diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9576 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 6807,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00044072278536800354,
+      "grad_norm": 0.24354467550812778,
+      "learning_rate": 2.936857562408223e-07,
+      "loss": 2.1339,
+      "step": 1
+    },
+    {
+      "epoch": 0.0022036139268400176,
+      "grad_norm": 0.27040776216996587,
+      "learning_rate": 1.4684287812041115e-06,
+      "loss": 2.3535,
+      "step": 5
+    },
+    {
+      "epoch": 0.004407227853680035,
+      "grad_norm": 0.23978713638500837,
+      "learning_rate": 2.936857562408223e-06,
+      "loss": 2.0659,
+      "step": 10
+    },
+    {
+      "epoch": 0.006610841780520053,
+      "grad_norm": 0.2562898248528363,
+      "learning_rate": 4.4052863436123355e-06,
+      "loss": 2.0742,
+      "step": 15
+    },
+    {
+      "epoch": 0.00881445570736007,
+      "grad_norm": 0.2671940707327157,
+      "learning_rate": 5.873715124816446e-06,
+      "loss": 2.0397,
+      "step": 20
+    },
+    {
+      "epoch": 0.011018069634200088,
+      "grad_norm": 0.2840547143688825,
+      "learning_rate": 7.3421439060205585e-06,
+      "loss": 2.3034,
+      "step": 25
+    },
+    {
+      "epoch": 0.013221683561040106,
+      "grad_norm": 0.2739646963997833,
+      "learning_rate": 8.810572687224671e-06,
+      "loss": 2.0168,
+      "step": 30
+    },
+    {
+      "epoch": 0.015425297487880123,
+      "grad_norm": 0.36046684768779125,
+      "learning_rate": 1.0279001468428782e-05,
+      "loss": 2.1073,
+      "step": 35
+    },
+    {
+      "epoch": 0.01762891141472014,
+      "grad_norm": 0.29214076993878346,
+      "learning_rate": 1.1747430249632892e-05,
+      "loss": 1.9053,
+      "step": 40
+    },
+    {
+      "epoch": 0.01983252534156016,
+      "grad_norm": 0.4413639637555924,
+      "learning_rate": 1.3215859030837005e-05,
+      "loss": 2.272,
+      "step": 45
+    },
+    {
+      "epoch": 0.022036139268400177,
+      "grad_norm": 0.3645215252816923,
+      "learning_rate": 1.4684287812041117e-05,
+      "loss": 2.0033,
+      "step": 50
+    },
+    {
+      "epoch": 0.024239753195240195,
+      "grad_norm": 0.26241461964383006,
+      "learning_rate": 1.615271659324523e-05,
+      "loss": 2.0168,
+      "step": 55
+    },
+    {
+      "epoch": 0.026443367122080213,
+      "grad_norm": 0.3108898378679939,
+      "learning_rate": 1.7621145374449342e-05,
+      "loss": 2.1111,
+      "step": 60
+    },
+    {
+      "epoch": 0.02864698104892023,
+      "grad_norm": 0.2662326669897319,
+      "learning_rate": 1.9089574155653454e-05,
+      "loss": 1.8806,
+      "step": 65
+    },
+    {
+      "epoch": 0.030850594975760245,
+      "grad_norm": 0.16963519927101947,
+      "learning_rate": 2.0558002936857563e-05,
+      "loss": 2.1257,
+      "step": 70
+    },
+    {
+      "epoch": 0.03305420890260027,
+      "grad_norm": 0.23310716143802493,
+      "learning_rate": 2.2026431718061676e-05,
+      "loss": 1.8759,
+      "step": 75
+    },
+    {
+      "epoch": 0.03525782282944028,
+      "grad_norm": 0.2575993518919339,
+      "learning_rate": 2.3494860499265785e-05,
+      "loss": 1.9053,
+      "step": 80
+    },
+    {
+      "epoch": 0.0374614367562803,
+      "grad_norm": 0.2746073869104198,
+      "learning_rate": 2.4963289280469897e-05,
+      "loss": 1.7651,
+      "step": 85
+    },
+    {
+      "epoch": 0.03966505068312032,
+      "grad_norm": 0.2279037029835098,
+      "learning_rate": 2.643171806167401e-05,
+      "loss": 1.8627,
+      "step": 90
+    },
+    {
+      "epoch": 0.04186866460996033,
+      "grad_norm": 0.27296809706241965,
+      "learning_rate": 2.7900146842878122e-05,
+      "loss": 1.9057,
+      "step": 95
+    },
+    {
+      "epoch": 0.044072278536800354,
+      "grad_norm": 0.2670117144218064,
+      "learning_rate": 2.9368575624082234e-05,
+      "loss": 1.8911,
+      "step": 100
+    },
+    {
+      "epoch": 0.04627589246364037,
+      "grad_norm": 0.23511621219550868,
+      "learning_rate": 3.0837004405286347e-05,
+      "loss": 2.024,
+      "step": 105
+    },
+    {
+      "epoch": 0.04847950639048039,
+      "grad_norm": 0.24414856722360948,
+      "learning_rate": 3.230543318649046e-05,
+      "loss": 1.9756,
+      "step": 110
+    },
+    {
+      "epoch": 0.050683120317320404,
+      "grad_norm": 0.25077238640232496,
+      "learning_rate": 3.377386196769457e-05,
+      "loss": 1.7627,
+      "step": 115
+    },
+    {
+      "epoch": 0.052886734244160426,
+      "grad_norm": 0.30541323094735956,
+      "learning_rate": 3.5242290748898684e-05,
+      "loss": 1.9455,
+      "step": 120
+    },
+    {
+      "epoch": 0.05509034817100044,
+      "grad_norm": 0.2593735959377697,
+      "learning_rate": 3.6710719530102796e-05,
+      "loss": 1.9408,
+      "step": 125
+    },
+    {
+      "epoch": 0.05729396209784046,
+      "grad_norm": 0.2356435173973878,
+      "learning_rate": 3.817914831130691e-05,
+      "loss": 1.9746,
+      "step": 130
+    },
+    {
+      "epoch": 0.059497576024680476,
+      "grad_norm": 0.2556575372905199,
+      "learning_rate": 3.9647577092511014e-05,
+      "loss": 1.9133,
+      "step": 135
+    },
+    {
+      "epoch": 0.06170118995152049,
+      "grad_norm": 0.23671492386855494,
+      "learning_rate": 4.1116005873715127e-05,
+      "loss": 1.7228,
+      "step": 140
+    },
+    {
+      "epoch": 0.0639048038783605,
+      "grad_norm": 0.2522128793647996,
+      "learning_rate": 4.258443465491924e-05,
+      "loss": 1.9416,
+      "step": 145
+    },
+    {
+      "epoch": 0.06610841780520053,
+      "grad_norm": 0.259439737149894,
+      "learning_rate": 4.405286343612335e-05,
+      "loss": 2.0295,
+      "step": 150
+    },
+    {
+      "epoch": 0.06831203173204055,
+      "grad_norm": 0.2632402024665515,
+      "learning_rate": 4.5521292217327464e-05,
+      "loss": 1.7906,
+      "step": 155
+    },
+    {
+      "epoch": 0.07051564565888056,
+      "grad_norm": 0.28869506877919027,
+      "learning_rate": 4.698972099853157e-05,
+      "loss": 1.73,
+      "step": 160
+    },
+    {
+      "epoch": 0.07271925958572058,
+      "grad_norm": 0.2586750438465526,
+      "learning_rate": 4.845814977973568e-05,
+      "loss": 1.8038,
+      "step": 165
+    },
+    {
+      "epoch": 0.0749228735125606,
+      "grad_norm": 0.2506125082680772,
+      "learning_rate": 4.9926578560939794e-05,
+      "loss": 1.8558,
+      "step": 170
+    },
+    {
+      "epoch": 0.07712648743940062,
+      "grad_norm": 0.2876243539987618,
+      "learning_rate": 5.1395007342143906e-05,
+      "loss": 1.9784,
+      "step": 175
+    },
+    {
+      "epoch": 0.07933010136624064,
+      "grad_norm": 0.23288536708042795,
+      "learning_rate": 5.286343612334802e-05,
+      "loss": 1.7534,
+      "step": 180
+    },
+    {
+      "epoch": 0.08153371529308065,
+      "grad_norm": 0.25062968804040664,
+      "learning_rate": 5.433186490455213e-05,
+      "loss": 1.8032,
+      "step": 185
+    },
+    {
+      "epoch": 0.08373732921992066,
+      "grad_norm": 0.3358849445093002,
+      "learning_rate": 5.5800293685756244e-05,
+      "loss": 1.9325,
+      "step": 190
+    },
+    {
+      "epoch": 0.08594094314676069,
+      "grad_norm": 0.25234625121100573,
+      "learning_rate": 5.7268722466960356e-05,
+      "loss": 1.8855,
+      "step": 195
+    },
+    {
+      "epoch": 0.08814455707360071,
+      "grad_norm": 0.3365993483895123,
+      "learning_rate": 5.873715124816447e-05,
+      "loss": 1.8118,
+      "step": 200
+    },
+    {
+      "epoch": 0.09034817100044072,
+      "grad_norm": 0.311599846916865,
+      "learning_rate": 6.020558002936858e-05,
+      "loss": 2.0721,
+      "step": 205
+    },
+    {
+      "epoch": 0.09255178492728074,
+      "grad_norm": 0.3363293627242514,
+      "learning_rate": 6.167400881057269e-05,
+      "loss": 1.804,
+      "step": 210
+    },
+    {
+      "epoch": 0.09475539885412076,
+      "grad_norm": 0.3414057038117249,
+      "learning_rate": 6.31424375917768e-05,
+      "loss": 1.871,
+      "step": 215
+    },
+    {
+      "epoch": 0.09695901278096078,
+      "grad_norm": 0.3051539417193126,
+      "learning_rate": 6.461086637298092e-05,
+      "loss": 2.0932,
+      "step": 220
+    },
+    {
+      "epoch": 0.0991626267078008,
+      "grad_norm": 0.2859223655623353,
+      "learning_rate": 6.607929515418503e-05,
+      "loss": 1.8641,
+      "step": 225
+    },
+    {
+      "epoch": 0.10136624063464081,
+      "grad_norm": 0.30959368455808256,
+      "learning_rate": 6.754772393538914e-05,
+      "loss": 1.8437,
+      "step": 230
+    },
+    {
+      "epoch": 0.10356985456148082,
+      "grad_norm": 0.3544528067642694,
+      "learning_rate": 6.901615271659326e-05,
+      "loss": 1.9002,
+      "step": 235
+    },
+    {
+      "epoch": 0.10577346848832085,
+      "grad_norm": 0.34632827275641465,
+      "learning_rate": 7.048458149779737e-05,
+      "loss": 1.8398,
+      "step": 240
+    },
+    {
+      "epoch": 0.10797708241516087,
+      "grad_norm": 0.597473977518495,
+      "learning_rate": 7.195301027900148e-05,
+      "loss": 1.9357,
+      "step": 245
+    },
+    {
+      "epoch": 0.11018069634200088,
+      "grad_norm": 0.35953743798846133,
+      "learning_rate": 7.342143906020559e-05,
+      "loss": 1.944,
+      "step": 250
+    },
+    {
+      "epoch": 0.1123843102688409,
+      "grad_norm": 0.32165430097730124,
+      "learning_rate": 7.48898678414097e-05,
+      "loss": 1.8974,
+      "step": 255
+    },
+    {
+      "epoch": 0.11458792419568092,
+      "grad_norm": 0.2616578656705798,
+      "learning_rate": 7.635829662261382e-05,
+      "loss": 1.6988,
+      "step": 260
+    },
+    {
+      "epoch": 0.11679153812252094,
+      "grad_norm": 0.37507900741140277,
+      "learning_rate": 7.782672540381793e-05,
+      "loss": 1.9061,
+      "step": 265
+    },
+    {
+      "epoch": 0.11899515204936095,
+      "grad_norm": 0.3505810167092092,
+      "learning_rate": 7.929515418502203e-05,
+      "loss": 1.7829,
+      "step": 270
+    },
+    {
+      "epoch": 0.12119876597620097,
+      "grad_norm": 0.2905187770243884,
+      "learning_rate": 8.076358296622614e-05,
+      "loss": 1.896,
+      "step": 275
+    },
+    {
+      "epoch": 0.12340237990304098,
+      "grad_norm": 0.31667074390598227,
+      "learning_rate": 8.223201174743025e-05,
+      "loss": 1.8549,
+      "step": 280
+    },
+    {
+      "epoch": 0.125605993829881,
+      "grad_norm": 0.33256810111534474,
+      "learning_rate": 8.370044052863437e-05,
+      "loss": 1.9184,
+      "step": 285
+    },
+    {
+      "epoch": 0.127809607756721,
+      "grad_norm": 0.41094013265883617,
+      "learning_rate": 8.516886930983848e-05,
+      "loss": 1.9407,
+      "step": 290
+    },
+    {
+      "epoch": 0.13001322168356105,
+      "grad_norm": 0.3268179907564926,
+      "learning_rate": 8.663729809104259e-05,
+      "loss": 1.858,
+      "step": 295
+    },
+    {
+      "epoch": 0.13221683561040107,
+      "grad_norm": 0.4082234533551937,
+      "learning_rate": 8.81057268722467e-05,
+      "loss": 1.8172,
+      "step": 300
+    },
+    {
+      "epoch": 0.13442044953724108,
+      "grad_norm": 0.3303954200472522,
+      "learning_rate": 8.957415565345081e-05,
+      "loss": 1.7879,
+      "step": 305
+    },
+    {
+      "epoch": 0.1366240634640811,
+      "grad_norm": 0.3785434588284624,
+      "learning_rate": 9.104258443465493e-05,
+      "loss": 1.7931,
+      "step": 310
+    },
+    {
+      "epoch": 0.1388276773909211,
+      "grad_norm": 0.3603897034728447,
+      "learning_rate": 9.251101321585903e-05,
+      "loss": 1.7793,
+      "step": 315
+    },
+    {
+      "epoch": 0.14103129131776113,
+      "grad_norm": 0.2967904662976913,
+      "learning_rate": 9.397944199706314e-05,
+      "loss": 1.7092,
+      "step": 320
+    },
+    {
+      "epoch": 0.14323490524460114,
+      "grad_norm": 0.2976559195453983,
+      "learning_rate": 9.544787077826725e-05,
+      "loss": 1.8302,
+      "step": 325
+    },
+    {
+      "epoch": 0.14543851917144116,
+      "grad_norm": 0.34525724393698276,
+      "learning_rate": 9.691629955947136e-05,
+      "loss": 1.9488,
+      "step": 330
+    },
+    {
+      "epoch": 0.14764213309828117,
+      "grad_norm": 0.3473487500731534,
+      "learning_rate": 9.838472834067548e-05,
+      "loss": 1.8169,
+      "step": 335
+    },
+    {
+      "epoch": 0.1498457470251212,
+      "grad_norm": 0.35776737697830296,
+      "learning_rate": 9.985315712187959e-05,
+      "loss": 1.7562,
+      "step": 340
+    },
+    {
+      "epoch": 0.15204936095196123,
+      "grad_norm": 0.35067754753032443,
+      "learning_rate": 0.00010132158590308371,
+      "loss": 1.9597,
+      "step": 345
+    },
+    {
+      "epoch": 0.15425297487880124,
+      "grad_norm": 0.31111407532049806,
+      "learning_rate": 0.00010279001468428781,
+      "loss": 1.6977,
+      "step": 350
+    },
+    {
+      "epoch": 0.15645658880564126,
+      "grad_norm": 0.350081092372353,
+      "learning_rate": 0.00010425844346549194,
+      "loss": 1.7585,
+      "step": 355
+    },
+    {
+      "epoch": 0.15866020273248127,
+      "grad_norm": 0.3445787779571548,
+      "learning_rate": 0.00010572687224669604,
+      "loss": 1.7956,
+      "step": 360
+    },
+    {
+      "epoch": 0.16086381665932128,
+      "grad_norm": 0.2992117867602888,
+      "learning_rate": 0.00010719530102790014,
+      "loss": 1.7752,
+      "step": 365
+    },
+    {
+      "epoch": 0.1630674305861613,
+      "grad_norm": 0.3754242349038873,
+      "learning_rate": 0.00010866372980910426,
+      "loss": 1.8151,
+      "step": 370
+    },
+    {
+      "epoch": 0.16527104451300131,
+      "grad_norm": 0.36786720092684805,
+      "learning_rate": 0.00011013215859030836,
+      "loss": 1.8172,
+      "step": 375
+    },
+    {
+      "epoch": 0.16747465843984133,
+      "grad_norm": 0.3572694028170277,
+      "learning_rate": 0.00011160058737151249,
+      "loss": 1.8668,
+      "step": 380
+    },
+    {
+      "epoch": 0.16967827236668137,
+      "grad_norm": 0.34400588996372305,
+      "learning_rate": 0.00011306901615271659,
+      "loss": 1.9685,
+      "step": 385
+    },
+    {
+      "epoch": 0.17188188629352139,
+      "grad_norm": 0.3484332919699417,
+      "learning_rate": 0.00011453744493392071,
+      "loss": 1.999,
+      "step": 390
+    },
+    {
+      "epoch": 0.1740855002203614,
+      "grad_norm": 0.3167720096524829,
+      "learning_rate": 0.00011600587371512481,
+      "loss": 1.7002,
+      "step": 395
+    },
+    {
+      "epoch": 0.17628911414720141,
+      "grad_norm": 0.30865191358312394,
+      "learning_rate": 0.00011747430249632894,
+      "loss": 1.7808,
+      "step": 400
+    },
+    {
+      "epoch": 0.17849272807404143,
+      "grad_norm": 0.3635805195870158,
+      "learning_rate": 0.00011894273127753304,
+      "loss": 1.8711,
+      "step": 405
+    },
+    {
+      "epoch": 0.18069634200088144,
+      "grad_norm": 0.3113426681083048,
+      "learning_rate": 0.00012041116005873716,
+      "loss": 1.7559,
+      "step": 410
+    },
+    {
+      "epoch": 0.18289995592772146,
+      "grad_norm": 0.3500754402872819,
+      "learning_rate": 0.00012187958883994126,
+      "loss": 2.0623,
+      "step": 415
+    },
+    {
+      "epoch": 0.18510356985456147,
+      "grad_norm": 0.3002611868025747,
+      "learning_rate": 0.00012334801762114539,
+      "loss": 1.6155,
+      "step": 420
+    },
+    {
+      "epoch": 0.1873071837814015,
+      "grad_norm": 0.30728017118253237,
+      "learning_rate": 0.00012481644640234947,
+      "loss": 1.7865,
+      "step": 425
+    },
+    {
+      "epoch": 0.18951079770824153,
+      "grad_norm": 0.30435853108098426,
+      "learning_rate": 0.0001262848751835536,
+      "loss": 1.8101,
+      "step": 430
+    },
+    {
+      "epoch": 0.19171441163508154,
+      "grad_norm": 0.33728305221904875,
+      "learning_rate": 0.0001277533039647577,
+      "loss": 1.7445,
+      "step": 435
+    },
+    {
+      "epoch": 0.19391802556192156,
+      "grad_norm": 0.3269188152820616,
+      "learning_rate": 0.00012922173274596184,
+      "loss": 1.8443,
+      "step": 440
+    },
+    {
+      "epoch": 0.19612163948876157,
+      "grad_norm": 0.30824761799725153,
+      "learning_rate": 0.00013069016152716592,
+      "loss": 1.8385,
+      "step": 445
+    },
+    {
+      "epoch": 0.1983252534156016,
+      "grad_norm": 0.2692779779339547,
+      "learning_rate": 0.00013215859030837006,
+      "loss": 1.7256,
+      "step": 450
+    },
+    {
+      "epoch": 0.2005288673424416,
+      "grad_norm": 0.29633902366885784,
+      "learning_rate": 0.00013362701908957415,
+      "loss": 1.8496,
+      "step": 455
+    },
+    {
+      "epoch": 0.20273248126928162,
+      "grad_norm": 0.31921033190555953,
+      "learning_rate": 0.00013509544787077829,
+      "loss": 1.9118,
+      "step": 460
+    },
+    {
+      "epoch": 0.20493609519612163,
+      "grad_norm": 0.38557464681376724,
+      "learning_rate": 0.00013656387665198237,
+      "loss": 1.7992,
+      "step": 465
+    },
+    {
+      "epoch": 0.20713970912296165,
+      "grad_norm": 0.28175106471056965,
+      "learning_rate": 0.0001380323054331865,
+      "loss": 1.8768,
+      "step": 470
+    },
+    {
+      "epoch": 0.2093433230498017,
+      "grad_norm": 0.31789045276124145,
+      "learning_rate": 0.0001395007342143906,
+      "loss": 1.7628,
+      "step": 475
+    },
+    {
+      "epoch": 0.2115469369766417,
+      "grad_norm": 0.2412861323934925,
+      "learning_rate": 0.00014096916299559473,
+      "loss": 1.7988,
+      "step": 480
+    },
+    {
+      "epoch": 0.21375055090348172,
+      "grad_norm": 0.24858109179997467,
+      "learning_rate": 0.00014243759177679882,
+      "loss": 1.8673,
+      "step": 485
+    },
+    {
+      "epoch": 0.21595416483032173,
+      "grad_norm": 0.30755969433310765,
+      "learning_rate": 0.00014390602055800296,
+      "loss": 1.7105,
+      "step": 490
+    },
+    {
+      "epoch": 0.21815777875716175,
+      "grad_norm": 0.2249398504391689,
+      "learning_rate": 0.00014537444933920705,
+      "loss": 1.7837,
+      "step": 495
+    },
+    {
+      "epoch": 0.22036139268400176,
+      "grad_norm": 0.3335603893570301,
