad_norm": 1.65277099609375,
+      "learning_rate": 0.00019506351310306664,
+      "loss": 1.2686,
+      "step": 5225
+    },
+    {
+      "epoch": 0.20134744947064484,
+      "grad_norm": 1.0601050853729248,
+      "learning_rate": 0.00019505412507777442,
+      "loss": 1.4066,
+      "step": 5230
+    },
+    {
+      "epoch": 0.20153994225216554,
+      "grad_norm": 0.9429787397384644,
+      "learning_rate": 0.00019504472836041217,
+      "loss": 1.208,
+      "step": 5235
+    },
+    {
+      "epoch": 0.20173243503368624,
+      "grad_norm": 0.9101033806800842,
+      "learning_rate": 0.00019503532295183908,
+      "loss": 1.3172,
+      "step": 5240
+    },
+    {
+      "epoch": 0.20192492781520693,
+      "grad_norm": 1.1404805183410645,
+      "learning_rate": 0.0001950259088529153,
+      "loss": 1.1539,
+      "step": 5245
+    },
+    {
+      "epoch": 0.20211742059672763,
+      "grad_norm": 1.1555522680282593,
+      "learning_rate": 0.00019501648606450161,
+      "loss": 1.3754,
+      "step": 5250
+    },
+    {
+      "epoch": 0.20230991337824833,
+      "grad_norm": 1.5473912954330444,
+      "learning_rate": 0.00019500705458745974,
+      "loss": 1.1878,
+      "step": 5255
+    },
+    {
+      "epoch": 0.202502406159769,
+      "grad_norm": 1.8766716718673706,
+      "learning_rate": 0.00019499761442265208,
+      "loss": 1.2445,
+      "step": 5260
+    },
+    {
+      "epoch": 0.2026948989412897,
+      "grad_norm": 1.7951183319091797,
+      "learning_rate": 0.00019498816557094188,
+      "loss": 1.3496,
+      "step": 5265
+    },
+    {
+      "epoch": 0.2028873917228104,
+      "grad_norm": 1.6615973711013794,
+      "learning_rate": 0.00019497870803319317,
+      "loss": 1.2919,
+      "step": 5270
+    },
+    {
+      "epoch": 0.2030798845043311,
+      "grad_norm": 1.2885236740112305,
+      "learning_rate": 0.00019496924181027078,
+      "loss": 1.1807,
+      "step": 5275
+    },
+    {
+      "epoch": 0.2032723772858518,
+      "grad_norm": 0.9546861052513123,
+      "learning_rate": 0.00019495976690304034,
+      "loss": 1.309,
+      "step": 5280
+    },
+    {
+      "epoch": 0.20346487006737246,
+      "grad_norm": 1.6904189586639404,
+      "learning_rate": 0.0001949502833123683,
+      "loss": 1.2244,
+      "step": 5285
+    },
+    {
+      "epoch": 0.20365736284889316,
+      "grad_norm": 1.394254446029663,
+      "learning_rate": 0.0001949407910391218,
+      "loss": 1.2877,
+      "step": 5290
+    },
+    {
+      "epoch": 0.20384985563041386,
+      "grad_norm": 0.8937919735908508,
+      "learning_rate": 0.0001949312900841689,
+      "loss": 1.2389,
+      "step": 5295
+    },
+    {
+      "epoch": 0.20404234841193455,
+      "grad_norm": 1.1096867322921753,
+      "learning_rate": 0.00019492178044837837,
+      "loss": 1.3766,
+      "step": 5300
+    },
+    {
+      "epoch": 0.20423484119345525,
+      "grad_norm": 1.009758472442627,
+      "learning_rate": 0.00019491226213261983,
+      "loss": 1.2281,
+      "step": 5305
+    },
+    {
+      "epoch": 0.20442733397497595,
+      "grad_norm": 1.4888296127319336,
+      "learning_rate": 0.00019490273513776365,
+      "loss": 1.0624,
+      "step": 5310
+    },
+    {
+      "epoch": 0.20461982675649662,
+      "grad_norm": 1.4901612997055054,
+      "learning_rate": 0.00019489319946468104,
+      "loss": 1.1554,
+      "step": 5315
+    },
+    {
+      "epoch": 0.20481231953801732,
+      "grad_norm": 1.2920863628387451,
+      "learning_rate": 0.0001948836551142439,
+      "loss": 1.2103,
+      "step": 5320
+    },
+    {
+      "epoch": 0.20500481231953802,
+      "grad_norm": 1.3616580963134766,
+      "learning_rate": 0.00019487410208732508,
+      "loss": 1.3246,
+      "step": 5325
+    },
+    {
+      "epoch": 0.2051973051010587,
+      "grad_norm": 1.0202921628952026,
+      "learning_rate": 0.0001948645403847981,
+      "loss": 1.3046,
+      "step": 5330
+    },
+    {
+      "epoch": 0.2053897978825794,
+      "grad_norm": 1.0083186626434326,
+      "learning_rate": 0.00019485497000753735,
+      "loss": 1.2541,
+      "step": 5335
+    },
+    {
+      "epoch": 0.2055822906641001,
+      "grad_norm": 1.137617588043213,
+      "learning_rate": 0.0001948453909564179,
+      "loss": 1.3143,
+      "step": 5340
+    },
+    {
+      "epoch": 0.20577478344562078,
+      "grad_norm": 1.6331067085266113,
+      "learning_rate": 0.00019483580323231578,
+      "loss": 1.1129,
+      "step": 5345
+    },
+    {
+      "epoch": 0.20596727622714148,
+      "grad_norm": 1.4032361507415771,
+      "learning_rate": 0.00019482620683610767,
+      "loss": 1.3412,
+      "step": 5350
+    },
+    {
+      "epoch": 0.20615976900866217,
+      "grad_norm": 1.3207452297210693,
+      "learning_rate": 0.00019481660176867108,
+      "loss": 1.4614,
+      "step": 5355
+    },
+    {
+      "epoch": 0.20635226179018287,
+      "grad_norm": 0.9236577749252319,
+      "learning_rate": 0.0001948069880308844,
+      "loss": 1.3131,
+      "step": 5360
+    },
+    {
+      "epoch": 0.20654475457170357,
+      "grad_norm": 2.2021703720092773,
+      "learning_rate": 0.0001947973656236267,
+      "loss": 1.2434,
+      "step": 5365
+    },
+    {
+      "epoch": 0.20673724735322424,
+      "grad_norm": 1.5074305534362793,
+      "learning_rate": 0.00019478773454777789,
+      "loss": 1.4204,
+      "step": 5370
+    },
+    {
+      "epoch": 0.20692974013474494,
+      "grad_norm": 1.5073877573013306,
+      "learning_rate": 0.00019477809480421865,
+      "loss": 1.4193,
+      "step": 5375
+    },
+    {
+      "epoch": 0.20712223291626564,
+      "grad_norm": 1.0522600412368774,
+      "learning_rate": 0.00019476844639383049,
+      "loss": 1.228,
+      "step": 5380
+    },
+    {
+      "epoch": 0.20731472569778633,
+      "grad_norm": 1.1478843688964844,
+      "learning_rate": 0.0001947587893174957,
+      "loss": 1.2315,
+      "step": 5385
+    },
+    {
+      "epoch": 0.20750721847930703,
+      "grad_norm": 0.922837495803833,
+      "learning_rate": 0.00019474912357609733,
+      "loss": 1.2567,
+      "step": 5390
+    },
+    {
+      "epoch": 0.20769971126082773,
+      "grad_norm": 1.156615972518921,
+      "learning_rate": 0.0001947394491705193,
+      "loss": 1.443,
+      "step": 5395
+    },
+    {
+      "epoch": 0.2078922040423484,
+      "grad_norm": 1.909555435180664,
+      "learning_rate": 0.0001947297661016462,
+      "loss": 1.1625,
+      "step": 5400
+    },
+    {
+      "epoch": 0.2080846968238691,
+      "grad_norm": 1.8379411697387695,
+      "learning_rate": 0.00019472007437036352,
+      "loss": 1.3015,
+      "step": 5405
+    },
+    {
+      "epoch": 0.2082771896053898,
+      "grad_norm": 1.188402771949768,
+      "learning_rate": 0.00019471037397755754,
+      "loss": 1.3294,
+      "step": 5410
+    },
+    {
+      "epoch": 0.2084696823869105,
+      "grad_norm": 1.597538948059082,
+      "learning_rate": 0.00019470066492411521,
+      "loss": 1.3824,
+      "step": 5415
+    },
+    {
+      "epoch": 0.2086621751684312,
+      "grad_norm": 1.0081026554107666,
+      "learning_rate": 0.00019469094721092444,
+      "loss": 1.2914,
+      "step": 5420
+    },
+    {
+      "epoch": 0.2088546679499519,
+      "grad_norm": 1.3790476322174072,
+      "learning_rate": 0.0001946812208388738,
+      "loss": 1.2817,
+      "step": 5425
+    },
+    {
+      "epoch": 0.20904716073147256,
+      "grad_norm": 1.777570128440857,
+      "learning_rate": 0.00019467148580885272,
+      "loss": 1.2253,
+      "step": 5430
+    },
+    {
+      "epoch": 0.20923965351299326,
+      "grad_norm": 1.1196024417877197,
+      "learning_rate": 0.00019466174212175142,
+      "loss": 1.2956,
+      "step": 5435
+    },
+    {
+      "epoch": 0.20943214629451395,
+      "grad_norm": 2.940906524658203,
+      "learning_rate": 0.00019465198977846086,
+      "loss": 1.3912,
+      "step": 5440
+    },
+    {
+      "epoch": 0.20962463907603465,
+      "grad_norm": 1.9075424671173096,
+      "learning_rate": 0.00019464222877987286,
+      "loss": 1.2518,
+      "step": 5445
+    },
+    {
+      "epoch": 0.20981713185755535,
+      "grad_norm": 1.0282469987869263,
+      "learning_rate": 0.00019463245912687996,
+      "loss": 1.2569,
+      "step": 5450
+    },
+    {
+      "epoch": 0.21000962463907605,
+      "grad_norm": 1.1651009321212769,
+      "learning_rate": 0.0001946226808203756,
+      "loss": 1.4676,
+      "step": 5455
+    },
+    {
+      "epoch": 0.21020211742059672,
+      "grad_norm": 1.1911680698394775,
+      "learning_rate": 0.00019461289386125388,
+      "loss": 1.3822,
+      "step": 5460
+    },
+    {
+      "epoch": 0.21039461020211742,
+      "grad_norm": 0.7187578082084656,
+      "learning_rate": 0.00019460309825040974,
+      "loss": 1.1462,
+      "step": 5465
+    },
+    {
+      "epoch": 0.2105871029836381,
+      "grad_norm": 2.401764154434204,
+      "learning_rate": 0.000194593293988739,
+      "loss": 1.3187,
+      "step": 5470
+    },
+    {
+      "epoch": 0.2107795957651588,
+      "grad_norm": 1.783333659172058,
+      "learning_rate": 0.0001945834810771381,
+      "loss": 1.3539,
+      "step": 5475
+    },
+    {
+      "epoch": 0.2109720885466795,
+      "grad_norm": 0.9923986196517944,
+      "learning_rate": 0.00019457365951650445,
+      "loss": 1.4837,
+      "step": 5480
+    },
+    {
+      "epoch": 0.21116458132820018,
+      "grad_norm": 1.0704642534255981,
+      "learning_rate": 0.00019456382930773612,
+      "loss": 1.2345,
+      "step": 5485
+    },
+    {
+      "epoch": 0.21135707410972088,
+      "grad_norm": 1.5242959260940552,
+      "learning_rate": 0.000194553990451732,
+      "loss": 1.2113,
+      "step": 5490
+    },
+    {
+      "epoch": 0.21154956689124157,
+      "grad_norm": 1.3185608386993408,
+      "learning_rate": 0.00019454414294939185,
+      "loss": 1.4083,
+      "step": 5495
+    },
+    {
+      "epoch": 0.21174205967276227,
+      "grad_norm": 1.1448662281036377,
+      "learning_rate": 0.00019453428680161615,
+      "loss": 1.4091,
+      "step": 5500
+    },
+    {
+      "epoch": 0.21193455245428297,
+      "grad_norm": 1.172396183013916,
+      "learning_rate": 0.0001945244220093061,
+      "loss": 1.1414,
+      "step": 5505
+    },
+    {
+      "epoch": 0.21212704523580367,
+      "grad_norm": 2.988346576690674,
+      "learning_rate": 0.00019451454857336383,
+      "loss": 1.3968,
+      "step": 5510
+    },
+    {
+      "epoch": 0.21231953801732434,
+      "grad_norm": 0.8824801445007324,
+      "learning_rate": 0.00019450466649469222,
+      "loss": 1.2229,
+      "step": 5515
+    },
+    {
+      "epoch": 0.21251203079884504,
+      "grad_norm": 1.7703745365142822,
+      "learning_rate": 0.00019449477577419488,
+      "loss": 1.3073,
+      "step": 5520
+    },
+    {
+      "epoch": 0.21270452358036573,
+      "grad_norm": 1.3374749422073364,
+      "learning_rate": 0.00019448487641277629,
+      "loss": 1.3908,
+      "step": 5525
+    },
+    {
+      "epoch": 0.21289701636188643,
+      "grad_norm": 1.2366503477096558,
+      "learning_rate": 0.00019447496841134163,
+      "loss": 1.2764,
+      "step": 5530
+    },
+    {
+      "epoch": 0.21308950914340713,
+      "grad_norm": 1.242353081703186,
+      "learning_rate": 0.00019446505177079696,
+      "loss": 1.3136,
+      "step": 5535
+    },
+    {
+      "epoch": 0.21328200192492783,
+      "grad_norm": 1.046583652496338,
+      "learning_rate": 0.00019445512649204907,
+      "loss": 1.1483,
+      "step": 5540
+    },
+    {
+      "epoch": 0.2134744947064485,
+      "grad_norm": 1.6280517578125,
+      "learning_rate": 0.00019444519257600558,
+      "loss": 1.4076,
+      "step": 5545
+    },
+    {
+      "epoch": 0.2136669874879692,
+      "grad_norm": 1.7472679615020752,
+      "learning_rate": 0.00019443525002357486,
+      "loss": 1.2842,
+      "step": 5550
+    },
+    {
+      "epoch": 0.2138594802694899,
+      "grad_norm": 1.101185917854309,
+      "learning_rate": 0.00019442529883566612,
+      "loss": 1.3037,
+      "step": 5555
+    },
+    {
+      "epoch": 0.2140519730510106,
+      "grad_norm": 1.8548834323883057,
+      "learning_rate": 0.0001944153390131893,
+      "loss": 1.4081,
+      "step": 5560
+    },
+    {
+      "epoch": 0.2142444658325313,
+      "grad_norm": 1.4205219745635986,
+      "learning_rate": 0.00019440537055705515,
+      "loss": 1.3419,
+      "step": 5565
+    },
+    {
+      "epoch": 0.214436958614052,
+      "grad_norm": 1.135933756828308,
+      "learning_rate": 0.0001943953934681753,
+      "loss": 0.9906,
+      "step": 5570
+    },
+    {
+      "epoch": 0.21462945139557266,
+      "grad_norm": 1.7350742816925049,
+      "learning_rate": 0.00019438540774746198,
+      "loss": 1.1193,
+      "step": 5575
+    },
+    {
+      "epoch": 0.21482194417709335,
+      "grad_norm": 1.891998291015625,
+      "learning_rate": 0.00019437541339582836,
+      "loss": 1.2271,
+      "step": 5580
+    },
+    {
+      "epoch": 0.21501443695861405,
+      "grad_norm": 1.2564722299575806,
+      "learning_rate": 0.0001943654104141884,
+      "loss": 1.5134,
+      "step": 5585
+    },
+    {
+      "epoch": 0.21520692974013475,
+      "grad_norm": 1.3632197380065918,
+      "learning_rate": 0.00019435539880345673,
+      "loss": 1.1772,
+      "step": 5590
+    },
+    {
+      "epoch": 0.21539942252165545,
+      "grad_norm": 1.8670414686203003,
+      "learning_rate": 0.00019434537856454894,
+      "loss": 1.2685,
+      "step": 5595
+    },
+    {
+      "epoch": 0.21559191530317612,
+      "grad_norm": 2.5948314666748047,
+      "learning_rate": 0.00019433534969838122,
+      "loss": 1.487,
+      "step": 5600
+    },
+    {
+      "epoch": 0.21578440808469682,
+      "grad_norm": 1.2312328815460205,
+      "learning_rate": 0.00019432531220587071,
+      "loss": 1.3394,
+      "step": 5605
+    },
+    {
+      "epoch": 0.2159769008662175,
+      "grad_norm": 0.9402896165847778,
+      "learning_rate": 0.0001943152660879352,
+      "loss": 1.1471,
+      "step": 5610
+    },
+    {
+      "epoch": 0.2161693936477382,
+      "grad_norm": 0.3871050477027893,
+      "learning_rate": 0.00019430521134549346,
+      "loss": 0.9597,
+      "step": 5615
+    },
+    {
+      "epoch": 0.2163618864292589,
+      "grad_norm": 0.9395222067832947,
+      "learning_rate": 0.0001942951479794648,
+      "loss": 1.3055,
+      "step": 5620
+    },
+    {
+      "epoch": 0.2165543792107796,
+      "grad_norm": 0.8928638696670532,
+      "learning_rate": 0.00019428507599076955,
+      "loss": 1.4099,
+      "step": 5625
+    },
+    {
+      "epoch": 0.21674687199230028,
+      "grad_norm": 1.8891551494598389,
+      "learning_rate": 0.00019427499538032865,
+      "loss": 1.5009,
+      "step": 5630
+    },
+    {
+      "epoch": 0.21693936477382098,
+      "grad_norm": 0.6684243679046631,
+      "learning_rate": 0.00019426490614906394,
+      "loss": 1.2251,
+      "step": 5635
+    },
+    {
+      "epoch": 0.21713185755534167,
+      "grad_norm": 1.5765355825424194,
+      "learning_rate": 0.00019425480829789803,
+      "loss": 1.1114,
+      "step": 5640
+    },
+    {
+      "epoch": 0.21732435033686237,
+      "grad_norm": 0.9966096878051758,
+      "learning_rate": 0.00019424470182775427,
+      "loss": 1.2907,
+      "step": 5645
+    },
+    {
+      "epoch": 0.21751684311838307,
+      "grad_norm": 1.263469934463501,
+      "learning_rate": 0.00019423458673955684,
+      "loss": 1.1443,
+      "step": 5650
+    },
+    {
+      "epoch": 0.21770933589990377,
+      "grad_norm": 1.5138813257217407,
+      "learning_rate": 0.0001942244630342307,
+      "loss": 1.2699,
+      "step": 5655
+    },
+    {
+      "epoch": 0.21790182868142444,
+      "grad_norm": 1.0215526819229126,
+      "learning_rate": 0.00019421433071270156,
+      "loss": 1.4265,
+      "step": 5660
+    },
+    {
+      "epoch": 0.21809432146294513,
+      "grad_norm": 0.7587301731109619,
+      "learning_rate": 0.00019420418977589605,
+      "loss": 1.1706,
+      "step": 5665
+    },
+    {
+      "epoch": 0.21828681424446583,
+      "grad_norm": 0.9531148672103882,
+      "learning_rate": 0.0001941940402247414,
+      "loss": 1.4041,
+      "step": 5670
+    },
+    {
+      "epoch": 0.21847930702598653,
+      "grad_norm": 1.098739743232727,
+      "learning_rate": 0.00019418388206016575,
+      "loss": 1.3476,
+      "step": 5675
+    },
+    {
+      "epoch": 0.21867179980750723,
+      "grad_norm": 1.0307271480560303,
+      "learning_rate": 0.000194173715283098,
+      "loss": 1.2333,
+      "step": 5680
+    },
+    {
+      "epoch": 0.2188642925890279,
+      "grad_norm": 1.538256049156189,
+      "learning_rate": 0.00019416353989446785,
+      "loss": 1.4489,
+      "step": 5685
+    },
+    {
+      "epoch": 0.2190567853705486,
+      "grad_norm": 1.5411714315414429,
+      "learning_rate": 0.00019415335589520574,
+      "loss": 1.2597,
+      "step": 5690
+    },
+    {
+      "epoch": 0.2192492781520693,
+      "grad_norm": 1.3543205261230469,
+      "learning_rate": 0.00019414316328624293,
+      "loss": 1.265,
+      "step": 5695
+    },
+    {
+      "epoch": 0.21944177093359,
+      "grad_norm": 0.7644770741462708,
+      "learning_rate": 0.0001941329620685115,
+      "loss": 1.1888,
+      "step": 5700
+    },
+    {
+      "epoch": 0.2196342637151107,
+      "grad_norm": 2.1122093200683594,
+      "learning_rate": 0.00019412275224294423,
+      "loss": 1.1301,
+      "step": 5705
+    },
+    {
+      "epoch": 0.2198267564966314,
+      "grad_norm": 1.4159448146820068,
+      "learning_rate": 0.00019411253381047477,
+      "loss": 1.209,
+      "step": 5710
+    },
+    {
+      "epoch": 0.22001924927815206,
+      "grad_norm": 1.4212615489959717,
+      "learning_rate": 0.00019410230677203755,
+      "loss": 1.3268,
+      "step": 5715
+    },
+    {
+      "epoch": 0.22021174205967275,
+      "grad_norm": 1.2042075395584106,
+      "learning_rate": 0.00019409207112856778,
+      "loss": 1.1976,
+      "step": 5720
+    },
+    {
+      "epoch": 0.22040423484119345,
+      "grad_norm": 1.5765044689178467,
+      "learning_rate": 0.00019408182688100136,
+      "loss": 1.3631,
+      "step": 5725
+    },
+    {
+      "epoch": 0.22059672762271415,
+      "grad_norm": 2.197000026702881,
+      "learning_rate": 0.00019407157403027514,
+      "loss": 1.2964,
+      "step": 5730
+    },
+    {
+      "epoch": 0.22078922040423485,
+      "grad_norm": 1.3434042930603027,
+      "learning_rate": 0.00019406131257732664,
+      "loss": 1.244,
+      "step": 5735
+    },
+    {
+      "epoch": 0.22098171318575555,
+      "grad_norm": 1.2889900207519531,
+      "learning_rate": 0.0001940510425230942,
+      "loss": 1.1333,
+      "step": 5740
+    },
+    {
+      "epoch": 0.22117420596727622,
+      "grad_norm": 0.8795220851898193,
+      "learning_rate": 0.00019404076386851692,
+      "loss": 1.2635,
+      "step": 5745
+    },
+    {
+      "epoch": 0.22136669874879691,
+      "grad_norm": 1.0312747955322266,
+      "learning_rate": 0.00019403047661453477,
+      "loss": 1.3195,
+      "step": 5750
+    },
+    {
+      "epoch": 0.2215591915303176,
+      "grad_norm": 1.5083264112472534,
+      "learning_rate": 0.00019402018076208845,
+      "loss": 1.3417,
+      "step": 5755
+    },
+    {
+      "epoch": 0.2217516843118383,
+      "grad_norm": 1.1538232564926147,
+      "learning_rate": 0.00019400987631211936,
+      "loss": 1.2956,
+      "step": 5760
+    },
+    {
+      "epoch": 0.221944177093359,
+      "grad_norm": 1.975381851196289,
+      "learning_rate": 0.0001939995632655699,
+      "loss": 1.4641,
+      "step": 5765
+    },
+    {
+      "epoch": 0.2221366698748797,
+      "grad_norm": 1.3251721858978271,
+      "learning_rate": 0.00019398924162338305,
+      "loss": 1.3429,
+      "step": 5770
+    },
+    {
+      "epoch": 0.22232916265640038,
+      "grad_norm": 1.1281229257583618,
+      "learning_rate": 0.0001939789113865027,
+      "loss": 1.2155,
+      "step": 5775
+    },
+    {
+      "epoch": 0.22252165543792107,
+      "grad_norm": 2.6070075035095215,
+      "learning_rate": 0.00019396857255587344,
+      "loss": 1.2634,
+      "step": 5780
+    },
+    {
+      "epoch": 0.22271414821944177,
+      "grad_norm": 1.0815184116363525,
+      "learning_rate": 0.00019395822513244067,
+      "loss": 1.1176,
+      "step": 5785
+    },
+    {
+      "epoch": 0.22290664100096247,
+      "grad_norm": 2.819180965423584,
+      "learning_rate": 0.0001939478691171507,
+      "loss": 1.2624,
+      "step": 5790
+    },
+    {
+      "epoch": 0.22309913378248317,
+      "grad_norm": 1.180055022239685,
+      "learning_rate": 0.0001939375045109504,
+      "loss": 1.3433,
+      "step": 5795
+    },
+    {
+      "epoch": 0.22329162656400384,
+      "grad_norm": 1.1582396030426025,
+      "learning_rate": 0.0001939271313147876,
+      "loss": 1.2815,
+      "step": 5800
+    },
+    {
+      "epoch": 0.22348411934552453,
+      "grad_norm": 2.32379412651062,
+      "learning_rate": 0.00019391674952961085,
+      "loss": 1.4095,
+      "step": 5805
+    },
+    {
+      "epoch": 0.22367661212704523,
+      "grad_norm": 1.5146657228469849,
+      "learning_rate": 0.0001939063591563695,
+      "loss": 1.2434,
+      "step": 5810
+    },
+    {
+      "epoch": 0.22386910490856593,
+      "grad_norm": 1.6434500217437744,
+      "learning_rate": 0.00019389596019601365,
+      "loss": 1.1739,
+      "step": 5815
+    },
+    {
+      "epoch": 0.22406159769008663,
+      "grad_norm": 1.7917993068695068,
+      "learning_rate": 0.0001938855526494943,