+      "learning_rate": 0.00014684287812041118,
+      "loss": 1.8333,
+      "step": 500
+    },
+    {
+      "epoch": 0.22256500661084178,
+      "grad_norm": 0.3038176355961198,
+      "learning_rate": 0.00014831130690161527,
+      "loss": 1.8275,
+      "step": 505
+    },
+    {
+      "epoch": 0.2247686205376818,
+      "grad_norm": 0.30793332193972345,
+      "learning_rate": 0.0001497797356828194,
+      "loss": 1.8177,
+      "step": 510
+    },
+    {
+      "epoch": 0.2269722344645218,
+      "grad_norm": 0.3072544700688745,
+      "learning_rate": 0.0001512481644640235,
+      "loss": 1.9828,
+      "step": 515
+    },
+    {
+      "epoch": 0.22917584839136185,
+      "grad_norm": 0.34999895392965585,
+      "learning_rate": 0.00015271659324522763,
+      "loss": 1.9117,
+      "step": 520
+    },
+    {
+      "epoch": 0.23137946231820186,
+      "grad_norm": 0.32911766969763456,
+      "learning_rate": 0.00015418502202643172,
+      "loss": 1.6697,
+      "step": 525
+    },
+    {
+      "epoch": 0.23358307624504188,
+      "grad_norm": 0.2617607674469746,
+      "learning_rate": 0.00015565345080763586,
+      "loss": 1.6296,
+      "step": 530
+    },
+    {
+      "epoch": 0.2357866901718819,
+      "grad_norm": 0.43182139308340584,
+      "learning_rate": 0.00015712187958883994,
+      "loss": 1.9794,
+      "step": 535
+    },
+    {
+      "epoch": 0.2379903040987219,
+      "grad_norm": 0.3127396374267501,
+      "learning_rate": 0.00015859030837004406,
+      "loss": 1.7436,
+      "step": 540
+    },
+    {
+      "epoch": 0.24019391802556192,
+      "grad_norm": 0.21415922561789702,
+      "learning_rate": 0.00016005873715124817,
+      "loss": 1.7742,
+      "step": 545
+    },
+    {
+      "epoch": 0.24239753195240193,
+      "grad_norm": 0.31852170762507637,
+      "learning_rate": 0.00016152716593245228,
+      "loss": 1.7969,
+      "step": 550
+    },
+    {
+      "epoch": 0.24460114587924195,
+      "grad_norm": 0.29928941424884764,
+      "learning_rate": 0.0001629955947136564,
+      "loss": 1.8178,
+      "step": 555
+    },
+    {
+      "epoch": 0.24680475980608196,
+      "grad_norm": 0.2145836014852306,
+      "learning_rate": 0.0001644640234948605,
+      "loss": 1.6149,
+      "step": 560
+    },
+    {
+      "epoch": 0.249008373732922,
+      "grad_norm": 0.2765877443446801,
+      "learning_rate": 0.00016593245227606462,
+      "loss": 1.859,
+      "step": 565
+    },
+    {
+      "epoch": 0.251211987659762,
+      "grad_norm": 0.26872348364190873,
+      "learning_rate": 0.00016740088105726873,
+      "loss": 1.8757,
+      "step": 570
+    },
+    {
+      "epoch": 0.25341560158660204,
+      "grad_norm": 0.3149952835357651,
+      "learning_rate": 0.00016886930983847284,
+      "loss": 1.869,
+      "step": 575
+    },
+    {
+      "epoch": 0.255619215513442,
+      "grad_norm": 0.2304679465113612,
+      "learning_rate": 0.00017033773861967696,
+      "loss": 1.7151,
+      "step": 580
+    },
+    {
+      "epoch": 0.25782282944028206,
+      "grad_norm": 0.24264862789474126,
+      "learning_rate": 0.00017180616740088107,
+      "loss": 2.0695,
+      "step": 585
+    },
+    {
+      "epoch": 0.2600264433671221,
+      "grad_norm": 0.2598067252093709,
+      "learning_rate": 0.00017327459618208518,
+      "loss": 1.839,
+      "step": 590
+    },
+    {
+      "epoch": 0.2622300572939621,
+      "grad_norm": 0.26705538323533523,
+      "learning_rate": 0.0001747430249632893,
+      "loss": 1.8008,
+      "step": 595
+    },
+    {
+      "epoch": 0.26443367122080214,
+      "grad_norm": 0.28479994342274373,
+      "learning_rate": 0.0001762114537444934,
+      "loss": 1.7284,
+      "step": 600
+    },
+    {
+      "epoch": 0.2666372851476421,
+      "grad_norm": 0.28497317726898896,
+      "learning_rate": 0.00017767988252569752,
+      "loss": 1.9214,
+      "step": 605
+    },
+    {
+      "epoch": 0.26884089907448216,
+      "grad_norm": 0.33639476801612694,
+      "learning_rate": 0.00017914831130690163,
+      "loss": 1.7516,
+      "step": 610
+    },
+    {
+      "epoch": 0.27104451300132215,
+      "grad_norm": 0.25526934350033054,
+      "learning_rate": 0.00018061674008810574,
+      "loss": 1.7688,
+      "step": 615
+    },
+    {
+      "epoch": 0.2732481269281622,
+      "grad_norm": 0.28869758623973935,
+      "learning_rate": 0.00018208516886930985,
+      "loss": 1.7843,
+      "step": 620
+    },
+    {
+      "epoch": 0.2754517408550022,
+      "grad_norm": 0.2782854068624019,
+      "learning_rate": 0.00018355359765051397,
+      "loss": 1.8081,
+      "step": 625
+    },
+    {
+      "epoch": 0.2776553547818422,
+      "grad_norm": 0.28975241284668296,
+      "learning_rate": 0.00018502202643171805,
+      "loss": 2.0243,
+      "step": 630
+    },
+    {
+      "epoch": 0.27985896870868227,
+      "grad_norm": 0.3308791482681284,
+      "learning_rate": 0.0001864904552129222,
+      "loss": 1.9301,
+      "step": 635
+    },
+    {
+      "epoch": 0.28206258263552225,
+      "grad_norm": 0.2725800531393519,
+      "learning_rate": 0.00018795888399412628,
+      "loss": 1.8126,
+      "step": 640
+    },
+    {
+      "epoch": 0.2842661965623623,
+      "grad_norm": 0.2347435274751105,
+      "learning_rate": 0.00018942731277533042,
+      "loss": 1.9526,
+      "step": 645
+    },
+    {
+      "epoch": 0.2864698104892023,
+      "grad_norm": 0.2594593727241724,
+      "learning_rate": 0.0001908957415565345,
+      "loss": 1.7362,
+      "step": 650
+    },
+    {
+      "epoch": 0.2886734244160423,
+      "grad_norm": 0.20428739622605385,
+      "learning_rate": 0.00019236417033773864,
+      "loss": 1.9143,
+      "step": 655
+    },
+    {
+      "epoch": 0.2908770383428823,
+      "grad_norm": 0.35231366470248277,
+      "learning_rate": 0.00019383259911894273,
+      "loss": 1.882,
+      "step": 660
+    },
+    {
+      "epoch": 0.29308065226972235,
+      "grad_norm": 0.28303949235132847,
+      "learning_rate": 0.00019530102790014687,
+      "loss": 1.9956,
+      "step": 665
+    },
+    {
+      "epoch": 0.29528426619656234,
+      "grad_norm": 0.28380137271318484,
+      "learning_rate": 0.00019676945668135095,
+      "loss": 1.9298,
+      "step": 670
+    },
+    {
+      "epoch": 0.2974878801234024,
+      "grad_norm": 0.29197603939206423,
+      "learning_rate": 0.0001982378854625551,
+      "loss": 1.9936,
+      "step": 675
+    },
+    {
+      "epoch": 0.2996914940502424,
+      "grad_norm": 0.3192279737304851,
+      "learning_rate": 0.00019970631424375918,
+      "loss": 1.856,
+      "step": 680
+    },
+    {
+      "epoch": 0.3018951079770824,
+      "grad_norm": 0.31555391042987874,
+      "learning_rate": 0.00019999978960491256,
+      "loss": 1.9492,
+      "step": 685
+    },
+    {
+      "epoch": 0.30409872190392245,
+      "grad_norm": 0.26468213545329267,
+      "learning_rate": 0.0001999989348763872,
+      "loss": 1.867,
+      "step": 690
+    },
+    {
+      "epoch": 0.30630233583076244,
+      "grad_norm": 0.2591553540599883,
+      "learning_rate": 0.0001999974226703463,
+      "loss": 1.7565,
+      "step": 695
+    },
+    {
+      "epoch": 0.3085059497576025,
+      "grad_norm": 0.33857027825097774,
+      "learning_rate": 0.00019999525299673244,
+      "loss": 1.8407,
+      "step": 700
+    },
+    {
+      "epoch": 0.31070956368444247,
+      "grad_norm": 0.27554046406812405,
+      "learning_rate": 0.0001999924258698108,
+      "loss": 1.8449,
+      "step": 705
+    },
+    {
+      "epoch": 0.3129131776112825,
+      "grad_norm": 0.2906036501277061,
+      "learning_rate": 0.0001999889413081694,
+      "loss": 1.9425,
+      "step": 710
+    },
+    {
+      "epoch": 0.3151167915381225,
+      "grad_norm": 0.24516499633235853,
+      "learning_rate": 0.00019998479933471862,
+      "loss": 1.8373,
+      "step": 715
+    },
+    {
+      "epoch": 0.31732040546496254,
+      "grad_norm": 0.261325768775238,
+      "learning_rate": 0.0001999799999766913,
+      "loss": 1.8655,
+      "step": 720
+    },
+    {
+      "epoch": 0.3195240193918026,
+      "grad_norm": 0.26292913191780404,
+      "learning_rate": 0.00019997454326564252,
+      "loss": 1.8011,
+      "step": 725
+    },
+    {
+      "epoch": 0.32172763331864257,
+      "grad_norm": 0.24941214673161538,
+      "learning_rate": 0.0001999684292374493,
+      "loss": 1.7063,
+      "step": 730
+    },
+    {
+      "epoch": 0.3239312472454826,
+      "grad_norm": 0.285047883815678,
+      "learning_rate": 0.00019996165793231038,
+      "loss": 1.9537,
+      "step": 735
+    },
+    {
+      "epoch": 0.3261348611723226,
+      "grad_norm": 0.2853612548244496,
+      "learning_rate": 0.0001999542293947461,
+      "loss": 1.7641,
+      "step": 740
+    },
+    {
+      "epoch": 0.32833847509916264,
+      "grad_norm": 0.2600200330765211,
+      "learning_rate": 0.00019994614367359792,
+      "loss": 1.8886,
+      "step": 745
+    },
+    {
+      "epoch": 0.33054208902600263,
+      "grad_norm": 0.36999348537839605,
+      "learning_rate": 0.00019993740082202818,
+      "loss": 1.798,
+      "step": 750
+    },
+    {
+      "epoch": 0.33274570295284267,
+      "grad_norm": 0.24647791783433215,
+      "learning_rate": 0.00019992800089751984,
+      "loss": 1.8922,
+      "step": 755
+    },
+    {
+      "epoch": 0.33494931687968266,
+      "grad_norm": 0.23414809097297196,
+      "learning_rate": 0.0001999179439618759,
+      "loss": 1.6675,
+      "step": 760
+    },
+    {
+      "epoch": 0.3371529308065227,
+      "grad_norm": 0.24171495213943817,
+      "learning_rate": 0.00019990723008121917,
+      "loss": 1.5054,
+      "step": 765
+    },
+    {
+      "epoch": 0.33935654473336274,
+      "grad_norm": 0.2462163957540956,
+      "learning_rate": 0.00019989585932599172,
+      "loss": 1.8441,
+      "step": 770
+    },
+    {
+      "epoch": 0.34156015866020273,
+      "grad_norm": 0.24192478872496498,
+      "learning_rate": 0.00019988383177095459,
+      "loss": 1.8535,
+      "step": 775
+    },
+    {
+      "epoch": 0.34376377258704277,
+      "grad_norm": 0.21554799200025276,
+      "learning_rate": 0.000199871147495187,
+      "loss": 1.7299,
+      "step": 780
+    },
+    {
+      "epoch": 0.34596738651388276,
+      "grad_norm": 0.29240426650240625,
+      "learning_rate": 0.00019985780658208618,
+      "loss": 1.9846,
+      "step": 785
+    },
+    {
+      "epoch": 0.3481710004407228,
+      "grad_norm": 0.32582827846868756,
+      "learning_rate": 0.00019984380911936648,
+      "loss": 1.5922,
+      "step": 790
+    },
+    {
+      "epoch": 0.3503746143675628,
+      "grad_norm": 0.3253623043496682,
+      "learning_rate": 0.00019982915519905912,
+      "loss": 1.7138,
+      "step": 795
+    },
+    {
+      "epoch": 0.35257822829440283,
+      "grad_norm": 0.3012253411690825,
+      "learning_rate": 0.00019981384491751133,
+      "loss": 1.9526,
+      "step": 800
+    },
+    {
+      "epoch": 0.3547818422212428,
+      "grad_norm": 0.32373710599771705,
+      "learning_rate": 0.00019979787837538587,
+      "loss": 2.0799,
+      "step": 805
+    },
+    {
+      "epoch": 0.35698545614808286,
+      "grad_norm": 0.2736459891711902,
+      "learning_rate": 0.00019978125567766023,
+      "loss": 1.8422,
+      "step": 810
+    },
+    {
+      "epoch": 0.3591890700749229,
+      "grad_norm": 0.28970333253517644,
+      "learning_rate": 0.00019976397693362614,
+      "loss": 1.8309,
+      "step": 815
+    },
+    {
+      "epoch": 0.3613926840017629,
+      "grad_norm": 0.271249926943477,
+      "learning_rate": 0.0001997460422568886,
+      "loss": 1.6581,
+      "step": 820
+    },
+    {
+      "epoch": 0.36359629792860293,
+      "grad_norm": 0.25582927997161137,
+      "learning_rate": 0.00019972745176536537,
+      "loss": 1.9441,
+      "step": 825
+    },
+    {
+      "epoch": 0.3657999118554429,
+      "grad_norm": 0.27987061803178176,
+      "learning_rate": 0.00019970820558128604,
+      "loss": 1.8015,
+      "step": 830
+    },
+    {
+      "epoch": 0.36800352578228296,
+      "grad_norm": 0.2956257318754398,
+      "learning_rate": 0.0001996883038311913,
+      "loss": 1.7853,
+      "step": 835
+    },
+    {
+      "epoch": 0.37020713970912295,
+      "grad_norm": 0.25727871315384043,
+      "learning_rate": 0.00019966774664593206,
+      "loss": 1.7594,
+      "step": 840
+    },
+    {
+      "epoch": 0.372410753635963,
+      "grad_norm": 0.3069722747545403,
+      "learning_rate": 0.00019964653416066868,
+      "loss": 1.9102,
+      "step": 845
+    },
+    {
+      "epoch": 0.374614367562803,
+      "grad_norm": 0.362312858514461,
+      "learning_rate": 0.0001996246665148699,
+      "loss": 1.8419,
+      "step": 850
+    },
+    {
+      "epoch": 0.376817981489643,
+      "grad_norm": 0.30043599837768675,
+      "learning_rate": 0.00019960214385231217,
+      "loss": 1.9281,
+      "step": 855
+    },
+    {
+      "epoch": 0.37902159541648306,
+      "grad_norm": 0.25742900200071334,
+      "learning_rate": 0.00019957896632107845,
+      "loss": 1.8382,
+      "step": 860
+    },
+    {
+      "epoch": 0.38122520934332305,
+      "grad_norm": 0.22904042108335546,
+      "learning_rate": 0.00019955513407355743,
+      "loss": 1.585,
+      "step": 865
+    },
+    {
+      "epoch": 0.3834288232701631,
+      "grad_norm": 0.35248924984262536,
+      "learning_rate": 0.0001995306472664425,
+      "loss": 1.8779,
+      "step": 870
+    },
+    {
+      "epoch": 0.3856324371970031,
+      "grad_norm": 0.21394107686214808,
+      "learning_rate": 0.00019950550606073056,
+      "loss": 1.6203,
+      "step": 875
+    },
+    {
+      "epoch": 0.3878360511238431,
+      "grad_norm": 0.23210191109497874,
+      "learning_rate": 0.00019947971062172118,
+      "loss": 1.7579,
+      "step": 880
+    },
+    {
+      "epoch": 0.3900396650506831,
+      "grad_norm": 0.280197585305159,
+      "learning_rate": 0.00019945326111901542,
+      "loss": 1.8697,
+      "step": 885
+    },
+    {
+      "epoch": 0.39224327897752315,
+      "grad_norm": 0.2007938963214883,
+      "learning_rate": 0.00019942615772651455,
+      "loss": 1.6718,
+      "step": 890
+    },
+    {
+      "epoch": 0.39444689290436313,
+      "grad_norm": 0.28396842059026434,
+      "learning_rate": 0.0001993984006224193,
+      "loss": 1.8261,
+      "step": 895
+    },
+    {
+      "epoch": 0.3966505068312032,
+      "grad_norm": 0.2970710682132302,
+      "learning_rate": 0.00019936998998922826,
+      "loss": 1.8988,
+      "step": 900
+    },
+    {
+      "epoch": 0.3988541207580432,
+      "grad_norm": 0.4794659287163716,
+      "learning_rate": 0.00019934092601373694,
+      "loss": 1.8387,
+      "step": 905
+    },
+    {
+      "epoch": 0.4010577346848832,
+      "grad_norm": 0.2810211439045603,
+      "learning_rate": 0.00019931120888703652,
+      "loss": 1.7516,
+      "step": 910
+    },
+    {
+      "epoch": 0.40326134861172325,
+      "grad_norm": 0.28984779802870314,
+      "learning_rate": 0.0001992808388045125,
+      "loss": 1.9212,
+      "step": 915
+    },
+    {
+      "epoch": 0.40546496253856323,
+      "grad_norm": 0.26294811280148017,
+      "learning_rate": 0.00019924981596584345,
+      "loss": 1.8798,
+      "step": 920
+    },
+    {
+      "epoch": 0.4076685764654033,
+      "grad_norm": 0.27546864810479915,
+      "learning_rate": 0.00019921814057499978,
+      "loss": 1.7595,
+      "step": 925
+    },
+    {
+      "epoch": 0.40987219039224326,
+      "grad_norm": 0.2506019656377615,
+      "learning_rate": 0.0001991858128402422,
+      "loss": 1.7625,
+      "step": 930
+    },
+    {
+      "epoch": 0.4120758043190833,
+      "grad_norm": 0.35470941281384444,
+      "learning_rate": 0.0001991528329741206,
+      "loss": 1.8644,
+      "step": 935
+    },
+    {
+      "epoch": 0.4142794182459233,
+      "grad_norm": 0.30456048523132856,
+      "learning_rate": 0.00019911920119347254,
+      "loss": 1.8427,
+      "step": 940
+    },
+    {
+      "epoch": 0.41648303217276333,
+      "grad_norm": 0.21180617855630868,
+      "learning_rate": 0.0001990849177194217,
+      "loss": 1.947,
+      "step": 945
+    },
+    {
+      "epoch": 0.4186866460996034,
+      "grad_norm": 0.27175217869149726,
+      "learning_rate": 0.00019904998277737668,
+      "loss": 1.5794,
+      "step": 950
+    },
+    {
+      "epoch": 0.42089026002644336,
+      "grad_norm": 0.25340188979799705,
+      "learning_rate": 0.00019901439659702924,
+      "loss": 1.655,
+      "step": 955
+    },
+    {
+      "epoch": 0.4230938739532834,
+      "grad_norm": 0.32039092541913483,
+      "learning_rate": 0.00019897815941235307,
+      "loss": 1.9448,
+      "step": 960
+    },
+    {
+      "epoch": 0.4252974878801234,
+      "grad_norm": 0.28203099326236186,
+      "learning_rate": 0.00019894127146160204,
+      "loss": 1.73,
+      "step": 965
+    },
+    {
+      "epoch": 0.42750110180696343,
+      "grad_norm": 0.790698073004524,
+      "learning_rate": 0.00019890373298730868,
+      "loss": 2.0466,
+      "step": 970
+    },
+    {
+      "epoch": 0.4297047157338034,
+      "grad_norm": 0.23107801878483133,
+      "learning_rate": 0.0001988655442362827,
+      "loss": 1.5805,
+      "step": 975
+    },
+    {
+      "epoch": 0.43190832966064346,
+      "grad_norm": 0.2396293478769625,
+      "learning_rate": 0.00019882670545960914,
+      "loss": 1.7482,
+      "step": 980
+    },
+    {
+      "epoch": 0.43411194358748345,
+      "grad_norm": 0.29159214942433453,
+      "learning_rate": 0.00019878721691264704,
+      "loss": 1.9851,
+      "step": 985
+    },
+    {
+      "epoch": 0.4363155575143235,
+      "grad_norm": 0.2825006648123125,
+      "learning_rate": 0.00019874707885502745,
+      "loss": 1.7534,
+      "step": 990
+    },
+    {
+      "epoch": 0.43851917144116354,
+      "grad_norm": 0.28034425756379244,
+      "learning_rate": 0.00019870629155065186,
+      "loss": 1.7489,
+      "step": 995
+    },
+    {
+      "epoch": 0.4407227853680035,
+      "grad_norm": 0.3181059895374247,
+      "learning_rate": 0.0001986648552676905,
+      "loss": 1.8798,
+      "step": 1000
+    },
+    {
+      "epoch": 0.44292639929484356,
+      "grad_norm": 0.20367681253555586,
+      "learning_rate": 0.0001986227702785805,
+      "loss": 1.8065,
+      "step": 1005
+    },
+    {
+      "epoch": 0.44513001322168355,
+      "grad_norm": 0.25873539379499144,
+      "learning_rate": 0.0001985800368600242,
+      "loss": 1.687,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4473336271485236,
+      "grad_norm": 0.2342074611961214,
+      "learning_rate": 0.0001985366552929871,
+      "loss": 1.9431,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4495372410753636,
+      "grad_norm": 0.48215863294928724,
+      "learning_rate": 0.00019849262586269642,
+      "loss": 1.8137,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4517408550022036,
+      "grad_norm": 0.27724392120608005,
+      "learning_rate": 0.00019844794885863877,
+      "loss": 1.8311,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4539444689290436,
+      "grad_norm": 0.22655546528476625,
+      "learning_rate": 0.00019840262457455855,
+      "loss": 1.6968,
+      "step": 1030
+    },
+    {
+      "epoch": 0.45614808285588365,
+      "grad_norm": 0.24804519754733267,
+      "learning_rate": 0.00019835665330845595,
+      "loss": 1.844,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4583516967827237,
+      "grad_norm": 0.3053870909241666,
+      "learning_rate": 0.00019831003536258487,
+      "loss": 1.6674,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4605553107095637,
+      "grad_norm": 0.2677054196356507,
+      "learning_rate": 0.00019826277104345109,
+      "loss": 1.994,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4627589246364037,
+      "grad_norm": 0.2332936772139347,
+      "learning_rate": 0.0001982148606618102,
+      "loss": 1.7963,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4649625385632437,
+      "grad_norm": 0.290304703836439,
+      "learning_rate": 0.00019816630453266555,
+      "loss": 1.8278,
+      "step": 1055
+    },
+    {
+      "epoch": 0.46716615249008375,
+      "grad_norm": 0.31440944408786714,
+      "learning_rate": 0.0001981171029752662,
+      "loss": 1.7053,
+      "step": 1060
+    },
+    {
+      "epoch": 0.46936976641692374,
+      "grad_norm": 0.31837742235142563,
+      "learning_rate": 0.00019806725631310476,
+      "loss": 1.8377,
+      "step": 1065
+    },
+    {
+      "epoch": 0.4715733803437638,
+      "grad_norm": 0.31835369363351046,
+      "learning_rate": 0.00019801676487391529,
+      "loss": 1.7635,
+      "step": 1070
+    },
+    {
+      "epoch": 0.47377699427060377,
+      "grad_norm": 0.32008261758346795,
+      "learning_rate": 0.0001979656289896712,
+      "loss": 1.8322,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4759806081974438,
+      "grad_norm": 0.2436306554476478,
+      "learning_rate": 0.000197913848996583,
+      "loss": 1.7057,
+      "step": 1080
+    },
+    {
+      "epoch": 0.47818422212428385,
+      "grad_norm": 0.24420764132400904,
+      "learning_rate": 0.00019786142523509615,
+      "loss": 1.7756,
+      "step": 1085
+    },
+    {
+      "epoch": 0.48038783605112384,
+      "grad_norm": 0.24156752527695532,
+      "learning_rate": 0.00019780835804988876,
+      "loss": 1.7788,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4825914499779639,
+      "grad_norm": 0.3341728235303617,
+      "learning_rate": 0.00019775464778986934,
+      "loss": 1.9,
+      "step": 1095
+    },
+    {
+      "epoch": 0.48479506390480387,
+      "grad_norm": 0.23920551280340352,
+      "learning_rate": 0.00019770029480817454,
+      "loss": 1.8956,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4869986778316439,
+      "grad_norm": 0.2463312946508512,
+      "learning_rate": 0.00019764529946216682,
+      "loss": 1.5345,
+      "step": 1105
+    },
+    {
+      "epoch": 0.4892022917584839,
+      "grad_norm": 0.26366246938923554,
+      "learning_rate": 0.00019758966211343206,
+      "loss": 1.7621,
+      "step": 1110
+    },
+    {
+      "epoch": 0.49140590568532394,
+      "grad_norm": 0.3123943209000793,
+      "learning_rate": 0.00019753338312777718,
+      "loss": 1.815,
+      "step": 1115
+    },
+    {
+      "epoch": 0.4936095196121639,
+      "grad_norm": 0.2551443475914167,
+      "learning_rate": 0.00019747646287522784,
+      "loss": 1.8611,
+      "step": 1120
+    },
+    {
+      "epoch": 0.49581313353900397,
+      "grad_norm": 0.2332525414967985,
+      "learning_rate": 0.0001974189017300259,
+      "loss": 1.5384,
+      "step": 1125
+    },
+    {
+      "epoch": 0.498016747465844,
+      "grad_norm": 0.2861706387721841,
+      "learning_rate": 0.00019736070007062692,
+      "loss": 2.029,
+      "step": 1130
+    },
+    {
+      "epoch": 0.500220361392684,
+      "grad_norm": 0.4583730418300565,
+      "learning_rate": 0.00019730185827969784,
+      "loss": 1.6826,
+      "step": 1135
+    },
+    {
+      "epoch": 0.502423975319524,
+      "grad_norm": 0.1931367375627417,
+      "learning_rate": 0.00019724237674411432,
+      "loss": 1.6877,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5046275892463641,
+      "grad_norm": 0.3082508866109686,
+      "learning_rate": 0.00019718225585495824,
+      "loss": 1.8148,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5068312031732041,
+      "grad_norm": 0.18106453992956664,
+      "learning_rate": 0.00019712149600751517,
+      "loss": 1.6556,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5090348171000441,
+      "grad_norm": 0.23000297672552616,
+      "learning_rate": 0.00019706009760127164,
+      "loss": 2.0152,
+      "step": 1155
+    },
+    {
+      "epoch": 0.511238431026884,
+      "grad_norm": 0.29387352965288377,
+      "learning_rate": 0.00019699806103991272,
+      "loss": 1.7962,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5134420449537241,
+      "grad_norm": 0.23060861917057546,
+      "learning_rate": 0.00019693538673131917,
+      "loss": 1.8123,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5156456588805641,
+      "grad_norm": 0.25329490761690543,
+      "learning_rate": 0.00019687207508756486,
+      "loss": 1.7052,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5178492728074041,
+      "grad_norm": 0.3356944723720528,
+      "learning_rate": 0.00019680812652491408,
+      "loss": 1.7985,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5200528867342442,
+      "grad_norm": 0.31492967037029806,
+      "learning_rate": 0.0001967435414638187,
+      "loss": 1.7971,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5222565006610842,
+      "grad_norm": 0.2915314540239125,
+      "learning_rate": 0.00019667832032891554,
+      "loss": 1.9571,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5244601145879242,
+      "grad_norm": 0.28065572919249093,
+      "learning_rate": 0.00019661246354902342,
+      "loss": 1.9185,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5266637285147642,
+      "grad_norm": 0.22888871376259906,
+      "learning_rate": 0.00019654597155714044,
+      "loss": 1.7367,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5288673424416043,
+      "grad_norm": 0.23069615225682083,
+      "learning_rate": 0.00019647884479044123,
+      "loss": 1.7333,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5310709563684443,
+      "grad_norm": 0.25845750908289505,
+      "learning_rate": 0.00019641108369027385,
+      "loss": 1.5907,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5332745702952842,
+      "grad_norm": 0.3069614304651339,
+      "learning_rate": 0.00019634268870215703,
+      "loss": 1.9282,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5354781842221242,
+      "grad_norm": 0.3005722535622308,
+      "learning_rate": 0.00019627366027577726,
+      "loss": 1.6378,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5376817981489643,
+      "grad_norm": 0.270244780455395,
+      "learning_rate": 0.00019620399886498578,
+      "loss": 1.6499,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5398854120758043,
+      "grad_norm": 0.2796374681033303,
+      "learning_rate": 0.0001961337049277955,
+      "loss": 1.7962,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5420890260026443,
+      "grad_norm": 0.2608673103569785,
+      "learning_rate": 0.00019606277892637823,
+      "loss": 1.6946,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5442926399294844,
+      "grad_norm": 0.35197474426731085,
+      "learning_rate": 0.00019599122132706146,
+      "loss": 1.9751,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5464962538563244,
+      "grad_norm": 0.26530267535084306,
+      "learning_rate": 0.0001959190326003253,
+      "loss": 1.7257,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5486998677831644,
+      "grad_norm": 0.2540337035705083,
+      "learning_rate": 0.00019584621322079942,
+      "loss": 1.8693,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5509034817100044,
+      "grad_norm": 0.2530867344602867,
+      "learning_rate": 0.00019577276366726003,
+      "loss": 1.6761,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5531070956368445,
+      "grad_norm": 0.2549205571103006,
+      "learning_rate": 0.00019569868442262655,
+      "loss": 1.8729,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5553107095636844,
+      "grad_norm": 0.2871173332255517,
+      "learning_rate": 0.00019562397597395857,
+      "loss": 1.8347,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5575143234905244,
+      "grad_norm": 0.28369117286467893,
+      "learning_rate": 0.0001955486388124525,
+      "loss": 1.8547,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5597179374173645,
+      "grad_norm": 0.2852722364768899,
+      "learning_rate": 0.00019547267343343857,
+      "loss": 1.6552,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5619215513442045,
+      "grad_norm": 0.34585189639791736,
+      "learning_rate": 0.0001953960803363774,
+      "loss": 1.6727,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5641251652710445,
+      "grad_norm": 0.2767216642372831,
+      "learning_rate": 0.00019531886002485674,
+      "loss": 1.8886,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5663287791978845,
+      "grad_norm": 0.20036600537602942,
+      "learning_rate": 0.00019524101300658813,
+      "loss": 1.8685,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5685323931247246,
+      "grad_norm": 0.261316473836974,
+      "learning_rate": 0.0001951625397934037,
+      "loss": 1.782,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5707360070515646,
+      "grad_norm": 0.4236076570583356,
+      "learning_rate": 0.0001950834409012527,
+      "loss": 1.8318,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5729396209784046,
+      "grad_norm": 0.28925306291568,
+      "learning_rate": 0.00019500371685019806,
+      "loss": 1.6012,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5751432349052445,
+      "grad_norm": 0.2945702905493002,
+      "learning_rate": 0.0001949233681644131,
+      "loss": 1.9158,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5773468488320846,
+      "grad_norm": 0.27644920036772963,
+      "learning_rate": 0.00019484239537217798,
+      "loss": 1.8232,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5795504627589246,
+      "grad_norm": 0.24910322311175717,
+      "learning_rate": 0.00019476079900587626,
+      "loss": 2.0731,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5817540766857646,
+      "grad_norm": 0.2746087574116961,
+      "learning_rate": 0.00019467857960199142,
+      "loss": 1.8429,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5839576906126047,
+      "grad_norm": 0.30262920301398166,
+      "learning_rate": 0.00019459573770110335,
+      "loss": 1.7647,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5861613045394447,
+      "grad_norm": 0.2570054670626858,
+      "learning_rate": 0.0001945122738478847,
+      "loss": 1.7511,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5883649184662847,
+      "grad_norm": 0.2506934757685486,
+      "learning_rate": 0.00019442818859109737,
+      "loss": 1.9036,
+      "step": 1335
+    },
+    {
+      "epoch": 0.5905685323931247,
+      "grad_norm": 0.28984253571835894,
+      "learning_rate": 0.00019434348248358892,
+      "loss": 1.7763,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5927721463199648,
+      "grad_norm": 0.28120650964379307,
+      "learning_rate": 0.00019425815608228888,
+      "loss": 1.8062,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5949757602468048,
+      "grad_norm": 0.26796794914129696,
+      "learning_rate": 0.00019417220994820514,
+      "loss": 1.7886,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5971793741736448,
+      "grad_norm": 0.27855604023858827,
+      "learning_rate": 0.00019408564464642024,
+      "loss": 1.6226,
+      "step": 1355
+    },
+    {
+      "epoch": 0.5993829881004848,
+      "grad_norm": 0.23996517928921976,
+      "learning_rate": 0.00019399846074608757,
+      "loss": 1.8206,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6015866020273248,
+      "grad_norm": 0.2718103628870133,
+      "learning_rate": 0.00019391065882042786,
+      "loss": 1.622,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6037902159541648,
+      "grad_norm": 0.40326599118637163,
+      "learning_rate": 0.00019382223944672516,
+      "loss": 1.8595,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6059938298810048,
+      "grad_norm": 0.26188146699351833,
+      "learning_rate": 0.00019373320320632313,
+      "loss": 1.719,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6081974438078449,
+      "grad_norm": 0.26353376595142103,
+      "learning_rate": 0.00019364355068462126,
+      "loss": 1.7599,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6104010577346849,
+      "grad_norm": 0.27506310894040453,
+      "learning_rate": 0.00019355328247107106,
+      "loss": 1.6895,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6126046716615249,
+      "grad_norm": 0.23293714707305346,
+      "learning_rate": 0.00019346239915917204,
+      "loss": 1.9199,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6148082855883649,
+      "grad_norm": 0.2728024319863431,
+      "learning_rate": 0.00019337090134646787,
+      "loss": 1.6137,
+      "step": 1395
+    },
+    {
+      "epoch": 0.617011899515205,
+      "grad_norm": 0.2580080135345881,
+      "learning_rate": 0.00019327878963454253,
+      "loss": 1.9251,
+      "step": 1400
+    },
+    {
+      "epoch": 0.619215513442045,
+      "grad_norm": 0.18292171734629373,
+      "learning_rate": 0.00019318606462901625,
+      "loss": 1.6127,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6214191273688849,
+      "grad_norm": 0.2540991707755816,
+      "learning_rate": 0.0001930927269395416,
+      "loss": 1.716,
+      "step": 1410
+    },
+    {
+      "epoch": 0.623622741295725,
+      "grad_norm": 0.2222252074542422,
+      "learning_rate": 0.00019299877717979944,
+      "loss": 1.649,
+      "step": 1415
+    },
+    {
+      "epoch": 0.625826355222565,
+      "grad_norm": 0.23656133580313402,
+      "learning_rate": 0.00019290421596749487,
+      "loss": 1.7321,
+      "step": 1420
+    },
+    {
+      "epoch": 0.628029969149405,
+      "grad_norm": 0.2770878731693278,
+      "learning_rate": 0.00019280904392435328,
+      "loss": 1.7982,
+      "step": 1425
+    },
+    {
+      "epoch": 0.630233583076245,
+      "grad_norm": 0.38713553349288304,
+      "learning_rate": 0.00019271326167611606,
+      "loss": 1.757,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6324371970030851,
+      "grad_norm": 0.21826448168974197,
+      "learning_rate": 0.00019261686985253668,
+      "loss": 1.6568,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6346408109299251,
+      "grad_norm": 0.28392855935252787,
+      "learning_rate": 0.00019251986908737646,
+      "loss": 1.6995,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6368444248567651,
+      "grad_norm": 0.3335345636849735,
+      "learning_rate": 0.00019242226001840043,
+      "loss": 1.6445,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6390480387836052,
+      "grad_norm": 0.23367317521663916,
+      "learning_rate": 0.0001923240432873731,
+      "loss": 1.7995,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6412516527104452,
+      "grad_norm": 0.26851942754392183,
+      "learning_rate": 0.00019222521954005424,
+      "loss": 1.8078,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6434552666372851,
+      "grad_norm": 0.28393938915488615,
+      "learning_rate": 0.00019212578942619474,
+      "loss": 1.7108,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6456588805641251,
+      "grad_norm": 0.2481400915816766,
+      "learning_rate": 0.00019202575359953213,
+      "loss": 1.7509,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6478624944909652,
+      "grad_norm": 0.22200215432596082,
+      "learning_rate": 0.00019192511271778656,
+      "loss": 1.6549,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6500661084178052,
+      "grad_norm": 0.24388551441860462,
+      "learning_rate": 0.00019182386744265623,
+      "loss": 1.9977,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6522697223446452,
+      "grad_norm": 0.28231311563927136,
+      "learning_rate": 0.00019172201843981314,
+      "loss": 1.7473,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6544733362714852,
+      "grad_norm": 0.2911468005732327,
+      "learning_rate": 0.00019161956637889872,
+      "loss": 1.8572,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6566769501983253,
+      "grad_norm": 0.2809891282005259,
+      "learning_rate": 0.0001915165119335194,
+      "loss": 1.6363,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6588805641251653,
+      "grad_norm": 0.257092783900919,
+      "learning_rate": 0.0001914128557812422,
+      "loss": 1.6894,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6610841780520053,
+      "grad_norm": 0.25492249438744313,
+      "learning_rate": 0.00019130859860359026,
+      "loss": 1.9549,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6632877919788454,
+      "grad_norm": 0.29656417671091373,
+      "learning_rate": 0.00019120374108603843,
+      "loss": 1.882,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6654914059056853,
+      "grad_norm": 0.26795937749586346,
+      "learning_rate": 0.0001910982839180086,
+      "loss": 1.7532,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6676950198325253,