+      "loss": 1.5106,
+      "step": 5820
+    },
+    {
+      "epoch": 0.22425409047160733,
+      "grad_norm": 1.10679030418396,
+      "learning_rate": 0.00019387513651776303,
+      "loss": 1.284,
+      "step": 5825
+    },
+    {
+      "epoch": 0.224446583253128,
+      "grad_norm": 1.521506905555725,
+      "learning_rate": 0.00019386471180177247,
+      "loss": 1.4129,
+      "step": 5830
+    },
+    {
+      "epoch": 0.2246390760346487,
+      "grad_norm": 1.4055581092834473,
+      "learning_rate": 0.00019385427850247572,
+      "loss": 1.2476,
+      "step": 5835
+    },
+    {
+      "epoch": 0.2248315688161694,
+      "grad_norm": 0.9506363868713379,
+      "learning_rate": 0.00019384383662082703,
+      "loss": 1.3105,
+      "step": 5840
+    },
+    {
+      "epoch": 0.2250240615976901,
+      "grad_norm": 1.354658842086792,
+      "learning_rate": 0.00019383338615778107,
+      "loss": 1.29,
+      "step": 5845
+    },
+    {
+      "epoch": 0.2252165543792108,
+      "grad_norm": 0.8972203135490417,
+      "learning_rate": 0.00019382292711429353,
+      "loss": 1.3407,
+      "step": 5850
+    },
+    {
+      "epoch": 0.22540904716073148,
+      "grad_norm": 0.9989115595817566,
+      "learning_rate": 0.00019381245949132085,
+      "loss": 1.1662,
+      "step": 5855
+    },
+    {
+      "epoch": 0.22560153994225216,
+      "grad_norm": 1.1133052110671997,
+      "learning_rate": 0.0001938019832898202,
+      "loss": 1.2674,
+      "step": 5860
+    },
+    {
+      "epoch": 0.22579403272377285,
+      "grad_norm": 1.3640556335449219,
+      "learning_rate": 0.00019379149851074957,
+      "loss": 1.1989,
+      "step": 5865
+    },
+    {
+      "epoch": 0.22598652550529355,
+      "grad_norm": 1.2812589406967163,
+      "learning_rate": 0.0001937810051550677,
+      "loss": 1.4749,
+      "step": 5870
+    },
+    {
+      "epoch": 0.22617901828681425,
+      "grad_norm": 1.223944902420044,
+      "learning_rate": 0.00019377050322373412,
+      "loss": 1.305,
+      "step": 5875
+    },
+    {
+      "epoch": 0.22637151106833495,
+      "grad_norm": 1.3493690490722656,
+      "learning_rate": 0.00019375999271770925,
+      "loss": 1.458,
+      "step": 5880
+    },
+    {
+      "epoch": 0.22656400384985564,
+      "grad_norm": 1.4042202234268188,
+      "learning_rate": 0.0001937494736379541,
+      "loss": 1.1714,
+      "step": 5885
+    },
+    {
+      "epoch": 0.22675649663137631,
+      "grad_norm": 1.6239880323410034,
+      "learning_rate": 0.00019373894598543066,
+      "loss": 1.3224,
+      "step": 5890
+    },
+    {
+      "epoch": 0.226948989412897,
+      "grad_norm": 1.096960425376892,
+      "learning_rate": 0.00019372840976110154,
+      "loss": 1.128,
+      "step": 5895
+    },
+    {
+      "epoch": 0.2271414821944177,
+      "grad_norm": 1.6740233898162842,
+      "learning_rate": 0.00019371786496593028,
+      "loss": 1.195,
+      "step": 5900
+    },
+    {
+      "epoch": 0.2273339749759384,
+      "grad_norm": 1.454030156135559,
+      "learning_rate": 0.00019370731160088105,
+      "loss": 1.2641,
+      "step": 5905
+    },
+    {
+      "epoch": 0.2275264677574591,
+      "grad_norm": 1.4465221166610718,
+      "learning_rate": 0.00019369674966691897,
+      "loss": 1.331,
+      "step": 5910
+    },
+    {
+      "epoch": 0.22771896053897978,
+      "grad_norm": 1.6115851402282715,
+      "learning_rate": 0.00019368617916500978,
+      "loss": 1.4061,
+      "step": 5915
+    },
+    {
+      "epoch": 0.22791145332050047,
+      "grad_norm": 1.0165706872940063,
+      "learning_rate": 0.00019367560009612013,
+      "loss": 1.177,
+      "step": 5920
+    },
+    {
+      "epoch": 0.22810394610202117,
+      "grad_norm": 1.5200728178024292,
+      "learning_rate": 0.00019366501246121737,
+      "loss": 1.1323,
+      "step": 5925
+    },
+    {
+      "epoch": 0.22829643888354187,
+      "grad_norm": 1.4613386392593384,
+      "learning_rate": 0.00019365441626126976,
+      "loss": 1.4626,
+      "step": 5930
+    },
+    {
+      "epoch": 0.22848893166506257,
+      "grad_norm": 1.2502466440200806,
+      "learning_rate": 0.00019364381149724613,
+      "loss": 1.2797,
+      "step": 5935
+    },
+    {
+      "epoch": 0.22868142444658326,
+      "grad_norm": 1.2946960926055908,
+      "learning_rate": 0.0001936331981701163,
+      "loss": 1.3844,
+      "step": 5940
+    },
+    {
+      "epoch": 0.22887391722810393,
+      "grad_norm": 1.2478231191635132,
+      "learning_rate": 0.00019362257628085074,
+      "loss": 1.2855,
+      "step": 5945
+    },
+    {
+      "epoch": 0.22906641000962463,
+      "grad_norm": 1.0097830295562744,
+      "learning_rate": 0.0001936119458304208,
+      "loss": 1.1223,
+      "step": 5950
+    },
+    {
+      "epoch": 0.22925890279114533,
+      "grad_norm": 1.3235141038894653,
+      "learning_rate": 0.00019360130681979852,
+      "loss": 1.284,
+      "step": 5955
+    },
+    {
+      "epoch": 0.22945139557266603,
+      "grad_norm": 1.6869986057281494,
+      "learning_rate": 0.00019359065924995678,
+      "loss": 1.517,
+      "step": 5960
+    },
+    {
+      "epoch": 0.22964388835418673,
+      "grad_norm": 0.9644334316253662,
+      "learning_rate": 0.00019358000312186925,
+      "loss": 1.0607,
+      "step": 5965
+    },
+    {
+      "epoch": 0.22983638113570742,
+      "grad_norm": 1.063192367553711,
+      "learning_rate": 0.0001935693384365103,
+      "loss": 0.9187,
+      "step": 5970
+    },
+    {
+      "epoch": 0.2300288739172281,
+      "grad_norm": 1.0339081287384033,
+      "learning_rate": 0.00019355866519485523,
+      "loss": 1.2946,
+      "step": 5975
+    },
+    {
+      "epoch": 0.2302213666987488,
+      "grad_norm": 1.3194791078567505,
+      "learning_rate": 0.00019354798339788,
+      "loss": 1.4293,
+      "step": 5980
+    },
+    {
+      "epoch": 0.2304138594802695,
+      "grad_norm": 1.8870794773101807,
+      "learning_rate": 0.00019353729304656136,
+      "loss": 1.4124,
+      "step": 5985
+    },
+    {
+      "epoch": 0.2306063522617902,
+      "grad_norm": 1.132385015487671,
+      "learning_rate": 0.00019352659414187694,
+      "loss": 1.1949,
+      "step": 5990
+    },
+    {
+      "epoch": 0.23079884504331089,
+      "grad_norm": 2.763613700866699,
+      "learning_rate": 0.000193515886684805,
+      "loss": 1.2341,
+      "step": 5995
+    },
+    {
+      "epoch": 0.23099133782483156,
+      "grad_norm": 1.6793404817581177,
+      "learning_rate": 0.00019350517067632473,
+      "loss": 1.3597,
+      "step": 6000
+    },
+    {
+      "epoch": 0.23118383060635225,
+      "grad_norm": 1.1538963317871094,
+      "learning_rate": 0.000193494446117416,
+      "loss": 1.1981,
+      "step": 6005
+    },
+    {
+      "epoch": 0.23137632338787295,
+      "grad_norm": 1.0233584642410278,
+      "learning_rate": 0.00019348371300905955,
+      "loss": 1.2821,
+      "step": 6010
+    },
+    {
+      "epoch": 0.23156881616939365,
+      "grad_norm": 1.3905096054077148,
+      "learning_rate": 0.0001934729713522368,
+      "loss": 1.3471,
+      "step": 6015
+    },
+    {
+      "epoch": 0.23176130895091435,
+      "grad_norm": 1.345563292503357,
+      "learning_rate": 0.00019346222114793,
+      "loss": 1.0454,
+      "step": 6020
+    },
+    {
+      "epoch": 0.23195380173243504,
+      "grad_norm": 0.739811897277832,
+      "learning_rate": 0.00019345146239712225,
+      "loss": 1.3125,
+      "step": 6025
+    },
+    {
+      "epoch": 0.23214629451395571,
+      "grad_norm": 1.977918028831482,
+      "learning_rate": 0.0001934406951007973,
+      "loss": 1.3328,
+      "step": 6030
+    },
+    {
+      "epoch": 0.2323387872954764,
+      "grad_norm": 0.9505223035812378,
+      "learning_rate": 0.00019342991925993977,
+      "loss": 1.1388,
+      "step": 6035
+    },
+    {
+      "epoch": 0.2325312800769971,
+      "grad_norm": 1.257755160331726,
+      "learning_rate": 0.00019341913487553502,
+      "loss": 1.3064,
+      "step": 6040
+    },
+    {
+      "epoch": 0.2327237728585178,
+      "grad_norm": 1.2003203630447388,
+      "learning_rate": 0.00019340834194856926,
+      "loss": 1.4369,
+      "step": 6045
+    },
+    {
+      "epoch": 0.2329162656400385,
+      "grad_norm": 1.2289738655090332,
+      "learning_rate": 0.0001933975404800294,
+      "loss": 1.1462,
+      "step": 6050
+    },
+    {
+      "epoch": 0.2331087584215592,
+      "grad_norm": 1.227171540260315,
+      "learning_rate": 0.00019338673047090317,
+      "loss": 1.1829,
+      "step": 6055
+    },
+    {
+      "epoch": 0.23330125120307987,
+      "grad_norm": 1.2766560316085815,
+      "learning_rate": 0.00019337591192217904,
+      "loss": 1.2572,
+      "step": 6060
+    },
+    {
+      "epoch": 0.23349374398460057,
+      "grad_norm": 2.6716904640197754,
+      "learning_rate": 0.00019336508483484634,
+      "loss": 1.0195,
+      "step": 6065
+    },
+    {
+      "epoch": 0.23368623676612127,
+      "grad_norm": 1.1586931943893433,
+      "learning_rate": 0.00019335424920989512,
+      "loss": 1.4932,
+      "step": 6070
+    },
+    {
+      "epoch": 0.23387872954764197,
+      "grad_norm": 1.0196670293807983,
+      "learning_rate": 0.00019334340504831624,
+      "loss": 1.3497,
+      "step": 6075
+    },
+    {
+      "epoch": 0.23407122232916266,
+      "grad_norm": 1.6527109146118164,
+      "learning_rate": 0.00019333255235110127,
+      "loss": 1.1239,
+      "step": 6080
+    },
+    {
+      "epoch": 0.23426371511068336,
+      "grad_norm": 0.9913870096206665,
+      "learning_rate": 0.00019332169111924271,
+      "loss": 1.2757,
+      "step": 6085
+    },
+    {
+      "epoch": 0.23445620789220403,
+      "grad_norm": 1.1027697324752808,
+      "learning_rate": 0.00019331082135373367,
+      "loss": 1.2512,
+      "step": 6090
+    },
+    {
+      "epoch": 0.23464870067372473,
+      "grad_norm": 1.9269218444824219,
+      "learning_rate": 0.00019329994305556815,
+      "loss": 1.4698,
+      "step": 6095
+    },
+    {
+      "epoch": 0.23484119345524543,
+      "grad_norm": 1.1504942178726196,
+      "learning_rate": 0.00019328905622574086,
+      "loss": 1.4844,
+      "step": 6100
+    },
+    {
+      "epoch": 0.23503368623676613,
+      "grad_norm": 1.1164321899414062,
+      "learning_rate": 0.0001932781608652474,
+      "loss": 1.2972,
+      "step": 6105
+    },
+    {
+      "epoch": 0.23522617901828682,
+      "grad_norm": 1.283000111579895,
+      "learning_rate": 0.00019326725697508407,
+      "loss": 1.3117,
+      "step": 6110
+    },
+    {
+      "epoch": 0.2354186717998075,
+      "grad_norm": 1.3553595542907715,
+      "learning_rate": 0.00019325634455624787,
+      "loss": 1.027,
+      "step": 6115
+    },
+    {
+      "epoch": 0.2356111645813282,
+      "grad_norm": 2.1605517864227295,
+      "learning_rate": 0.00019324542360973674,
+      "loss": 1.2211,
+      "step": 6120
+    },
+    {
+      "epoch": 0.2358036573628489,
+      "grad_norm": 1.1028283834457397,
+      "learning_rate": 0.00019323449413654933,
+      "loss": 1.3034,
+      "step": 6125
+    },
+    {
+      "epoch": 0.2359961501443696,
+      "grad_norm": 1.1728841066360474,
+      "learning_rate": 0.00019322355613768505,
+      "loss": 1.3135,
+      "step": 6130
+    },
+    {
+      "epoch": 0.23618864292589029,
+      "grad_norm": 1.7304178476333618,
+      "learning_rate": 0.0001932126096141441,
+      "loss": 1.3516,
+      "step": 6135
+    },
+    {
+      "epoch": 0.23638113570741098,
+      "grad_norm": 1.3326451778411865,
+      "learning_rate": 0.00019320165456692748,
+      "loss": 1.3371,
+      "step": 6140
+    },
+    {
+      "epoch": 0.23657362848893165,
+      "grad_norm": 1.6894330978393555,
+      "learning_rate": 0.00019319069099703697,
+      "loss": 1.2126,
+      "step": 6145
+    },
+    {
+      "epoch": 0.23676612127045235,
+      "grad_norm": 1.7248213291168213,
+      "learning_rate": 0.0001931797189054751,
+      "loss": 1.193,
+      "step": 6150
+    },
+    {
+      "epoch": 0.23695861405197305,
+      "grad_norm": 1.1517174243927002,
+      "learning_rate": 0.0001931687382932452,
+      "loss": 1.1472,
+      "step": 6155
+    },
+    {
+      "epoch": 0.23715110683349375,
+      "grad_norm": 2.4606590270996094,
+      "learning_rate": 0.00019315774916135134,
+      "loss": 1.524,
+      "step": 6160
+    },
+    {
+      "epoch": 0.23734359961501444,
+      "grad_norm": 1.6130386590957642,
+      "learning_rate": 0.00019314675151079844,
+      "loss": 1.052,
+      "step": 6165
+    },
+    {
+      "epoch": 0.23753609239653514,
+      "grad_norm": 1.3845412731170654,
+      "learning_rate": 0.00019313574534259216,
+      "loss": 1.2557,
+      "step": 6170
+    },
+    {
+      "epoch": 0.2377285851780558,
+      "grad_norm": 1.3509567975997925,
+      "learning_rate": 0.00019312473065773893,
+      "loss": 1.3083,
+      "step": 6175
+    },
+    {
+      "epoch": 0.2379210779595765,
+      "grad_norm": 1.358113408088684,
+      "learning_rate": 0.000193113707457246,
+      "loss": 1.2226,
+      "step": 6180
+    },
+    {
+      "epoch": 0.2381135707410972,
+      "grad_norm": 0.9598337411880493,
+      "learning_rate": 0.00019310267574212134,
+      "loss": 1.1861,
+      "step": 6185
+    },
+    {
+      "epoch": 0.2383060635226179,
+      "grad_norm": 1.347159743309021,
+      "learning_rate": 0.0001930916355133737,
+      "loss": 1.2782,
+      "step": 6190
+    },
+    {
+      "epoch": 0.2384985563041386,
+      "grad_norm": 1.0227164030075073,
+      "learning_rate": 0.0001930805867720127,
+      "loss": 1.2909,
+      "step": 6195
+    },
+    {
+      "epoch": 0.2386910490856593,
+      "grad_norm": 1.8373135328292847,
+      "learning_rate": 0.00019306952951904865,
+      "loss": 1.3371,
+      "step": 6200
+    },
+    {
+      "epoch": 0.23888354186717997,
+      "grad_norm": 2.130218267440796,
+      "learning_rate": 0.00019305846375549263,
+      "loss": 1.3275,
+      "step": 6205
+    },
+    {
+      "epoch": 0.23907603464870067,
+      "grad_norm": 1.3699109554290771,
+      "learning_rate": 0.00019304738948235656,
+      "loss": 1.172,
+      "step": 6210
+    },
+    {
+      "epoch": 0.23926852743022137,
+      "grad_norm": 1.8254964351654053,
+      "learning_rate": 0.0001930363067006531,
+      "loss": 1.166,
+      "step": 6215
+    },
+    {
+      "epoch": 0.23946102021174206,
+      "grad_norm": 2.6475026607513428,
+      "learning_rate": 0.00019302521541139571,
+      "loss": 1.3168,
+      "step": 6220
+    },
+    {
+      "epoch": 0.23965351299326276,
+      "grad_norm": 1.4869440793991089,
+      "learning_rate": 0.0001930141156155986,
+      "loss": 1.1112,
+      "step": 6225
+    },
+    {
+      "epoch": 0.23984600577478343,
+      "grad_norm": 1.0316526889801025,
+      "learning_rate": 0.00019300300731427678,
+      "loss": 1.3845,
+      "step": 6230
+    },
+    {
+      "epoch": 0.24003849855630413,
+      "grad_norm": 1.1549556255340576,
+      "learning_rate": 0.00019299189050844603,
+      "loss": 1.378,
+      "step": 6235
+    },
+    {
+      "epoch": 0.24023099133782483,
+      "grad_norm": 1.9833987951278687,
+      "learning_rate": 0.00019298076519912294,
+      "loss": 1.2631,
+      "step": 6240
+    },
+    {
+      "epoch": 0.24042348411934553,
+      "grad_norm": 1.1354988813400269,
+      "learning_rate": 0.00019296963138732478,
+      "loss": 1.6525,
+      "step": 6245
+    },
+    {
+      "epoch": 0.24061597690086622,
+      "grad_norm": 1.6483670473098755,
+      "learning_rate": 0.0001929584890740697,
+      "loss": 0.9828,
+      "step": 6250
+    },
+    {
+      "epoch": 0.24080846968238692,
+      "grad_norm": 1.537610650062561,
+      "learning_rate": 0.00019294733826037659,
+      "loss": 1.3566,
+      "step": 6255
+    },
+    {
+      "epoch": 0.2410009624639076,
+      "grad_norm": 1.207406759262085,
+      "learning_rate": 0.0001929361789472651,
+      "loss": 1.3306,
+      "step": 6260
+    },
+    {
+      "epoch": 0.2411934552454283,
+      "grad_norm": 1.4772666692733765,
+      "learning_rate": 0.00019292501113575572,
+      "loss": 1.3117,
+      "step": 6265
+    },
+    {
+      "epoch": 0.241385948026949,
+      "grad_norm": 1.8285613059997559,
+      "learning_rate": 0.00019291383482686962,
+      "loss": 1.3711,
+      "step": 6270
+    },
+    {
+      "epoch": 0.24157844080846969,
+      "grad_norm": 0.9223503470420837,
+      "learning_rate": 0.00019290265002162884,
+      "loss": 1.1712,
+      "step": 6275
+    },
+    {
+      "epoch": 0.24177093358999038,
+      "grad_norm": 2.1818087100982666,
+      "learning_rate": 0.00019289145672105612,
+      "loss": 1.1596,
+      "step": 6280
+    },
+    {
+      "epoch": 0.24196342637151108,
+      "grad_norm": 0.8749092817306519,
+      "learning_rate": 0.00019288025492617504,
+      "loss": 1.0726,
+      "step": 6285
+    },
+    {
+      "epoch": 0.24215591915303175,
+      "grad_norm": 1.1598855257034302,
+      "learning_rate": 0.00019286904463800995,
+      "loss": 1.2931,
+      "step": 6290
+    },
+    {
+      "epoch": 0.24234841193455245,
+      "grad_norm": 1.4357101917266846,
+      "learning_rate": 0.0001928578258575859,
+      "loss": 1.2612,
+      "step": 6295
+    },
+    {
+      "epoch": 0.24254090471607315,
+      "grad_norm": 0.9731203317642212,
+      "learning_rate": 0.0001928465985859288,
+      "loss": 1.178,
+      "step": 6300
+    },
+    {
+      "epoch": 0.24273339749759384,
+      "grad_norm": 1.1217381954193115,
+      "learning_rate": 0.00019283536282406534,
+      "loss": 1.285,
+      "step": 6305
+    },
+    {
+      "epoch": 0.24292589027911454,
+      "grad_norm": 1.415860891342163,
+      "learning_rate": 0.0001928241185730229,
+      "loss": 1.399,
+      "step": 6310
+    },
+    {
+      "epoch": 0.2431183830606352,
+      "grad_norm": 0.9067175388336182,
+      "learning_rate": 0.00019281286583382973,
+      "loss": 1.2336,
+      "step": 6315
+    },
+    {
+      "epoch": 0.2433108758421559,
+      "grad_norm": 1.6320233345031738,
+      "learning_rate": 0.0001928016046075148,
+      "loss": 1.4348,
+      "step": 6320
+    },
+    {
+      "epoch": 0.2435033686236766,
+      "grad_norm": 1.3945854902267456,
+      "learning_rate": 0.0001927903348951079,
+      "loss": 1.1614,
+      "step": 6325
+    },
+    {
+      "epoch": 0.2436958614051973,
+      "grad_norm": 1.37948477268219,
+      "learning_rate": 0.00019277905669763952,
+      "loss": 1.2058,
+      "step": 6330
+    },
+    {
+      "epoch": 0.243888354186718,
+      "grad_norm": 1.3325083255767822,
+      "learning_rate": 0.00019276777001614104,
+      "loss": 1.2737,
+      "step": 6335
+    },
+    {
+      "epoch": 0.2440808469682387,
+      "grad_norm": 1.5902581214904785,
+      "learning_rate": 0.00019275647485164453,
+      "loss": 1.3706,
+      "step": 6340
+    },
+    {
+      "epoch": 0.24427333974975937,
+      "grad_norm": 1.1309142112731934,
+      "learning_rate": 0.00019274517120518284,
+      "loss": 1.2408,
+      "step": 6345
+    },
+    {
+      "epoch": 0.24446583253128007,
+      "grad_norm": 1.9998489618301392,
+      "learning_rate": 0.0001927338590777896,
+      "loss": 1.3079,
+      "step": 6350
+    },
+    {
+      "epoch": 0.24465832531280077,
+      "grad_norm": 1.569667100906372,
+      "learning_rate": 0.00019272253847049927,
+      "loss": 1.2365,
+      "step": 6355
+    },
+    {
+      "epoch": 0.24485081809432147,
+      "grad_norm": 1.2294694185256958,
+      "learning_rate": 0.00019271120938434702,
+      "loss": 1.3544,
+      "step": 6360
+    },
+    {
+      "epoch": 0.24504331087584216,
+      "grad_norm": 1.9876806735992432,
+      "learning_rate": 0.00019269987182036883,
+      "loss": 1.3675,
+      "step": 6365
+    },
+    {
+      "epoch": 0.24523580365736286,
+      "grad_norm": 1.3317819833755493,
+      "learning_rate": 0.0001926885257796015,
+      "loss": 1.0949,
+      "step": 6370
+    },
+    {
+      "epoch": 0.24542829643888353,
+      "grad_norm": 1.7602546215057373,
+      "learning_rate": 0.00019267717126308242,
+      "loss": 1.3168,
+      "step": 6375
+    },
+    {
+      "epoch": 0.24562078922040423,
+      "grad_norm": 1.5651274919509888,
+      "learning_rate": 0.00019266580827184996,
+      "loss": 1.2802,
+      "step": 6380
+    },
+    {
+      "epoch": 0.24581328200192493,
+      "grad_norm": 0.9537544846534729,
+      "learning_rate": 0.0001926544368069432,
+      "loss": 1.1876,
+      "step": 6385
+    },
+    {
+      "epoch": 0.24600577478344562,
+      "grad_norm": 0.9649773240089417,
+      "learning_rate": 0.000192643056869402,
+      "loss": 1.1378,
+      "step": 6390
+    },
+    {
+      "epoch": 0.24619826756496632,
+      "grad_norm": 1.6363686323165894,
+      "learning_rate": 0.00019263166846026692,
+      "loss": 1.3284,
+      "step": 6395
+    },
+    {
+      "epoch": 0.24639076034648702,
+      "grad_norm": 1.748897910118103,
+      "learning_rate": 0.00019262027158057943,
+      "loss": 1.4314,
+      "step": 6400
+    },
+    {
+      "epoch": 0.2465832531280077,
+      "grad_norm": 2.138967990875244,
+      "learning_rate": 0.00019260886623138164,
+      "loss": 1.2244,
+      "step": 6405
+    },
+    {
+      "epoch": 0.2467757459095284,
+      "grad_norm": 2.517312526702881,
+      "learning_rate": 0.0001925974524137165,
+      "loss": 1.3394,
+      "step": 6410
+    },
+    {
+      "epoch": 0.24696823869104909,
+      "grad_norm": 1.7510714530944824,
+      "learning_rate": 0.00019258603012862772,
+      "loss": 1.3369,
+      "step": 6415
+    },
+    {
+      "epoch": 0.24716073147256978,