+      "grad_norm": 0.30019068058678056,
+      "learning_rate": 0.0001909922277928654,
+      "loss": 1.9185,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6698986337593653,
+      "grad_norm": 0.2908491561313871,
+      "learning_rate": 0.00019088557340791136,
+      "loss": 1.8659,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6721022476862054,
+      "grad_norm": 0.2664246491795904,
+      "learning_rate": 0.00019077832146438257,
+      "loss": 1.618,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6743058616130454,
+      "grad_norm": 0.2670150987011056,
+      "learning_rate": 0.00019067047266744396,
+      "loss": 1.82,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6765094755398854,
+      "grad_norm": 0.26014739460054054,
+      "learning_rate": 0.0001905620277261847,
+      "loss": 1.8267,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6787130894667255,
+      "grad_norm": 0.39761269473071104,
+      "learning_rate": 0.00019045298735361345,
+      "loss": 1.7682,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6809167033935655,
+      "grad_norm": 0.25574765659361764,
+      "learning_rate": 0.0001903433522666538,
+      "loss": 1.8709,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6831203173204055,
+      "grad_norm": 0.2633667660696579,
+      "learning_rate": 0.00019023312318613945,
+      "loss": 1.7439,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6853239312472454,
+      "grad_norm": 0.33365458542367915,
+      "learning_rate": 0.00019012230083680954,
+      "loss": 1.8991,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6875275451740855,
+      "grad_norm": 0.26233949374382926,
+      "learning_rate": 0.0001900108859473039,
+      "loss": 1.7681,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6897311591009255,
+      "grad_norm": 0.27517789876709425,
+      "learning_rate": 0.00018989887925015814,
+      "loss": 1.8164,
+      "step": 1565
+    },
+    {
+      "epoch": 0.6919347730277655,
+      "grad_norm": 0.22352922332415057,
+      "learning_rate": 0.00018978628148179897,
+      "loss": 1.6674,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6941383869546055,
+      "grad_norm": 0.2705784713837994,
+      "learning_rate": 0.0001896730933825393,
+      "loss": 1.883,
+      "step": 1575
+    },
+    {
+      "epoch": 0.6963420008814456,
+      "grad_norm": 0.2333580239686404,
+      "learning_rate": 0.00018955931569657333,
+      "loss": 1.6889,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6985456148082856,
+      "grad_norm": 0.29573837903023903,
+      "learning_rate": 0.00018944494917197172,
+      "loss": 1.8473,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7007492287351256,
+      "grad_norm": 0.24394341802838163,
+      "learning_rate": 0.00018932999456067675,
+      "loss": 1.844,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7029528426619657,
+      "grad_norm": 0.2454925329534158,
+      "learning_rate": 0.0001892144526184971,
+      "loss": 1.7824,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7051564565888057,
+      "grad_norm": 0.23532373440650545,
+      "learning_rate": 0.00018909832410510315,
+      "loss": 1.8537,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7073600705156456,
+      "grad_norm": 0.2602720785746293,
+      "learning_rate": 0.00018898160978402198,
+      "loss": 1.8717,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7095636844424856,
+      "grad_norm": 0.23292617880883168,
+      "learning_rate": 0.00018886431042263208,
+      "loss": 1.703,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7117672983693257,
+      "grad_norm": 0.32473590479932374,
+      "learning_rate": 0.0001887464267921587,
+      "loss": 1.7075,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7139709122961657,
+      "grad_norm": 0.21131060430922607,
+      "learning_rate": 0.00018862795966766833,
+      "loss": 1.6993,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7161745262230057,
+      "grad_norm": 0.2945260951578129,
+      "learning_rate": 0.0001885089098280641,
+      "loss": 1.7315,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7183781401498458,
+      "grad_norm": 0.2988718316963067,
+      "learning_rate": 0.0001883892780560802,
+      "loss": 1.7079,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7205817540766858,
+      "grad_norm": 0.2671861990469183,
+      "learning_rate": 0.00018826906513827704,
+      "loss": 1.8816,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7227853680035258,
+      "grad_norm": 0.3539590759381266,
+      "learning_rate": 0.00018814827186503595,
+      "loss": 1.7559,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7249889819303658,
+      "grad_norm": 0.2098483762554333,
+      "learning_rate": 0.00018802689903055396,
+      "loss": 1.8296,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7271925958572059,
+      "grad_norm": 0.24342758941079398,
+      "learning_rate": 0.0001879049474328387,
+      "loss": 1.8845,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7293962097840458,
+      "grad_norm": 0.30213591141833546,
+      "learning_rate": 0.00018778241787370303,
+      "loss": 1.7739,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7315998237108858,
+      "grad_norm": 0.2796229215884198,
+      "learning_rate": 0.00018765931115875985,
+      "loss": 1.7238,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7338034376377258,
+      "grad_norm": 0.2818539631618944,
+      "learning_rate": 0.00018753562809741673,
+      "loss": 1.7833,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7360070515645659,
+      "grad_norm": 0.2534113717910187,
+      "learning_rate": 0.00018741136950287067,
+      "loss": 1.781,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7382106654914059,
+      "grad_norm": 0.25817838823410116,
+      "learning_rate": 0.0001872865361921027,
+      "loss": 1.6845,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7404142794182459,
+      "grad_norm": 0.25172352057271447,
+      "learning_rate": 0.00018716112898587247,
+      "loss": 1.9169,
+      "step": 1680
+    },
+    {
+      "epoch": 0.742617893345086,
+      "grad_norm": 0.257829203989891,
+      "learning_rate": 0.000187035148708713,
+      "loss": 1.7977,
+      "step": 1685
+    },
+    {
+      "epoch": 0.744821507271926,
+      "grad_norm": 0.21408758544817338,
+      "learning_rate": 0.00018690859618892506,
+      "loss": 1.6934,
+      "step": 1690
+    },
+    {
+      "epoch": 0.747025121198766,
+      "grad_norm": 0.32305529721640136,
+      "learning_rate": 0.0001867814722585719,
+      "loss": 1.811,
+      "step": 1695
+    },
+    {
+      "epoch": 0.749228735125606,
+      "grad_norm": 0.27366670016595385,
+      "learning_rate": 0.0001866537777534737,
+      "loss": 1.6083,
+      "step": 1700
+    },
+    {
+      "epoch": 0.751432349052446,
+      "grad_norm": 0.2978907514821807,
+      "learning_rate": 0.00018652551351320198,
+      "loss": 1.7621,
+      "step": 1705
+    },
+    {
+      "epoch": 0.753635962979286,
+      "grad_norm": 0.2682049213061407,
+      "learning_rate": 0.00018639668038107437,
+      "loss": 1.8008,
+      "step": 1710
+    },
+    {
+      "epoch": 0.755839576906126,
+      "grad_norm": 0.2422117187271709,
+      "learning_rate": 0.0001862672792041487,
+      "loss": 1.9899,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7580431908329661,
+      "grad_norm": 0.27006478070278994,
+      "learning_rate": 0.0001861373108332177,
+      "loss": 1.7577,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7602468047598061,
+      "grad_norm": 0.31176719361390687,
+      "learning_rate": 0.0001860067761228033,
+      "loss": 1.7494,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7624504186866461,
+      "grad_norm": 0.2965711368621697,
+      "learning_rate": 0.00018587567593115098,
+      "loss": 1.9554,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7646540326134861,
+      "grad_norm": 0.26922434332877426,
+      "learning_rate": 0.0001857440111202242,
+      "loss": 1.7415,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7668576465403262,
+      "grad_norm": 0.25558172886185604,
+      "learning_rate": 0.00018561178255569879,
+      "loss": 1.7389,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7690612604671662,
+      "grad_norm": 0.24853091753480136,
+      "learning_rate": 0.000185478991106957,
+      "loss": 1.9658,
+      "step": 1745
+    },
+    {
+      "epoch": 0.7712648743940062,
+      "grad_norm": 0.29324751667885357,
+      "learning_rate": 0.00018534563764708206,
+      "loss": 1.8161,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7734684883208461,
+      "grad_norm": 0.2517432199347208,
+      "learning_rate": 0.00018521172305285236,
+      "loss": 1.6512,
+      "step": 1755
+    },
+    {
+      "epoch": 0.7756721022476862,
+      "grad_norm": 0.25142671678387746,
+      "learning_rate": 0.00018507724820473556,
+      "loss": 1.7221,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7778757161745262,
+      "grad_norm": 0.2851848328431096,
+      "learning_rate": 0.00018494221398688307,
+      "loss": 1.9137,
+      "step": 1765
+    },
+    {
+      "epoch": 0.7800793301013662,
+      "grad_norm": 0.27922286690055653,
+      "learning_rate": 0.00018480662128712389,
+      "loss": 1.7529,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7822829440282063,
+      "grad_norm": 0.24188014554995038,
+      "learning_rate": 0.00018467047099695905,
+      "loss": 1.7036,
+      "step": 1775
+    },
+    {
+      "epoch": 0.7844865579550463,
+      "grad_norm": 0.21832286860860883,
+      "learning_rate": 0.00018453376401155562,
+      "loss": 1.8127,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7866901718818863,
+      "grad_norm": 0.26225286533808284,
+      "learning_rate": 0.00018439650122974087,
+      "loss": 1.7398,
+      "step": 1785
+    },
+    {
+      "epoch": 0.7888937858087263,
+      "grad_norm": 0.2585197080539731,
+      "learning_rate": 0.0001842586835539964,
+      "loss": 1.8645,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7910973997355664,
+      "grad_norm": 0.24265804391287996,
+      "learning_rate": 0.00018412031189045196,
+      "loss": 1.7356,
+      "step": 1795
+    },
+    {
+      "epoch": 0.7933010136624064,
+      "grad_norm": 0.2590383769624523,
+      "learning_rate": 0.00018398138714887993,
+      "loss": 1.6518,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7955046275892463,
+      "grad_norm": 0.23488722647726504,
+      "learning_rate": 0.00018384191024268894,
+      "loss": 1.8054,
+      "step": 1805
+    },
+    {
+      "epoch": 0.7977082415160864,
+      "grad_norm": 0.23606528165986015,
+      "learning_rate": 0.00018370188208891803,
+      "loss": 1.6994,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7999118554429264,
+      "grad_norm": 0.29081075890767155,
+      "learning_rate": 0.00018356130360823068,
+      "loss": 1.987,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8021154693697664,
+      "grad_norm": 0.24366570601075402,
+      "learning_rate": 0.00018342017572490858,
+      "loss": 1.5363,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8043190832966064,
+      "grad_norm": 0.2677025615653137,
+      "learning_rate": 0.0001832784993668458,
+      "loss": 1.7781,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8065226972234465,
+      "grad_norm": 0.22776397402134876,
+      "learning_rate": 0.0001831362754655424,
+      "loss": 1.8064,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8087263111502865,
+      "grad_norm": 0.21327160658864644,
+      "learning_rate": 0.0001829935049560985,
+      "loss": 1.5278,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8109299250771265,
+      "grad_norm": 0.4311681362534179,
+      "learning_rate": 0.0001828501887772081,
+      "loss": 1.9316,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8131335390039665,
+      "grad_norm": 0.3127936675415962,
+      "learning_rate": 0.00018270632787115295,
+      "loss": 1.9393,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8153371529308066,
+      "grad_norm": 0.2872735828645766,
+      "learning_rate": 0.0001825619231837962,
+      "loss": 1.8913,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8175407668576465,
+      "grad_norm": 0.23659304750850696,
+      "learning_rate": 0.0001824169756645763,
+      "loss": 1.79,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8197443807844865,
+      "grad_norm": 0.22502963250778782,
+      "learning_rate": 0.00018227148626650072,
+      "loss": 1.7616,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8219479947113266,
+      "grad_norm": 0.22975193518596113,
+      "learning_rate": 0.00018212545594613978,
+      "loss": 1.7862,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8241516086381666,
+      "grad_norm": 0.22813094138391649,
+      "learning_rate": 0.00018197888566362023,
+      "loss": 1.6909,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8263552225650066,
+      "grad_norm": 0.3170639776475332,
+      "learning_rate": 0.00018183177638261895,
+      "loss": 1.8876,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8285588364918466,
+      "grad_norm": 0.25433638493270494,
+      "learning_rate": 0.00018168412907035672,
+      "loss": 1.7447,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8307624504186867,
+      "grad_norm": 0.2527115922098035,
+      "learning_rate": 0.00018153594469759175,
+      "loss": 1.7288,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8329660643455267,
+      "grad_norm": 0.24448176152726306,
+      "learning_rate": 0.00018138722423861333,
+      "loss": 1.8385,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8351696782723667,
+      "grad_norm": 0.2797043743850419,
+      "learning_rate": 0.00018123796867123548,
+      "loss": 1.8353,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8373732921992068,
+      "grad_norm": 0.26948384291872024,
+      "learning_rate": 0.00018108817897679043,
+      "loss": 1.6995,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8395769061260467,
+      "grad_norm": 0.23050209288787957,
+      "learning_rate": 0.00018093785614012228,
+      "loss": 1.7752,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8417805200528867,
+      "grad_norm": 0.28186670258249874,
+      "learning_rate": 0.0001807870011495803,
+      "loss": 1.8608,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8439841339797267,
+      "grad_norm": 0.29105838653881044,
+      "learning_rate": 0.00018063561499701282,
+      "loss": 1.8997,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8461877479065668,
+      "grad_norm": 0.29426186229081736,
+      "learning_rate": 0.00018048369867776029,
+      "loss": 1.6416,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8483913618334068,
+      "grad_norm": 0.21168547827802603,
+      "learning_rate": 0.00018033125319064902,
+      "loss": 1.8158,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8505949757602468,
+      "grad_norm": 0.35188920795977724,
+      "learning_rate": 0.00018017827953798444,
+      "loss": 1.7531,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8527985896870868,
+      "grad_norm": 0.2979347408479712,
+      "learning_rate": 0.0001800247787255447,
+      "loss": 1.9657,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8550022036139269,
+      "grad_norm": 0.3102746795792285,
+      "learning_rate": 0.00017987075176257382,
+      "loss": 1.5273,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8572058175407669,
+      "grad_norm": 0.2141203314519693,
+      "learning_rate": 0.00017971619966177524,
+      "loss": 1.7978,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8594094314676068,
+      "grad_norm": 0.3010217472666534,
+      "learning_rate": 0.00017956112343930512,
+      "loss": 1.8066,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8616130453944469,
+      "grad_norm": 0.2685511500450377,
+      "learning_rate": 0.00017940552411476566,
+      "loss": 1.8096,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8638166593212869,
+      "grad_norm": 0.3648269512255727,
+      "learning_rate": 0.00017924940271119827,
+      "loss": 1.9212,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8660202732481269,
+      "grad_norm": 0.23673592232409962,
+      "learning_rate": 0.00017909276025507696,
+      "loss": 1.9925,
+      "step": 1965
+    },
+    {
+      "epoch": 0.8682238871749669,
+      "grad_norm": 0.22051543871837528,
+      "learning_rate": 0.00017893559777630173,
+      "loss": 1.8895,
+      "step": 1970
+    },
+    {
+      "epoch": 0.870427501101807,
+      "grad_norm": 0.25128906373836973,
+      "learning_rate": 0.00017877791630819149,
+      "loss": 1.7637,
+      "step": 1975
+    },
+    {
+      "epoch": 0.872631115028647,
+      "grad_norm": 0.2982588862018805,
+      "learning_rate": 0.00017861971688747747,
+      "loss": 1.865,
+      "step": 1980
+    },
+    {
+      "epoch": 0.874834728955487,
+      "grad_norm": 0.2927080940082808,
+      "learning_rate": 0.00017846100055429642,
+      "loss": 1.742,
+      "step": 1985
+    },
+    {
+      "epoch": 0.8770383428823271,
+      "grad_norm": 0.30412581475879524,
+      "learning_rate": 0.00017830176835218368,
+      "loss": 1.6706,
+      "step": 1990
+    },
+    {
+      "epoch": 0.8792419568091671,
+      "grad_norm": 0.2902194881075315,
+      "learning_rate": 0.0001781420213280662,
+      "loss": 1.8014,
+      "step": 1995
+    },
+    {
+      "epoch": 0.881445570736007,
+      "grad_norm": 0.3378104042442177,
+      "learning_rate": 0.00017798176053225606,
+      "loss": 1.8318,
+      "step": 2000
+    },
+    {
+      "epoch": 0.883649184662847,
+      "grad_norm": 0.2465500138547731,
+      "learning_rate": 0.0001778209870184431,
+      "loss": 1.6756,
+      "step": 2005
+    },
+    {
+      "epoch": 0.8858527985896871,
+      "grad_norm": 0.27205493094420413,
+      "learning_rate": 0.00017765970184368835,
+      "loss": 1.7398,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8880564125165271,
+      "grad_norm": 0.28881718844442433,
+      "learning_rate": 0.0001774979060684168,
+      "loss": 1.8652,
+      "step": 2015
+    },
+    {
+      "epoch": 0.8902600264433671,
+      "grad_norm": 0.2930186193777016,
+      "learning_rate": 0.0001773356007564107,
+      "loss": 1.7748,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8924636403702071,
+      "grad_norm": 0.28319545050901546,
+      "learning_rate": 0.0001771727869748023,
+      "loss": 1.7198,
+      "step": 2025
+    },
+    {
+      "epoch": 0.8946672542970472,
+      "grad_norm": 0.26485391899745814,
+      "learning_rate": 0.000177009465794067,
+      "loss": 1.7109,
+      "step": 2030
+    },
+    {
+      "epoch": 0.8968708682238872,
+      "grad_norm": 0.2367084302014588,
+      "learning_rate": 0.0001768456382880163,
+      "loss": 1.773,
+      "step": 2035
+    },
+    {
+      "epoch": 0.8990744821507272,
+      "grad_norm": 0.28707015291179266,
+      "learning_rate": 0.00017668130553379063,
+      "loss": 1.8698,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9012780960775673,
+      "grad_norm": 0.26521748140973644,
+      "learning_rate": 0.00017651646861185252,
+      "loss": 1.5433,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9034817100044072,
+      "grad_norm": 0.25186929013428017,
+      "learning_rate": 0.0001763511286059791,
+      "loss": 1.7003,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9056853239312472,
+      "grad_norm": 0.2646341657457682,
+      "learning_rate": 0.0001761852866032554,
+      "loss": 1.8017,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9078889378580872,
+      "grad_norm": 0.27426709503702656,
+      "learning_rate": 0.0001760189436940669,
+      "loss": 1.717,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9100925517849273,
+      "grad_norm": 0.2909513812455515,
+      "learning_rate": 0.00017585210097209242,
+      "loss": 1.8286,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9122961657117673,
+      "grad_norm": 0.3195261689470233,
+      "learning_rate": 0.00017568475953429706,
+      "loss": 1.9248,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9144997796386073,
+      "grad_norm": 0.2559454772629871,
+      "learning_rate": 0.00017551692048092487,
+      "loss": 1.9235,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9167033935654474,
+      "grad_norm": 0.2787085951940099,
+      "learning_rate": 0.00017534858491549167,
+      "loss": 1.5563,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9189070074922874,
+      "grad_norm": 0.27430346968172775,
+      "learning_rate": 0.00017517975394477765,
+      "loss": 1.7408,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9211106214191274,
+      "grad_norm": 0.3057530869061833,
+      "learning_rate": 0.00017501042867882043,
+      "loss": 1.8029,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9233142353459673,
+      "grad_norm": 0.28792933063158205,
+      "learning_rate": 0.0001748406102309073,
+      "loss": 1.7174,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9255178492728074,
+      "grad_norm": 0.2926998767200136,
+      "learning_rate": 0.00017467029971756837,
+      "loss": 1.7753,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9277214631996474,
+      "grad_norm": 0.3202640628143745,
+      "learning_rate": 0.00017449949825856881,
+      "loss": 1.6815,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9299250771264874,
+      "grad_norm": 0.2898987783743886,
+      "learning_rate": 0.00017432820697690183,
+      "loss": 1.5471,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9321286910533274,
+      "grad_norm": 0.25429358796976326,
+      "learning_rate": 0.00017415642699878108,
+      "loss": 1.801,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9343323049801675,
+      "grad_norm": 0.2906871474145296,
+      "learning_rate": 0.00017398415945363326,
+      "loss": 1.7255,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9365359189070075,
+      "grad_norm": 0.24941394640440692,
+      "learning_rate": 0.00017381140547409091,
+      "loss": 1.7265,
+      "step": 2125
+    },
+    {
+      "epoch": 0.9387395328338475,
+      "grad_norm": 0.2680278555985492,
+      "learning_rate": 0.00017363816619598462,
+      "loss": 1.8507,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9409431467606876,
+      "grad_norm": 0.24326339969260075,
+      "learning_rate": 0.00017346444275833587,
+      "loss": 1.8278,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9431467606875276,
+      "grad_norm": 0.2881192799230667,
+      "learning_rate": 0.00017329023630334935,
+      "loss": 1.6301,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9453503746143676,
+      "grad_norm": 0.22324975257691354,
+      "learning_rate": 0.00017311554797640552,
+      "loss": 1.8182,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9475539885412075,
+      "grad_norm": 0.38310870823903137,
+      "learning_rate": 0.0001729403789260531,
+      "loss": 1.6758,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9497576024680476,
+      "grad_norm": 0.3082503271893105,
+      "learning_rate": 0.0001727647303040015,
+      "loss": 1.717,
+      "step": 2155
+    },
+    {
+      "epoch": 0.9519612163948876,
+      "grad_norm": 0.2910152870088472,
+      "learning_rate": 0.00017258860326511318,
+      "loss": 1.6762,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9541648303217276,
+      "grad_norm": 0.2751619266213339,
+      "learning_rate": 0.00017241199896739614,
+      "loss": 1.5402,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9563684442485677,
+      "grad_norm": 0.2633670365064433,
+      "learning_rate": 0.00017223491857199636,
+      "loss": 1.6089,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9585720581754077,
+      "grad_norm": 0.26253468282607706,
+      "learning_rate": 0.00017205736324318999,
+      "loss": 1.8698,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9607756721022477,
+      "grad_norm": 0.2623810146620032,
+      "learning_rate": 0.0001718793341483758,
+      "loss": 1.7996,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9629792860290877,
+      "grad_norm": 0.2541467303967672,
+      "learning_rate": 0.00017170083245806757,
+      "loss": 1.7066,
+      "step": 2185
+    },
+    {
+      "epoch": 0.9651828999559278,
+      "grad_norm": 0.2450837271489877,
+      "learning_rate": 0.00017152185934588623,
+      "loss": 1.9326,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9673865138827678,
+      "grad_norm": 0.3277904173770177,
+      "learning_rate": 0.00017134241598855236,
+      "loss": 1.941,
+      "step": 2195
+    },
+    {
+      "epoch": 0.9695901278096077,
+      "grad_norm": 0.26036896345598354,
+      "learning_rate": 0.0001711625035658782,
+      "loss": 1.7543,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9717937417364477,
+      "grad_norm": 0.294563677729738,
+      "learning_rate": 0.00017098212326076008,
+      "loss": 1.6402,
+      "step": 2205
+    },
+    {
+      "epoch": 0.9739973556632878,
+      "grad_norm": 0.3220804392466043,
+      "learning_rate": 0.0001708012762591706,
+      "loss": 1.7078,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9762009695901278,
+      "grad_norm": 0.2811686235748074,
+      "learning_rate": 0.00017061996375015078,
+      "loss": 1.9067,
+      "step": 2215
+    },
+    {
+      "epoch": 0.9784045835169678,
+      "grad_norm": 0.23864027527398188,
+      "learning_rate": 0.00017043818692580228,
+      "loss": 1.9146,
+      "step": 2220
+    },
+    {
+      "epoch": 0.9806081974438079,
+      "grad_norm": 0.2954583734368435,
+      "learning_rate": 0.00017025594698127965,
+      "loss": 1.791,
+      "step": 2225
+    },
+    {
+      "epoch": 0.9828118113706479,
+      "grad_norm": 0.2510099463630807,
+      "learning_rate": 0.00017007324511478223,
+      "loss": 1.5883,
+      "step": 2230
+    },
+    {
+      "epoch": 0.9850154252974879,
+      "grad_norm": 0.27874972856939056,
+      "learning_rate": 0.00016989008252754655,
+      "loss": 1.7865,
+      "step": 2235
+    },
+    {
+      "epoch": 0.9872190392243279,
+      "grad_norm": 0.2611091563705049,
+      "learning_rate": 0.00016970646042383826,
+      "loss": 1.8104,
+      "step": 2240
+    },
+    {
+      "epoch": 0.989422653151168,
+      "grad_norm": 0.28511909006315234,
+      "learning_rate": 0.00016952238001094428,
+      "loss": 1.6686,
+      "step": 2245
+    },
+    {
+      "epoch": 0.9916262670780079,
+      "grad_norm": 0.3381628494565006,
+      "learning_rate": 0.00016933784249916476,
+      "loss": 1.9412,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9938298810048479,
+      "grad_norm": 0.28520119611042416,
+      "learning_rate": 0.00016915284910180533,
+      "loss": 1.6889,
+      "step": 2255
+    },
+    {
+      "epoch": 0.996033494931688,
+      "grad_norm": 0.2883852687543481,
+      "learning_rate": 0.00016896740103516895,
+      "loss": 1.8003,
+      "step": 2260
+    },
+    {
+      "epoch": 0.998237108858528,
+      "grad_norm": 0.24451797378236065,
+      "learning_rate": 0.0001687814995185479,
+      "loss": 1.8752,
+      "step": 2265
+    },
+    {
+      "epoch": 1.000440722785368,
+      "grad_norm": 0.33508649606500057,
+      "learning_rate": 0.00016859514577421592,
+      "loss": 1.8299,
+      "step": 2270
+    },
+    {
+      "epoch": 1.002644336712208,
+      "grad_norm": 0.2636718535549577,
+      "learning_rate": 0.00016840834102741997,
+      "loss": 1.706,
+      "step": 2275
+    },
+    {
+      "epoch": 1.004847950639048,
+      "grad_norm": 0.27046399540212357,
+      "learning_rate": 0.00016822108650637238,
+      "loss": 1.6819,
+      "step": 2280
+    },
+    {
+      "epoch": 1.007051564565888,
+      "grad_norm": 0.30877609431387876,
+      "learning_rate": 0.00016803338344224266,
+      "loss": 1.7218,
+      "step": 2285
+    },
+    {
+      "epoch": 1.0092551784927282,
+      "grad_norm": 0.23492423335336904,
+      "learning_rate": 0.00016784523306914934,
+      "loss": 1.57,
+      "step": 2290
+    },
+    {
+      "epoch": 1.0114587924195682,
+      "grad_norm": 0.2795118641324036,
+      "learning_rate": 0.00016765663662415204,
+      "loss": 1.7023,
+      "step": 2295
+    },
+    {
+      "epoch": 1.0136624063464081,
+      "grad_norm": 0.34933397332267874,
+      "learning_rate": 0.00016746759534724316,
+      "loss": 1.9401,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0158660202732481,
+      "grad_norm": 0.2928363020374088,
+      "learning_rate": 0.00016727811048133985,
+      "loss": 1.7873,
+      "step": 2305
+    },
+    {
+      "epoch": 1.0180696342000881,
+      "grad_norm": 0.2765841121495588,
+      "learning_rate": 0.00016708818327227574,
+      "loss": 1.6925,
+      "step": 2310
+    },
+    {
+      "epoch": 1.020273248126928,
+      "grad_norm": 0.3407816261608907,
+      "learning_rate": 0.00016689781496879283,
+      "loss": 1.7583,
+      "step": 2315
+    },
+    {
+      "epoch": 1.022476862053768,
+      "grad_norm": 0.3837729171353601,
+      "learning_rate": 0.00016670700682253328,
+      "loss": 1.7058,
+      "step": 2320
+    },
+    {
+      "epoch": 1.0246804759806083,
+      "grad_norm": 0.2833490201778371,
+      "learning_rate": 0.00016651576008803112,
+      "loss": 1.7306,
+      "step": 2325
+    },
+    {
+      "epoch": 1.0268840899074483,
+      "grad_norm": 0.24052446562056756,
+      "learning_rate": 0.00016632407602270398,
+      "loss": 1.6612,
+      "step": 2330
+    },
+    {
+      "epoch": 1.0290877038342883,
+      "grad_norm": 0.2706767115179465,
+      "learning_rate": 0.00016613195588684488,
+      "loss": 1.5943,
+      "step": 2335
+    },
+    {
+      "epoch": 1.0312913177611283,
+      "grad_norm": 0.2694153548940927,
+      "learning_rate": 0.00016593940094361407,
+      "loss": 1.7072,
+      "step": 2340
+    },
+    {
+      "epoch": 1.0334949316879682,
+      "grad_norm": 0.2675415111083724,
+      "learning_rate": 0.0001657464124590304,
+      "loss": 1.7392,
+      "step": 2345
+    },
+    {
+      "epoch": 1.0356985456148082,
+      "grad_norm": 0.2783138545253644,
+      "learning_rate": 0.00016555299170196332,
+      "loss": 1.7264,
+      "step": 2350
+    },
+    {
+      "epoch": 1.0379021595416482,
+      "grad_norm": 0.29818023025744944,
+      "learning_rate": 0.00016535913994412436,
+      "loss": 1.6038,
+      "step": 2355
+    },
+    {
+      "epoch": 1.0401057734684884,
+      "grad_norm": 0.30521070526088534,
+      "learning_rate": 0.00016516485846005882,
+      "loss": 1.652,
+      "step": 2360
+    },
+    {
+      "epoch": 1.0423093873953284,
+      "grad_norm": 0.24990284589728196,
+      "learning_rate": 0.00016497014852713738,
+      "loss": 1.5745,
+      "step": 2365
+    },
+    {
+      "epoch": 1.0445130013221684,
+      "grad_norm": 0.36544530123344093,
+      "learning_rate": 0.0001647750114255477,
+      "loss": 1.8678,
+      "step": 2370
+    },
+    {
+      "epoch": 1.0467166152490084,
+      "grad_norm": 0.2984936690547218,
+      "learning_rate": 0.000164579448438286,
+      "loss": 1.8263,
+      "step": 2375
+    },
+    {
+      "epoch": 1.0489202291758484,
+      "grad_norm": 0.3138432337464576,
+      "learning_rate": 0.00016438346085114865,
+      "loss": 1.7807,
+      "step": 2380
+    },
+    {
+      "epoch": 1.0511238431026884,
+      "grad_norm": 0.409332779624452,
+      "learning_rate": 0.00016418704995272373,
+      "loss": 1.8612,
+      "step": 2385
+    },
+    {
+      "epoch": 1.0533274570295283,
+      "grad_norm": 0.27134598526523496,
+      "learning_rate": 0.00016399021703438247,
+      "loss": 1.7323,
+      "step": 2390
+    },
+    {
+      "epoch": 1.0555310709563686,
+      "grad_norm": 0.2852569861948395,
+      "learning_rate": 0.0001637929633902708,
+      "loss": 1.7619,
+      "step": 2395
+    },
+    {
+      "epoch": 1.0577346848832085,
+      "grad_norm": 0.2913405836362002,
+      "learning_rate": 0.00016359529031730093,
+      "loss": 1.8196,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0599382988100485,
+      "grad_norm": 0.3029325708755559,
+      "learning_rate": 0.00016339719911514272,
+      "loss": 1.7579,
+      "step": 2405
+    },
+    {
+      "epoch": 1.0621419127368885,
+      "grad_norm": 0.2738995593865038,
+      "learning_rate": 0.00016319869108621512,
+      "loss": 1.8309,
+      "step": 2410
+    },
+    {
+      "epoch": 1.0643455266637285,
+      "grad_norm": 0.26503420605103173,
+      "learning_rate": 0.00016299976753567772,
+      "loss": 1.708,
+      "step": 2415
+    },
+    {
+      "epoch": 1.0665491405905685,
+      "grad_norm": 0.32283303116470646,
+      "learning_rate": 0.00016280042977142204,
+      "loss": 1.6915,
+      "step": 2420
+    },
+    {
+      "epoch": 1.0687527545174085,
+      "grad_norm": 0.24107839513422735,
+      "learning_rate": 0.00016260067910406304,
+      "loss": 1.5685,
+      "step": 2425
+    },
+    {
+      "epoch": 1.0709563684442487,
+      "grad_norm": 0.29608435429464214,
+      "learning_rate": 0.00016240051684693042,
+      "loss": 1.7239,
+      "step": 2430
+    },
+    {
+      "epoch": 1.0731599823710887,
+      "grad_norm": 0.29986988301001716,
+      "learning_rate": 0.00016219994431606005,
+      "loss": 1.6816,
+      "step": 2435
+    },
+    {
+      "epoch": 1.0753635962979287,
+      "grad_norm": 0.24843757070636935,
+      "learning_rate": 0.00016199896283018527,
+      "loss": 1.5677,
+      "step": 2440
+    },
+    {
+      "epoch": 1.0775672102247686,
+      "grad_norm": 0.2983161128962509,
+      "learning_rate": 0.00016179757371072824,
+      "loss": 1.7859,
+      "step": 2445
+    },
+    {
+      "epoch": 1.0797708241516086,
+      "grad_norm": 0.2787284156730912,
+      "learning_rate": 0.00016159577828179123,
+      "loss": 1.562,
+      "step": 2450
+    },
+    {
+      "epoch": 1.0819744380784486,
+      "grad_norm": 0.3269743373279123,
+      "learning_rate": 0.0001613935778701479,
+      "loss": 1.8759,
+      "step": 2455
+    },
+    {
+      "epoch": 1.0841780520052886,
+      "grad_norm": 0.2802831939826356,
+      "learning_rate": 0.0001611909738052347,
+      "loss": 1.7401,
+      "step": 2460
+    },
+    {
+      "epoch": 1.0863816659321286,
+      "grad_norm": 0.23846388238067243,
+      "learning_rate": 0.000160987967419142,
+      "loss": 1.727,
+      "step": 2465
+    },
+    {
+      "epoch": 1.0885852798589688,
+      "grad_norm": 0.2659078946356941,
+      "learning_rate": 0.00016078456004660536,
+      "loss": 1.6454,
+      "step": 2470
+    },
+    {
+      "epoch": 1.0907888937858088,
+      "grad_norm": 0.29111397323788823,
+      "learning_rate": 0.00016058075302499673,
+      "loss": 1.7724,
+      "step": 2475
+    },
+    {
+      "epoch": 1.0929925077126488,
+      "grad_norm": 0.35649745669814514,
+      "learning_rate": 0.00016037654769431576,
+      "loss": 1.6527,
+      "step": 2480
+    },
+    {
+      "epoch": 1.0951961216394888,
+      "grad_norm": 0.28097148721584675,
+      "learning_rate": 0.00016017194539718086,
+      "loss": 1.7563,
+      "step": 2485
+    },
+    {
+      "epoch": 1.0973997355663287,
+      "grad_norm": 0.39237884989816185,
+      "learning_rate": 0.0001599669474788205,
+      "loss": 1.8656,
+      "step": 2490
+    },
+    {
+      "epoch": 1.0996033494931687,
+      "grad_norm": 0.2633010799321111,
+      "learning_rate": 0.00015976155528706415,
+      "loss": 1.7129,
+      "step": 2495
+    },
+    {
+      "epoch": 1.1018069634200087,
+      "grad_norm": 0.2719172135219212,
+      "learning_rate": 0.0001595557701723338,
+      "loss": 1.5688,
+      "step": 2500
+    },
+    {
+      "epoch": 1.104010577346849,
+      "grad_norm": 0.2731378192802766,
+      "learning_rate": 0.00015934959348763467,
+      "loss": 1.7727,
+      "step": 2505
+    },
+    {
+      "epoch": 1.106214191273689,
+      "grad_norm": 0.31808171740983593,
+      "learning_rate": 0.00015914302658854657,
+      "loss": 1.8461,
+      "step": 2510
+    },
+    {
+      "epoch": 1.108417805200529,
+      "grad_norm": 0.2869427974730242,
+      "learning_rate": 0.00015893607083321477,
+      "loss": 1.8664,
+      "step": 2515
+    },
+    {
+      "epoch": 1.110621419127369,
+      "grad_norm": 0.2187585282755932,
+      "learning_rate": 0.00015872872758234148,
+      "loss": 1.6029,
+      "step": 2520
+    },
+    {
+      "epoch": 1.1128250330542089,
+      "grad_norm": 0.2667619562054791,
+      "learning_rate": 0.00015852099819917639,