+      "grad_norm": 1.1651504039764404,
+      "learning_rate": 0.00019257459937715985,
+      "loss": 1.2953,
+      "step": 6420
+    },
+    {
+      "epoch": 0.24735322425409048,
+      "grad_norm": 1.325554609298706,
+      "learning_rate": 0.0001925631601603581,
+      "loss": 1.3062,
+      "step": 6425
+    },
+    {
+      "epoch": 0.24754571703561115,
+      "grad_norm": 1.0340043306350708,
+      "learning_rate": 0.00019255171247926852,
+      "loss": 1.337,
+      "step": 6430
+    },
+    {
+      "epoch": 0.24773820981713185,
+      "grad_norm": 1.677131175994873,
+      "learning_rate": 0.00019254025633493792,
+      "loss": 1.3179,
+      "step": 6435
+    },
+    {
+      "epoch": 0.24793070259865255,
+      "grad_norm": 2.475339651107788,
+      "learning_rate": 0.00019252879172841395,
+      "loss": 1.4765,
+      "step": 6440
+    },
+    {
+      "epoch": 0.24812319538017324,
+      "grad_norm": 1.1302917003631592,
+      "learning_rate": 0.00019251731866074486,
+      "loss": 1.3029,
+      "step": 6445
+    },
+    {
+      "epoch": 0.24831568816169394,
+      "grad_norm": 1.3425379991531372,
+      "learning_rate": 0.0001925058371329799,
+      "loss": 1.1263,
+      "step": 6450
+    },
+    {
+      "epoch": 0.24850818094321464,
+      "grad_norm": 1.0058633089065552,
+      "learning_rate": 0.0001924943471461689,
+      "loss": 1.1059,
+      "step": 6455
+    },
+    {
+      "epoch": 0.2487006737247353,
+      "grad_norm": 1.9793190956115723,
+      "learning_rate": 0.0001924828487013626,
+      "loss": 1.5268,
+      "step": 6460
+    },
+    {
+      "epoch": 0.248893166506256,
+      "grad_norm": 1.0673744678497314,
+      "learning_rate": 0.00019247134179961242,
+      "loss": 1.2199,
+      "step": 6465
+    },
+    {
+      "epoch": 0.2490856592877767,
+      "grad_norm": 1.1182838678359985,
+      "learning_rate": 0.00019245982644197057,
+      "loss": 1.5456,
+      "step": 6470
+    },
+    {
+      "epoch": 0.2492781520692974,
+      "grad_norm": 0.9264312982559204,
+      "learning_rate": 0.00019244830262949014,
+      "loss": 1.2367,
+      "step": 6475
+    },
+    {
+      "epoch": 0.2494706448508181,
+      "grad_norm": 1.2094528675079346,
+      "learning_rate": 0.00019243677036322478,
+      "loss": 1.2026,
+      "step": 6480
+    },
+    {
+      "epoch": 0.2496631376323388,
+      "grad_norm": 1.275902509689331,
+      "learning_rate": 0.00019242522964422917,
+      "loss": 1.206,
+      "step": 6485
+    },
+    {
+      "epoch": 0.24985563041385947,
+      "grad_norm": 1.515559434890747,
+      "learning_rate": 0.00019241368047355853,
+      "loss": 1.2222,
+      "step": 6490
+    },
+    {
+      "epoch": 0.25004812319538017,
+      "grad_norm": 0.9974495768547058,
+      "learning_rate": 0.000192402122852269,
+      "loss": 1.5274,
+      "step": 6495
+    },
+    {
+      "epoch": 0.2502406159769009,
+      "grad_norm": 1.8940407037734985,
+      "learning_rate": 0.00019239055678141746,
+      "loss": 1.3639,
+      "step": 6500
+    },
+    {
+      "epoch": 0.25043310875842156,
+      "grad_norm": 1.7484371662139893,
+      "learning_rate": 0.00019237898226206153,
+      "loss": 1.3517,
+      "step": 6505
+    },
+    {
+      "epoch": 0.25062560153994223,
+      "grad_norm": 1.004660725593567,
+      "learning_rate": 0.00019236739929525963,
+      "loss": 1.0603,
+      "step": 6510
+    },
+    {
+      "epoch": 0.25081809432146296,
+      "grad_norm": 0.9729489684104919,
+      "learning_rate": 0.00019235580788207093,
+      "loss": 1.3252,
+      "step": 6515
+    },
+    {
+      "epoch": 0.25101058710298363,
+      "grad_norm": 0.4645654857158661,
+      "learning_rate": 0.00019234420802355539,
+      "loss": 1.1804,
+      "step": 6520
+    },
+    {
+      "epoch": 0.25120307988450435,
+      "grad_norm": 1.0810743570327759,
+      "learning_rate": 0.00019233259972077378,
+      "loss": 1.3045,
+      "step": 6525
+    },
+    {
+      "epoch": 0.251395572666025,
+      "grad_norm": 1.1666224002838135,
+      "learning_rate": 0.00019232098297478756,
+      "loss": 1.324,
+      "step": 6530
+    },
+    {
+      "epoch": 0.2515880654475457,
+      "grad_norm": 1.06947660446167,
+      "learning_rate": 0.000192309357786659,
+      "loss": 1.3131,
+      "step": 6535
+    },
+    {
+      "epoch": 0.2517805582290664,
+      "grad_norm": 1.1774028539657593,
+      "learning_rate": 0.0001922977241574512,
+      "loss": 1.301,
+      "step": 6540
+    },
+    {
+      "epoch": 0.2519730510105871,
+      "grad_norm": 1.528041958808899,
+      "learning_rate": 0.0001922860820882279,
+      "loss": 1.2542,
+      "step": 6545
+    },
+    {
+      "epoch": 0.2521655437921078,
+      "grad_norm": 1.1932915449142456,
+      "learning_rate": 0.00019227443158005377,
+      "loss": 1.125,
+      "step": 6550
+    },
+    {
+      "epoch": 0.2523580365736285,
+      "grad_norm": 1.3258370161056519,
+      "learning_rate": 0.0001922627726339941,
+      "loss": 1.3776,
+      "step": 6555
+    },
+    {
+      "epoch": 0.25255052935514916,
+      "grad_norm": 0.994076132774353,
+      "learning_rate": 0.0001922511052511151,
+      "loss": 1.0908,
+      "step": 6560
+    },
+    {
+      "epoch": 0.2527430221366699,
+      "grad_norm": 1.0820032358169556,
+      "learning_rate": 0.00019223942943248358,
+      "loss": 1.215,
+      "step": 6565
+    },
+    {
+      "epoch": 0.25293551491819055,
+      "grad_norm": 0.9792138338088989,
+      "learning_rate": 0.00019222774517916734,
+      "loss": 1.2413,
+      "step": 6570
+    },
+    {
+      "epoch": 0.2531280076997113,
+      "grad_norm": 1.1704801321029663,
+      "learning_rate": 0.0001922160524922347,
+      "loss": 1.5203,
+      "step": 6575
+    },
+    {
+      "epoch": 0.25332050048123195,
+      "grad_norm": 1.6249198913574219,
+      "learning_rate": 0.00019220435137275494,
+      "loss": 1.2771,
+      "step": 6580
+    },
+    {
+      "epoch": 0.2535129932627527,
+      "grad_norm": 1.3218034505844116,
+      "learning_rate": 0.00019219264182179804,
+      "loss": 1.4433,
+      "step": 6585
+    },
+    {
+      "epoch": 0.25370548604427334,
+      "grad_norm": 1.7230724096298218,
+      "learning_rate": 0.0001921809238404348,
+      "loss": 1.1069,
+      "step": 6590
+    },
+    {
+      "epoch": 0.253897978825794,
+      "grad_norm": 1.3148738145828247,
+      "learning_rate": 0.00019216919742973669,
+      "loss": 1.2386,
+      "step": 6595
+    },
+    {
+      "epoch": 0.25409047160731474,
+      "grad_norm": 1.257513403892517,
+      "learning_rate": 0.00019215746259077605,
+      "loss": 1.3476,
+      "step": 6600
+    },
+    {
+      "epoch": 0.2542829643888354,
+      "grad_norm": 0.965403139591217,
+      "learning_rate": 0.00019214571932462592,
+      "loss": 1.1045,
+      "step": 6605
+    },
+    {
+      "epoch": 0.25447545717035613,
+      "grad_norm": 0.8903887867927551,
+      "learning_rate": 0.0001921339676323602,
+      "loss": 1.1481,
+      "step": 6610
+    },
+    {
+      "epoch": 0.2546679499518768,
+      "grad_norm": 1.284529209136963,
+      "learning_rate": 0.00019212220751505345,
+      "loss": 1.3179,
+      "step": 6615
+    },
+    {
+      "epoch": 0.2548604427333975,
+      "grad_norm": 2.3491082191467285,
+      "learning_rate": 0.0001921104389737811,
+      "loss": 1.3042,
+      "step": 6620
+    },
+    {
+      "epoch": 0.2550529355149182,
+      "grad_norm": 1.4170057773590088,
+      "learning_rate": 0.00019209866200961927,
+      "loss": 1.3775,
+      "step": 6625
+    },
+    {
+      "epoch": 0.25524542829643887,
+      "grad_norm": 1.4182847738265991,
+      "learning_rate": 0.00019208687662364488,
+      "loss": 1.3895,
+      "step": 6630
+    },
+    {
+      "epoch": 0.2554379210779596,
+      "grad_norm": 1.2162110805511475,
+      "learning_rate": 0.00019207508281693568,
+      "loss": 1.0754,
+      "step": 6635
+    },
+    {
+      "epoch": 0.25563041385948027,
+      "grad_norm": 1.473873257637024,
+      "learning_rate": 0.00019206328059057006,
+      "loss": 1.3323,
+      "step": 6640
+    },
+    {
+      "epoch": 0.25582290664100094,
+      "grad_norm": 1.2990386486053467,
+      "learning_rate": 0.0001920514699456273,
+      "loss": 1.2304,
+      "step": 6645
+    },
+    {
+      "epoch": 0.25601539942252166,
+      "grad_norm": 1.2828303575515747,
+      "learning_rate": 0.00019203965088318743,
+      "loss": 1.2566,
+      "step": 6650
+    },
+    {
+      "epoch": 0.25620789220404233,
+      "grad_norm": 0.9165570735931396,
+      "learning_rate": 0.00019202782340433115,
+      "loss": 1.2186,
+      "step": 6655
+    },
+    {
+      "epoch": 0.25640038498556306,
+      "grad_norm": 2.0381886959075928,
+      "learning_rate": 0.00019201598751014006,
+      "loss": 1.114,
+      "step": 6660
+    },
+    {
+      "epoch": 0.2565928777670837,
+      "grad_norm": 1.252790093421936,
+      "learning_rate": 0.00019200414320169647,
+      "loss": 1.2354,
+      "step": 6665
+    },
+    {
+      "epoch": 0.25678537054860445,
+      "grad_norm": 1.1557594537734985,
+      "learning_rate": 0.00019199229048008347,
+      "loss": 1.3652,
+      "step": 6670
+    },
+    {
+      "epoch": 0.2569778633301251,
+      "grad_norm": 1.356181025505066,
+      "learning_rate": 0.0001919804293463849,
+      "loss": 1.1026,
+      "step": 6675
+    },
+    {
+      "epoch": 0.2571703561116458,
+      "grad_norm": 1.2493314743041992,
+      "learning_rate": 0.00019196855980168536,
+      "loss": 1.2225,
+      "step": 6680
+    },
+    {
+      "epoch": 0.2573628488931665,
+      "grad_norm": 1.7480677366256714,
+      "learning_rate": 0.00019195668184707025,
+      "loss": 1.2898,
+      "step": 6685
+    },
+    {
+      "epoch": 0.2575553416746872,
+      "grad_norm": 1.0522620677947998,
+      "learning_rate": 0.00019194479548362577,
+      "loss": 1.1404,
+      "step": 6690
+    },
+    {
+      "epoch": 0.2577478344562079,
+      "grad_norm": 1.4085676670074463,
+      "learning_rate": 0.00019193290071243882,
+      "loss": 1.5024,
+      "step": 6695
+    },
+    {
+      "epoch": 0.2579403272377286,
+      "grad_norm": 1.393096923828125,
+      "learning_rate": 0.0001919209975345971,
+      "loss": 1.2555,
+      "step": 6700
+    },
+    {
+      "epoch": 0.25813282001924925,
+      "grad_norm": 1.5740808248519897,
+      "learning_rate": 0.00019190908595118907,
+      "loss": 1.2362,
+      "step": 6705
+    },
+    {
+      "epoch": 0.25832531280077,
+      "grad_norm": 1.3243273496627808,
+      "learning_rate": 0.00019189716596330395,
+      "loss": 1.2517,
+      "step": 6710
+    },
+    {
+      "epoch": 0.25851780558229065,
+      "grad_norm": 2.5867626667022705,
+      "learning_rate": 0.00019188523757203177,
+      "loss": 1.3509,
+      "step": 6715
+    },
+    {
+      "epoch": 0.2587102983638114,
+      "grad_norm": 1.450181484222412,
+      "learning_rate": 0.00019187330077846334,
+      "loss": 1.3451,
+      "step": 6720
+    },
+    {
+      "epoch": 0.25890279114533205,
+      "grad_norm": 1.4387754201889038,
+      "learning_rate": 0.0001918613555836901,
+      "loss": 1.2518,
+      "step": 6725
+    },
+    {
+      "epoch": 0.25909528392685277,
+      "grad_norm": 1.427882432937622,
+      "learning_rate": 0.00019184940198880448,
+      "loss": 1.235,
+      "step": 6730
+    },
+    {
+      "epoch": 0.25928777670837344,
+      "grad_norm": 1.060436487197876,
+      "learning_rate": 0.00019183743999489947,
+      "loss": 1.4583,
+      "step": 6735
+    },
+    {
+      "epoch": 0.2594802694898941,
+      "grad_norm": 1.0780494213104248,
+      "learning_rate": 0.00019182546960306893,
+      "loss": 1.1134,
+      "step": 6740
+    },
+    {
+      "epoch": 0.25967276227141484,
+      "grad_norm": 1.3795710802078247,
+      "learning_rate": 0.0001918134908144075,
+      "loss": 1.2979,
+      "step": 6745
+    },
+    {
+      "epoch": 0.2598652550529355,
+      "grad_norm": 2.0972957611083984,
+      "learning_rate": 0.00019180150363001051,
+      "loss": 1.6512,
+      "step": 6750
+    },
+    {
+      "epoch": 0.26005774783445623,
+      "grad_norm": 1.129204273223877,
+      "learning_rate": 0.00019178950805097416,
+      "loss": 1.2263,
+      "step": 6755
+    },
+    {
+      "epoch": 0.2602502406159769,
+      "grad_norm": 0.8816843628883362,
+      "learning_rate": 0.00019177750407839536,
+      "loss": 1.2265,
+      "step": 6760
+    },
+    {
+      "epoch": 0.26044273339749757,
+      "grad_norm": 1.5167860984802246,
+      "learning_rate": 0.00019176549171337178,
+      "loss": 1.226,
+      "step": 6765
+    },
+    {
+      "epoch": 0.2606352261790183,
+      "grad_norm": 1.329172968864441,
+      "learning_rate": 0.00019175347095700188,
+      "loss": 1.3375,
+      "step": 6770
+    },
+    {
+      "epoch": 0.26082771896053897,
+      "grad_norm": 1.8215051889419556,
+      "learning_rate": 0.00019174144181038485,
+      "loss": 1.2453,
+      "step": 6775
+    },
+    {
+      "epoch": 0.2610202117420597,
+      "grad_norm": 1.147878646850586,
+      "learning_rate": 0.00019172940427462072,
+      "loss": 1.3137,
+      "step": 6780
+    },
+    {
+      "epoch": 0.26121270452358036,
+      "grad_norm": 1.5783206224441528,
+      "learning_rate": 0.0001917173583508102,
+      "loss": 1.1803,
+      "step": 6785
+    },
+    {
+      "epoch": 0.26140519730510103,
+      "grad_norm": 1.7433182001113892,
+      "learning_rate": 0.00019170530404005485,
+      "loss": 1.171,
+      "step": 6790
+    },
+    {
+      "epoch": 0.26159769008662176,
+      "grad_norm": 1.5278960466384888,
+      "learning_rate": 0.0001916932413434569,
+      "loss": 1.2274,
+      "step": 6795
+    },
+    {
+      "epoch": 0.26179018286814243,
+      "grad_norm": 1.375710368156433,
+      "learning_rate": 0.00019168117026211948,
+      "loss": 1.241,
+      "step": 6800
+    },
+    {
+      "epoch": 0.26198267564966315,
+      "grad_norm": 2.146165370941162,
+      "learning_rate": 0.00019166909079714636,
+      "loss": 1.2778,
+      "step": 6805
+    },
+    {
+      "epoch": 0.2621751684311838,
+      "grad_norm": 1.7670506238937378,
+      "learning_rate": 0.00019165700294964216,
+      "loss": 1.3293,
+      "step": 6810
+    },
+    {
+      "epoch": 0.26236766121270455,
+      "grad_norm": 1.5492186546325684,
+      "learning_rate": 0.00019164490672071217,
+      "loss": 1.2808,
+      "step": 6815
+    },
+    {
+      "epoch": 0.2625601539942252,
+      "grad_norm": 1.4138727188110352,
+      "learning_rate": 0.00019163280211146257,
+      "loss": 1.2352,
+      "step": 6820
+    },
+    {
+      "epoch": 0.2627526467757459,
+      "grad_norm": 1.185674786567688,
+      "learning_rate": 0.00019162068912300024,
+      "loss": 1.1883,
+      "step": 6825
+    },
+    {
+      "epoch": 0.2629451395572666,
+      "grad_norm": 1.717349886894226,
+      "learning_rate": 0.0001916085677564328,
+      "loss": 1.1329,
+      "step": 6830
+    },
+    {
+      "epoch": 0.2631376323387873,
+      "grad_norm": 1.1391080617904663,
+      "learning_rate": 0.00019159643801286872,
+      "loss": 1.4104,
+      "step": 6835
+    },
+    {
+      "epoch": 0.263330125120308,
+      "grad_norm": 1.0915690660476685,
+      "learning_rate": 0.00019158429989341716,
+      "loss": 1.2813,
+      "step": 6840
+    },
+    {
+      "epoch": 0.2635226179018287,
+      "grad_norm": 1.120492696762085,
+      "learning_rate": 0.000191572153399188,
+      "loss": 1.2669,
+      "step": 6845
+    },
+    {
+      "epoch": 0.26371511068334935,
+      "grad_norm": 1.0648150444030762,
+      "learning_rate": 0.0001915599985312921,
+      "loss": 1.2581,
+      "step": 6850
+    },
+    {
+      "epoch": 0.2639076034648701,
+      "grad_norm": 1.7173513174057007,
+      "learning_rate": 0.0001915478352908408,
+      "loss": 1.2081,
+      "step": 6855
+    },
+    {
+      "epoch": 0.26410009624639075,
+      "grad_norm": 1.3801002502441406,
+      "learning_rate": 0.00019153566367894644,
+      "loss": 1.4625,
+      "step": 6860
+    },
+    {
+      "epoch": 0.2642925890279115,
+      "grad_norm": 2.5863940715789795,
+      "learning_rate": 0.00019152348369672203,
+      "loss": 1.4777,
+      "step": 6865
+    },
+    {
+      "epoch": 0.26448508180943214,
+      "grad_norm": 1.5995707511901855,
+      "learning_rate": 0.0001915112953452813,
+      "loss": 1.2089,
+      "step": 6870
+    },
+    {
+      "epoch": 0.2646775745909528,
+      "grad_norm": 1.2661023139953613,
+      "learning_rate": 0.0001914990986257388,
+      "loss": 1.1937,
+      "step": 6875
+    },
+    {
+      "epoch": 0.26487006737247354,
+      "grad_norm": 1.4782702922821045,
+      "learning_rate": 0.00019148689353920987,
+      "loss": 1.2462,
+      "step": 6880
+    },
+    {
+      "epoch": 0.2650625601539942,
+      "grad_norm": 1.8557063341140747,
+      "learning_rate": 0.0001914746800868106,
+      "loss": 1.425,
+      "step": 6885
+    },
+    {
+      "epoch": 0.26525505293551493,
+      "grad_norm": 2.825359582901001,
+      "learning_rate": 0.00019146245826965775,
+      "loss": 1.3628,
+      "step": 6890
+    },
+    {
+      "epoch": 0.2654475457170356,
+      "grad_norm": 1.7262654304504395,
+      "learning_rate": 0.00019145022808886902,
+      "loss": 1.2902,
+      "step": 6895
+    },
+    {
+      "epoch": 0.26564003849855633,
+      "grad_norm": 0.9676236510276794,
+      "learning_rate": 0.00019143798954556268,
+      "loss": 1.3342,
+      "step": 6900
+    },
+    {
+      "epoch": 0.265832531280077,
+      "grad_norm": 1.4607850313186646,
+      "learning_rate": 0.00019142574264085797,
+      "loss": 1.3084,
+      "step": 6905
+    },
+    {
+      "epoch": 0.26602502406159767,
+      "grad_norm": 2.181511878967285,
+      "learning_rate": 0.0001914134873758747,
+      "loss": 1.1746,
+      "step": 6910
+    },
+    {
+      "epoch": 0.2662175168431184,
+      "grad_norm": 1.4534579515457153,
+      "learning_rate": 0.00019140122375173362,
+      "loss": 1.3071,
+      "step": 6915
+    },
+    {
+      "epoch": 0.26641000962463907,
+      "grad_norm": 1.607039213180542,
+      "learning_rate": 0.00019138895176955604,
+      "loss": 1.2883,
+      "step": 6920
+    },
+    {
+      "epoch": 0.2666025024061598,
+      "grad_norm": 0.9929762482643127,
+      "learning_rate": 0.00019137667143046425,
+      "loss": 1.1122,
+      "step": 6925
+    },
+    {
+      "epoch": 0.26679499518768046,
+      "grad_norm": 1.6732393503189087,
+      "learning_rate": 0.0001913643827355812,
+      "loss": 1.149,
+      "step": 6930
+    },
+    {
+      "epoch": 0.26698748796920113,
+      "grad_norm": 1.3785120248794556,
+      "learning_rate": 0.0001913520856860305,
+      "loss": 1.3759,
+      "step": 6935
+    },
+    {
+      "epoch": 0.26717998075072186,
+      "grad_norm": 1.8252770900726318,
+      "learning_rate": 0.0001913397802829368,
+      "loss": 1.2633,
+      "step": 6940
+    },
+    {
+      "epoch": 0.2673724735322425,
+      "grad_norm": 1.6789536476135254,
+      "learning_rate": 0.0001913274665274252,
+      "loss": 1.2741,
+      "step": 6945
+    },
+    {
+      "epoch": 0.26756496631376325,
+      "grad_norm": 2.0153861045837402,
+      "learning_rate": 0.00019131514442062184,
+      "loss": 1.196,
+      "step": 6950
+    },
+    {
+      "epoch": 0.2677574590952839,
+      "grad_norm": 1.0000704526901245,
+      "learning_rate": 0.0001913028139636534,
+      "loss": 1.1872,
+      "step": 6955
+    },
+    {
+      "epoch": 0.2679499518768046,
+      "grad_norm": 1.2803142070770264,
+      "learning_rate": 0.00019129047515764743,
+      "loss": 1.2655,
+      "step": 6960
+    },
+    {
+      "epoch": 0.2681424446583253,
+      "grad_norm": 0.9827659130096436,
+      "learning_rate": 0.00019127812800373225,
+      "loss": 1.3503,
+      "step": 6965
+    },
+    {
+      "epoch": 0.268334937439846,
+      "grad_norm": 1.3766348361968994,
+      "learning_rate": 0.00019126577250303697,
+      "loss": 1.2851,
+      "step": 6970
+    },
+    {
+      "epoch": 0.2685274302213667,
+      "grad_norm": 2.285708427429199,
+      "learning_rate": 0.00019125340865669134,
+      "loss": 1.3247,
+      "step": 6975
+    },
+    {
+      "epoch": 0.2687199230028874,
+      "grad_norm": 1.79937744140625,
+      "learning_rate": 0.000191241036465826,
+      "loss": 1.0306,
+      "step": 6980
+    },
+    {
+      "epoch": 0.2689124157844081,
+      "grad_norm": 1.6062885522842407,
+      "learning_rate": 0.0001912286559315723,
+      "loss": 1.2068,
+      "step": 6985
+    },
+    {
+      "epoch": 0.2691049085659288,
+      "grad_norm": 1.9590744972229004,
+      "learning_rate": 0.00019121626705506233,
+      "loss": 1.2195,
+      "step": 6990
+    },
+    {
+      "epoch": 0.26929740134744945,
+      "grad_norm": 1.366186261177063,
+      "learning_rate": 0.000191203869837429,
+      "loss": 1.1627,
+      "step": 6995
+    },
+    {
+      "epoch": 0.2694898941289702,
+      "grad_norm": 0.9655261635780334,
+      "learning_rate": 0.00019119146427980593,
+      "loss": 1.053,
+      "step": 7000
+    },
+    {
+      "epoch": 0.26968238691049085,
+      "grad_norm": 1.4636151790618896,
+      "learning_rate": 0.00019117905038332756,
+      "loss": 1.0954,
+      "step": 7005
+    },
+    {
+      "epoch": 0.26987487969201157,
+      "grad_norm": 1.4435783624649048,
+      "learning_rate": 0.00019116662814912903,
+      "loss": 1.2102,
+      "step": 7010
+    },
+    {
+      "epoch": 0.27006737247353224,
+      "grad_norm": 0.9880768060684204,
+      "learning_rate": 0.00019115419757834628,
+      "loss": 1.0698,
+      "step": 7015
+    },
+    {
+      "epoch": 0.2702598652550529,
+      "grad_norm": 1.516515851020813,
+      "learning_rate": 0.000191141758672116,
+      "loss": 1.3894,
+      "step": 7020
+    },
+    {
+      "epoch": 0.27045235803657364,
+      "grad_norm": 2.1763806343078613,
+      "learning_rate": 0.00019112931143157563,
+      "loss": 1.3794,
+      "step": 7025
+    },
+    {
+      "epoch": 0.2706448508180943,
+      "grad_norm": 1.2275705337524414,
+      "learning_rate": 0.00019111685585786344,
+      "loss": 1.2897,
+      "step": 7030
+    },
+    {
+      "epoch": 0.27083734359961503,
+      "grad_norm": 0.966526985168457,
+      "learning_rate": 0.00019110439195211835,
+      "loss": 1.2112,
+      "step": 7035
+    },
+    {
+      "epoch": 0.2710298363811357,
+      "grad_norm": 1.251911997795105,
+      "learning_rate": 0.00019109191971548016,
+      "loss": 1.2481,
+      "step": 7040
+    },
+    {
+      "epoch": 0.27122232916265643,
+      "grad_norm": 2.3555140495300293,
+      "learning_rate": 0.0001910794391490893,
+      "loss": 1.3372,
+      "step": 7045
+    },
+    {
+      "epoch": 0.2714148219441771,
+      "grad_norm": 1.229268193244934,
+      "learning_rate": 0.0001910669502540871,
+      "loss": 1.4362,
+      "step": 7050
+    },
+    {
+      "epoch": 0.27160731472569777,
+      "grad_norm": 1.2356593608856201,
+      "learning_rate": 0.00019105445303161555,
+      "loss": 1.379,
+      "step": 7055
+    },
+    {
+      "epoch": 0.2717998075072185,
+      "grad_norm": 1.910232424736023,
+      "learning_rate": 0.00019104194748281747,
+      "loss": 1.2902,
+      "step": 7060
+    },
+    {
+      "epoch": 0.27199230028873916,
+      "grad_norm": 1.9058904647827148,
+      "learning_rate": 0.0001910294336088364,
+      "loss": 1.3313,
+      "step": 7065
+    },
+    {
+      "epoch": 0.2721847930702599,
+      "grad_norm": 0.8631892800331116,
+      "learning_rate": 0.0001910169114108166,
+      "loss": 1.2843,
+      "step": 7070
+    },
+    {
+      "epoch": 0.27237728585178056,
+      "grad_norm": 1.2212119102478027,
+      "learning_rate": 0.0001910043808899032,
+      "loss": 1.2588,
+      "step": 7075
+    },
+    {
+      "epoch": 0.27256977863330123,
+      "grad_norm": 2.3140738010406494,
+      "learning_rate": 0.00019099184204724202,
+      "loss": 1.1781,
+      "step": 7080
+    },
+    {
+      "epoch": 0.27276227141482196,
+      "grad_norm": 1.0162906646728516,
+      "learning_rate": 0.00019097929488397965,
+      "loss": 1.3433,
+      "step": 7085
+    },
+    {
+      "epoch": 0.2729547641963426,
+      "grad_norm": 1.719766616821289,
+      "learning_rate": 0.00019096673940126343,
+      "loss": 1.1469,
+      "step": 7090
+    },
+    {
+      "epoch": 0.27314725697786335,
+      "grad_norm": 1.5173147916793823,
+      "learning_rate": 0.00019095417560024153,
+      "loss": 1.1663,
+      "step": 7095
+    },
+    {
+      "epoch": 0.273339749759384,
+      "grad_norm": 2.1228654384613037,
+      "learning_rate": 0.00019094160348206277,
+      "loss": 1.3433,
+      "step": 7100
+    },
+    {
+      "epoch": 0.2735322425409047,
+      "grad_norm": 1.3896198272705078,
+      "learning_rate": 0.00019092902304787679,
+      "loss": 1.1782,
+      "step": 7105
+    },
+    {
+      "epoch": 0.2737247353224254,
+      "grad_norm": 1.6935322284698486,
+      "learning_rate": 0.00019091643429883402,
+      "loss": 1.1867,
+      "step": 7110
+    },
+    {
+      "epoch": 0.2739172281039461,
+      "grad_norm": 1.5454139709472656,
+      "learning_rate": 0.00019090383723608558,
+      "loss": 1.3938,
+      "step": 7115
+    },
+    {
+      "epoch": 0.2741097208854668,
+      "grad_norm": 1.1493245363235474,
+      "learning_rate": 0.00019089123186078342,
+      "loss": 1.2127,
+      "step": 7120
+    },
+    {
+      "epoch": 0.2743022136669875,
+      "grad_norm": 1.7321335077285767,
+      "learning_rate": 0.00019087861817408021,
+      "loss": 1.3068,
+      "step": 7125
+    },
+    {
+      "epoch": 0.2744947064485082,
+      "grad_norm": 1.7654987573623657,
+      "learning_rate": 0.00019086599617712936,
+      "loss": 1.3236,
+      "step": 7130
+    },
+    {
+      "epoch": 0.2746871992300289,
+      "grad_norm": 1.0047959089279175,
+      "learning_rate": 0.0001908533658710851,
+      "loss": 1.404,
+      "step": 7135
+    },
+    {
+      "epoch": 0.27487969201154955,
+      "grad_norm": 1.9708582162857056,
+      "learning_rate": 0.0001908407272571024,
+      "loss": 1.2387,
+      "step": 7140
+    },
+    {
+      "epoch": 0.2750721847930703,
+      "grad_norm": 2.097369432449341,
+      "learning_rate": 0.00019082808033633696,
+      "loss": 1.189,
+      "step": 7145
+    },
+    {
+      "epoch": 0.27526467757459094,
+      "grad_norm": 1.1789932250976562,
+      "learning_rate": 0.00019081542510994523,
+      "loss": 1.4815,
+      "step": 7150
+    },
+    {
+      "epoch": 0.27545717035611167,
+      "grad_norm": 1.7205069065093994,
+      "learning_rate": 0.00019080276157908447,
+      "loss": 1.2906,
+      "step": 7155
+    },
+    {
+      "epoch": 0.27564966313763234,
+      "grad_norm": 1.7320606708526611,
+      "learning_rate": 0.0001907900897449127,
+      "loss": 1.339,
+      "step": 7160
+    },
+    {
+      "epoch": 0.275842155919153,
+      "grad_norm": 2.100649356842041,
+      "learning_rate": 0.00019077740960858863,
+      "loss": 1.3145,
+      "step": 7165
+    },
+    {
+      "epoch": 0.27603464870067373,
+      "grad_norm": 1.9302312135696411,
+      "learning_rate": 0.00019076472117127182,
+      "loss": 1.3082,
+      "step": 7170
+    },
+    {
+      "epoch": 0.2762271414821944,
+      "grad_norm": 0.5863549113273621,
+      "learning_rate": 0.0001907520244341225,
+      "loss": 1.0183,
+      "step": 7175
+    },
+    {
+      "epoch": 0.27641963426371513,
+      "grad_norm": 1.0428977012634277,
+      "learning_rate": 0.00019073931939830174,
+      "loss": 1.2488,
+      "step": 7180
+    },
+    {
+      "epoch": 0.2766121270452358,
+      "grad_norm": 1.1643081903457642,
+      "learning_rate": 0.0001907266060649713,
+      "loss": 1.476,
+      "step": 7185
+    },
+    {
+      "epoch": 0.27680461982675647,
+      "grad_norm": 1.0771207809448242,
+      "learning_rate": 0.00019071388443529376,
+      "loss": 1.3134,
+      "step": 7190
+    },
+    {
+      "epoch": 0.2769971126082772,
+      "grad_norm": 1.9787309169769287,
+      "learning_rate": 0.00019070115451043238,
+      "loss": 1.3884,
+      "step": 7195
+    },
+    {
+      "epoch": 0.27718960538979787,
+      "grad_norm": 2.095546245574951,
+      "learning_rate": 0.0001906884162915513,
+      "loss": 1.1221,
+      "step": 7200
+    },
+    {
+      "epoch": 0.2773820981713186,
+      "grad_norm": 2.0389225482940674,
+      "learning_rate": 0.00019067566977981528,
+      "loss": 1.0463,
+      "step": 7205
+    },
+    {
+      "epoch": 0.27757459095283926,
+      "grad_norm": 0.9991855621337891,
+      "learning_rate": 0.00019066291497638993,
+      "loss": 1.341,
+      "step": 7210
+    },
+    {
+      "epoch": 0.27776708373436,
+      "grad_norm": 1.411401391029358,
+      "learning_rate": 0.0001906501518824416,
+      "loss": 1.434,
+      "step": 7215
+    },
+    {
+      "epoch": 0.27795957651588066,
+      "grad_norm": 1.61775803565979,
+      "learning_rate": 0.0001906373804991374,
+      "loss": 1.1553,
+      "step": 7220
+    },
+    {
+      "epoch": 0.2781520692974013,
+      "grad_norm": 2.546022653579712,
+      "learning_rate": 0.00019062460082764515,
+      "loss": 1.2496,
+      "step": 7225
+    },
+    {
+      "epoch": 0.27834456207892205,
+      "grad_norm": 1.2731270790100098,
+      "learning_rate": 0.00019061181286913348,
+      "loss": 1.3236,
+      "step": 7230
+    },
+    {
+      "epoch": 0.2785370548604427,
+      "grad_norm": 1.0163904428482056,
+      "learning_rate": 0.00019059901662477177,
+      "loss": 1.2854,
+      "step": 7235
+    },
+    {
+      "epoch": 0.27872954764196345,
+      "grad_norm": 1.0653849840164185,
+      "learning_rate": 0.0001905862120957302,
+      "loss": 1.6351,
+      "step": 7240
+    },
+    {
+      "epoch": 0.2789220404234841,
+      "grad_norm": 1.081264853477478,
+      "learning_rate": 0.00019057339928317958,
+      "loss": 1.2466,
+      "step": 7245
+    },
+    {
+      "epoch": 0.2791145332050048,
+      "grad_norm": 1.3285462856292725,
+      "learning_rate": 0.00019056057818829156,
+      "loss": 1.2087,
+      "step": 7250
+    },
+    {
+      "epoch": 0.2793070259865255,
+      "grad_norm": 1.067254900932312,
+      "learning_rate": 0.0001905477488122386,
+      "loss": 1.3877,
+      "step": 7255
+    },
+    {
+      "epoch": 0.2794995187680462,
+      "grad_norm": 0.9383085370063782,
+      "learning_rate": 0.0001905349111561938,
+      "loss": 1.0643,
+      "step": 7260
+    },
+    {
+      "epoch": 0.2796920115495669,
+      "grad_norm": 2.7797493934631348,
+      "learning_rate": 0.00019052206522133117,
+      "loss": 1.3828,
+      "step": 7265
+    },
+    {
+      "epoch": 0.2798845043310876,
+      "grad_norm": 1.410261631011963,
+      "learning_rate": 0.0001905092110088253,
+      "loss": 1.3019,
+      "step": 7270
+    },
+    {
+      "epoch": 0.28007699711260825,
+      "grad_norm": 2.313541889190674,
+      "learning_rate": 0.0001904963485198517,
+      "loss": 1.2058,
+      "step": 7275
+    },
+    {
+      "epoch": 0.280269489894129,
+      "grad_norm": 1.4474842548370361,
+      "learning_rate": 0.00019048347775558645,
+      "loss": 1.2187,
+      "step": 7280
+    },
+    {
+      "epoch": 0.28046198267564965,
+      "grad_norm": 1.5846171379089355,
+      "learning_rate": 0.00019047059871720657,
+      "loss": 1.0326,
+      "step": 7285
+    },
+    {
+      "epoch": 0.28065447545717037,
+      "grad_norm": 1.1118413209915161,
+      "learning_rate": 0.00019045771140588976,
+      "loss": 1.2881,
+      "step": 7290
+    },
+    {
+      "epoch": 0.28084696823869104,
+      "grad_norm": 2.5894134044647217,
+      "learning_rate": 0.00019044481582281448,
+      "loss": 1.3885,
+      "step": 7295
+    },
+    {
+      "epoch": 0.28103946102021177,
+      "grad_norm": 1.6019679307937622,
+      "learning_rate": 0.00019043191196915993,
+      "loss": 1.3247,
+      "step": 7300
+    },
+    {
+      "epoch": 0.28123195380173244,
+      "grad_norm": 1.3384417295455933,
+      "learning_rate": 0.00019041899984610606,
+      "loss": 1.346,
+      "step": 7305
+    },
+    {
+      "epoch": 0.2814244465832531,
+      "grad_norm": 1.3584142923355103,
+      "learning_rate": 0.00019040607945483367,
+      "loss": 1.3418,
+      "step": 7310
+    },
+    {
+      "epoch": 0.28161693936477383,
+      "grad_norm": 1.379162073135376,
+      "learning_rate": 0.00019039315079652416,
+      "loss": 1.293,
+      "step": 7315
+    },
+    {
+      "epoch": 0.2818094321462945,
+      "grad_norm": 1.499841570854187,
+      "learning_rate": 0.00019038021387235982,
+      "loss": 1.2131,
+      "step": 7320
+    },
+    {
+      "epoch": 0.28200192492781523,
+      "grad_norm": 1.9813991785049438,
+      "learning_rate": 0.00019036726868352366,
+      "loss": 1.3282,
+      "step": 7325
+    },
+    {
+      "epoch": 0.2821944177093359,
+      "grad_norm": 1.404096245765686,
+      "learning_rate": 0.00019035431523119938,
+      "loss": 1.2238,
+      "step": 7330
+    },
+    {
+      "epoch": 0.28238691049085657,
+      "grad_norm": 1.1089609861373901,
+      "learning_rate": 0.00019034135351657152,
+      "loss": 1.1705,
+      "step": 7335
+    },
+    {
+      "epoch": 0.2825794032723773,
+      "grad_norm": 1.0567266941070557,
+      "learning_rate": 0.00019032838354082535,
+      "loss": 1.1228,
+      "step": 7340
+    },
+    {
+      "epoch": 0.28277189605389796,
+      "grad_norm": 1.2407151460647583,
+      "learning_rate": 0.00019031540530514685,
+      "loss": 1.1154,
+      "step": 7345
+    },
+    {
+      "epoch": 0.2829643888354187,
+      "grad_norm": 1.3094842433929443,
+      "learning_rate": 0.00019030241881072283,
+      "loss": 1.2251,
+      "step": 7350
+    },
+    {
+      "epoch": 0.28315688161693936,
+      "grad_norm": 0.9434831142425537,
+      "learning_rate": 0.00019028942405874082,
+      "loss": 1.0644,
+      "step": 7355
+    },
+    {
+      "epoch": 0.2833493743984601,
+      "grad_norm": 1.107958197593689,
+      "learning_rate": 0.0001902764210503891,
+      "loss": 1.295,
+      "step": 7360
+    },
+    {
+      "epoch": 0.28354186717998076,
+      "grad_norm": 1.4402803182601929,
+      "learning_rate": 0.00019026340978685666,
+      "loss": 1.3339,
+      "step": 7365
+    },
+    {
+      "epoch": 0.2837343599615014,
+      "grad_norm": 1.1564158201217651,
+      "learning_rate": 0.0001902503902693334,
+      "loss": 1.252,
+      "step": 7370
+    },
+    {
+      "epoch": 0.28392685274302215,
+      "grad_norm": 1.8258494138717651,
+      "learning_rate": 0.00019023736249900973,
+      "loss": 1.3495,
+      "step": 7375
+    },
+    {
+      "epoch": 0.2841193455245428,
+      "grad_norm": 1.1436362266540527,
+      "learning_rate": 0.00019022432647707708,
+      "loss": 1.4295,
+      "step": 7380
+    },
+    {
+      "epoch": 0.28431183830606355,
+      "grad_norm": 1.1649361848831177,
+      "learning_rate": 0.00019021128220472747,
+      "loss": 1.3438,
+      "step": 7385
+    },
+    {
+      "epoch": 0.2845043310875842,
+      "grad_norm": 1.7044711112976074,
+      "learning_rate": 0.00019019822968315364,
+      "loss": 1.2735,
+      "step": 7390
+    },
+    {
+      "epoch": 0.2846968238691049,
+      "grad_norm": 0.8998376727104187,
+      "learning_rate": 0.00019018516891354924,
+      "loss": 1.1817,
+      "step": 7395
+    },
+    {
+      "epoch": 0.2848893166506256,
+      "grad_norm": 1.8617538213729858,
+      "learning_rate": 0.00019017209989710855,
+      "loss": 1.3235,
+      "step": 7400
+    },
+    {
+      "epoch": 0.2850818094321463,
+      "grad_norm": 0.9981639981269836,
+      "learning_rate": 0.00019015902263502669,
+      "loss": 1.1171,
+      "step": 7405
+    },
+    {
+      "epoch": 0.285274302213667,
+      "grad_norm": 0.935457170009613,
+      "learning_rate": 0.00019014593712849944,
+      "loss": 1.1926,
+      "step": 7410
+    },
+    {
+      "epoch": 0.2854667949951877,
+      "grad_norm": 1.3465532064437866,
+      "learning_rate": 0.00019013284337872341,
+      "loss": 1.5102,
+      "step": 7415
+    },
+    {
+      "epoch": 0.28565928777670835,
+      "grad_norm": 1.3213337659835815,
+      "learning_rate": 0.00019011974138689595,
+      "loss": 1.2597,
+      "step": 7420
+    },
+    {
+      "epoch": 0.2858517805582291,
+      "grad_norm": 1.655229091644287,
+      "learning_rate": 0.0001901066311542151,
+      "loss": 1.0345,
+      "step": 7425
+    },
+    {
+      "epoch": 0.28604427333974974,
+      "grad_norm": 1.0165207386016846,
+      "learning_rate": 0.00019009351268187974,
+      "loss": 1.2854,
+      "step": 7430
+    },
+    {
+      "epoch": 0.28623676612127047,
+      "grad_norm": 1.3425116539001465,
+      "learning_rate": 0.00019008038597108945,
+      "loss": 1.381,
+      "step": 7435
+    },
+    {
+      "epoch": 0.28642925890279114,
+      "grad_norm": 1.2017732858657837,
+      "learning_rate": 0.0001900672510230446,
+      "loss": 1.2171,
+      "step": 7440
+    },
+    {
+      "epoch": 0.28662175168431187,
+      "grad_norm": 1.4958349466323853,
+      "learning_rate": 0.00019005410783894626,
+      "loss": 1.3524,
+      "step": 7445
+    },
+    {
+      "epoch": 0.28681424446583254,
+      "grad_norm": 1.1109000444412231,
+      "learning_rate": 0.00019004095641999636,
+      "loss": 1.2046,
+      "step": 7450
+    },
+    {
+      "epoch": 0.2870067372473532,
+      "grad_norm": 1.5347834825515747,
+      "learning_rate": 0.00019002779676739745,
+      "loss": 1.2295,
+      "step": 7455
+    },
+    {
+      "epoch": 0.28719923002887393,
+      "grad_norm": 1.5204600095748901,
+      "learning_rate": 0.00019001462888235286,
+      "loss": 1.0319,
+      "step": 7460
+    },
+    {
+      "epoch": 0.2873917228103946,
+      "grad_norm": 2.0644850730895996,
+      "learning_rate": 0.00019000145276606677,
+      "loss": 1.2371,
+      "step": 7465
+    },
+    {
+      "epoch": 0.2875842155919153,
+      "grad_norm": 1.5903024673461914,
+      "learning_rate": 0.00018998826841974407,
+      "loss": 1.3781,
+      "step": 7470
+    },
+    {
+      "epoch": 0.287776708373436,
+      "grad_norm": 1.045086145401001,
+      "learning_rate": 0.00018997507584459032,
+      "loss": 1.0918,
+      "step": 7475
+    },
+    {
+      "epoch": 0.28796920115495667,
+      "grad_norm": 1.499211311340332,
+      "learning_rate": 0.0001899618750418119,
+      "loss": 1.2377,
+      "step": 7480
+    },
+    {
+      "epoch": 0.2881616939364774,
+      "grad_norm": 1.2885223627090454,
+      "learning_rate": 0.00018994866601261597,
+      "loss": 1.2936,
+      "step": 7485
+    },
+    {
+      "epoch": 0.28835418671799806,
+      "grad_norm": 1.9687073230743408,
+      "learning_rate": 0.00018993544875821035,
+      "loss": 1.2043,
+      "step": 7490
+    },
+    {
+      "epoch": 0.2885466794995188,
+      "grad_norm": 0.9758608937263489,
+      "learning_rate": 0.00018992222327980375,
+      "loss": 1.0775,
+      "step": 7495
+    },
+    {
+      "epoch": 0.28873917228103946,
+      "grad_norm": 1.4256442785263062,
+      "learning_rate": 0.00018990898957860547,
+      "loss": 1.2608,
+      "step": 7500
+    },
+    {
+      "epoch": 0.28893166506256013,
+      "grad_norm": 1.267991304397583,
+      "learning_rate": 0.00018989574765582572,
+      "loss": 1.3826,
+      "step": 7505
+    },
+    {
+      "epoch": 0.28912415784408085,
+      "grad_norm": 1.4104158878326416,
+      "learning_rate": 0.00018988249751267534,
+      "loss": 1.1589,
+      "step": 7510
+    },
+    {
+      "epoch": 0.2893166506256015,
+      "grad_norm": 0.9540778994560242,
+      "learning_rate": 0.000189869239150366,
+      "loss": 1.196,
+      "step": 7515
+    },
+    {
+      "epoch": 0.28950914340712225,
+      "grad_norm": 4.175881385803223,
+      "learning_rate": 0.00018985597257011006,
+      "loss": 1.3408,
+      "step": 7520
+    },
+    {
+      "epoch": 0.2897016361886429,
+      "grad_norm": 1.79558527469635,
+      "learning_rate": 0.00018984269777312066,
+      "loss": 1.0596,
+      "step": 7525
+    },
+    {
+      "epoch": 0.28989412897016364,
+      "grad_norm": 1.5449460744857788,
+      "learning_rate": 0.0001898294147606117,
+      "loss": 1.2628,
+      "step": 7530
+    },
+    {
+      "epoch": 0.2900866217516843,
+      "grad_norm": 1.5056041479110718,
+      "learning_rate": 0.00018981612353379784,
+      "loss": 1.132,
+      "step": 7535
+    },
+    {
+      "epoch": 0.290279114533205,
+      "grad_norm": 1.7045507431030273,
+      "learning_rate": 0.00018980282409389445,
+      "loss": 1.1663,
+      "step": 7540
+    },
+    {
+      "epoch": 0.2904716073147257,
+      "grad_norm": 1.203892469406128,
+      "learning_rate": 0.00018978951644211766,
+      "loss": 1.1168,
+      "step": 7545
+    },
+    {
+      "epoch": 0.2906641000962464,
+      "grad_norm": 0.9239038228988647,
+      "learning_rate": 0.0001897762005796844,
+      "loss": 1.3328,
+      "step": 7550
+    },
+    {
+      "epoch": 0.2908565928777671,
+      "grad_norm": 1.3521167039871216,
+      "learning_rate": 0.00018976287650781238,
+      "loss": 1.2766,
+      "step": 7555
+    },
+    {
+      "epoch": 0.2910490856592878,
+      "grad_norm": 1.3824992179870605,
+      "learning_rate": 0.00018974954422771987,
+      "loss": 1.0153,
+      "step": 7560
+    },
+    {
+      "epoch": 0.29124157844080845,
+      "grad_norm": 0.9183006286621094,
+      "learning_rate": 0.00018973620374062607,
+      "loss": 1.0558,
+      "step": 7565
+    },
+    {
+      "epoch": 0.29143407122232917,
+      "grad_norm": 1.7128045558929443,
+      "learning_rate": 0.0001897228550477509,
+      "loss": 1.316,
+      "step": 7570
+    },
+    {
+      "epoch": 0.29162656400384984,
+      "grad_norm": 1.3998011350631714,
+      "learning_rate": 0.000189709498150315,
+      "loss": 1.2359,
+      "step": 7575
+    },
+    {
+      "epoch": 0.29181905678537057,
+      "grad_norm": 1.2251836061477661,
+      "learning_rate": 0.00018969613304953975,
+      "loss": 1.2464,
+      "step": 7580
+    },
+    {
+      "epoch": 0.29201154956689124,
+      "grad_norm": 1.3014954328536987,
+      "learning_rate": 0.00018968275974664734,
+      "loss": 1.0624,
+      "step": 7585
+    },
+    {
+      "epoch": 0.2922040423484119,
+      "grad_norm": 1.8785862922668457,
+      "learning_rate": 0.00018966937824286062,
+      "loss": 1.3491,
+      "step": 7590
+    },
+    {
+      "epoch": 0.29239653512993263,
+      "grad_norm": 1.0634154081344604,
+      "learning_rate": 0.00018965598853940327,
+      "loss": 1.1012,
+      "step": 7595
+    },
+    {
+      "epoch": 0.2925890279114533,
+      "grad_norm": 0.9114715456962585,
+      "learning_rate": 0.00018964259063749967,
+      "loss": 1.3738,
+      "step": 7600
+    },
+    {
+      "epoch": 0.29278152069297403,
+      "grad_norm": 1.9063506126403809,
+      "learning_rate": 0.00018962918453837503,
+      "loss": 1.1161,
+      "step": 7605
+    },
+    {
+      "epoch": 0.2929740134744947,
+      "grad_norm": 1.12264084815979,
+      "learning_rate": 0.00018961577024325516,
+      "loss": 1.4191,
+      "step": 7610
+    },
+    {
+      "epoch": 0.2931665062560154,
+      "grad_norm": 1.4751306772232056,