+      "loss": 1.8981,
+      "step": 2525
+    },
+    {
+      "epoch": 1.1150286469810489,
+      "grad_norm": 0.3372014415136272,
+      "learning_rate": 0.00015831288404950802,
+      "loss": 1.7639,
+      "step": 2530
+    },
+    {
+      "epoch": 1.1172322609078889,
+      "grad_norm": 0.3048158336194716,
+      "learning_rate": 0.0001581043865016547,
+      "loss": 1.6387,
+      "step": 2535
+    },
+    {
+      "epoch": 1.1194358748347288,
+      "grad_norm": 0.2971544124275263,
+      "learning_rate": 0.00015789550692645556,
+      "loss": 1.7692,
+      "step": 2540
+    },
+    {
+      "epoch": 1.121639488761569,
+      "grad_norm": 0.2518439420968902,
+      "learning_rate": 0.00015768624669726145,
+      "loss": 1.5533,
+      "step": 2545
+    },
+    {
+      "epoch": 1.123843102688409,
+      "grad_norm": 0.28947857100850594,
+      "learning_rate": 0.00015747660718992598,
+      "loss": 1.6443,
+      "step": 2550
+    },
+    {
+      "epoch": 1.126046716615249,
+      "grad_norm": 0.27744579032481387,
+      "learning_rate": 0.00015726658978279642,
+      "loss": 1.7146,
+      "step": 2555
+    },
+    {
+      "epoch": 1.128250330542089,
+      "grad_norm": 0.28434343297956843,
+      "learning_rate": 0.00015705619585670478,
+      "loss": 1.8686,
+      "step": 2560
+    },
+    {
+      "epoch": 1.130453944468929,
+      "grad_norm": 0.26116640553101533,
+      "learning_rate": 0.00015684542679495847,
+      "loss": 1.7831,
+      "step": 2565
+    },
+    {
+      "epoch": 1.132657558395769,
+      "grad_norm": 0.31380511545232626,
+      "learning_rate": 0.00015663428398333157,
+      "loss": 1.6778,
+      "step": 2570
+    },
+    {
+      "epoch": 1.134861172322609,
+      "grad_norm": 0.2747114435356579,
+      "learning_rate": 0.0001564227688100552,
+      "loss": 1.6324,
+      "step": 2575
+    },
+    {
+      "epoch": 1.1370647862494492,
+      "grad_norm": 0.2678076483406117,
+      "learning_rate": 0.00015621088266580904,
+      "loss": 1.4946,
+      "step": 2580
+    },
+    {
+      "epoch": 1.1392684001762892,
+      "grad_norm": 0.33262666113926864,
+      "learning_rate": 0.00015599862694371157,
+      "loss": 1.882,
+      "step": 2585
+    },
+    {
+      "epoch": 1.1414720141031292,
+      "grad_norm": 0.30861283361311576,
+      "learning_rate": 0.00015578600303931136,
+      "loss": 1.6738,
+      "step": 2590
+    },
+    {
+      "epoch": 1.1436756280299691,
+      "grad_norm": 0.30927635505724477,
+      "learning_rate": 0.00015557301235057767,
+      "loss": 1.7005,
+      "step": 2595
+    },
+    {
+      "epoch": 1.1458792419568091,
+      "grad_norm": 0.28688292761286854,
+      "learning_rate": 0.00015535965627789126,
+      "loss": 1.6462,
+      "step": 2600
+    },
+    {
+      "epoch": 1.1480828558836491,
+      "grad_norm": 0.27171604870426835,
+      "learning_rate": 0.00015514593622403532,
+      "loss": 1.585,
+      "step": 2605
+    },
+    {
+      "epoch": 1.150286469810489,
+      "grad_norm": 0.3092542662127764,
+      "learning_rate": 0.0001549318535941861,
+      "loss": 1.9096,
+      "step": 2610
+    },
+    {
+      "epoch": 1.1524900837373293,
+      "grad_norm": 0.2716117253477657,
+      "learning_rate": 0.00015471740979590377,
+      "loss": 1.7352,
+      "step": 2615
+    },
+    {
+      "epoch": 1.1546936976641693,
+      "grad_norm": 0.3066935084879198,
+      "learning_rate": 0.0001545026062391231,
+      "loss": 1.8141,
+      "step": 2620
+    },
+    {
+      "epoch": 1.1568973115910093,
+      "grad_norm": 0.346363671435922,
+      "learning_rate": 0.00015428744433614415,
+      "loss": 1.5573,
+      "step": 2625
+    },
+    {
+      "epoch": 1.1591009255178493,
+      "grad_norm": 0.29157425354464905,
+      "learning_rate": 0.00015407192550162318,
+      "loss": 1.5464,
+      "step": 2630
+    },
+    {
+      "epoch": 1.1613045394446893,
+      "grad_norm": 0.2806374114105095,
+      "learning_rate": 0.0001538560511525632,
+      "loss": 1.6386,
+      "step": 2635
+    },
+    {
+      "epoch": 1.1635081533715292,
+      "grad_norm": 0.2982356223064332,
+      "learning_rate": 0.0001536398227083046,
+      "loss": 1.7813,
+      "step": 2640
+    },
+    {
+      "epoch": 1.1657117672983692,
+      "grad_norm": 0.33202427377524757,
+      "learning_rate": 0.00015342324159051587,
+      "loss": 1.9532,
+      "step": 2645
+    },
+    {
+      "epoch": 1.1679153812252094,
+      "grad_norm": 0.33897582151852124,
+      "learning_rate": 0.00015320630922318444,
+      "loss": 1.7746,
+      "step": 2650
+    },
+    {
+      "epoch": 1.1701189951520494,
+      "grad_norm": 0.3321654404402584,
+      "learning_rate": 0.00015298902703260692,
+      "loss": 2.0143,
+      "step": 2655
+    },
+    {
+      "epoch": 1.1723226090788894,
+      "grad_norm": 0.21833177585011845,
+      "learning_rate": 0.0001527713964473802,
+      "loss": 1.6702,
+      "step": 2660
+    },
+    {
+      "epoch": 1.1745262230057294,
+      "grad_norm": 0.26595318714741284,
+      "learning_rate": 0.00015255341889839157,
+      "loss": 1.683,
+      "step": 2665
+    },
+    {
+      "epoch": 1.1767298369325694,
+      "grad_norm": 0.2880553889858748,
+      "learning_rate": 0.00015233509581880973,
+      "loss": 1.6248,
+      "step": 2670
+    },
+    {
+      "epoch": 1.1789334508594094,
+      "grad_norm": 0.2966201251141622,
+      "learning_rate": 0.0001521164286440751,
+      "loss": 1.5788,
+      "step": 2675
+    },
+    {
+      "epoch": 1.1811370647862494,
+      "grad_norm": 0.32819026600076084,
+      "learning_rate": 0.00015189741881189054,
+      "loss": 1.6132,
+      "step": 2680
+    },
+    {
+      "epoch": 1.1833406787130896,
+      "grad_norm": 0.25598509051489227,
+      "learning_rate": 0.00015167806776221178,
+      "loss": 1.7194,
+      "step": 2685
+    },
+    {
+      "epoch": 1.1855442926399296,
+      "grad_norm": 0.32216524743892266,
+      "learning_rate": 0.000151458376937238,
+      "loss": 1.6499,
+      "step": 2690
+    },
+    {
+      "epoch": 1.1877479065667695,
+      "grad_norm": 0.25107175983686675,
+      "learning_rate": 0.00015123834778140233,
+      "loss": 1.6059,
+      "step": 2695
+    },
+    {
+      "epoch": 1.1899515204936095,
+      "grad_norm": 0.26100544568346645,
+      "learning_rate": 0.00015101798174136247,
+      "loss": 1.7677,
+      "step": 2700
+    },
+    {
+      "epoch": 1.1921551344204495,
+      "grad_norm": 0.28164900490660866,
+      "learning_rate": 0.000150797280265991,
+      "loss": 1.6141,
+      "step": 2705
+    },
+    {
+      "epoch": 1.1943587483472895,
+      "grad_norm": 0.25606567943961145,
+      "learning_rate": 0.00015057624480636594,
+      "loss": 1.6868,
+      "step": 2710
+    },
+    {
+      "epoch": 1.1965623622741295,
+      "grad_norm": 0.30536561643444826,
+      "learning_rate": 0.0001503548768157612,
+      "loss": 1.515,
+      "step": 2715
+    },
+    {
+      "epoch": 1.1987659762009697,
+      "grad_norm": 0.2495814195185963,
+      "learning_rate": 0.00015013317774963708,
+      "loss": 1.5754,
+      "step": 2720
+    },
+    {
+      "epoch": 1.2009695901278097,
+      "grad_norm": 0.3292329862356163,
+      "learning_rate": 0.00014991114906563055,
+      "loss": 1.6599,
+      "step": 2725
+    },
+    {
+      "epoch": 1.2031732040546497,
+      "grad_norm": 0.28736997494401945,
+      "learning_rate": 0.00014968879222354597,
+      "loss": 1.6939,
+      "step": 2730
+    },
+    {
+      "epoch": 1.2053768179814897,
+      "grad_norm": 0.28066545139959265,
+      "learning_rate": 0.00014946610868534502,
+      "loss": 1.5954,
+      "step": 2735
+    },
+    {
+      "epoch": 1.2075804319083296,
+      "grad_norm": 0.27963157670324124,
+      "learning_rate": 0.00014924309991513757,
+      "loss": 1.6816,
+      "step": 2740
+    },
+    {
+      "epoch": 1.2097840458351696,
+      "grad_norm": 0.3074722093066757,
+      "learning_rate": 0.0001490197673791717,
+      "loss": 1.5102,
+      "step": 2745
+    },
+    {
+      "epoch": 1.2119876597620096,
+      "grad_norm": 0.5175039724428968,
+      "learning_rate": 0.00014879611254582428,
+      "loss": 1.8587,
+      "step": 2750
+    },
+    {
+      "epoch": 1.2141912736888498,
+      "grad_norm": 0.4607644456138267,
+      "learning_rate": 0.00014857213688559124,
+      "loss": 1.8861,
+      "step": 2755
+    },
+    {
+      "epoch": 1.2163948876156898,
+      "grad_norm": 0.203471979843397,
+      "learning_rate": 0.00014834784187107785,
+      "loss": 1.5549,
+      "step": 2760
+    },
+    {
+      "epoch": 1.2185985015425298,
+      "grad_norm": 0.27836500512738993,
+      "learning_rate": 0.00014812322897698912,
+      "loss": 1.6677,
+      "step": 2765
+    },
+    {
+      "epoch": 1.2208021154693698,
+      "grad_norm": 0.2681432382916657,
+      "learning_rate": 0.00014789829968012,
+      "loss": 1.8601,
+      "step": 2770
+    },
+    {
+      "epoch": 1.2230057293962098,
+      "grad_norm": 0.24822408823990583,
+      "learning_rate": 0.00014767305545934588,
+      "loss": 1.9008,
+      "step": 2775
+    },
+    {
+      "epoch": 1.2252093433230498,
+      "grad_norm": 0.26961514947075566,
+      "learning_rate": 0.00014744749779561258,
+      "loss": 1.7573,
+      "step": 2780
+    },
+    {
+      "epoch": 1.2274129572498897,
+      "grad_norm": 0.40650037835816966,
+      "learning_rate": 0.0001472216281719269,
+      "loss": 1.6177,
+      "step": 2785
+    },
+    {
+      "epoch": 1.22961657117673,
+      "grad_norm": 0.27436209441284687,
+      "learning_rate": 0.0001469954480733465,
+      "loss": 1.6021,
+      "step": 2790
+    },
+    {
+      "epoch": 1.23182018510357,
+      "grad_norm": 0.2563256590060921,
+      "learning_rate": 0.00014676895898697062,
+      "loss": 1.9842,
+      "step": 2795
+    },
+    {
+      "epoch": 1.23402379903041,
+      "grad_norm": 0.35591941682342815,
+      "learning_rate": 0.00014654216240192995,
+      "loss": 1.6028,
+      "step": 2800
+    },
+    {
+      "epoch": 1.23622741295725,
+      "grad_norm": 0.33349441263276575,
+      "learning_rate": 0.00014631505980937688,
+      "loss": 1.852,
+      "step": 2805
+    },
+    {
+      "epoch": 1.23843102688409,
+      "grad_norm": 0.31528740240587627,
+      "learning_rate": 0.0001460876527024758,
+      "loss": 1.5587,
+      "step": 2810
+    },
+    {
+      "epoch": 1.24063464081093,
+      "grad_norm": 0.25877213395041015,
+      "learning_rate": 0.00014585994257639324,
+      "loss": 1.5482,
+      "step": 2815
+    },
+    {
+      "epoch": 1.2428382547377699,
+      "grad_norm": 0.2910006006185105,
+      "learning_rate": 0.00014563193092828803,
+      "loss": 1.6998,
+      "step": 2820
+    },
+    {
+      "epoch": 1.24504186866461,
+      "grad_norm": 0.37486054420050446,
+      "learning_rate": 0.00014540361925730147,
+      "loss": 1.8516,
+      "step": 2825
+    },
+    {
+      "epoch": 1.24724548259145,
+      "grad_norm": 0.34335913235176224,
+      "learning_rate": 0.00014517500906454742,
+      "loss": 1.6384,
+      "step": 2830
+    },
+    {
+      "epoch": 1.24944909651829,
+      "grad_norm": 0.2930060418053195,
+      "learning_rate": 0.00014494610185310252,
+      "loss": 1.8508,
+      "step": 2835
+    },
+    {
+      "epoch": 1.25165271044513,
+      "grad_norm": 0.27045205482633095,
+      "learning_rate": 0.00014471689912799626,
+      "loss": 1.5935,
+      "step": 2840
+    },
+    {
+      "epoch": 1.25385632437197,
+      "grad_norm": 0.30870157000744336,
+      "learning_rate": 0.00014448740239620108,
+      "loss": 1.8287,
+      "step": 2845
+    },
+    {
+      "epoch": 1.25605993829881,
+      "grad_norm": 0.27087387170107313,
+      "learning_rate": 0.00014425761316662241,
+      "loss": 1.9209,
+      "step": 2850
+    },
+    {
+      "epoch": 1.25826355222565,
+      "grad_norm": 0.32362518237545235,
+      "learning_rate": 0.0001440275329500889,
+      "loss": 1.7297,
+      "step": 2855
+    },
+    {
+      "epoch": 1.2604671661524902,
+      "grad_norm": 0.3473479960796504,
+      "learning_rate": 0.00014379716325934236,
+      "loss": 1.7847,
+      "step": 2860
+    },
+    {
+      "epoch": 1.26267078007933,
+      "grad_norm": 0.2985625193084418,
+      "learning_rate": 0.0001435665056090278,
+      "loss": 1.7811,
+      "step": 2865
+    },
+    {
+      "epoch": 1.2648743940061702,
+      "grad_norm": 0.2549368072217001,
+      "learning_rate": 0.00014333556151568364,
+      "loss": 1.8424,
+      "step": 2870
+    },
+    {
+      "epoch": 1.2670780079330102,
+      "grad_norm": 0.3031064940015495,
+      "learning_rate": 0.00014310433249773146,
+      "loss": 1.8502,
+      "step": 2875
+    },
+    {
+      "epoch": 1.2692816218598502,
+      "grad_norm": 0.2398504180714546,
+      "learning_rate": 0.00014287282007546627,
+      "loss": 1.648,
+      "step": 2880
+    },
+    {
+      "epoch": 1.2714852357866901,
+      "grad_norm": 0.33873410733492354,
+      "learning_rate": 0.00014264102577104645,
+      "loss": 1.6617,
+      "step": 2885
+    },
+    {
+      "epoch": 1.2736888497135301,
+      "grad_norm": 0.2655552963090036,
+      "learning_rate": 0.00014240895110848365,
+      "loss": 1.7205,
+      "step": 2890
+    },
+    {
+      "epoch": 1.2758924636403703,
+      "grad_norm": 0.30714278749605195,
+      "learning_rate": 0.0001421765976136328,
+      "loss": 1.7343,
+      "step": 2895
+    },
+    {
+      "epoch": 1.27809607756721,
+      "grad_norm": 0.31410869299454564,
+      "learning_rate": 0.0001419439668141822,
+      "loss": 1.7369,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2802996914940503,
+      "grad_norm": 0.29231072094243427,
+      "learning_rate": 0.0001417110602396434,
+      "loss": 1.5317,
+      "step": 2905
+    },
+    {
+      "epoch": 1.2825033054208903,
+      "grad_norm": 0.26156390622111436,
+      "learning_rate": 0.00014147787942134089,
+      "loss": 1.4907,
+      "step": 2910
+    },
+    {
+      "epoch": 1.2847069193477303,
+      "grad_norm": 0.2432071100976,
+      "learning_rate": 0.00014124442589240265,
+      "loss": 1.7181,
+      "step": 2915
+    },
+    {
+      "epoch": 1.2869105332745703,
+      "grad_norm": 0.28813004538064096,
+      "learning_rate": 0.00014101070118774936,
+      "loss": 1.7243,
+      "step": 2920
+    },
+    {
+      "epoch": 1.2891141472014103,
+      "grad_norm": 0.29339284132217475,
+      "learning_rate": 0.00014077670684408485,
+      "loss": 1.7679,
+      "step": 2925
+    },
+    {
+      "epoch": 1.2913177611282503,
+      "grad_norm": 0.2588877094640161,
+      "learning_rate": 0.00014054244439988566,
+      "loss": 1.72,
+      "step": 2930
+    },
+    {
+      "epoch": 1.2935213750550902,
+      "grad_norm": 0.3011003249431555,
+      "learning_rate": 0.0001403079153953911,
+      "loss": 1.879,
+      "step": 2935
+    },
+    {
+      "epoch": 1.2957249889819304,
+      "grad_norm": 0.29129639306491034,
+      "learning_rate": 0.00014007312137259307,
+      "loss": 1.7124,
+      "step": 2940
+    },
+    {
+      "epoch": 1.2979286029087704,
+      "grad_norm": 0.30248111079542994,
+      "learning_rate": 0.00013983806387522592,
+      "loss": 1.5669,
+      "step": 2945
+    },
+    {
+      "epoch": 1.3001322168356104,
+      "grad_norm": 0.39466981086485026,
+      "learning_rate": 0.00013960274444875628,
+      "loss": 1.7579,
+      "step": 2950
+    },
+    {
+      "epoch": 1.3023358307624504,
+      "grad_norm": 0.30172725615707247,
+      "learning_rate": 0.000139367164640373,
+      "loss": 1.8218,
+      "step": 2955
+    },
+    {
+      "epoch": 1.3045394446892904,
+      "grad_norm": 0.286453848416204,
+      "learning_rate": 0.00013913132599897683,
+      "loss": 1.9354,
+      "step": 2960
+    },
+    {
+      "epoch": 1.3067430586161304,
+      "grad_norm": 0.3902213358624171,
+      "learning_rate": 0.00013889523007517028,
+      "loss": 1.7235,
+      "step": 2965
+    },
+    {
+      "epoch": 1.3089466725429704,
+      "grad_norm": 0.2589601846721671,
+      "learning_rate": 0.00013865887842124755,
+      "loss": 1.5088,
+      "step": 2970
+    },
+    {
+      "epoch": 1.3111502864698106,
+      "grad_norm": 0.2838598126846581,
+      "learning_rate": 0.0001384222725911842,
+      "loss": 1.6694,
+      "step": 2975
+    },
+    {
+      "epoch": 1.3133539003966506,
+      "grad_norm": 0.2985112357585295,
+      "learning_rate": 0.00013818541414062683,
+      "loss": 1.8195,
+      "step": 2980
+    },
+    {
+      "epoch": 1.3155575143234906,
+      "grad_norm": 0.344496826298518,
+      "learning_rate": 0.0001379483046268832,
+      "loss": 1.7105,
+      "step": 2985
+    },
+    {
+      "epoch": 1.3177611282503305,
+      "grad_norm": 0.2832321651335367,
+      "learning_rate": 0.00013771094560891155,
+      "loss": 1.6398,
+      "step": 2990
+    },
+    {
+      "epoch": 1.3199647421771705,
+      "grad_norm": 0.3166967200246273,
+      "learning_rate": 0.00013747333864731073,
+      "loss": 1.8804,
+      "step": 2995
+    },
+    {
+      "epoch": 1.3221683561040105,
+      "grad_norm": 0.29889345173462817,
+      "learning_rate": 0.00013723548530430974,
+      "loss": 1.5327,
+      "step": 3000
+    },
+    {
+      "epoch": 1.3243719700308505,
+      "grad_norm": 0.2989561597186618,
+      "learning_rate": 0.00013699738714375748,
+      "loss": 1.8312,
+      "step": 3005
+    },
+    {
+      "epoch": 1.3265755839576907,
+      "grad_norm": 0.27767734489817053,
+      "learning_rate": 0.00013675904573111247,
+      "loss": 1.7797,
+      "step": 3010
+    },
+    {
+      "epoch": 1.3287791978845307,
+      "grad_norm": 0.3263812983049982,
+      "learning_rate": 0.00013652046263343262,
+      "loss": 1.7061,
+      "step": 3015
+    },
+    {
+      "epoch": 1.3309828118113707,
+      "grad_norm": 0.24589188706441673,
+      "learning_rate": 0.00013628163941936485,
+      "loss": 1.7644,
+      "step": 3020
+    },
+    {
+      "epoch": 1.3331864257382107,
+      "grad_norm": 0.277800302556096,
+      "learning_rate": 0.00013604257765913484,
+      "loss": 1.8151,
+      "step": 3025
+    },
+    {
+      "epoch": 1.3353900396650507,
+      "grad_norm": 0.23353363973323982,
+      "learning_rate": 0.0001358032789245366,
+      "loss": 1.7236,
+      "step": 3030