+      "learning_rate": 0.00018960234775336677,
+      "loss": 1.2153,
+      "step": 7615
+    },
+    {
+      "epoch": 0.2933589990375361,
+      "grad_norm": 1.4374860525131226,
+      "learning_rate": 0.00018958891706993724,
+      "loss": 1.1999,
+      "step": 7620
+    },
+    {
+      "epoch": 0.29355149181905676,
+      "grad_norm": 1.5792250633239746,
+      "learning_rate": 0.0001895754781941947,
+      "loss": 1.266,
+      "step": 7625
+    },
+    {
+      "epoch": 0.2937439846005775,
+      "grad_norm": 1.3390734195709229,
+      "learning_rate": 0.00018956203112736807,
+      "loss": 1.2703,
+      "step": 7630
+    },
+    {
+      "epoch": 0.29393647738209816,
+      "grad_norm": 1.2470978498458862,
+      "learning_rate": 0.00018954857587068701,
+      "loss": 1.0415,
+      "step": 7635
+    },
+    {
+      "epoch": 0.2941289701636189,
+      "grad_norm": 1.6102235317230225,
+      "learning_rate": 0.00018953511242538186,
+      "loss": 1.2707,
+      "step": 7640
+    },
+    {
+      "epoch": 0.29432146294513956,
+      "grad_norm": 1.334554672241211,
+      "learning_rate": 0.0001895216407926838,
+      "loss": 1.2672,
+      "step": 7645
+    },
+    {
+      "epoch": 0.2945139557266602,
+      "grad_norm": 1.2881218194961548,
+      "learning_rate": 0.00018950816097382475,
+      "loss": 1.1641,
+      "step": 7650
+    },
+    {
+      "epoch": 0.29470644850818095,
+      "grad_norm": 1.2150179147720337,
+      "learning_rate": 0.00018949467297003732,
+      "loss": 1.2636,
+      "step": 7655
+    },
+    {
+      "epoch": 0.2948989412897016,
+      "grad_norm": 1.1388130187988281,
+      "learning_rate": 0.00018948117678255485,
+      "loss": 1.2354,
+      "step": 7660
+    },
+    {
+      "epoch": 0.29509143407122235,
+      "grad_norm": 0.785776674747467,
+      "learning_rate": 0.0001894676724126115,
+      "loss": 1.2621,
+      "step": 7665
+    },
+    {
+      "epoch": 0.295283926852743,
+      "grad_norm": 1.005819320678711,
+      "learning_rate": 0.00018945415986144223,
+      "loss": 1.1175,
+      "step": 7670
+    },
+    {
+      "epoch": 0.29547641963426374,
+      "grad_norm": 2.2892065048217773,
+      "learning_rate": 0.00018944063913028264,
+      "loss": 1.148,
+      "step": 7675
+    },
+    {
+      "epoch": 0.2956689124157844,
+      "grad_norm": 2.0920302867889404,
+      "learning_rate": 0.00018942711022036903,
+      "loss": 1.178,
+      "step": 7680
+    },
+    {
+      "epoch": 0.2958614051973051,
+      "grad_norm": 1.228538155555725,
+      "learning_rate": 0.00018941357313293863,
+      "loss": 1.2499,
+      "step": 7685
+    },
+    {
+      "epoch": 0.2960538979788258,
+      "grad_norm": 1.8671079874038696,
+      "learning_rate": 0.00018940002786922925,
+      "loss": 1.2361,
+      "step": 7690
+    },
+    {
+      "epoch": 0.2962463907603465,
+      "grad_norm": 1.7283247709274292,
+      "learning_rate": 0.00018938647443047957,
+      "loss": 1.2695,
+      "step": 7695
+    },
+    {
+      "epoch": 0.2964388835418672,
+      "grad_norm": 1.9629713296890259,
+      "learning_rate": 0.0001893729128179289,
+      "loss": 1.5226,
+      "step": 7700
+    },
+    {
+      "epoch": 0.2966313763233879,
+      "grad_norm": 1.2868784666061401,
+      "learning_rate": 0.00018935934303281743,
+      "loss": 1.3237,
+      "step": 7705
+    },
+    {
+      "epoch": 0.29682386910490854,
+      "grad_norm": 1.3925827741622925,
+      "learning_rate": 0.000189345765076386,
+      "loss": 1.4075,
+      "step": 7710
+    },
+    {
+      "epoch": 0.29701636188642927,
+      "grad_norm": 1.1560002565383911,
+      "learning_rate": 0.0001893321789498762,
+      "loss": 1.3212,
+      "step": 7715
+    },
+    {
+      "epoch": 0.29720885466794994,
+      "grad_norm": 1.207263708114624,
+      "learning_rate": 0.0001893185846545304,
+      "loss": 1.3106,
+      "step": 7720
+    },
+    {
+      "epoch": 0.29740134744947067,
+      "grad_norm": Infinity,
+      "learning_rate": 0.00018930770333752716,
+      "loss": 1.5499,
+      "step": 7725
+    },
+    {
+      "epoch": 0.29759384023099134,
+      "grad_norm": 1.2437909841537476,
+      "learning_rate": 0.0001892940943414097,
+      "loss": 1.2797,
+      "step": 7730
+    },
+    {
+      "epoch": 0.297786333012512,
+      "grad_norm": 0.8919286131858826,
+      "learning_rate": 0.00018928047717993885,
+      "loss": 1.1074,
+      "step": 7735
+    },
+    {
+      "epoch": 0.29797882579403273,
+      "grad_norm": 1.219995379447937,
+      "learning_rate": 0.00018926685185435978,
+      "loss": 1.0856,
+      "step": 7740
+    },
+    {
+      "epoch": 0.2981713185755534,
+      "grad_norm": 0.8819857835769653,
+      "learning_rate": 0.00018925321836591846,
+      "loss": 1.3518,
+      "step": 7745
+    },
+    {
+      "epoch": 0.2983638113570741,
+      "grad_norm": 1.2268033027648926,
+      "learning_rate": 0.00018923957671586154,
+      "loss": 1.3786,
+      "step": 7750
+    },
+    {
+      "epoch": 0.2985563041385948,
+      "grad_norm": 0.9456066489219666,
+      "learning_rate": 0.0001892259269054365,
+      "loss": 1.3424,
+      "step": 7755
+    },
+    {
+      "epoch": 0.2987487969201155,
+      "grad_norm": 1.5397047996520996,
+      "learning_rate": 0.0001892122689358915,
+      "loss": 1.3618,
+      "step": 7760
+    },
+    {
+      "epoch": 0.2989412897016362,
+      "grad_norm": 1.3874872922897339,
+      "learning_rate": 0.0001891986028084755,
+      "loss": 1.2717,
+      "step": 7765
+    },
+    {
+      "epoch": 0.29913378248315686,
+      "grad_norm": 1.1725342273712158,
+      "learning_rate": 0.00018918492852443817,
+      "loss": 1.4347,
+      "step": 7770
+    },
+    {
+      "epoch": 0.2993262752646776,
+      "grad_norm": 1.2135777473449707,
+      "learning_rate": 0.0001891712460850299,
+      "loss": 1.1892,
+      "step": 7775
+    },
+    {
+      "epoch": 0.29951876804619826,
+      "grad_norm": 1.549715280532837,
+      "learning_rate": 0.00018915755549150188,
+      "loss": 1.2041,
+      "step": 7780
+    },
+    {
+      "epoch": 0.299711260827719,
+      "grad_norm": 0.9927541613578796,
+      "learning_rate": 0.00018914385674510605,
+      "loss": 1.2198,
+      "step": 7785
+    },
+    {
+      "epoch": 0.29990375360923965,
+      "grad_norm": 1.3314557075500488,
+      "learning_rate": 0.00018913014984709502,
+      "loss": 1.1805,
+      "step": 7790
+    },
+    {
+      "epoch": 0.3000962463907603,
+      "grad_norm": 1.4021222591400146,
+      "learning_rate": 0.00018911643479872225,
+      "loss": 1.3375,
+      "step": 7795
+    },
+    {
+      "epoch": 0.30028873917228105,
+      "grad_norm": 1.0226534605026245,
+      "learning_rate": 0.00018910271160124182,
+      "loss": 1.329,
+      "step": 7800
+    },
+    {
+      "epoch": 0.3004812319538017,
+      "grad_norm": 0.8493847846984863,
+      "learning_rate": 0.0001890889802559087,
+      "loss": 1.4581,
+      "step": 7805
+    },
+    {
+      "epoch": 0.30067372473532245,
+      "grad_norm": 1.0437967777252197,
+      "learning_rate": 0.00018907524076397847,
+      "loss": 1.409,
+      "step": 7810
+    },
+    {
+      "epoch": 0.3008662175168431,
+      "grad_norm": 2.574695110321045,
+      "learning_rate": 0.00018906149312670754,
+      "loss": 1.3962,
+      "step": 7815
+    },
+    {
+      "epoch": 0.3010587102983638,
+      "grad_norm": 1.3757768869400024,
+      "learning_rate": 0.00018904773734535306,
+      "loss": 1.4098,
+      "step": 7820
+    },
+    {
+      "epoch": 0.3012512030798845,
+      "grad_norm": 1.2249635457992554,
+      "learning_rate": 0.0001890339734211729,
+      "loss": 1.1643,
+      "step": 7825
+    },
+    {
+      "epoch": 0.3014436958614052,
+      "grad_norm": 1.6329936981201172,
+      "learning_rate": 0.00018902020135542564,
+      "loss": 1.1914,
+      "step": 7830
+    },
+    {
+      "epoch": 0.3016361886429259,
+      "grad_norm": 1.0217385292053223,
+      "learning_rate": 0.0001890064211493707,
+      "loss": 1.043,
+      "step": 7835
+    },
+    {
+      "epoch": 0.3018286814244466,
+      "grad_norm": 1.448754072189331,
+      "learning_rate": 0.0001889926328042681,
+      "loss": 1.0953,
+      "step": 7840
+    },
+    {
+      "epoch": 0.3020211742059673,
+      "grad_norm": 0.9284221529960632,
+      "learning_rate": 0.00018897883632137881,
+      "loss": 1.321,
+      "step": 7845
+    },
+    {
+      "epoch": 0.30221366698748797,
+      "grad_norm": 1.4679608345031738,
+      "learning_rate": 0.00018896503170196435,
+      "loss": 1.2266,
+      "step": 7850
+    },
+    {
+      "epoch": 0.30240615976900864,
+      "grad_norm": 1.1148631572723389,
+      "learning_rate": 0.00018895121894728709,
+      "loss": 1.1666,
+      "step": 7855
+    },
+    {
+      "epoch": 0.30259865255052937,
+      "grad_norm": 1.0431932210922241,
+      "learning_rate": 0.00018893739805861008,
+      "loss": 1.2986,
+      "step": 7860
+    },
+    {
+      "epoch": 0.30279114533205004,
+      "grad_norm": 1.5691524744033813,
+      "learning_rate": 0.00018892356903719718,
+      "loss": 1.3928,
+      "step": 7865
+    },
+    {
+      "epoch": 0.30298363811357076,
+      "grad_norm": 1.6849128007888794,
+      "learning_rate": 0.000188909731884313,
+      "loss": 1.3569,
+      "step": 7870
+    },
+    {
+      "epoch": 0.30317613089509143,
+      "grad_norm": 1.1832456588745117,
+      "learning_rate": 0.00018889588660122276,
+      "loss": 1.2984,
+      "step": 7875
+    },
+    {
+      "epoch": 0.3033686236766121,
+      "grad_norm": 1.3270272016525269,
+      "learning_rate": 0.0001888820331891926,
+      "loss": 1.1498,
+      "step": 7880
+    },
+    {
+      "epoch": 0.30356111645813283,
+      "grad_norm": 1.6383373737335205,
+      "learning_rate": 0.0001888681716494893,
+      "loss": 1.4725,
+      "step": 7885
+    },
+    {
+      "epoch": 0.3037536092396535,
+      "grad_norm": 1.1068469285964966,
+      "learning_rate": 0.00018885430198338038,
+      "loss": 1.3326,
+      "step": 7890
+    },
+    {
+      "epoch": 0.3039461020211742,
+      "grad_norm": 1.8454192876815796,
+      "learning_rate": 0.00018884042419213412,
+      "loss": 1.2307,
+      "step": 7895
+    },
+    {
+      "epoch": 0.3041385948026949,
+      "grad_norm": 1.160762906074524,
+      "learning_rate": 0.00018882653827701965,
+      "loss": 1.6025,
+      "step": 7900
+    },
+    {
+      "epoch": 0.30433108758421556,
+      "grad_norm": 1.9325065612792969,
+      "learning_rate": 0.00018881264423930663,
+      "loss": 1.3071,
+      "step": 7905
+    },
+    {
+      "epoch": 0.3045235803657363,
+      "grad_norm": 0.9047966003417969,
+      "learning_rate": 0.00018879874208026562,
+      "loss": 1.3166,
+      "step": 7910
+    },
+    {
+      "epoch": 0.30471607314725696,
+      "grad_norm": 0.9753623008728027,
+      "learning_rate": 0.00018878483180116793,
+      "loss": 1.3702,
+      "step": 7915
+    },
+    {
+      "epoch": 0.3049085659287777,
+      "grad_norm": 1.210321307182312,
+      "learning_rate": 0.00018877091340328549,
+      "loss": 1.3775,
+      "step": 7920
+    },
+    {
+      "epoch": 0.30510105871029836,
+      "grad_norm": 1.287484049797058,
+      "learning_rate": 0.00018875698688789106,
+      "loss": 1.3534,
+      "step": 7925
+    },
+    {
+      "epoch": 0.3052935514918191,
+      "grad_norm": 1.1604797840118408,
+      "learning_rate": 0.00018874305225625814,
+      "loss": 1.2154,
+      "step": 7930
+    },
+    {
+      "epoch": 0.30548604427333975,
+      "grad_norm": 1.4771429300308228,
+      "learning_rate": 0.00018872910950966097,
+      "loss": 1.2438,
+      "step": 7935
+    },
+    {
+      "epoch": 0.3056785370548604,
+      "grad_norm": 1.1472980976104736,
+      "learning_rate": 0.00018871515864937453,
+      "loss": 1.0805,
+      "step": 7940
+    },
+    {
+      "epoch": 0.30587102983638115,
+      "grad_norm": 1.1015262603759766,
+      "learning_rate": 0.0001887011996766745,
+      "loss": 1.0594,
+      "step": 7945
+    },
+    {
+      "epoch": 0.3060635226179018,
+      "grad_norm": 1.5410771369934082,
+      "learning_rate": 0.00018868723259283737,
+      "loss": 1.2624,
+      "step": 7950
+    },
+    {
+      "epoch": 0.30625601539942254,
+      "grad_norm": 1.2014496326446533,
+      "learning_rate": 0.0001886732573991403,
+      "loss": 1.2259,
+      "step": 7955
+    },
+    {
+      "epoch": 0.3064485081809432,
+      "grad_norm": 2.0007143020629883,
+      "learning_rate": 0.0001886592740968612,
+      "loss": 1.3877,
+      "step": 7960
+    },
+    {
+      "epoch": 0.3066410009624639,
+      "grad_norm": 1.2455111742019653,
+      "learning_rate": 0.00018864528268727887,
+      "loss": 1.3254,
+      "step": 7965
+    },
+    {
+      "epoch": 0.3068334937439846,
+      "grad_norm": 1.2766424417495728,
+      "learning_rate": 0.00018863128317167264,
+      "loss": 1.2663,
+      "step": 7970
+    },
+    {
+      "epoch": 0.3070259865255053,
+      "grad_norm": 1.2151165008544922,
+      "learning_rate": 0.0001886172755513227,
+      "loss": 1.3597,
+      "step": 7975
+    },
+    {
+      "epoch": 0.307218479307026,
+      "grad_norm": 1.1774568557739258,
+      "learning_rate": 0.0001886032598275099,
+      "loss": 1.1311,
+      "step": 7980
+    },
+    {
+      "epoch": 0.3074109720885467,
+      "grad_norm": 1.43276846408844,
+      "learning_rate": 0.00018858923600151596,
+      "loss": 1.1123,
+      "step": 7985
+    },
+    {
+      "epoch": 0.3076034648700674,
+      "grad_norm": 1.691684603691101,
+      "learning_rate": 0.00018857520407462326,
+      "loss": 1.4089,
+      "step": 7990
+    },
+    {
+      "epoch": 0.30779595765158807,
+      "grad_norm": 1.7944872379302979,
+      "learning_rate": 0.00018856116404811487,
+      "loss": 1.3098,
+      "step": 7995
+    },
+    {
+      "epoch": 0.30798845043310874,
+      "grad_norm": 1.2894377708435059,
+      "learning_rate": 0.00018854711592327473,
+      "loss": 1.2128,
+      "step": 8000
+    },
+    {
+      "epoch": 0.30818094321462947,
+      "grad_norm": 2.52504301071167,
+      "learning_rate": 0.00018853305970138737,
+      "loss": 1.4214,
+      "step": 8005
+    },
+    {
+      "epoch": 0.30837343599615014,
+      "grad_norm": 1.0757540464401245,
+      "learning_rate": 0.0001885189953837382,
+      "loss": 1.1836,
+      "step": 8010
+    },
+    {
+      "epoch": 0.30856592877767086,
+      "grad_norm": 0.9253488183021545,
+      "learning_rate": 0.0001885049229716133,
+      "loss": 1.0756,
+      "step": 8015
+    },
+    {
+      "epoch": 0.30875842155919153,
+      "grad_norm": 2.042194366455078,
+      "learning_rate": 0.00018849084246629945,
+      "loss": 1.4017,
+      "step": 8020
+    },
+    {
+      "epoch": 0.3089509143407122,
+      "grad_norm": 1.750023603439331,
+      "learning_rate": 0.00018847675386908427,
+      "loss": 1.2352,
+      "step": 8025
+    },
+    {
+      "epoch": 0.3091434071222329,
+      "grad_norm": 1.5334408283233643,
+      "learning_rate": 0.00018846265718125605,
+      "loss": 1.3053,
+      "step": 8030
+    },
+    {
+      "epoch": 0.3093358999037536,
+      "grad_norm": 1.262428641319275,
+      "learning_rate": 0.00018844855240410387,
+      "loss": 1.28,
+      "step": 8035
+    },
+    {
+      "epoch": 0.3095283926852743,
+      "grad_norm": 1.1430000066757202,
+      "learning_rate": 0.0001884344395389175,
+      "loss": 1.2133,
+      "step": 8040
+    },
+    {
+      "epoch": 0.309720885466795,
+      "grad_norm": 1.792740821838379,
+      "learning_rate": 0.0001884203185869874,
+      "loss": 1.3004,
+      "step": 8045
+    },
+    {
+      "epoch": 0.30991337824831566,
+      "grad_norm": 1.7067112922668457,
+      "learning_rate": 0.00018840618954960495,
+      "loss": 1.4131,
+      "step": 8050
+    },
+    {
+      "epoch": 0.3101058710298364,
+      "grad_norm": 1.5428810119628906,
+      "learning_rate": 0.00018839205242806206,
+      "loss": 1.2361,
+      "step": 8055
+    },
+    {
+      "epoch": 0.31029836381135706,
+      "grad_norm": 1.078902244567871,
+      "learning_rate": 0.00018837790722365152,
+      "loss": 1.2126,
+      "step": 8060
+    },
+    {
+      "epoch": 0.3104908565928778,
+      "grad_norm": 1.5348985195159912,
+      "learning_rate": 0.00018836375393766684,
+      "loss": 1.2591,
+      "step": 8065
+    },
+    {
+      "epoch": 0.31068334937439845,
+      "grad_norm": 1.2026286125183105,
+      "learning_rate": 0.00018834959257140222,
+      "loss": 1.3059,
+      "step": 8070
+    },
+    {
+      "epoch": 0.3108758421559192,
+      "grad_norm": 1.3559043407440186,
+      "learning_rate": 0.0001883354231261526,
+      "loss": 1.2006,
+      "step": 8075
+    },
+    {
+      "epoch": 0.31106833493743985,
+      "grad_norm": 1.2358171939849854,
+      "learning_rate": 0.00018832124560321374,
+      "loss": 1.2656,
+      "step": 8080
+    },
+    {
+      "epoch": 0.3112608277189605,
+      "grad_norm": 1.720358967781067,
+      "learning_rate": 0.00018830706000388202,
+      "loss": 1.3493,
+      "step": 8085
+    },
+    {
+      "epoch": 0.31145332050048125,
+      "grad_norm": 1.4281798601150513,
+      "learning_rate": 0.00018829286632945463,
+      "loss": 1.1485,
+      "step": 8090
+    },
+    {
+      "epoch": 0.3116458132820019,
+      "grad_norm": 1.6174485683441162,
+      "learning_rate": 0.00018827866458122951,
+      "loss": 1.4384,
+      "step": 8095
+    },
+    {
+      "epoch": 0.31183830606352264,
+      "grad_norm": 1.0020065307617188,
+      "learning_rate": 0.00018826445476050532,
+      "loss": 1.0489,
+      "step": 8100
+    },
+    {
+      "epoch": 0.3120307988450433,
+      "grad_norm": 1.8663140535354614,
+      "learning_rate": 0.0001882502368685814,
+      "loss": 1.3252,
+      "step": 8105
+    },
+    {
+      "epoch": 0.312223291626564,
+      "grad_norm": 1.4404470920562744,
+      "learning_rate": 0.00018823601090675796,
+      "loss": 1.1452,
+      "step": 8110
+    },
+    {
+      "epoch": 0.3124157844080847,
+      "grad_norm": 1.3358442783355713,
+      "learning_rate": 0.00018822177687633583,
+      "loss": 1.1581,
+      "step": 8115
+    },
+    {
+      "epoch": 0.3126082771896054,
+      "grad_norm": 1.6938860416412354,
+      "learning_rate": 0.00018820753477861662,
+      "loss": 1.5378,
+      "step": 8120
+    },
+    {
+      "epoch": 0.3128007699711261,
+      "grad_norm": 1.1914762258529663,
+      "learning_rate": 0.00018819328461490268,
+      "loss": 1.172,
+      "step": 8125
+    },
+    {
+      "epoch": 0.3129932627526468,
+      "grad_norm": 2.0504634380340576,
+      "learning_rate": 0.0001881790263864971,
+      "loss": 1.2462,
+      "step": 8130
+    },
+    {
+      "epoch": 0.31318575553416744,
+      "grad_norm": 1.548021912574768,
+      "learning_rate": 0.00018816476009470367,
+      "loss": 1.271,
+      "step": 8135
+    },
+    {
+      "epoch": 0.31337824831568817,
+      "grad_norm": 1.2875434160232544,
+      "learning_rate": 0.00018815048574082698,
+      "loss": 1.2484,
+      "step": 8140
+    },
+    {
+      "epoch": 0.31357074109720884,
+      "grad_norm": 0.936850905418396,
+      "learning_rate": 0.00018813620332617227,
+      "loss": 1.2765,
+      "step": 8145
+    },
+    {
+      "epoch": 0.31376323387872956,
+      "grad_norm": 1.2823413610458374,
+      "learning_rate": 0.00018812191285204566,
+      "loss": 1.1859,
+      "step": 8150
+    },
+    {
+      "epoch": 0.31395572666025023,
+      "grad_norm": 2.052490472793579,
+      "learning_rate": 0.00018810761431975386,
+      "loss": 1.2033,
+      "step": 8155
+    },
+    {
+      "epoch": 0.31414821944177096,
+      "grad_norm": 2.4439830780029297,
+      "learning_rate": 0.00018809330773060442,
+      "loss": 1.3678,
+      "step": 8160
+    },
+    {
+      "epoch": 0.31434071222329163,
+      "grad_norm": 1.9978455305099487,
+      "learning_rate": 0.0001880789930859055,
+      "loss": 1.25,
+      "step": 8165
+    },
+    {
+      "epoch": 0.3145332050048123,
+      "grad_norm": 1.2606321573257446,
+      "learning_rate": 0.00018806467038696615,
+      "loss": 1.4966,
+      "step": 8170
+    },
+    {
+      "epoch": 0.314725697786333,
+      "grad_norm": 1.4588353633880615,
+      "learning_rate": 0.00018805033963509605,
+      "loss": 1.1843,
+      "step": 8175
+    },
+    {
+      "epoch": 0.3149181905678537,
+      "grad_norm": 2.8686156272888184,
+      "learning_rate": 0.00018803600083160574,
+      "loss": 1.3017,
+      "step": 8180
+    },
+    {
+      "epoch": 0.3151106833493744,
+      "grad_norm": 1.812328815460205,
+      "learning_rate": 0.00018802165397780626,
+      "loss": 1.4141,
+      "step": 8185
+    },
+    {
+      "epoch": 0.3153031761308951,
+      "grad_norm": 1.4686119556427002,
+      "learning_rate": 0.00018800729907500968,
+      "loss": 1.4522,
+      "step": 8190
+    },
+    {
+      "epoch": 0.31549566891241576,
+      "grad_norm": 1.766160249710083,
+      "learning_rate": 0.00018799293612452856,
+      "loss": 1.1501,
+      "step": 8195
+    },
+    {
+      "epoch": 0.3156881616939365,
+      "grad_norm": 1.5843030214309692,
+      "learning_rate": 0.00018797856512767634,
+      "loss": 1.2997,
+      "step": 8200
+    },
+    {
+      "epoch": 0.31588065447545716,
+      "grad_norm": 1.2028679847717285,
+      "learning_rate": 0.00018796418608576712,
+      "loss": 1.108,
+      "step": 8205
+    },
+    {
+      "epoch": 0.3160731472569779,
+      "grad_norm": 1.4626559019088745,
+      "learning_rate": 0.0001879497990001158,
+      "loss": 1.116,
+      "step": 8210
+    },
+    {
+      "epoch": 0.31626564003849855,
+      "grad_norm": 1.956745982170105,
+      "learning_rate": 0.000187935403872038,
+      "loss": 1.2741,
+      "step": 8215
+    },
+    {
+      "epoch": 0.3164581328200192,
+      "grad_norm": 1.1932622194290161,
+      "learning_rate": 0.00018792100070285002,
+      "loss": 1.1966,
+      "step": 8220
+    },
+    {
+      "epoch": 0.31665062560153995,
+      "grad_norm": 2.212184429168701,
+      "learning_rate": 0.00018790658949386892,
+      "loss": 1.1485,
+      "step": 8225
+    },
+    {
+      "epoch": 0.3168431183830606,
+      "grad_norm": 0.867708146572113,
+      "learning_rate": 0.00018789217024641256,
+      "loss": 1.2457,
+      "step": 8230
+    },
+    {
+      "epoch": 0.31703561116458134,
+      "grad_norm": 2.4929304122924805,
+      "learning_rate": 0.0001878777429617995,
+      "loss": 1.1819,
+      "step": 8235
+    },
+    {
+      "epoch": 0.317228103946102,
+      "grad_norm": 1.4232670068740845,
+      "learning_rate": 0.00018786330764134897,
+      "loss": 1.2189,
+      "step": 8240
+    },
+    {
+      "epoch": 0.31742059672762274,
+      "grad_norm": 1.8306447267532349,
+      "learning_rate": 0.00018784886428638094,
+      "loss": 1.2939,
+      "step": 8245
+    },
+    {
+      "epoch": 0.3176130895091434,
+      "grad_norm": 0.9103988409042358,
+      "learning_rate": 0.00018783441289821627,
+      "loss": 1.2982,
+      "step": 8250
+    },
+    {
+      "epoch": 0.3178055822906641,
+      "grad_norm": 1.08035409450531,
+      "learning_rate": 0.0001878199534781764,
+      "loss": 1.2777,
+      "step": 8255
+    },
+    {
+      "epoch": 0.3179980750721848,
+      "grad_norm": 1.1342133283615112,
+      "learning_rate": 0.0001878054860275835,
+      "loss": 1.1476,
+      "step": 8260
+    },
+    {
+      "epoch": 0.3181905678537055,
+      "grad_norm": 1.7727190256118774,
+      "learning_rate": 0.0001877910105477606,
+      "loss": 1.1887,
+      "step": 8265
+    },
+    {
+      "epoch": 0.3183830606352262,
+      "grad_norm": 2.5168001651763916,
+      "learning_rate": 0.0001877765270400313,
+      "loss": 1.0494,
+      "step": 8270
+    },
+    {
+      "epoch": 0.31857555341674687,
+      "grad_norm": 1.2397305965423584,
+      "learning_rate": 0.0001877620355057201,
+      "loss": 1.321,
+      "step": 8275
+    },
+    {
+      "epoch": 0.31876804619826754,
+      "grad_norm": 1.3002814054489136,
+      "learning_rate": 0.0001877475359461521,
+      "loss": 1.1543,
+      "step": 8280
+    },
+    {
+      "epoch": 0.31896053897978827,
+      "grad_norm": 1.5683960914611816,
+      "learning_rate": 0.00018773302836265322,
+      "loss": 1.1987,
+      "step": 8285
+    },
+    {
+      "epoch": 0.31915303176130894,
+      "grad_norm": 1.6934245824813843,
+      "learning_rate": 0.00018771851275655008,
+      "loss": 1.2946,
+      "step": 8290
+    },
+    {
+      "epoch": 0.31934552454282966,
+      "grad_norm": 1.4387637376785278,
+      "learning_rate": 0.00018770398912917004,
+      "loss": 1.2151,
+      "step": 8295
+    },
+    {
+      "epoch": 0.31953801732435033,
+      "grad_norm": 1.3155730962753296,
+      "learning_rate": 0.00018768945748184117,
+      "loss": 1.1692,
+      "step": 8300
+    },
+    {
+      "epoch": 0.31973051010587106,
+      "grad_norm": 1.039670467376709,
+      "learning_rate": 0.0001876749178158923,
+      "loss": 1.2783,
+      "step": 8305
+    },
+    {
+      "epoch": 0.3199230028873917,
+      "grad_norm": 1.1988794803619385,
+      "learning_rate": 0.00018766037013265302,
+      "loss": 1.1775,
+      "step": 8310
+    },
+    {
+      "epoch": 0.3201154956689124,
+      "grad_norm": 1.39814031124115,
+      "learning_rate": 0.00018764581443345355,
+      "loss": 1.2256,
+      "step": 8315
+    },
+    {
+      "epoch": 0.3203079884504331,
+      "grad_norm": 1.7934690713882446,
+      "learning_rate": 0.00018763125071962495,
+      "loss": 1.3505,
+      "step": 8320
+    },
+    {
+      "epoch": 0.3205004812319538,
+      "grad_norm": 1.5974578857421875,
+      "learning_rate": 0.00018761667899249899,
+      "loss": 1.1725,
+      "step": 8325
+    },
+    {
+      "epoch": 0.3206929740134745,
+      "grad_norm": 0.9480400085449219,
+      "learning_rate": 0.00018760209925340818,
+      "loss": 1.2059,
+      "step": 8330
+    },
+    {
+      "epoch": 0.3208854667949952,
+      "grad_norm": 1.9734187126159668,
+      "learning_rate": 0.00018758751150368564,
+      "loss": 1.2116,
+      "step": 8335
+    },
+    {
+      "epoch": 0.32107795957651586,
+      "grad_norm": 0.9984979033470154,
+      "learning_rate": 0.00018757291574466543,
+      "loss": 1.1347,
+      "step": 8340
+    },
+    {
+      "epoch": 0.3212704523580366,
+      "grad_norm": 0.96681147813797,
+      "learning_rate": 0.00018755831197768215,
+      "loss": 1.2824,
+      "step": 8345
+    },
+    {
+      "epoch": 0.32146294513955725,
+      "grad_norm": 1.5365724563598633,
+      "learning_rate": 0.00018754370020407127,
+      "loss": 1.3718,
+      "step": 8350
+    },
+    {
+      "epoch": 0.321655437921078,
+      "grad_norm": 1.6202696561813354,
+      "learning_rate": 0.00018752908042516897,
+      "loss": 1.3233,
+      "step": 8355
+    },
+    {
+      "epoch": 0.32184793070259865,
+      "grad_norm": 2.0272514820098877,
+      "learning_rate": 0.00018751445264231207,
+      "loss": 1.3406,
+      "step": 8360
+    },
+    {
+      "epoch": 0.3220404234841193,
+      "grad_norm": 1.1724604368209839,
+      "learning_rate": 0.0001874998168568382,
+      "loss": 1.2649,
+      "step": 8365
+    },
+    {
+      "epoch": 0.32223291626564005,
+      "grad_norm": 1.0908805131912231,
+      "learning_rate": 0.00018748517307008573,
+      "loss": 1.2924,
+      "step": 8370
+    },
+    {
+      "epoch": 0.3224254090471607,
+      "grad_norm": 1.0658169984817505,
+      "learning_rate": 0.0001874705212833937,
+      "loss": 1.1266,
+      "step": 8375
+    },
+    {
+      "epoch": 0.32261790182868144,
+      "grad_norm": 1.2267755270004272,
+      "learning_rate": 0.00018745586149810194,
+      "loss": 1.172,
+      "step": 8380
+    },
+    {
+      "epoch": 0.3228103946102021,
+      "grad_norm": 0.9808927178382874,
+      "learning_rate": 0.000187441193715551,
+      "loss": 1.1241,
+      "step": 8385
+    },
+    {
+      "epoch": 0.32300288739172284,
+      "grad_norm": 1.2251529693603516,
+      "learning_rate": 0.00018742651793708212,
+      "loss": 1.1649,
+      "step": 8390
+    },
+    {
+      "epoch": 0.3231953801732435,
+      "grad_norm": 1.7396290302276611,
+      "learning_rate": 0.00018741183416403734,
+      "loss": 1.173,
+      "step": 8395
+    },
+    {
+      "epoch": 0.3233878729547642,
+      "grad_norm": 1.1498087644577026,
+      "learning_rate": 0.00018739714239775936,
+      "loss": 1.266,
+      "step": 8400
+    },
+    {
+      "epoch": 0.3235803657362849,
+      "grad_norm": 0.9458256959915161,
+      "learning_rate": 0.0001873824426395917,
+      "loss": 1.1651,
+      "step": 8405
+    },
+    {
+      "epoch": 0.3237728585178056,
+      "grad_norm": 1.701441764831543,
+      "learning_rate": 0.00018736773489087845,
+      "loss": 1.4314,
+      "step": 8410
+    },
+    {
+      "epoch": 0.3239653512993263,
+      "grad_norm": 1.3168058395385742,
+      "learning_rate": 0.00018735301915296466,
+      "loss": 1.3837,
+      "step": 8415
+    },
+    {
+      "epoch": 0.32415784408084697,
+      "grad_norm": 1.2277673482894897,
+      "learning_rate": 0.0001873382954271959,
+      "loss": 1.2433,
+      "step": 8420
+    },
+    {
+      "epoch": 0.32435033686236764,
+      "grad_norm": 1.3443776369094849,
+      "learning_rate": 0.00018732356371491858,
+      "loss": 1.1514,
+      "step": 8425
+    },
+    {
+      "epoch": 0.32454282964388836,
+      "grad_norm": 1.3421462774276733,
+      "learning_rate": 0.00018730882401747984,
+      "loss": 1.2908,
+      "step": 8430
+    },
+    {
+      "epoch": 0.32473532242540903,
+      "grad_norm": 2.7043700218200684,
+      "learning_rate": 0.0001872940763362275,
+      "loss": 1.426,
+      "step": 8435
+    },
+    {
+      "epoch": 0.32492781520692976,
+      "grad_norm": 1.2363086938858032,
+      "learning_rate": 0.00018727932067251016,
+      "loss": 1.2172,
+      "step": 8440
+    },
+    {
+      "epoch": 0.32512030798845043,
+      "grad_norm": 1.7551484107971191,
+      "learning_rate": 0.00018726455702767713,
+      "loss": 1.2379,
+      "step": 8445
+    },
+    {
+      "epoch": 0.3253128007699711,
+      "grad_norm": 1.2935433387756348,
+      "learning_rate": 0.00018724978540307844,
+      "loss": 1.2109,
+      "step": 8450
+    },
+    {
+      "epoch": 0.3255052935514918,
+      "grad_norm": 1.723219871520996,
+      "learning_rate": 0.00018723500580006483,
+      "loss": 1.3996,
+      "step": 8455
+    },
+    {
+      "epoch": 0.3256977863330125,
+      "grad_norm": 1.1455639600753784,
+      "learning_rate": 0.0001872202182199878,
+      "loss": 1.1223,
+      "step": 8460
+    },
+    {
+      "epoch": 0.3258902791145332,
+      "grad_norm": 1.194926381111145,
+      "learning_rate": 0.0001872054226641996,
+      "loss": 1.3301,
+      "step": 8465
+    },
+    {
+      "epoch": 0.3260827718960539,
+      "grad_norm": 1.9672341346740723,
+      "learning_rate": 0.00018719061913405322,
+      "loss": 1.3884,
+      "step": 8470
+    },
+    {
+      "epoch": 0.3262752646775746,
+      "grad_norm": 1.5594457387924194,
+      "learning_rate": 0.0001871758076309023,
+      "loss": 1.1862,
+      "step": 8475
+    },
+    {
+      "epoch": 0.3264677574590953,
+      "grad_norm": 1.141787052154541,
+      "learning_rate": 0.0001871609881561012,
+      "loss": 1.2375,
+      "step": 8480
+    },
+    {
+      "epoch": 0.32666025024061596,
+      "grad_norm": 1.1914411783218384,
+      "learning_rate": 0.0001871461607110052,
+      "loss": 1.397,
+      "step": 8485
+    },
+    {
+      "epoch": 0.3268527430221367,
+      "grad_norm": 1.2841687202453613,
+      "learning_rate": 0.00018713132529697007,
+      "loss": 1.3052,
+      "step": 8490
+    },
+    {
+      "epoch": 0.32704523580365735,
+      "grad_norm": 2.2977144718170166,
+      "learning_rate": 0.0001871164819153524,
+      "loss": 1.2819,
+      "step": 8495
+    },
+    {
+      "epoch": 0.3272377285851781,
+      "grad_norm": 1.62446928024292,
+      "learning_rate": 0.00018710163056750957,
+      "loss": 1.1739,
+      "step": 8500
+    },
+    {
+      "epoch": 0.32743022136669875,
+      "grad_norm": 1.471348524093628,
+      "learning_rate": 0.00018708677125479963,
+      "loss": 1.0684,
+      "step": 8505
+    },
+    {
+      "epoch": 0.3276227141482194,
+      "grad_norm": 1.0703455209732056,
+      "learning_rate": 0.00018707190397858133,
+      "loss": 1.0832,
+      "step": 8510
+    },
+    {
+      "epoch": 0.32781520692974014,
+      "grad_norm": 1.3942466974258423,
+      "learning_rate": 0.00018705702874021425,
+      "loss": 1.1855,
+      "step": 8515
+    },
+    {
+      "epoch": 0.3280076997112608,
+      "grad_norm": 1.1790398359298706,
+      "learning_rate": 0.00018704214554105856,
+      "loss": 1.1459,
+      "step": 8520
+    },
+    {
+      "epoch": 0.32820019249278154,
+      "grad_norm": 1.2982394695281982,
+      "learning_rate": 0.00018702725438247527,
+      "loss": 1.2642,
+      "step": 8525
+    },
+    {
+      "epoch": 0.3283926852743022,
+      "grad_norm": 1.4757968187332153,
+      "learning_rate": 0.00018701235526582608,
+      "loss": 1.291,
+      "step": 8530
+    },
+    {
+      "epoch": 0.3285851780558229,
+      "grad_norm": 1.6837409734725952,
+      "learning_rate": 0.0001870004302436148,
+      "loss": 1.3796,
+      "step": 8535
+    },
+    {
+      "epoch": 0.3287776708373436,
+      "grad_norm": 1.1914480924606323,
+      "learning_rate": 0.00018698551680588075,
+      "loss": 1.2608,
+      "step": 8540
+    },
+    {
+      "epoch": 0.3289701636188643,
+      "grad_norm": 1.2581427097320557,
+      "learning_rate": 0.00018697059541389742,
+      "loss": 1.3011,
+      "step": 8545
+    },
+    {
+      "epoch": 0.329162656400385,
+      "grad_norm": 1.5642743110656738,
+      "learning_rate": 0.0001869556660690293,
+      "loss": 1.2273,
+      "step": 8550
+    },
+    {
+      "epoch": 0.32935514918190567,
+      "grad_norm": 1.621721863746643,
+      "learning_rate": 0.0001869407287726415,
+      "loss": 1.1648,
+      "step": 8555
+    },
+    {
+      "epoch": 0.3295476419634264,
+      "grad_norm": 0.9840386509895325,
+      "learning_rate": 0.00018692578352610002,
+      "loss": 1.2741,
+      "step": 8560
+    },
+    {
+      "epoch": 0.32974013474494707,
+      "grad_norm": 1.5852268934249878,
+      "learning_rate": 0.00018691083033077144,
+      "loss": 1.2913,
+      "step": 8565
+    },
+    {
+      "epoch": 0.32993262752646774,
+      "grad_norm": 1.280247688293457,
+      "learning_rate": 0.00018689586918802314,
+      "loss": 1.172,
+      "step": 8570
+    },
+    {
+      "epoch": 0.33012512030798846,
+      "grad_norm": 1.3940321207046509,
+      "learning_rate": 0.0001868809000992233,
+      "loss": 1.175,
+      "step": 8575
+    },
+    {
+      "epoch": 0.33031761308950913,
+      "grad_norm": 1.0753341913223267,
+      "learning_rate": 0.00018686592306574063,
+      "loss": 1.3922,
+      "step": 8580
+    },
+    {
+      "epoch": 0.33051010587102986,
+      "grad_norm": 1.5959515571594238,
+      "learning_rate": 0.00018685093808894476,
+      "loss": 1.2741,
+      "step": 8585
+    },
+    {
+      "epoch": 0.33070259865255053,
+      "grad_norm": 1.1567896604537964,
+      "learning_rate": 0.00018683594517020593,
+      "loss": 1.1325,
+      "step": 8590
+    },
+    {
+      "epoch": 0.3308950914340712,
+      "grad_norm": 1.202486276626587,
+      "learning_rate": 0.0001868209443108951,
+      "loss": 1.1915,
+      "step": 8595
+    },
+    {
+      "epoch": 0.3310875842155919,
+      "grad_norm": 1.6866669654846191,
+      "learning_rate": 0.00018680593551238412,
+      "loss": 1.2806,
+      "step": 8600
+    },
+    {
+      "epoch": 0.3312800769971126,
+      "grad_norm": 1.1932209730148315,
+      "learning_rate": 0.00018679091877604536,
+      "loss": 1.2254,
+      "step": 8605
+    },
+    {
+      "epoch": 0.3314725697786333,
+      "grad_norm": 1.5348761081695557,
+      "learning_rate": 0.000186775894103252,
+      "loss": 1.1519,
+      "step": 8610
+    },
+    {
+      "epoch": 0.331665062560154,
+      "grad_norm": 1.908500075340271,
+      "learning_rate": 0.00018676086149537792,
+      "loss": 1.3105,
+      "step": 8615
+    },
+    {
+      "epoch": 0.3318575553416747,
+      "grad_norm": 2.0427961349487305,
+      "learning_rate": 0.00018674582095379788,
+      "loss": 1.1415,
+      "step": 8620
+    },
+    {
+      "epoch": 0.3320500481231954,
+      "grad_norm": 1.0964915752410889,
+      "learning_rate": 0.00018673077247988707,
+      "loss": 1.2041,
+      "step": 8625
+    },
+    {
+      "epoch": 0.33224254090471605,
+      "grad_norm": 1.2229498624801636,
+      "learning_rate": 0.00018671571607502168,
+      "loss": 1.2975,
+      "step": 8630
+    },
+    {
+      "epoch": 0.3324350336862368,
+      "grad_norm": 1.3551470041275024,
+      "learning_rate": 0.00018670065174057854,
+      "loss": 1.1592,
+      "step": 8635
+    },
+    {
+      "epoch": 0.33262752646775745,
+      "grad_norm": 0.8810299634933472,
+      "learning_rate": 0.0001866855794779351,
+      "loss": 1.1414,
+      "step": 8640
+    },
+    {
+      "epoch": 0.3328200192492782,
+      "grad_norm": 1.5907199382781982,
+      "learning_rate": 0.00018667049928846967,
+      "loss": 1.2191,
+      "step": 8645
+    },
+    {
+      "epoch": 0.33301251203079885,
+      "grad_norm": 2.042478561401367,
+      "learning_rate": 0.0001866554111735612,
+      "loss": 1.1619,
+      "step": 8650
+    },
+    {
+      "epoch": 0.3332050048123195,
+      "grad_norm": 1.6686564683914185,
+      "learning_rate": 0.00018664031513458942,
+      "loss": 1.2534,
+      "step": 8655
+    },
+    {
+      "epoch": 0.33339749759384024,
+      "grad_norm": 1.7643070220947266,
+      "learning_rate": 0.0001866252111729348,
+      "loss": 1.2631,
+      "step": 8660
+    },
+    {
+      "epoch": 0.3335899903753609,
+      "grad_norm": 1.4883722066879272,
+      "learning_rate": 0.0001866100992899784,
+      "loss": 1.1786,
+      "step": 8665
+    },
+    {
+      "epoch": 0.33378248315688164,
+      "grad_norm": 0.9850770235061646,
+      "learning_rate": 0.00018659497948710218,
+      "loss": 1.4181,
+      "step": 8670
+    },
+    {
+      "epoch": 0.3339749759384023,
+      "grad_norm": 0.9056932926177979,
+      "learning_rate": 0.00018657985176568875,
+      "loss": 1.0365,
+      "step": 8675
+    },
+    {
+      "epoch": 0.334167468719923,
+      "grad_norm": 1.9456449747085571,
+      "learning_rate": 0.00018656471612712137,
+      "loss": 1.227,
+      "step": 8680
+    },
+    {
+      "epoch": 0.3343599615014437,
+      "grad_norm": 1.289870262145996,
+      "learning_rate": 0.00018654957257278415,
+      "loss": 1.32,
+      "step": 8685
+    },
+    {
+      "epoch": 0.3345524542829644,
+      "grad_norm": 1.048143744468689,
+      "learning_rate": 0.00018653442110406189,
+      "loss": 1.2123,
+      "step": 8690
+    },
+    {
+      "epoch": 0.3347449470644851,
+      "grad_norm": 1.1696733236312866,
+      "learning_rate": 0.00018651926172234004,
+      "loss": 1.0226,
+      "step": 8695
+    },
+    {
+      "epoch": 0.33493743984600577,
+      "grad_norm": 1.4806257486343384,
+      "learning_rate": 0.00018650409442900486,
+      "loss": 1.1715,
+      "step": 8700
+    },
+    {
+      "epoch": 0.3351299326275265,
+      "grad_norm": 1.525719404220581,
+      "learning_rate": 0.00018648891922544325,
+      "loss": 1.2037,
+      "step": 8705
+    },
+    {
+      "epoch": 0.33532242540904716,
+      "grad_norm": 1.3378442525863647,
+      "learning_rate": 0.00018647373611304293,
+      "loss": 1.2188,
+      "step": 8710
+    },
+    {
+      "epoch": 0.33551491819056783,
+      "grad_norm": 0.870988130569458,
+      "learning_rate": 0.00018645854509319226,
+      "loss": 1.2153,
+      "step": 8715
+    },
+    {
+      "epoch": 0.33570741097208856,
+      "grad_norm": 1.5496007204055786,
+      "learning_rate": 0.00018644334616728042,
+      "loss": 1.1974,
+      "step": 8720
+    },
+    {
+      "epoch": 0.33589990375360923,
+      "grad_norm": 1.0248416662216187,
+      "learning_rate": 0.00018642813933669717,
+      "loss": 1.2845,
+      "step": 8725
+    },
+    {
+      "epoch": 0.33609239653512996,
+      "grad_norm": 1.9984816312789917,
+      "learning_rate": 0.00018641292460283313,
+      "loss": 1.3144,
+      "step": 8730
+    },
+    {
+      "epoch": 0.3362848893166506,
+      "grad_norm": 1.3114112615585327,
+      "learning_rate": 0.00018639770196707955,
+      "loss": 1.209,
+      "step": 8735
+    },
+    {
+      "epoch": 0.3364773820981713,
+      "grad_norm": 1.1683485507965088,
+      "learning_rate": 0.00018638247143082848,
+      "loss": 1.2688,
+      "step": 8740
+    },
+    {
+      "epoch": 0.336669874879692,
+      "grad_norm": 1.507900595664978,
+      "learning_rate": 0.0001863672329954726,
+      "loss": 1.1325,
+      "step": 8745
+    },
+    {
+      "epoch": 0.3368623676612127,
+      "grad_norm": 1.3393852710723877,
+      "learning_rate": 0.00018635198666240542,
+      "loss": 1.1573,
+      "step": 8750
+    },
+    {
+      "epoch": 0.3370548604427334,
+      "grad_norm": 1.0203709602355957,
+      "learning_rate": 0.00018633673243302108,
+      "loss": 1.2922,
+      "step": 8755
+    },
+    {
+      "epoch": 0.3372473532242541,
+      "grad_norm": 0.8483877778053284,
+      "learning_rate": 0.00018632147030871448,
+      "loss": 1.2252,
+      "step": 8760
+    },
+    {
+      "epoch": 0.33743984600577476,
+      "grad_norm": 0.983748197555542,
+      "learning_rate": 0.00018630620029088125,
+      "loss": 1.2027,
+      "step": 8765
+    },
+    {
+      "epoch": 0.3376323387872955,
+      "grad_norm": 1.2489101886749268,
+      "learning_rate": 0.00018629092238091775,
+      "loss": 1.1962,
+      "step": 8770
+    },
+    {
+      "epoch": 0.33782483156881615,
+      "grad_norm": 1.4553676843643188,
+      "learning_rate": 0.000186275636580221,
+      "loss": 1.3698,
+      "step": 8775
+    },
+    {
+      "epoch": 0.3380173243503369,
+      "grad_norm": 0.9494854807853699,
+      "learning_rate": 0.0001862603428901888,
+      "loss": 1.25,
+      "step": 8780
+    },
+    {
+      "epoch": 0.33820981713185755,
+      "grad_norm": 0.8667522072792053,
+      "learning_rate": 0.00018624504131221968,
+      "loss": 1.222,
+      "step": 8785
+    },
+    {
+      "epoch": 0.3384023099133783,
+      "grad_norm": 1.4215630292892456,
+      "learning_rate": 0.00018622973184771285,
+      "loss": 1.2592,
+      "step": 8790
+    },
+    {
+      "epoch": 0.33859480269489894,
+      "grad_norm": 0.9913888573646545,
+      "learning_rate": 0.00018621441449806828,
+      "loss": 1.2904,
+      "step": 8795
+    },
+    {
+      "epoch": 0.3387872954764196,
+      "grad_norm": 0.9612273573875427,
+      "learning_rate": 0.00018619908926468664,
+      "loss": 1.24,
+      "step": 8800
+    },
+    {
+      "epoch": 0.33897978825794034,
+      "grad_norm": 1.656568169593811,
+      "learning_rate": 0.00018618375614896926,
+      "loss": 1.1763,
+      "step": 8805
+    },
+    {
+      "epoch": 0.339172281039461,
+      "grad_norm": 1.4496088027954102,
+      "learning_rate": 0.0001861684151523183,
+      "loss": 1.2045,
+      "step": 8810