+    },
+    {
+      "epoch": 1.3375936535918906,
+      "grad_norm": 0.34178889147907426,
+      "learning_rate": 0.00013556374478892232,
+      "loss": 1.7669,
+      "step": 3035
+    },
+    {
+      "epoch": 1.3397972675187306,
+      "grad_norm": 0.28321134962453065,
+      "learning_rate": 0.00013532397682719185,
+      "loss": 1.6165,
+      "step": 3040
+    },
+    {
+      "epoch": 1.3420008814455708,
+      "grad_norm": 0.3160168756862356,
+      "learning_rate": 0.00013508397661578242,
+      "loss": 1.8131,
+      "step": 3045
+    },
+    {
+      "epoch": 1.3442044953724108,
+      "grad_norm": 0.2800833181613442,
+      "learning_rate": 0.0001348437457326582,
+      "loss": 1.9182,
+      "step": 3050
+    },
+    {
+      "epoch": 1.3464081092992508,
+      "grad_norm": 0.2679976561310916,
+      "learning_rate": 0.00013460328575730019,
+      "loss": 1.8312,
+      "step": 3055
+    },
+    {
+      "epoch": 1.3486117232260908,
+      "grad_norm": 0.2899158112172882,
+      "learning_rate": 0.00013436259827069534,
+      "loss": 1.8217,
+      "step": 3060
+    },
+    {
+      "epoch": 1.3508153371529308,
+      "grad_norm": 0.26525449374755994,
+      "learning_rate": 0.00013412168485532676,
+      "loss": 1.7636,
+      "step": 3065
+    },
+    {
+      "epoch": 1.3530189510797708,
+      "grad_norm": 0.2643909185128004,
+      "learning_rate": 0.00013388054709516272,
+      "loss": 1.6257,
+      "step": 3070
+    },
+    {
+      "epoch": 1.3552225650066108,
+      "grad_norm": 0.3217683133395989,
+      "learning_rate": 0.0001336391865756468,
+      "loss": 1.8385,
+      "step": 3075
+    },
+    {
+      "epoch": 1.357426178933451,
+      "grad_norm": 0.30256367251501726,
+      "learning_rate": 0.00013339760488368695,
+      "loss": 1.5994,
+      "step": 3080
+    },
+    {
+      "epoch": 1.359629792860291,
+      "grad_norm": 0.33945781722722157,
+      "learning_rate": 0.00013315580360764542,
+      "loss": 1.6502,
+      "step": 3085
+    },
+    {
+      "epoch": 1.361833406787131,
+      "grad_norm": 0.2343801676740979,
+      "learning_rate": 0.00013291378433732818,
+      "loss": 1.7302,
+      "step": 3090
+    },
+    {
+      "epoch": 1.364037020713971,
+      "grad_norm": 0.3789024872984378,
+      "learning_rate": 0.00013267154866397447,
+      "loss": 1.9092,
+      "step": 3095
+    },
+    {
+      "epoch": 1.366240634640811,
+      "grad_norm": 0.27745717625968813,
+      "learning_rate": 0.00013242909818024628,
+      "loss": 1.6587,
+      "step": 3100
+    },
+    {
+      "epoch": 1.368444248567651,
+      "grad_norm": 0.26534498515018917,
+      "learning_rate": 0.0001321864344802181,
+      "loss": 1.7184,
+      "step": 3105
+    },
+    {
+      "epoch": 1.3706478624944909,
+      "grad_norm": 0.25912670390615655,
+      "learning_rate": 0.00013194355915936611,
+      "loss": 1.7708,
+      "step": 3110
+    },
+    {
+      "epoch": 1.372851476421331,
+      "grad_norm": 0.23803080715278022,
+      "learning_rate": 0.000131700473814558,
+      "loss": 1.7224,
+      "step": 3115
+    },
+    {
+      "epoch": 1.375055090348171,
+      "grad_norm": 0.30746363012665606,
+      "learning_rate": 0.00013145718004404223,
+      "loss": 1.754,
+      "step": 3120
+    },
+    {
+      "epoch": 1.377258704275011,
+      "grad_norm": 0.2672636477164184,
+      "learning_rate": 0.00013121367944743777,
+      "loss": 1.6989,
+      "step": 3125
+    },
+    {
+      "epoch": 1.379462318201851,
+      "grad_norm": 0.28413083026015534,
+      "learning_rate": 0.0001309699736257232,
+      "loss": 1.6421,
+      "step": 3130
+    },
+    {
+      "epoch": 1.381665932128691,
+      "grad_norm": 0.3193377255035563,
+      "learning_rate": 0.00013072606418122667,
+      "loss": 1.8467,
+      "step": 3135
+    },
+    {
+      "epoch": 1.383869546055531,
+      "grad_norm": 0.2833583691387121,
+      "learning_rate": 0.00013048195271761498,
+      "loss": 1.6013,
+      "step": 3140
+    },
+    {
+      "epoch": 1.386073159982371,
+      "grad_norm": 0.25616581325290944,
+      "learning_rate": 0.00013023764083988323,
+      "loss": 1.7542,
+      "step": 3145
+    },
+    {
+      "epoch": 1.3882767739092112,
+      "grad_norm": 0.34179796686003233,
+      "learning_rate": 0.0001299931301543442,
+      "loss": 1.6674,
+      "step": 3150
+    },
+    {
+      "epoch": 1.390480387836051,
+      "grad_norm": 0.2612422912889042,
+      "learning_rate": 0.00012974842226861773,
+      "loss": 1.3979,
+      "step": 3155
+    },
+    {
+      "epoch": 1.3926840017628912,
+      "grad_norm": 0.30556335921079647,
+      "learning_rate": 0.0001295035187916204,
+      "loss": 1.7775,
+      "step": 3160
+    },
+    {
+      "epoch": 1.3948876156897312,
+      "grad_norm": 0.34927377227523054,
+      "learning_rate": 0.00012925842133355454,
+      "loss": 1.7384,
+      "step": 3165
+    },
+    {
+      "epoch": 1.3970912296165712,
+      "grad_norm": 0.35826503954646516,
+      "learning_rate": 0.00012901313150589806,
+      "loss": 1.8279,
+      "step": 3170
+    },
+    {
+      "epoch": 1.3992948435434112,
+      "grad_norm": 0.3558470664702752,
+      "learning_rate": 0.0001287676509213936,
+      "loss": 1.6467,
+      "step": 3175
+    },
+    {
+      "epoch": 1.4014984574702511,
+      "grad_norm": 0.2897608843662268,
+      "learning_rate": 0.00012852198119403798,
+      "loss": 1.6509,
+      "step": 3180
+    },
+    {
+      "epoch": 1.4037020713970914,
+      "grad_norm": 0.32428607006990234,
+      "learning_rate": 0.00012827612393907163,
+      "loss": 1.7118,
+      "step": 3185
+    },
+    {
+      "epoch": 1.4059056853239311,
+      "grad_norm": 0.29054441869310144,
+      "learning_rate": 0.0001280300807729679,
+      "loss": 1.6328,
+      "step": 3190
+    },
+    {
+      "epoch": 1.4081092992507713,
+      "grad_norm": 0.2672079347199706,
+      "learning_rate": 0.0001277838533134226,
+      "loss": 1.7875,
+      "step": 3195
+    },
+    {
+      "epoch": 1.4103129131776113,
+      "grad_norm": 0.3317588448314954,
+      "learning_rate": 0.00012753744317934307,
+      "loss": 1.9754,
+      "step": 3200
+    },
+    {
+      "epoch": 1.4125165271044513,
+      "grad_norm": 0.30976057441678767,
+      "learning_rate": 0.0001272908519908379,
+      "loss": 1.7292,
+      "step": 3205
+    },
+    {
+      "epoch": 1.4147201410312913,
+      "grad_norm": 0.2588899590749228,
+      "learning_rate": 0.00012704408136920585,
+      "loss": 1.661,
+      "step": 3210
+    },
+    {
+      "epoch": 1.4169237549581313,
+      "grad_norm": 0.34042631761749226,
+      "learning_rate": 0.0001267971329369256,
+      "loss": 1.7688,
+      "step": 3215
+    },
+    {
+      "epoch": 1.4191273688849715,
+      "grad_norm": 0.2613604846991468,
+      "learning_rate": 0.00012655000831764495,
+      "loss": 1.7979,
+      "step": 3220
+    },
+    {
+      "epoch": 1.4213309828118112,
+      "grad_norm": 0.30387612038339795,
+      "learning_rate": 0.00012630270913616985,
+      "loss": 1.6008,
+      "step": 3225
+    },
+    {
+      "epoch": 1.4235345967386515,
+      "grad_norm": 0.3245461529092582,
+      "learning_rate": 0.00012605523701845431,
+      "loss": 1.7394,
+      "step": 3230
+    },
+    {
+      "epoch": 1.4257382106654914,
+      "grad_norm": 0.23964868596701627,
+      "learning_rate": 0.00012580759359158905,
+      "loss": 1.5526,
+      "step": 3235
+    },
+    {
+      "epoch": 1.4279418245923314,
+      "grad_norm": 0.281150127445612,
+      "learning_rate": 0.00012555978048379133,
+      "loss": 1.6581,
+      "step": 3240
+    },
+    {
+      "epoch": 1.4301454385191714,
+      "grad_norm": 0.31208534783523834,
+      "learning_rate": 0.00012531179932439397,
+      "loss": 1.6698,
+      "step": 3245
+    },
+    {
+      "epoch": 1.4323490524460114,
+      "grad_norm": 0.3197645332854783,
+      "learning_rate": 0.00012506365174383467,
+      "loss": 1.8493,
+      "step": 3250
+    },
+    {
+      "epoch": 1.4345526663728516,
+      "grad_norm": 0.29747281096057276,
+      "learning_rate": 0.0001248153393736454,
+      "loss": 1.923,
+      "step": 3255
+    },
+    {
+      "epoch": 1.4367562802996914,
+      "grad_norm": 0.2706957926203667,
+      "learning_rate": 0.00012456686384644148,
+      "loss": 1.7219,
+      "step": 3260
+    },
+    {
+      "epoch": 1.4389598942265316,
+      "grad_norm": 0.3015008988665459,
+      "learning_rate": 0.00012431822679591112,
+      "loss": 1.6334,
+      "step": 3265
+    },
+    {
+      "epoch": 1.4411635081533716,
+      "grad_norm": 0.28824055515626146,
+      "learning_rate": 0.00012406942985680437,
+      "loss": 1.7096,
+      "step": 3270
+    },
+    {
+      "epoch": 1.4433671220802116,
+      "grad_norm": 0.28792375010811966,
+      "learning_rate": 0.00012382047466492262,
+      "loss": 1.6993,
+      "step": 3275
+    },
+    {
+      "epoch": 1.4455707360070515,
+      "grad_norm": 0.2778794221727809,
+      "learning_rate": 0.0001235713628571077,
+      "loss": 1.699,
+      "step": 3280
+    },
+    {
+      "epoch": 1.4477743499338915,
+      "grad_norm": 0.3173174516544841,
+      "learning_rate": 0.00012332209607123117,
+      "loss": 1.6214,
+      "step": 3285
+    },
+    {
+      "epoch": 1.4499779638607315,
+      "grad_norm": 0.30655650928697775,
+      "learning_rate": 0.0001230726759461836,
+      "loss": 1.7923,
+      "step": 3290
+    },
+    {
+      "epoch": 1.4521815777875715,
+      "grad_norm": 0.2517175305379352,
+      "learning_rate": 0.00012282310412186365,
+      "loss": 1.8434,
+      "step": 3295
+    },
+    {
+      "epoch": 1.4543851917144117,
+      "grad_norm": 0.24920356879351888,
+      "learning_rate": 0.0001225733822391675,
+      "loss": 1.6146,
+      "step": 3300
+    },
+    {
+      "epoch": 1.4565888056412517,
+      "grad_norm": 0.2950517285567546,
+      "learning_rate": 0.00012232351193997774,
+      "loss": 1.6819,
+      "step": 3305
+    },
+    {
+      "epoch": 1.4587924195680917,
+      "grad_norm": 0.3058013022960617,
+      "learning_rate": 0.000122073494867153,
+      "loss": 1.579,
+      "step": 3310
+    },
+    {
+      "epoch": 1.4609960334949317,
+      "grad_norm": 0.46566643268469327,
+      "learning_rate": 0.00012182333266451684,
+      "loss": 1.6713,
+      "step": 3315
+    },
+    {
+      "epoch": 1.4631996474217717,
+      "grad_norm": 0.3269081430168815,
+      "learning_rate": 0.00012157302697684695,
+      "loss": 1.6608,
+      "step": 3320
+    },
+    {
+      "epoch": 1.4654032613486117,
+      "grad_norm": 0.2472677464376836,
+      "learning_rate": 0.00012132257944986454,
+      "loss": 1.6504,
+      "step": 3325
+    },
+    {
+      "epoch": 1.4676068752754516,
+      "grad_norm": 0.3189526409165166,
+      "learning_rate": 0.00012107199173022327,
+      "loss": 1.6308,
+      "step": 3330
+    },
+    {
+      "epoch": 1.4698104892022918,
+      "grad_norm": 0.2439726953786154,
+      "learning_rate": 0.00012082126546549864,
+      "loss": 1.6694,
+      "step": 3335
+    },
+    {
+      "epoch": 1.4720141031291318,
+      "grad_norm": 0.28416189235796196,
+      "learning_rate": 0.000120570402304177,
+      "loss": 1.9048,
+      "step": 3340
+    },
+    {
+      "epoch": 1.4742177170559718,
+      "grad_norm": 0.25976522560441834,
+      "learning_rate": 0.00012031940389564478,
+      "loss": 1.7083,
+      "step": 3345
+    },
+    {
+      "epoch": 1.4764213309828118,
+      "grad_norm": 0.2874680056323443,
+      "learning_rate": 0.00012006827189017773,
+      "loss": 1.7914,
+      "step": 3350
+    },
+    {
+      "epoch": 1.4786249449096518,
+      "grad_norm": 0.35212840385267163,
+      "learning_rate": 0.00011981700793892982,
+      "loss": 1.8617,
+      "step": 3355
+    },
+    {
+      "epoch": 1.4808285588364918,
+      "grad_norm": 0.3258646795205973,
+      "learning_rate": 0.00011956561369392274,
+      "loss": 1.8569,
+      "step": 3360
+    },
+    {
+      "epoch": 1.4830321727633318,
+      "grad_norm": 0.25990120954046436,
+      "learning_rate": 0.0001193140908080346,
+      "loss": 1.7778,
+      "step": 3365
+    },
+    {
+      "epoch": 1.485235786690172,
+      "grad_norm": 0.2916499249746569,
+      "learning_rate": 0.00011906244093498955,
+      "loss": 1.7442,
+      "step": 3370
+    },
+    {
+      "epoch": 1.487439400617012,
+      "grad_norm": 0.38118475856684764,
+      "learning_rate": 0.00011881066572934644,
+      "loss": 1.6281,
+      "step": 3375
+    },
+    {
+      "epoch": 1.489643014543852,
+      "grad_norm": 0.33602332943649665,
+      "learning_rate": 0.00011855876684648837,
+      "loss": 1.6655,
+      "step": 3380
+    },
+    {
+      "epoch": 1.491846628470692,
+      "grad_norm": 0.3091891353046593,
+      "learning_rate": 0.00011830674594261145,
+      "loss": 1.818,
+      "step": 3385
+    },
+    {
+      "epoch": 1.494050242397532,
+      "grad_norm": 0.2675107541956203,
+      "learning_rate": 0.0001180546046747141,
+      "loss": 1.9917,
+      "step": 3390
+    },
+    {
+      "epoch": 1.496253856324372,
+      "grad_norm": 0.24171145502227592,
+      "learning_rate": 0.00011780234470058613,
+      "loss": 1.747,
+      "step": 3395
+    },
+    {
+      "epoch": 1.498457470251212,
+      "grad_norm": 0.31043683691075824,
+      "learning_rate": 0.0001175499676787978,
+      "loss": 1.7863,
+      "step": 3400
+    },
+    {
+      "epoch": 1.500661084178052,
+      "grad_norm": 0.25938236019167105,
+      "learning_rate": 0.000117297475268689,
+      "loss": 1.6216,
+      "step": 3405
+    },
+    {
+      "epoch": 1.5028646981048919,
+      "grad_norm": 0.2889407665894309,
+      "learning_rate": 0.00011704486913035819,
+      "loss": 1.7023,
+      "step": 3410
+    },
+    {
+      "epoch": 1.505068312031732,
+      "grad_norm": 0.2595635779433101,
+      "learning_rate": 0.00011679215092465163,
+      "loss": 1.6651,
+      "step": 3415
+    },
+    {
+      "epoch": 1.507271925958572,
+      "grad_norm": 0.3147360748379521,
+      "learning_rate": 0.00011653932231315245,
+      "loss": 1.6855,
+      "step": 3420
+    },
+    {
+      "epoch": 1.509475539885412,
+      "grad_norm": 0.2456935617451673,
+      "learning_rate": 0.00011628638495816955,
+      "loss": 1.6982,
+      "step": 3425
+    },
+    {
+      "epoch": 1.511679153812252,
+      "grad_norm": 0.2644287001406921,
+      "learning_rate": 0.00011603334052272696,
+      "loss": 1.7438,
+      "step": 3430
+    },
+    {
+      "epoch": 1.513882767739092,
+      "grad_norm": 0.3098487244790836,
+      "learning_rate": 0.0001157801906705526,
+      "loss": 1.7459,
+      "step": 3435
+    },
+    {
+      "epoch": 1.5160863816659322,
+      "grad_norm": 0.2635907160016382,
+      "learning_rate": 0.00011552693706606758,
+      "loss": 1.5969,
+      "step": 3440
+    },
+    {
+      "epoch": 1.518289995592772,
+      "grad_norm": 0.293285003547433,
+      "learning_rate": 0.00011527358137437516,
+      "loss": 1.7899,
+      "step": 3445
+    },
+    {
+      "epoch": 1.5204936095196122,
+      "grad_norm": 0.3154800148422279,
+      "learning_rate": 0.00011502012526124978,
+      "loss": 1.7859,
+      "step": 3450
+    },
+    {
+      "epoch": 1.5226972234464522,
+      "grad_norm": 0.3029474817652635,
+      "learning_rate": 0.00011476657039312613,
+      "loss": 1.8433,
+      "step": 3455
+    },
+    {
+      "epoch": 1.5249008373732922,
+      "grad_norm": 0.3394061586641444,
+      "learning_rate": 0.00011451291843708824,
+      "loss": 1.8191,
+      "step": 3460
+    },
+    {
+      "epoch": 1.5271044513001322,
+      "grad_norm": 0.28588437415991474,
+      "learning_rate": 0.00011425917106085844,
+      "loss": 1.6528,
+      "step": 3465
+    },
+    {
+      "epoch": 1.5293080652269722,
+      "grad_norm": 0.2624098766559022,
+      "learning_rate": 0.00011400532993278643,
+      "loss": 1.8208,
+      "step": 3470
+    },
+    {
+      "epoch": 1.5315116791538124,
+      "grad_norm": 0.29050819773398057,
+      "learning_rate": 0.00011375139672183834,
+      "loss": 1.763,
+      "step": 3475
+    },
+    {
+      "epoch": 1.5337152930806521,
+      "grad_norm": 0.28595831334373306,
+      "learning_rate": 0.00011349737309758572,
+      "loss": 1.6389,
+      "step": 3480
+    },
+    {
+      "epoch": 1.5359189070074923,
+      "grad_norm": 0.310106838673627,
+      "learning_rate": 0.00011324326073019458,
+      "loss": 1.7008,
+      "step": 3485
+    },
+    {
+      "epoch": 1.5381225209343323,
+      "grad_norm": 0.3425497639069633,
+      "learning_rate": 0.0001129890612904144,
+      "loss": 1.7975,
+      "step": 3490
+    },
+    {
+      "epoch": 1.5403261348611723,
+      "grad_norm": 0.3049155101860015,
+      "learning_rate": 0.0001127347764495671,
+      "loss": 1.6302,
+      "step": 3495
+    },
+    {
+      "epoch": 1.5425297487880123,
+      "grad_norm": 0.3083723366063809,
+      "learning_rate": 0.00011248040787953622,
+      "loss": 1.8779,
+      "step": 3500
+    },
+    {
+      "epoch": 1.5447333627148523,
+      "grad_norm": 0.2714184317474351,
+      "learning_rate": 0.00011222595725275562,
+      "loss": 1.6655,
+      "step": 3505
+    },
+    {
+      "epoch": 1.5469369766416925,
+      "grad_norm": 0.3020528286222207,
+      "learning_rate": 0.00011197142624219887,
+      "loss": 1.5374,
+      "step": 3510
+    },
+    {
+      "epoch": 1.5491405905685323,
+      "grad_norm": 0.270325291856936,
+      "learning_rate": 0.00011171681652136793,
+      "loss": 1.7442,
+      "step": 3515
+    },
+    {
+      "epoch": 1.5513442044953725,
+      "grad_norm": 0.2718748140775875,
+      "learning_rate": 0.00011146212976428232,
+      "loss": 1.793,
+      "step": 3520
+    },
+    {
+      "epoch": 1.5535478184222125,
+      "grad_norm": 0.2684213723870114,
+      "learning_rate": 0.00011120736764546799,
+      "loss": 1.5847,
+      "step": 3525
+    },
+    {
+      "epoch": 1.5557514323490524,
+      "grad_norm": 0.3293563453835575,
+      "learning_rate": 0.00011095253183994645,
+      "loss": 1.5808,
+      "step": 3530
+    },
+    {
+      "epoch": 1.5579550462758924,
+      "grad_norm": 0.32122911366332685,
+      "learning_rate": 0.0001106976240232237,
+      "loss": 1.7343,
+      "step": 3535
+    },
+    {
+      "epoch": 1.5601586602027324,