+    },
+    {
+      "epoch": 0.33936477382098174,
+      "grad_norm": 1.3886058330535889,
+      "learning_rate": 0.0001861530662761366,
+      "loss": 1.3111,
+      "step": 8815
+    },
+    {
+      "epoch": 0.3395572666025024,
+      "grad_norm": 1.644887089729309,
+      "learning_rate": 0.0001861377095218277,
+      "loss": 1.3172,
+      "step": 8820
+    },
+    {
+      "epoch": 0.3397497593840231,
+      "grad_norm": 1.1925910711288452,
+      "learning_rate": 0.00018612234489079587,
+      "loss": 1.3268,
+      "step": 8825
+    },
+    {
+      "epoch": 0.3399422521655438,
+      "grad_norm": 1.1367309093475342,
+      "learning_rate": 0.0001861069723844461,
+      "loss": 1.1209,
+      "step": 8830
+    },
+    {
+      "epoch": 0.34013474494706447,
+      "grad_norm": 1.0649480819702148,
+      "learning_rate": 0.00018609159200418414,
+      "loss": 1.1514,
+      "step": 8835
+    },
+    {
+      "epoch": 0.3403272377285852,
+      "grad_norm": 1.1887884140014648,
+      "learning_rate": 0.00018607620375141637,
+      "loss": 1.1026,
+      "step": 8840
+    },
+    {
+      "epoch": 0.34051973051010587,
+      "grad_norm": 1.9125694036483765,
+      "learning_rate": 0.00018606080762754995,
+      "loss": 1.4718,
+      "step": 8845
+    },
+    {
+      "epoch": 0.34071222329162654,
+      "grad_norm": 1.1742594242095947,
+      "learning_rate": 0.00018604540363399282,
+      "loss": 1.3206,
+      "step": 8850
+    },
+    {
+      "epoch": 0.34090471607314726,
+      "grad_norm": 1.504146695137024,
+      "learning_rate": 0.0001860299917721535,
+      "loss": 1.1639,
+      "step": 8855
+    },
+    {
+      "epoch": 0.34109720885466793,
+      "grad_norm": 0.8869237899780273,
+      "learning_rate": 0.00018601457204344131,
+      "loss": 1.2674,
+      "step": 8860
+    },
+    {
+      "epoch": 0.34128970163618866,
+      "grad_norm": 0.8492304682731628,
+      "learning_rate": 0.00018599914444926636,
+      "loss": 1.2732,
+      "step": 8865
+    },
+    {
+      "epoch": 0.34148219441770933,
+      "grad_norm": 1.1681571006774902,
+      "learning_rate": 0.00018598370899103932,
+      "loss": 1.2995,
+      "step": 8870
+    },
+    {
+      "epoch": 0.34167468719923005,
+      "grad_norm": 1.6912837028503418,
+      "learning_rate": 0.00018596826567017166,
+      "loss": 1.3217,
+      "step": 8875
+    },
+    {
+      "epoch": 0.3418671799807507,
+      "grad_norm": 1.0427602529525757,
+      "learning_rate": 0.0001859528144880756,
+      "loss": 1.05,
+      "step": 8880
+    },
+    {
+      "epoch": 0.3420596727622714,
+      "grad_norm": 1.9644991159439087,
+      "learning_rate": 0.00018593735544616404,
+      "loss": 1.1087,
+      "step": 8885
+    },
+    {
+      "epoch": 0.3422521655437921,
+      "grad_norm": 1.966264247894287,
+      "learning_rate": 0.0001859218885458506,
+      "loss": 1.2221,
+      "step": 8890
+    },
+    {
+      "epoch": 0.3424446583253128,
+      "grad_norm": 1.9770557880401611,
+      "learning_rate": 0.00018590641378854965,
+      "loss": 1.2489,
+      "step": 8895
+    },
+    {
+      "epoch": 0.3426371511068335,
+      "grad_norm": 1.4175180196762085,
+      "learning_rate": 0.00018589093117567625,
+      "loss": 1.1292,
+      "step": 8900
+    },
+    {
+      "epoch": 0.3428296438883542,
+      "grad_norm": 1.066177487373352,
+      "learning_rate": 0.00018587544070864612,
+      "loss": 1.1182,
+      "step": 8905
+    },
+    {
+      "epoch": 0.34302213666987486,
+      "grad_norm": 2.6207172870635986,
+      "learning_rate": 0.00018585994238887586,
+      "loss": 1.1,
+      "step": 8910
+    },
+    {
+      "epoch": 0.3432146294513956,
+      "grad_norm": 1.6905888319015503,
+      "learning_rate": 0.0001858444362177826,
+      "loss": 1.3135,
+      "step": 8915
+    },
+    {
+      "epoch": 0.34340712223291625,
+      "grad_norm": 1.117883324623108,
+      "learning_rate": 0.00018582892219678435,
+      "loss": 1.3394,
+      "step": 8920
+    },
+    {
+      "epoch": 0.343599615014437,
+      "grad_norm": 1.549805760383606,
+      "learning_rate": 0.00018581340032729972,
+      "loss": 1.1957,
+      "step": 8925
+    },
+    {
+      "epoch": 0.34379210779595765,
+      "grad_norm": 1.165260672569275,
+      "learning_rate": 0.00018579787061074807,
+      "loss": 1.2406,
+      "step": 8930
+    },
+    {
+      "epoch": 0.34398460057747837,
+      "grad_norm": 1.1872533559799194,
+      "learning_rate": 0.00018578233304854952,
+      "loss": 1.1831,
+      "step": 8935
+    },
+    {
+      "epoch": 0.34417709335899904,
+      "grad_norm": 0.8727648854255676,
+      "learning_rate": 0.00018576678764212489,
+      "loss": 1.2645,
+      "step": 8940
+    },
+    {
+      "epoch": 0.3443695861405197,
+      "grad_norm": 1.1179304122924805,
+      "learning_rate": 0.00018575123439289567,
+      "loss": 1.297,
+      "step": 8945
+    },
+    {
+      "epoch": 0.34456207892204044,
+      "grad_norm": 1.9064927101135254,
+      "learning_rate": 0.0001857356733022841,
+      "loss": 1.3917,
+      "step": 8950
+    },
+    {
+      "epoch": 0.3447545717035611,
+      "grad_norm": 2.100154399871826,
+      "learning_rate": 0.00018572010437171315,
+      "loss": 1.1723,
+      "step": 8955
+    },
+    {
+      "epoch": 0.34494706448508183,
+      "grad_norm": 1.0105838775634766,
+      "learning_rate": 0.00018570452760260654,
+      "loss": 1.0851,
+      "step": 8960
+    },
+    {
+      "epoch": 0.3451395572666025,
+      "grad_norm": 1.760038137435913,
+      "learning_rate": 0.0001856889429963886,
+      "loss": 1.0612,
+      "step": 8965
+    },
+    {
+      "epoch": 0.3453320500481232,
+      "grad_norm": 1.5740501880645752,
+      "learning_rate": 0.00018567335055448444,
+      "loss": 1.117,
+      "step": 8970
+    },
+    {
+      "epoch": 0.3455245428296439,
+      "grad_norm": 1.4148597717285156,
+      "learning_rate": 0.00018565775027831993,
+      "loss": 1.2003,
+      "step": 8975
+    },
+    {
+      "epoch": 0.34571703561116457,
+      "grad_norm": 1.2243534326553345,
+      "learning_rate": 0.00018564214216932159,
+      "loss": 1.2106,
+      "step": 8980
+    },
+    {
+      "epoch": 0.3459095283926853,
+      "grad_norm": 1.3532603979110718,
+      "learning_rate": 0.00018562652622891666,
+      "loss": 1.1703,
+      "step": 8985
+    },
+    {
+      "epoch": 0.34610202117420596,
+      "grad_norm": 1.6701220273971558,
+      "learning_rate": 0.00018561090245853315,
+      "loss": 1.2409,
+      "step": 8990
+    },
+    {
+      "epoch": 0.34629451395572663,
+      "grad_norm": 1.6342322826385498,
+      "learning_rate": 0.00018559527085959968,
+      "loss": 1.2981,
+      "step": 8995
+    },
+    {
+      "epoch": 0.34648700673724736,
+      "grad_norm": 2.4354701042175293,
+      "learning_rate": 0.00018557963143354576,
+      "loss": 1.1021,
+      "step": 9000
+    },
+    {
+      "epoch": 0.34667949951876803,
+      "grad_norm": 1.5688186883926392,
+      "learning_rate": 0.00018556398418180146,
+      "loss": 1.2649,
+      "step": 9005
+    },
+    {
+      "epoch": 0.34687199230028876,
+      "grad_norm": 2.2158894538879395,
+      "learning_rate": 0.0001855483291057976,
+      "loss": 1.2335,
+      "step": 9010
+    },
+    {
+      "epoch": 0.3470644850818094,
+      "grad_norm": 1.7294437885284424,
+      "learning_rate": 0.00018553266620696573,
+      "loss": 1.3235,
+      "step": 9015
+    },
+    {
+      "epoch": 0.34725697786333015,
+      "grad_norm": 1.1023756265640259,
+      "learning_rate": 0.00018551699548673814,
+      "loss": 1.3515,
+      "step": 9020
+    },
+    {
+      "epoch": 0.3474494706448508,
+      "grad_norm": 1.4505863189697266,
+      "learning_rate": 0.00018550131694654784,
+      "loss": 1.3773,
+      "step": 9025
+    },
+    {
+      "epoch": 0.3476419634263715,
+      "grad_norm": 2.221957206726074,
+      "learning_rate": 0.00018548563058782847,
+      "loss": 1.0896,
+      "step": 9030
+    },
+    {
+      "epoch": 0.3478344562078922,
+      "grad_norm": 0.917010486125946,
+      "learning_rate": 0.0001854699364120145,
+      "loss": 1.1569,
+      "step": 9035
+    },
+    {
+      "epoch": 0.3480269489894129,
+      "grad_norm": 1.4631186723709106,
+      "learning_rate": 0.00018545423442054105,
+      "loss": 1.2169,
+      "step": 9040
+    },
+    {
+      "epoch": 0.3482194417709336,
+      "grad_norm": 1.0917268991470337,
+      "learning_rate": 0.0001854385246148439,
+      "loss": 1.2425,
+      "step": 9045
+    },
+    {
+      "epoch": 0.3484119345524543,
+      "grad_norm": 1.5985426902770996,
+      "learning_rate": 0.00018542280699635968,
+      "loss": 1.0944,
+      "step": 9050
+    },
+    {
+      "epoch": 0.34860442733397495,
+      "grad_norm": 1.5402495861053467,
+      "learning_rate": 0.0001854070815665256,
+      "loss": 1.1497,
+      "step": 9055
+    },
+    {
+      "epoch": 0.3487969201154957,
+      "grad_norm": 1.211295485496521,
+      "learning_rate": 0.00018539134832677972,
+      "loss": 1.0403,
+      "step": 9060
+    },
+    {
+      "epoch": 0.34898941289701635,
+      "grad_norm": 1.0569374561309814,
+      "learning_rate": 0.00018537560727856068,
+      "loss": 1.2886,
+      "step": 9065
+    },
+    {
+      "epoch": 0.3491819056785371,
+      "grad_norm": 1.550212025642395,
+      "learning_rate": 0.00018535985842330793,
+      "loss": 1.2654,
+      "step": 9070
+    },
+    {
+      "epoch": 0.34937439846005774,
+      "grad_norm": 1.7941083908081055,
+      "learning_rate": 0.00018534410176246154,
+      "loss": 1.2757,
+      "step": 9075
+    },
+    {
+      "epoch": 0.3495668912415784,
+      "grad_norm": 0.9004856944084167,
+      "learning_rate": 0.00018532833729746243,
+      "loss": 1.2045,
+      "step": 9080
+    },
+    {
+      "epoch": 0.34975938402309914,
+      "grad_norm": 0.9916037321090698,
+      "learning_rate": 0.00018531256502975216,
+      "loss": 1.1788,
+      "step": 9085
+    },
+    {
+      "epoch": 0.3499518768046198,
+      "grad_norm": 1.0524908304214478,
+      "learning_rate": 0.00018529678496077292,
+      "loss": 1.3298,
+      "step": 9090
+    },
+    {
+      "epoch": 0.35014436958614054,
+      "grad_norm": 2.7244019508361816,
+      "learning_rate": 0.00018528099709196774,
+      "loss": 1.3274,
+      "step": 9095
+    },
+    {
+      "epoch": 0.3503368623676612,
+      "grad_norm": 1.4286680221557617,
+      "learning_rate": 0.0001852652014247803,
+      "loss": 1.193,
+      "step": 9100
+    },
+    {
+      "epoch": 0.35052935514918193,
+      "grad_norm": 1.0943810939788818,
+      "learning_rate": 0.00018524939796065503,
+      "loss": 1.2953,
+      "step": 9105
+    },
+    {
+      "epoch": 0.3507218479307026,
+      "grad_norm": 1.1513092517852783,
+      "learning_rate": 0.00018523358670103704,
+      "loss": 1.3436,
+      "step": 9110
+    },
+    {
+      "epoch": 0.35091434071222327,
+      "grad_norm": 2.142829656600952,
+      "learning_rate": 0.00018521776764737218,
+      "loss": 1.2998,
+      "step": 9115
+    },
+    {
+      "epoch": 0.351106833493744,
+      "grad_norm": 0.9734616875648499,
+      "learning_rate": 0.00018520194080110699,
+      "loss": 1.2794,
+      "step": 9120
+    },
+    {
+      "epoch": 0.35129932627526467,
+      "grad_norm": 1.0793628692626953,
+      "learning_rate": 0.00018518610616368868,
+      "loss": 1.2574,
+      "step": 9125
+    },
+    {
+      "epoch": 0.3514918190567854,
+      "grad_norm": 2.409484386444092,
+      "learning_rate": 0.00018517026373656532,
+      "loss": 1.1601,
+      "step": 9130
+    },
+    {
+      "epoch": 0.35168431183830606,
+      "grad_norm": 1.1166318655014038,
+      "learning_rate": 0.0001851544135211855,
+      "loss": 1.2705,
+      "step": 9135
+    },
+    {
+      "epoch": 0.35187680461982673,
+      "grad_norm": 1.183131217956543,
+      "learning_rate": 0.0001851385555189987,
+      "loss": 1.132,
+      "step": 9140
+    },
+    {
+      "epoch": 0.35206929740134746,
+      "grad_norm": 1.3792176246643066,
+      "learning_rate": 0.00018512268973145497,
+      "loss": 1.1271,
+      "step": 9145
+    },
+    {
+      "epoch": 0.35226179018286813,
+      "grad_norm": 1.3978809118270874,
+      "learning_rate": 0.00018510681616000513,
+      "loss": 1.3828,
+      "step": 9150
+    },
+    {
+      "epoch": 0.35245428296438885,
+      "grad_norm": 1.0242118835449219,
+      "learning_rate": 0.00018509093480610078,
+      "loss": 1.1982,
+      "step": 9155
+    },
+    {
+      "epoch": 0.3526467757459095,
+      "grad_norm": 1.326621174812317,
+      "learning_rate": 0.00018507504567119408,
+      "loss": 1.0175,
+      "step": 9160
+    },
+    {
+      "epoch": 0.3528392685274302,
+      "grad_norm": 1.1905460357666016,
+      "learning_rate": 0.00018505914875673805,
+      "loss": 1.3367,
+      "step": 9165
+    },
+    {
+      "epoch": 0.3530317613089509,
+      "grad_norm": 1.5423171520233154,
+      "learning_rate": 0.0001850432440641863,
+      "loss": 1.1721,
+      "step": 9170
+    },
+    {
+      "epoch": 0.3532242540904716,
+      "grad_norm": 1.0577900409698486,
+      "learning_rate": 0.00018502733159499326,
+      "loss": 1.2173,
+      "step": 9175
+    },
+    {
+      "epoch": 0.3534167468719923,
+      "grad_norm": 0.8053417205810547,
+      "learning_rate": 0.000185011411350614,
+      "loss": 1.1492,
+      "step": 9180
+    },
+    {
+      "epoch": 0.353609239653513,
+      "grad_norm": 1.076053261756897,
+      "learning_rate": 0.0001849954833325043,
+      "loss": 1.2117,
+      "step": 9185
+    },
+    {
+      "epoch": 0.3538017324350337,
+      "grad_norm": 1.206359624862671,
+      "learning_rate": 0.0001849795475421207,
+      "loss": 1.1659,
+      "step": 9190
+    },
+    {
+      "epoch": 0.3539942252165544,
+      "grad_norm": 1.4652369022369385,
+      "learning_rate": 0.00018496360398092046,
+      "loss": 1.2605,
+      "step": 9195
+    },
+    {
+      "epoch": 0.35418671799807505,
+      "grad_norm": 1.158055067062378,
+      "learning_rate": 0.00018494765265036144,
+      "loss": 1.414,
+      "step": 9200
+    },
+    {
+      "epoch": 0.3543792107795958,
+      "grad_norm": 2.4634461402893066,
+      "learning_rate": 0.0001849316935519023,
+      "loss": 1.1982,
+      "step": 9205
+    },
+    {
+      "epoch": 0.35457170356111645,
+      "grad_norm": 1.875139594078064,
+      "learning_rate": 0.00018491572668700242,
+      "loss": 1.4133,
+      "step": 9210
+    },
+    {
+      "epoch": 0.3547641963426372,
+      "grad_norm": 1.0054875612258911,
+      "learning_rate": 0.00018489975205712185,
+      "loss": 1.2294,
+      "step": 9215
+    },
+    {
+      "epoch": 0.35495668912415784,
+      "grad_norm": 2.2620842456817627,
+      "learning_rate": 0.00018488376966372134,
+      "loss": 1.2672,
+      "step": 9220
+    },
+    {
+      "epoch": 0.3551491819056785,
+      "grad_norm": 1.584251880645752,
+      "learning_rate": 0.00018486777950826243,
+      "loss": 1.4366,
+      "step": 9225
+    },
+    {
+      "epoch": 0.35534167468719924,
+      "grad_norm": 1.6498923301696777,
+      "learning_rate": 0.00018485178159220725,
+      "loss": 1.3502,
+      "step": 9230
+    },
+    {
+      "epoch": 0.3555341674687199,
+      "grad_norm": 1.6700108051300049,
+      "learning_rate": 0.00018483577591701876,
+      "loss": 1.2462,
+      "step": 9235
+    },
+    {
+      "epoch": 0.35572666025024063,
+      "grad_norm": 1.6976680755615234,
+      "learning_rate": 0.00018481976248416052,
+      "loss": 1.4637,
+      "step": 9240
+    },
+    {
+      "epoch": 0.3559191530317613,
+      "grad_norm": 0.9686551094055176,
+      "learning_rate": 0.0001848037412950969,
+      "loss": 1.1902,
+      "step": 9245
+    },
+    {
+      "epoch": 0.35611164581328203,
+      "grad_norm": 1.2102336883544922,
+      "learning_rate": 0.00018478771235129292,
+      "loss": 1.586,
+      "step": 9250
+    },
+    {
+      "epoch": 0.3563041385948027,
+      "grad_norm": 1.7220674753189087,
+      "learning_rate": 0.0001847716756542143,
+      "loss": 1.2324,
+      "step": 9255
+    },
+    {
+      "epoch": 0.35649663137632337,
+      "grad_norm": 1.7433216571807861,
+      "learning_rate": 0.0001847556312053275,
+      "loss": 1.4454,
+      "step": 9260
+    },
+    {
+      "epoch": 0.3566891241578441,
+      "grad_norm": 0.9930455088615417,
+      "learning_rate": 0.0001847395790060997,
+      "loss": 1.1601,
+      "step": 9265
+    },
+    {
+      "epoch": 0.35688161693936477,
+      "grad_norm": 1.1169023513793945,
+      "learning_rate": 0.00018472351905799873,
+      "loss": 1.2534,
+      "step": 9270
+    },
+    {
+      "epoch": 0.3570741097208855,
+      "grad_norm": 1.238748550415039,
+      "learning_rate": 0.00018470745136249316,
+      "loss": 1.2174,
+      "step": 9275
+    },
+    {
+      "epoch": 0.35726660250240616,
+      "grad_norm": 2.130223035812378,
+      "learning_rate": 0.00018469137592105235,
+      "loss": 1.3975,
+      "step": 9280
+    },
+    {
+      "epoch": 0.35745909528392683,
+      "grad_norm": 1.4341787099838257,
+      "learning_rate": 0.0001846752927351462,
+      "loss": 1.1725,
+      "step": 9285
+    },
+    {
+      "epoch": 0.35765158806544756,
+      "grad_norm": 1.948145866394043,
+      "learning_rate": 0.00018465920180624548,
+      "loss": 1.2741,
+      "step": 9290
+    },
+    {
+      "epoch": 0.3578440808469682,
+      "grad_norm": 1.0314382314682007,
+      "learning_rate": 0.00018464310313582157,
+      "loss": 1.0998,
+      "step": 9295
+    },
+    {
+      "epoch": 0.35803657362848895,
+      "grad_norm": 1.0461472272872925,
+      "learning_rate": 0.0001846269967253466,
+      "loss": 1.1953,
+      "step": 9300
+    },
+    {
+      "epoch": 0.3582290664100096,
+      "grad_norm": 1.781084656715393,
+      "learning_rate": 0.00018461088257629334,
+      "loss": 1.3629,
+      "step": 9305
+    },
+    {
+      "epoch": 0.3584215591915303,
+      "grad_norm": 1.9082306623458862,
+      "learning_rate": 0.00018459476069013537,
+      "loss": 1.2675,
+      "step": 9310
+    },
+    {
+      "epoch": 0.358614051973051,
+      "grad_norm": 1.803348422050476,
+      "learning_rate": 0.00018457863106834693,
+      "loss": 1.2303,
+      "step": 9315
+    },
+    {
+      "epoch": 0.3588065447545717,
+      "grad_norm": 1.5346139669418335,
+      "learning_rate": 0.000184562493712403,
+      "loss": 1.3354,
+      "step": 9320
+    },
+    {
+      "epoch": 0.3589990375360924,
+      "grad_norm": 1.3731290102005005,
+      "learning_rate": 0.00018454634862377916,
+      "loss": 1.4874,
+      "step": 9325
+    },
+    {
+      "epoch": 0.3591915303176131,
+      "grad_norm": 1.186759352684021,
+      "learning_rate": 0.0001845301958039518,
+      "loss": 1.29,
+      "step": 9330
+    },
+    {
+      "epoch": 0.3593840230991338,
+      "grad_norm": 3.729174852371216,
+      "learning_rate": 0.00018451403525439802,
+      "loss": 1.2589,
+      "step": 9335
+    },
+    {
+      "epoch": 0.3595765158806545,
+      "grad_norm": 2.46051025390625,
+      "learning_rate": 0.00018449786697659554,
+      "loss": 1.1818,
+      "step": 9340
+    },
+    {
+      "epoch": 0.35976900866217515,
+      "grad_norm": 1.6652323007583618,
+      "learning_rate": 0.00018448169097202288,
+      "loss": 1.2719,
+      "step": 9345
+    },
+    {
+      "epoch": 0.3599615014436959,
+      "grad_norm": 1.375410556793213,
+      "learning_rate": 0.00018446550724215922,
+      "loss": 1.2687,
+      "step": 9350
+    },
+    {
+      "epoch": 0.36015399422521654,
+      "grad_norm": 1.9113675355911255,
+      "learning_rate": 0.00018444931578848447,
+      "loss": 1.2475,
+      "step": 9355
+    },
+    {
+      "epoch": 0.36034648700673727,
+      "grad_norm": 1.8949065208435059,
+      "learning_rate": 0.0001844331166124792,
+      "loss": 1.3439,
+      "step": 9360
+    },
+    {
+      "epoch": 0.36053897978825794,
+      "grad_norm": 1.0940630435943604,
+      "learning_rate": 0.00018441690971562476,
+      "loss": 1.203,
+      "step": 9365
+    },
+    {
+      "epoch": 0.3607314725697786,
+      "grad_norm": 1.2999101877212524,
+      "learning_rate": 0.00018440069509940315,
+      "loss": 1.2729,
+      "step": 9370
+    },
+    {
+      "epoch": 0.36092396535129934,
+      "grad_norm": 1.3675721883773804,
+      "learning_rate": 0.00018438447276529702,
+      "loss": 1.2024,
+      "step": 9375
+    },
+    {
+      "epoch": 0.36111645813282,
+      "grad_norm": 1.6651533842086792,
+      "learning_rate": 0.00018436824271478988,
+      "loss": 1.2235,
+      "step": 9380
+    },
+    {
+      "epoch": 0.36130895091434073,
+      "grad_norm": 2.16670823097229,
+      "learning_rate": 0.00018435200494936585,
+      "loss": 1.4486,
+      "step": 9385
+    },
+    {
+      "epoch": 0.3615014436958614,
+      "grad_norm": 1.3305730819702148,
+      "learning_rate": 0.00018433575947050972,
+      "loss": 1.2003,
+      "step": 9390
+    },
+    {
+      "epoch": 0.36169393647738207,
+      "grad_norm": 1.5913615226745605,
+      "learning_rate": 0.00018431950627970708,
+      "loss": 1.2722,
+      "step": 9395
+    },
+    {
+      "epoch": 0.3618864292589028,
+      "grad_norm": 0.9965779781341553,
+      "learning_rate": 0.00018430324537844415,
+      "loss": 1.0604,
+      "step": 9400
+    },
+    {
+      "epoch": 0.36207892204042347,
+      "grad_norm": 1.7614198923110962,
+      "learning_rate": 0.00018428697676820788,
+      "loss": 1.2734,
+      "step": 9405
+    },
+    {
+      "epoch": 0.3622714148219442,
+      "grad_norm": 1.190706491470337,
+      "learning_rate": 0.00018427070045048594,
+      "loss": 1.2309,
+      "step": 9410
+    },
+    {
+      "epoch": 0.36246390760346486,
+      "grad_norm": 1.1487165689468384,
+      "learning_rate": 0.00018425441642676667,
+      "loss": 1.2049,
+      "step": 9415
+    },
+    {
+      "epoch": 0.3626564003849856,
+      "grad_norm": 1.0437067747116089,
+      "learning_rate": 0.00018423812469853918,
+      "loss": 1.3632,
+      "step": 9420
+    },
+    {
+      "epoch": 0.36284889316650626,
+      "grad_norm": 1.7774686813354492,
+      "learning_rate": 0.00018422182526729318,
+      "loss": 1.1797,
+      "step": 9425
+    },
+    {
+      "epoch": 0.36304138594802693,
+      "grad_norm": 1.3748910427093506,
+      "learning_rate": 0.0001842055181345192,
+      "loss": 1.4438,
+      "step": 9430
+    },
+    {
+      "epoch": 0.36323387872954765,
+      "grad_norm": 0.891248881816864,
+      "learning_rate": 0.00018418920330170842,
+      "loss": 1.3017,
+      "step": 9435
+    },
+    {
+      "epoch": 0.3634263715110683,
+      "grad_norm": 1.5410393476486206,
+      "learning_rate": 0.00018417288077035267,
+      "loss": 1.2239,
+      "step": 9440
+    },
+    {
+      "epoch": 0.36361886429258905,
+      "grad_norm": 1.3638213872909546,
+      "learning_rate": 0.00018415655054194457,
+      "loss": 1.2245,
+      "step": 9445
+    },
+    {
+      "epoch": 0.3638113570741097,
+      "grad_norm": 1.84505033493042,
+      "learning_rate": 0.00018414021261797743,
+      "loss": 1.1362,
+      "step": 9450
+    },
+    {
+      "epoch": 0.3640038498556304,
+      "grad_norm": 1.5999794006347656,
+      "learning_rate": 0.00018412386699994518,
+      "loss": 1.1647,
+      "step": 9455
+    },
+    {
+      "epoch": 0.3641963426371511,
+      "grad_norm": 1.55308997631073,
+      "learning_rate": 0.0001841075136893426,
+      "loss": 1.2612,
+      "step": 9460
+    },
+    {
+      "epoch": 0.3643888354186718,
+      "grad_norm": 1.3549528121948242,
+      "learning_rate": 0.00018409115268766505,
+      "loss": 1.2095,
+      "step": 9465
+    },
+    {
+      "epoch": 0.3645813282001925,
+      "grad_norm": 1.123184323310852,
+      "learning_rate": 0.00018407478399640862,
+      "loss": 1.3047,
+      "step": 9470
+    },
+    {
+      "epoch": 0.3647738209817132,
+      "grad_norm": 1.3776748180389404,
+      "learning_rate": 0.00018405840761707016,
+      "loss": 1.1064,
+      "step": 9475
+    },
+    {
+      "epoch": 0.36496631376323385,
+      "grad_norm": 1.3778200149536133,
+      "learning_rate": 0.00018404202355114718,
+      "loss": 1.0956,
+      "step": 9480
+    },
+    {
+      "epoch": 0.3651588065447546,
+      "grad_norm": 0.9069898128509521,
+      "learning_rate": 0.00018402563180013783,
+      "loss": 1.141,
+      "step": 9485
+    },
+    {
+      "epoch": 0.36535129932627525,
+      "grad_norm": 1.3908804655075073,
+      "learning_rate": 0.0001840092323655411,
+      "loss": 1.2679,
+      "step": 9490
+    },
+    {
+      "epoch": 0.365543792107796,
+      "grad_norm": 1.3785732984542847,
+      "learning_rate": 0.00018399282524885654,
+      "loss": 1.22,
+      "step": 9495
+    },
+    {
+      "epoch": 0.36573628488931664,
+      "grad_norm": 1.1326193809509277,
+      "learning_rate": 0.00018397641045158453,
+      "loss": 1.2289,
+      "step": 9500
+    },
+    {
+      "epoch": 0.36592877767083737,
+      "grad_norm": 1.2267814874649048,
+      "learning_rate": 0.0001839599879752261,
+      "loss": 1.1337,
+      "step": 9505
+    },
+    {
+      "epoch": 0.36612127045235804,
+      "grad_norm": 0.8690314888954163,
+      "learning_rate": 0.00018394355782128295,
+      "loss": 1.2535,
+      "step": 9510
+    },
+    {
+      "epoch": 0.3663137632338787,
+      "grad_norm": 1.448415994644165,
+      "learning_rate": 0.00018392711999125748,
+      "loss": 1.1405,
+      "step": 9515
+    },
+    {
+      "epoch": 0.36650625601539943,
+      "grad_norm": 1.8989317417144775,
+      "learning_rate": 0.00018391067448665288,
+      "loss": 1.091,
+      "step": 9520
+    },
+    {
+      "epoch": 0.3666987487969201,
+      "grad_norm": 1.2263299226760864,
+      "learning_rate": 0.00018389422130897295,
+      "loss": 1.1925,
+      "step": 9525
+    },
+    {
+      "epoch": 0.36689124157844083,
+      "grad_norm": 0.8818153142929077,
+      "learning_rate": 0.00018387776045972225,
+      "loss": 1.2961,
+      "step": 9530
+    },
+    {
+      "epoch": 0.3670837343599615,
+      "grad_norm": 1.0975017547607422,
+      "learning_rate": 0.00018386129194040597,
+      "loss": 1.414,
+      "step": 9535
+    },
+    {
+      "epoch": 0.36727622714148217,
+      "grad_norm": 2.2097692489624023,
+      "learning_rate": 0.00018384481575253004,
+      "loss": 1.1941,
+      "step": 9540
+    },
+    {
+      "epoch": 0.3674687199230029,
+      "grad_norm": 1.2249376773834229,
+      "learning_rate": 0.0001838283318976012,
+      "loss": 1.4472,
+      "step": 9545
+    },
+    {
+      "epoch": 0.36766121270452357,
+      "grad_norm": 1.0000889301300049,
+      "learning_rate": 0.0001838118403771267,
+      "loss": 1.2399,
+      "step": 9550
+    },
+    {
+      "epoch": 0.3678537054860443,
+      "grad_norm": 1.0249544382095337,
+      "learning_rate": 0.00018379534119261458,
+      "loss": 1.3182,
+      "step": 9555
+    },
+    {
+      "epoch": 0.36804619826756496,
+      "grad_norm": 1.2347283363342285,
+      "learning_rate": 0.00018377883434557362,
+      "loss": 1.1313,
+      "step": 9560
+    },
+    {
+      "epoch": 0.3682386910490857,
+      "grad_norm": 1.1021714210510254,
+      "learning_rate": 0.0001837623198375132,
+      "loss": 1.2381,
+      "step": 9565
+    },
+    {
+      "epoch": 0.36843118383060636,
+      "grad_norm": 1.0923985242843628,
+      "learning_rate": 0.00018374579766994355,
+      "loss": 1.3386,
+      "step": 9570
+    },
+    {
+      "epoch": 0.368623676612127,
+      "grad_norm": 1.7709978818893433,
+      "learning_rate": 0.00018372926784437547,
+      "loss": 1.2405,
+      "step": 9575
+    },
+    {
+      "epoch": 0.36881616939364775,
+      "grad_norm": 1.316901683807373,
+      "learning_rate": 0.00018371273036232047,
+      "loss": 1.1244,
+      "step": 9580
+    },
+    {
+      "epoch": 0.3690086621751684,
+      "grad_norm": 1.7281345129013062,
+      "learning_rate": 0.00018369618522529085,
+      "loss": 1.2979,
+      "step": 9585
+    },
+    {
+      "epoch": 0.36920115495668915,
+      "grad_norm": 1.6363762617111206,
+      "learning_rate": 0.00018367963243479953,
+      "loss": 1.1528,
+      "step": 9590
+    },
+    {
+      "epoch": 0.3693936477382098,
+      "grad_norm": 1.7078179121017456,
+      "learning_rate": 0.00018366307199236013,
+      "loss": 1.2833,
+      "step": 9595
+    },
+    {
+      "epoch": 0.3695861405197305,
+      "grad_norm": 1.9110232591629028,
+      "learning_rate": 0.000183646503899487,
+      "loss": 1.4191,
+      "step": 9600
+    },
+    {
+      "epoch": 0.3697786333012512,
+      "grad_norm": 0.952301025390625,
+      "learning_rate": 0.00018362992815769525,
+      "loss": 1.1504,
+      "step": 9605
+    },
+    {
+      "epoch": 0.3699711260827719,
+      "grad_norm": 0.9142165780067444,
+      "learning_rate": 0.0001836133447685005,
+      "loss": 1.2617,
+      "step": 9610
+    },
+    {
+      "epoch": 0.3701636188642926,
+      "grad_norm": 1.5571134090423584,
+      "learning_rate": 0.0001835967537334193,
+      "loss": 1.3054,
+      "step": 9615
+    },
+    {
+      "epoch": 0.3703561116458133,
+      "grad_norm": 1.799795389175415,
+      "learning_rate": 0.00018358015505396877,
+      "loss": 1.0603,
+      "step": 9620
+    },
+    {
+      "epoch": 0.37054860442733395,
+      "grad_norm": 1.6660315990447998,
+      "learning_rate": 0.0001835635487316667,
+      "loss": 1.1757,
+      "step": 9625
+    },
+    {
+      "epoch": 0.3707410972088547,
+      "grad_norm": 0.9840423464775085,
+      "learning_rate": 0.00018354693476803168,
+      "loss": 0.9815,
+      "step": 9630
+    },
+    {
+      "epoch": 0.37093358999037535,
+      "grad_norm": 2.0538954734802246,
+      "learning_rate": 0.00018353031316458286,
+      "loss": 1.2396,
+      "step": 9635
+    },
+    {
+      "epoch": 0.37112608277189607,
+      "grad_norm": 1.2079198360443115,
+      "learning_rate": 0.0001835136839228403,
+      "loss": 1.2731,
+      "step": 9640
+    },
+    {
+      "epoch": 0.37131857555341674,
+      "grad_norm": 1.7076921463012695,
+      "learning_rate": 0.00018349704704432457,
+      "loss": 1.1388,
+      "step": 9645
+    },
+    {
+      "epoch": 0.37151106833493747,
+      "grad_norm": 1.0324435234069824,
+      "learning_rate": 0.00018348040253055698,
+      "loss": 0.9949,
+      "step": 9650
+    },
+    {
+      "epoch": 0.37170356111645814,
+      "grad_norm": 1.3635584115982056,
+      "learning_rate": 0.0001834637503830596,
+      "loss": 1.307,
+      "step": 9655
+    },
+    {
+      "epoch": 0.3718960538979788,
+      "grad_norm": 1.6683429479599,
+      "learning_rate": 0.00018344709060335513,
+      "loss": 1.1687,
+      "step": 9660
+    },
+    {
+      "epoch": 0.37208854667949953,
+      "grad_norm": 2.3687121868133545,
+      "learning_rate": 0.00018343042319296702,
+      "loss": 1.4163,
+      "step": 9665
+    },
+    {
+      "epoch": 0.3722810394610202,
+      "grad_norm": 1.9078242778778076,
+      "learning_rate": 0.00018341374815341937,
+      "loss": 1.2986,
+      "step": 9670
+    },
+    {
+      "epoch": 0.37247353224254093,
+      "grad_norm": 1.6381220817565918,
+      "learning_rate": 0.00018339706548623706,
+      "loss": 1.5092,
+      "step": 9675
+    },
+    {
+      "epoch": 0.3726660250240616,
+      "grad_norm": 1.3529161214828491,
+      "learning_rate": 0.00018338037519294553,
+      "loss": 1.2296,
+      "step": 9680
+    },
+    {
+      "epoch": 0.37285851780558227,
+      "grad_norm": 1.1034053564071655,
+      "learning_rate": 0.00018336367727507104,
+      "loss": 1.2774,
+      "step": 9685
+    },
+    {
+      "epoch": 0.373051010587103,
+      "grad_norm": 2.0935397148132324,
+      "learning_rate": 0.0001833469717341405,
+      "loss": 1.2247,
+      "step": 9690
+    },
+    {
+      "epoch": 0.37324350336862366,
+      "grad_norm": 1.6294866800308228,
+      "learning_rate": 0.0001833302585716815,
+      "loss": 1.3766,
+      "step": 9695
+    },
+    {
+      "epoch": 0.3734359961501444,
+      "grad_norm": 1.6927978992462158,
+      "learning_rate": 0.0001833135377892224,
+      "loss": 1.3069,
+      "step": 9700
+    },
+    {
+      "epoch": 0.37362848893166506,
+      "grad_norm": 0.8497247695922852,
+      "learning_rate": 0.00018329680938829212,
+      "loss": 1.0906,
+      "step": 9705
+    },
+    {
+      "epoch": 0.37382098171318573,
+      "grad_norm": 1.9347554445266724,
+      "learning_rate": 0.00018328007337042046,
+      "loss": 1.277,
+      "step": 9710
+    },
+    {
+      "epoch": 0.37401347449470645,
+      "grad_norm": 1.023130178451538,
+      "learning_rate": 0.00018326332973713776,
+      "loss": 1.254,
+      "step": 9715
+    },
+    {
+      "epoch": 0.3742059672762271,
+      "grad_norm": 1.7206385135650635,
+      "learning_rate": 0.0001832465784899751,
+      "loss": 1.2141,
+      "step": 9720
+    },
+    {
+      "epoch": 0.37439846005774785,
+      "grad_norm": 1.2445294857025146,
+      "learning_rate": 0.00018322981963046433,
+      "loss": 1.3817,
+      "step": 9725
+    },
+    {
+      "epoch": 0.3745909528392685,
+      "grad_norm": 1.832334280014038,
+      "learning_rate": 0.00018321305316013788,
+      "loss": 1.3584,
+      "step": 9730
+    },
+    {
+      "epoch": 0.37478344562078925,
+      "grad_norm": 1.2087010145187378,
+      "learning_rate": 0.00018319627908052898,
+      "loss": 1.116,
+      "step": 9735
+    },
+    {
+      "epoch": 0.3749759384023099,
+      "grad_norm": 1.286687970161438,
+      "learning_rate": 0.00018317949739317147,
+      "loss": 1.1913,
+      "step": 9740
+    },
+    {
+      "epoch": 0.3751684311838306,
+      "grad_norm": 1.44833242893219,
+      "learning_rate": 0.00018316270809959993,
+      "loss": 1.2713,
+      "step": 9745
+    },
+    {
+      "epoch": 0.3753609239653513,
+      "grad_norm": 1.1395667791366577,
+      "learning_rate": 0.00018314591120134963,
+      "loss": 1.2912,
+      "step": 9750
+    },
+    {
+      "epoch": 0.375553416746872,
+      "grad_norm": 1.1399837732315063,
+      "learning_rate": 0.00018312910669995654,
+      "loss": 1.2804,
+      "step": 9755
+    },
+    {
+      "epoch": 0.3757459095283927,
+      "grad_norm": 1.814249038696289,
+      "learning_rate": 0.00018311229459695735,
+      "loss": 1.1062,
+      "step": 9760
+    },
+    {
+      "epoch": 0.3759384023099134,
+      "grad_norm": 1.4851144552230835,
+      "learning_rate": 0.00018309547489388933,
+      "loss": 1.2826,
+      "step": 9765
+    },
+    {
+      "epoch": 0.37613089509143405,
+      "grad_norm": 0.9308827519416809,
+      "learning_rate": 0.00018307864759229065,
+      "loss": 1.3706,
+      "step": 9770
+    },
+    {
+      "epoch": 0.3763233878729548,
+      "grad_norm": 3.707566261291504,
+      "learning_rate": 0.00018306181269369998,
+      "loss": 1.2292,
+      "step": 9775
+    },
+    {
+      "epoch": 0.37651588065447544,
+      "grad_norm": 2.6666324138641357,
+      "learning_rate": 0.00018304497019965677,
+      "loss": 1.4645,
+      "step": 9780
+    },
+    {
+      "epoch": 0.37670837343599617,
+      "grad_norm": 1.5997512340545654,
+      "learning_rate": 0.00018302812011170114,
+      "loss": 1.2812,
+      "step": 9785
+    },
+    {
+      "epoch": 0.37690086621751684,
+      "grad_norm": 0.8998873233795166,
+      "learning_rate": 0.00018301126243137395,
+      "loss": 1.195,
+      "step": 9790
+    },
+    {
+      "epoch": 0.3770933589990375,
+      "grad_norm": 1.407524585723877,
+      "learning_rate": 0.0001829943971602167,
+      "loss": 1.1793,
+      "step": 9795
+    },
+    {
+      "epoch": 0.37728585178055823,
+      "grad_norm": 1.1469497680664062,
+      "learning_rate": 0.00018297752429977164,
+      "loss": 1.3624,
+      "step": 9800
+    },
+    {
+      "epoch": 0.3774783445620789,
+      "grad_norm": 1.4583423137664795,
+      "learning_rate": 0.00018296064385158164,
+      "loss": 1.2033,
+      "step": 9805
+    },
+    {
+      "epoch": 0.37767083734359963,
+      "grad_norm": 1.0782575607299805,
+      "learning_rate": 0.00018294375581719036,
+      "loss": 1.1823,
+      "step": 9810
+    },
+    {
+      "epoch": 0.3778633301251203,
+      "grad_norm": 1.1890922784805298,
+      "learning_rate": 0.00018292686019814202,
+      "loss": 1.2711,
+      "step": 9815
+    },
+    {
+      "epoch": 0.378055822906641,
+      "grad_norm": 0.854491651058197,
+      "learning_rate": 0.00018290995699598165,
+      "loss": 1.1953,
+      "step": 9820
+    },
+    {
+      "epoch": 0.3782483156881617,
+      "grad_norm": 1.2184374332427979,
+      "learning_rate": 0.00018289304621225497,
+      "loss": 1.2052,
+      "step": 9825
+    },
+    {
+      "epoch": 0.37844080846968237,
+      "grad_norm": 1.1952948570251465,
+      "learning_rate": 0.0001828761278485083,
+      "loss": 1.2516,
+      "step": 9830
+    },
+    {
+      "epoch": 0.3786333012512031,
+      "grad_norm": 2.1117265224456787,
+      "learning_rate": 0.00018285920190628879,
+      "loss": 1.2834,
+      "step": 9835
+    },
+    {
+      "epoch": 0.37882579403272376,
+      "grad_norm": 1.1815403699874878,
+      "learning_rate": 0.00018284226838714412,
+      "loss": 1.0574,
+      "step": 9840
+    },
+    {
+      "epoch": 0.3790182868142445,
+      "grad_norm": 1.3763145208358765,
+      "learning_rate": 0.00018282532729262278,
+      "loss": 1.2813,
+      "step": 9845
+    },
+    {
+      "epoch": 0.37921077959576516,
+      "grad_norm": 1.5308822393417358,
+      "learning_rate": 0.00018280837862427393,
+      "loss": 1.2118,
+      "step": 9850
+    },
+    {
+      "epoch": 0.3794032723772858,
+      "grad_norm": 1.1991111040115356,
+      "learning_rate": 0.00018279142238364745,
+      "loss": 1.0999,
+      "step": 9855
+    },
+    {
+      "epoch": 0.37959576515880655,
+      "grad_norm": 1.7062435150146484,
+      "learning_rate": 0.0001827744585722938,
+      "loss": 1.2103,
+      "step": 9860
+    },
+    {
+      "epoch": 0.3797882579403272,
+      "grad_norm": 1.5572453737258911,
+      "learning_rate": 0.00018275748719176425,
+      "loss": 1.112,
+      "step": 9865
+    },
+    {
+      "epoch": 0.37998075072184795,
+      "grad_norm": 0.9328321218490601,
+      "learning_rate": 0.00018274050824361072,
+      "loss": 1.2688,
+      "step": 9870
+    },
+    {
+      "epoch": 0.3801732435033686,
+      "grad_norm": 1.290634036064148,
+      "learning_rate": 0.0001827235217293858,
+      "loss": 1.1486,
+      "step": 9875
+    },
+    {
+      "epoch": 0.38036573628488934,
+      "grad_norm": 1.7471963167190552,
+      "learning_rate": 0.00018270652765064283,
+      "loss": 1.2584,
+      "step": 9880
+    },
+    {
+      "epoch": 0.38055822906641,
+      "grad_norm": 1.4827409982681274,
+      "learning_rate": 0.00018268952600893577,
+      "loss": 1.3655,
+      "step": 9885
+    },
+    {
+      "epoch": 0.3807507218479307,
+      "grad_norm": 1.0229063034057617,
+      "learning_rate": 0.00018267251680581935,
+      "loss": 1.1955,
+      "step": 9890
+    },
+    {
+      "epoch": 0.3809432146294514,
+      "grad_norm": 1.3075898885726929,
+      "learning_rate": 0.0001826555000428489,
+      "loss": 0.9779,
+      "step": 9895
+    },
+    {
+      "epoch": 0.3811357074109721,
+      "grad_norm": 1.5942119359970093,
+      "learning_rate": 0.00018263847572158053,
+      "loss": 1.2556,
+      "step": 9900
+    },
+    {
+      "epoch": 0.3813282001924928,
+      "grad_norm": 0.9223330616950989,
+      "learning_rate": 0.00018262144384357097,
+      "loss": 1.1109,
+      "step": 9905
+    },
+    {
+      "epoch": 0.3815206929740135,
+      "grad_norm": 1.7757457494735718,
+      "learning_rate": 0.00018260440441037766,
+      "loss": 1.2219,
+      "step": 9910
+    },
+    {
+      "epoch": 0.38171318575553415,
+      "grad_norm": 1.4870551824569702,
+      "learning_rate": 0.00018258735742355883,
+      "loss": 1.3312,
+      "step": 9915
+    },
+    {
+      "epoch": 0.38190567853705487,
+      "grad_norm": 1.2982031106948853,
+      "learning_rate": 0.00018257030288467322,
+      "loss": 1.2421,
+      "step": 9920
+    },
+    {
+      "epoch": 0.38209817131857554,
+      "grad_norm": 1.016822338104248,
+      "learning_rate": 0.0001825532407952804,
+      "loss": 1.3542,
+      "step": 9925
+    },
+    {
+      "epoch": 0.38229066410009627,
+      "grad_norm": 1.0763219594955444,
+      "learning_rate": 0.00018253617115694058,
+      "loss": 1.2579,
+      "step": 9930
+    },
+    {
+      "epoch": 0.38248315688161694,
+      "grad_norm": 1.7673341035842896,
+      "learning_rate": 0.00018251909397121464,
+      "loss": 1.1875,
+      "step": 9935
+    },
+    {
+      "epoch": 0.3826756496631376,
+      "grad_norm": 1.3719041347503662,
+      "learning_rate": 0.00018250200923966423,
+      "loss": 1.1493,
+      "step": 9940
+    },
+    {
+      "epoch": 0.38286814244465833,
+      "grad_norm": 1.8589760065078735,
+      "learning_rate": 0.00018248491696385157,
+      "loss": 1.2751,
+      "step": 9945
+    },
+    {
+      "epoch": 0.383060635226179,
+      "grad_norm": 1.6069539785385132,
+      "learning_rate": 0.0001824678171453397,
+      "loss": 1.415,
+      "step": 9950
+    },
+    {
+      "epoch": 0.38325312800769973,
+      "grad_norm": 1.7131226062774658,
+      "learning_rate": 0.0001824507097856922,
+      "loss": 1.1773,
+      "step": 9955
+    },
+    {
+      "epoch": 0.3834456207892204,
+      "grad_norm": 0.7622759342193604,
+      "learning_rate": 0.0001824335948864735,
+      "loss": 1.1588,
+      "step": 9960
+    },
+    {
+      "epoch": 0.3836381135707411,
+      "grad_norm": 1.6202800273895264,
+      "learning_rate": 0.0001824164724492486,
+      "loss": 1.3064,
+      "step": 9965
+    },
+    {
+      "epoch": 0.3838306063522618,
+      "grad_norm": 1.5452194213867188,
+      "learning_rate": 0.0001823993424755833,
+      "loss": 1.2993,
+      "step": 9970
+    },
+    {
+      "epoch": 0.38402309913378246,
+      "grad_norm": 1.013929009437561,
+      "learning_rate": 0.00018238220496704396,
+      "loss": 1.3123,
+      "step": 9975
+    },
+    {
+      "epoch": 0.3842155919153032,
+      "grad_norm": 0.9624648094177246,
+      "learning_rate": 0.0001823650599251977,
+      "loss": 1.0517,
+      "step": 9980
+    },
+    {
+      "epoch": 0.38440808469682386,
+      "grad_norm": 1.2065962553024292,
+      "learning_rate": 0.00018234790735161232,
+      "loss": 1.1954,
+      "step": 9985
+    },
+    {
+      "epoch": 0.3846005774783446,
+      "grad_norm": 1.425376057624817,
+      "learning_rate": 0.00018233074724785634,
+      "loss": 1.069,
+      "step": 9990
+    },
+    {
+      "epoch": 0.38479307025986526,
+      "grad_norm": 1.0355112552642822,
+      "learning_rate": 0.00018231357961549888,
+      "loss": 1.0839,
+      "step": 9995
+    },
+    {
+      "epoch": 0.3849855630413859,
+      "grad_norm": 1.7273633480072021,
+      "learning_rate": 0.00018229640445610988,
+      "loss": 1.1324,
+      "step": 10000
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 51950,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 10000,
+  "total_flos": 3.133033729973453e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}