+      "grad_norm": 0.31939212525307864,
+      "learning_rate": 0.0001104426458712791,
+      "loss": 1.7123,
+      "step": 3540
+    },
+    {
+      "epoch": 1.5623622741295726,
+      "grad_norm": 0.2676344457188956,
+      "learning_rate": 0.00011018759906055463,
+      "loss": 1.4029,
+      "step": 3545
+    },
+    {
+      "epoch": 1.5645658880564124,
+      "grad_norm": 0.33141673784681086,
+      "learning_rate": 0.00010993248526794347,
+      "loss": 1.8105,
+      "step": 3550
+    },
+    {
+      "epoch": 1.5667695019832526,
+      "grad_norm": 0.2715916392293134,
+      "learning_rate": 0.00010967730617077938,
+      "loss": 1.73,
+      "step": 3555
+    },
+    {
+      "epoch": 1.5689731159100926,
+      "grad_norm": 0.2615038650065928,
+      "learning_rate": 0.00010942206344682541,
+      "loss": 1.7547,
+      "step": 3560
+    },
+    {
+      "epoch": 1.5711767298369326,
+      "grad_norm": 0.2662856667093564,
+      "learning_rate": 0.00010916675877426296,
+      "loss": 1.6934,
+      "step": 3565
+    },
+    {
+      "epoch": 1.5733803437637726,
+      "grad_norm": 0.20249417258651908,
+      "learning_rate": 0.00010891139383168072,
+      "loss": 1.7876,
+      "step": 3570
+    },
+    {
+      "epoch": 1.5755839576906125,
+      "grad_norm": 0.30704028921333987,
+      "learning_rate": 0.00010865597029806365,
+      "loss": 1.7228,
+      "step": 3575
+    },
+    {
+      "epoch": 1.5777875716174528,
+      "grad_norm": 0.32055060377455574,
+      "learning_rate": 0.00010840048985278195,
+      "loss": 1.7169,
+      "step": 3580
+    },
+    {
+      "epoch": 1.5799911855442925,
+      "grad_norm": 0.31196324685842286,
+      "learning_rate": 0.00010814495417557997,
+      "loss": 1.875,
+      "step": 3585
+    },
+    {
+      "epoch": 1.5821947994711327,
+      "grad_norm": 0.3005168077317045,
+      "learning_rate": 0.00010788936494656523,
+      "loss": 1.8862,
+      "step": 3590
+    },
+    {
+      "epoch": 1.5843984133979727,
+      "grad_norm": 0.2856041438770126,
+      "learning_rate": 0.00010763372384619738,
+      "loss": 1.6419,
+      "step": 3595
+    },
+    {
+      "epoch": 1.5866020273248127,
+      "grad_norm": 0.27459499931453724,
+      "learning_rate": 0.00010737803255527702,
+      "loss": 1.7495,
+      "step": 3600
+    },
+    {
+      "epoch": 1.5888056412516527,
+      "grad_norm": 0.23142425230470423,
+      "learning_rate": 0.00010712229275493489,
+      "loss": 1.7615,
+      "step": 3605
+    },
+    {
+      "epoch": 1.5910092551784927,
+      "grad_norm": 0.26704359594443944,
+      "learning_rate": 0.00010686650612662048,
+      "loss": 1.7043,
+      "step": 3610
+    },
+    {
+      "epoch": 1.5932128691053329,
+      "grad_norm": 0.2644858878666521,
+      "learning_rate": 0.00010661067435209135,
+      "loss": 1.8665,
+      "step": 3615
+    },
+    {
+      "epoch": 1.5954164830321726,
+      "grad_norm": 0.306618968505366,
+      "learning_rate": 0.00010635479911340176,
+      "loss": 1.8191,
+      "step": 3620
+    },
+    {
+      "epoch": 1.5976200969590129,
+      "grad_norm": 0.32629807608957256,
+      "learning_rate": 0.00010609888209289183,
+      "loss": 1.781,
+      "step": 3625
+    },
+    {
+      "epoch": 1.5998237108858528,
+      "grad_norm": 0.2701943161629176,
+      "learning_rate": 0.00010584292497317633,
+      "loss": 1.6162,
+      "step": 3630
+    },
+    {
+      "epoch": 1.6020273248126928,
+      "grad_norm": 0.40336230029221404,
+      "learning_rate": 0.00010558692943713373,
+      "loss": 1.722,
+      "step": 3635
+    },
+    {
+      "epoch": 1.6042309387395328,
+      "grad_norm": 0.2767732465609723,
+      "learning_rate": 0.000105330897167895,
+      "loss": 1.6427,
+      "step": 3640
+    },
+    {
+      "epoch": 1.6064345526663728,
+      "grad_norm": 0.28371049131821663,
+      "learning_rate": 0.00010507482984883268,
+      "loss": 1.6872,
+      "step": 3645
+    },
+    {
+      "epoch": 1.608638166593213,
+      "grad_norm": 0.2975217481512648,
+      "learning_rate": 0.00010481872916354978,
+      "loss": 1.6807,
+      "step": 3650
+    },
+    {
+      "epoch": 1.6108417805200528,
+      "grad_norm": 0.30522288055794994,
+      "learning_rate": 0.00010456259679586862,
+      "loss": 1.6253,
+      "step": 3655
+    },
+    {
+      "epoch": 1.613045394446893,
+      "grad_norm": 0.2893887034344458,
+      "learning_rate": 0.00010430643442981986,
+      "loss": 1.6465,
+      "step": 3660
+    },
+    {
+      "epoch": 1.615249008373733,
+      "grad_norm": 0.28628624467329145,
+      "learning_rate": 0.0001040502437496315,
+      "loss": 1.6428,
+      "step": 3665
+    },
+    {
+      "epoch": 1.617452622300573,
+      "grad_norm": 0.31683631498188874,
+      "learning_rate": 0.00010379402643971746,
+      "loss": 1.7033,
+      "step": 3670
+    },
+    {
+      "epoch": 1.619656236227413,
+      "grad_norm": 0.2627512390977551,
+      "learning_rate": 0.00010353778418466697,
+      "loss": 1.8805,
+      "step": 3675
+    },
+    {
+      "epoch": 1.621859850154253,
+      "grad_norm": 0.2976244892863047,
+      "learning_rate": 0.00010328151866923316,
+      "loss": 1.8013,
+      "step": 3680
+    },
+    {
+      "epoch": 1.6240634640810931,
+      "grad_norm": 0.31860629887164105,
+      "learning_rate": 0.00010302523157832216,
+      "loss": 1.648,
+      "step": 3685
+    },
+    {
+      "epoch": 1.626267078007933,
+      "grad_norm": 0.331987729083311,
+      "learning_rate": 0.00010276892459698182,
+      "loss": 1.6325,
+      "step": 3690
+    },
+    {
+      "epoch": 1.6284706919347731,
+      "grad_norm": 0.2818827039809831,
+      "learning_rate": 0.00010251259941039098,
+      "loss": 1.7213,
+      "step": 3695
+    },
+    {
+      "epoch": 1.6306743058616129,
+      "grad_norm": 0.2511192996283762,
+      "learning_rate": 0.00010225625770384797,
+      "loss": 1.5629,
+      "step": 3700
+    },
+    {
+      "epoch": 1.632877919788453,
+      "grad_norm": 0.29741607234777423,
+      "learning_rate": 0.00010199990116275988,
+      "loss": 1.7834,
+      "step": 3705
+    },
+    {
+      "epoch": 1.635081533715293,
+      "grad_norm": 0.30036469948366823,
+      "learning_rate": 0.00010174353147263125,
+      "loss": 1.4849,
+      "step": 3710
+    },
+    {
+      "epoch": 1.637285147642133,
+      "grad_norm": 0.30522753547462433,
+      "learning_rate": 0.00010148715031905312,
+      "loss": 1.8071,
+      "step": 3715
+    },
+    {
+      "epoch": 1.6394887615689733,
+      "grad_norm": 0.2769872314517015,
+      "learning_rate": 0.00010123075938769187,
+      "loss": 1.685,
+      "step": 3720
+    },
+    {
+      "epoch": 1.641692375495813,
+      "grad_norm": 0.2709018320510704,
+      "learning_rate": 0.00010097436036427816,
+      "loss": 1.7853,
+      "step": 3725
+    },
+    {
+      "epoch": 1.6438959894226532,
+      "grad_norm": 0.3446719925528754,
+      "learning_rate": 0.00010071795493459591,
+      "loss": 1.7783,
+      "step": 3730
+    },
+    {
+      "epoch": 1.646099603349493,
+      "grad_norm": 0.3208180036756325,
+      "learning_rate": 0.00010046154478447114,
+      "loss": 1.8982,
+      "step": 3735
+    },
+    {
+      "epoch": 1.6483032172763332,
+      "grad_norm": 0.289239117609306,
+      "learning_rate": 0.00010020513159976084,
+      "loss": 1.7313,
+      "step": 3740
+    },
+    {
+      "epoch": 1.6505068312031732,
+      "grad_norm": 0.30122205784761724,
+      "learning_rate": 9.994871706634204e-05,
+      "loss": 1.6831,
+      "step": 3745
+    },
+    {
+      "epoch": 1.6527104451300132,
+      "grad_norm": 0.2835725442151606,
+      "learning_rate": 9.96923028701006e-05,
+      "loss": 1.6129,
+      "step": 3750
+    },
+    {
+      "epoch": 1.6549140590568534,
+      "grad_norm": 0.4849263642027594,
+      "learning_rate": 9.943589069692014e-05,
+      "loss": 1.783,
+      "step": 3755
+    },
+    {
+      "epoch": 1.6571176729836932,
+      "grad_norm": 0.34060901148734135,
+      "learning_rate": 9.917948223267105e-05,
+      "loss": 1.642,
+      "step": 3760
+    },
+    {
+      "epoch": 1.6593212869105334,
+      "grad_norm": 0.3001957536668968,
+      "learning_rate": 9.892307916319919e-05,
+      "loss": 1.6005,
+      "step": 3765
+    },
+    {
+      "epoch": 1.6615249008373731,
+      "grad_norm": 0.3141764748091563,
+      "learning_rate": 9.866668317431514e-05,
+      "loss": 1.8968,
+      "step": 3770
+    },
+    {
+      "epoch": 1.6637285147642134,
+      "grad_norm": 0.3066452663914771,
+      "learning_rate": 9.841029595178282e-05,
+      "loss": 1.6288,
+      "step": 3775
+    },
+    {
+      "epoch": 1.6659321286910533,
+      "grad_norm": 0.2737312482046674,
+      "learning_rate": 9.815391918130848e-05,
+      "loss": 1.6151,
+      "step": 3780
+    },
+    {
+      "epoch": 1.6681357426178933,
+      "grad_norm": 0.337368580313636,
+      "learning_rate": 9.789755454852971e-05,
+      "loss": 1.6298,
+      "step": 3785
+    },
+    {
+      "epoch": 1.6703393565447333,
+      "grad_norm": 0.22864544249672591,
+      "learning_rate": 9.764120373900436e-05,
+      "loss": 1.7166,
+      "step": 3790
+    },
+    {
+      "epoch": 1.6725429704715733,
+      "grad_norm": 0.3199596986088049,
+      "learning_rate": 9.738486843819919e-05,
+      "loss": 1.4291,
+      "step": 3795
+    },
+    {
+      "epoch": 1.6747465843984135,
+      "grad_norm": 0.484113228627448,
+      "learning_rate": 9.712855033147921e-05,
+      "loss": 1.8267,
+      "step": 3800
+    },
+    {
+      "epoch": 1.6769501983252533,
+      "grad_norm": 0.29066627384947913,
+      "learning_rate": 9.68722511040962e-05,
+      "loss": 1.7618,
+      "step": 3805
+    },
+    {
+      "epoch": 1.6791538122520935,
+      "grad_norm": 0.2756578416564992,
+      "learning_rate": 9.661597244117802e-05,
+      "loss": 1.7626,
+      "step": 3810
+    },
+    {
+      "epoch": 1.6813574261789335,
+      "grad_norm": 0.26262162877838774,
+      "learning_rate": 9.635971602771716e-05,
+      "loss": 1.6437,
+      "step": 3815
+    },
+    {
+      "epoch": 1.6835610401057735,
+      "grad_norm": 0.3485930873206027,
+      "learning_rate": 9.61034835485598e-05,
+      "loss": 1.9054,
+      "step": 3820
+    },
+    {
+      "epoch": 1.6857646540326134,
+      "grad_norm": 0.347720471228885,
+      "learning_rate": 9.584727668839487e-05,
+      "loss": 1.6653,
+      "step": 3825
+    },
+    {
+      "epoch": 1.6879682679594534,
+      "grad_norm": 0.259896835339376,
+      "learning_rate": 9.559109713174282e-05,
+      "loss": 1.8298,
+      "step": 3830
+    },
+    {
+      "epoch": 1.6901718818862936,
+      "grad_norm": 0.27688333027135853,
+      "learning_rate": 9.533494656294458e-05,
+      "loss": 1.5074,
+      "step": 3835
+    },
+    {
+      "epoch": 1.6923754958131334,
+      "grad_norm": 0.3186540043438054,
+      "learning_rate": 9.507882666615049e-05,
+      "loss": 1.666,
+      "step": 3840
+    },
+    {
+      "epoch": 1.6945791097399736,
+      "grad_norm": 0.34137430082378867,
+      "learning_rate": 9.482273912530913e-05,
+      "loss": 1.7862,
+      "step": 3845
+    },
+    {
+      "epoch": 1.6967827236668136,
+      "grad_norm": 0.31312228735788816,
+      "learning_rate": 9.456668562415657e-05,
+      "loss": 1.6969,
+      "step": 3850
+    },
+    {
+      "epoch": 1.6989863375936536,
+      "grad_norm": 0.3211377859121694,
+      "learning_rate": 9.431066784620486e-05,
+      "loss": 1.8653,
+      "step": 3855
+    },
+    {
+      "epoch": 1.7011899515204936,
+      "grad_norm": 0.3443229930133863,
+      "learning_rate": 9.405468747473127e-05,
+      "loss": 1.696,
+      "step": 3860
+    },
+    {
+      "epoch": 1.7033935654473336,
+      "grad_norm": 0.2900871769643087,
+      "learning_rate": 9.379874619276707e-05,
+      "loss": 1.7654,
+      "step": 3865
+    },
+    {
+      "epoch": 1.7055971793741738,
+      "grad_norm": 0.3109956258190758,
+      "learning_rate": 9.354284568308665e-05,
+      "loss": 1.6229,
+      "step": 3870
+    },
+    {
+      "epoch": 1.7078007933010135,
+      "grad_norm": 0.29096423007153205,
+      "learning_rate": 9.328698762819623e-05,
+      "loss": 1.7274,
+      "step": 3875
+    },
+    {
+      "epoch": 1.7100044072278537,
+      "grad_norm": 0.2674081631314747,
+      "learning_rate": 9.303117371032284e-05,
+      "loss": 1.7598,
+      "step": 3880
+    },
+    {
+      "epoch": 1.7122080211546937,
+      "grad_norm": 0.3370488411342742,
+      "learning_rate": 9.277540561140342e-05,
+      "loss": 1.7854,
+      "step": 3885
+    },
+    {
+      "epoch": 1.7144116350815337,
+      "grad_norm": 0.2236164025203487,
+      "learning_rate": 9.251968501307365e-05,
+      "loss": 1.735,
+      "step": 3890
+    },
+    {
+      "epoch": 1.7166152490083737,
+      "grad_norm": 0.31253584348251195,
+      "learning_rate": 9.226401359665686e-05,
+      "loss": 1.5912,
+      "step": 3895
+    },
+    {
+      "epoch": 1.7188188629352137,
+      "grad_norm": 0.28523913611604496,
+      "learning_rate": 9.2008393043153e-05,
+      "loss": 1.7252,
+      "step": 3900
+    },
+    {
+      "epoch": 1.721022476862054,
+      "grad_norm": 0.3641271132364838,
+      "learning_rate": 9.17528250332277e-05,
+      "loss": 1.7001,
+      "step": 3905
+    },
+    {
+      "epoch": 1.7232260907888937,
+      "grad_norm": 0.34337087912187736,
+      "learning_rate": 9.149731124720104e-05,
+      "loss": 1.941,
+      "step": 3910
+    },
+    {
+      "epoch": 1.7254297047157339,
+      "grad_norm": 0.3388940885148476,
+      "learning_rate": 9.124185336503656e-05,
+      "loss": 1.751,
+      "step": 3915
+    },
+    {
+      "epoch": 1.7276333186425739,
+      "grad_norm": 0.3980423482061792,
+      "learning_rate": 9.098645306633029e-05,
+      "loss": 1.7571,
+      "step": 3920
+    },
+    {
+      "epoch": 1.7298369325694138,
+      "grad_norm": 0.28139197000268995,
+      "learning_rate": 9.073111203029972e-05,
+      "loss": 1.5225,
+      "step": 3925
+    },
+    {
+      "epoch": 1.7320405464962538,
+      "grad_norm": 0.28174263964005763,
+      "learning_rate": 9.04758319357726e-05,
+      "loss": 1.6746,
+      "step": 3930
+    },
+    {
+      "epoch": 1.7342441604230938,
+      "grad_norm": 0.2461518030614674,
+      "learning_rate": 9.0220614461176e-05,
+      "loss": 1.5447,
+      "step": 3935
+    },
+    {
+      "epoch": 1.736447774349934,
+      "grad_norm": 0.3297581828947644,
+      "learning_rate": 8.99654612845253e-05,
+      "loss": 1.5802,
+      "step": 3940
+    },
+    {
+      "epoch": 1.7386513882767738,
+      "grad_norm": 0.3434764691584567,
+      "learning_rate": 8.971037408341319e-05,
+      "loss": 1.6836,
+      "step": 3945
+    },
+    {
+      "epoch": 1.740855002203614,
+      "grad_norm": 0.2967197456557559,
+      "learning_rate": 8.94553545349985e-05,
+      "loss": 1.6141,
+      "step": 3950
+    },
+    {
+      "epoch": 1.743058616130454,
+      "grad_norm": 0.3235657879123794,
+      "learning_rate": 8.92004043159953e-05,
+      "loss": 1.8539,
+      "step": 3955
+    },
+    {
+      "epoch": 1.745262230057294,
+      "grad_norm": 0.30230482414452203,
+      "learning_rate": 8.894552510266172e-05,
+      "loss": 1.6447,
+      "step": 3960
+    },
+    {
+      "epoch": 1.747465843984134,
+      "grad_norm": 0.25937552544500664,
+      "learning_rate": 8.869071857078926e-05,
+      "loss": 1.7132,
+      "step": 3965
+    },
+    {
+      "epoch": 1.749669457910974,
+      "grad_norm": 0.3504915271341817,
+      "learning_rate": 8.843598639569134e-05,
+      "loss": 1.616,
+      "step": 3970
+    },
+    {
+      "epoch": 1.7518730718378142,
+      "grad_norm": 0.3391539087204084,
+      "learning_rate": 8.818133025219258e-05,
+      "loss": 1.7726,
+      "step": 3975
+    },
+    {
+      "epoch": 1.754076685764654,
+      "grad_norm": 0.2945081027363211,
+      "learning_rate": 8.79267518146177e-05,
+      "loss": 1.6354,
+      "step": 3980
+    },
+    {
+      "epoch": 1.7562802996914941,
+      "grad_norm": 0.3122456807955038,
+      "learning_rate": 8.767225275678054e-05,
+      "loss": 1.7703,
+      "step": 3985
+    },
+    {
+      "epoch": 1.7584839136183341,
+      "grad_norm": 0.2911403528384371,
+      "learning_rate": 8.741783475197301e-05,
+      "loss": 1.6184,
+      "step": 3990
+    },
+    {
+      "epoch": 1.760687527545174,
+      "grad_norm": 0.28063561583988544,
+      "learning_rate": 8.716349947295406e-05,
+      "loss": 1.6723,
+      "step": 3995
+    },
+    {
+      "epoch": 1.762891141472014,
+      "grad_norm": 0.24899262558923158,
+      "learning_rate": 8.690924859193877e-05,
+      "loss": 1.5999,
+      "step": 4000
+    },
+    {
+      "epoch": 1.765094755398854,
+      "grad_norm": 0.21950111519024182,
+      "learning_rate": 8.665508378058737e-05,
+      "loss": 1.59,
+      "step": 4005
+    },
+    {
+      "epoch": 1.7672983693256943,
+      "grad_norm": 0.2812929263006604,
+      "learning_rate": 8.640100670999413e-05,
+      "loss": 1.7219,
+      "step": 4010
+    },
+    {
+      "epoch": 1.769501983252534,
+      "grad_norm": 0.26800092694380395,
+      "learning_rate": 8.614701905067648e-05,
+      "loss": 1.5878,
+      "step": 4015
+    },
+    {
+      "epoch": 1.7717055971793743,
+      "grad_norm": 0.3421307184861735,
+      "learning_rate": 8.589312247256385e-05,
+      "loss": 1.5289,
+      "step": 4020
+    },
+    {
+      "epoch": 1.7739092111062142,
+      "grad_norm": 0.2759648702531539,
+      "learning_rate": 8.563931864498709e-05,
+      "loss": 1.7232,
+      "step": 4025
+    },
+    {
+      "epoch": 1.7761128250330542,
+      "grad_norm": 0.3083755515457062,
+      "learning_rate": 8.538560923666697e-05,
+      "loss": 1.5333,
+      "step": 4030
+    },
+    {
+      "epoch": 1.7783164389598942,
+      "grad_norm": 0.32344968491148207,
+      "learning_rate": 8.51319959157036e-05,
+      "loss": 1.6531,
+      "step": 4035
+    },
+    {
+      "epoch": 1.7805200528867342,
+      "grad_norm": 0.2628449279485873,
+      "learning_rate": 8.487848034956527e-05,
+      "loss": 1.7176,
+      "step": 4040
+    },
+    {
+      "epoch": 1.7827236668135744,
+      "grad_norm": 0.30081231536696845,
+      "learning_rate": 8.462506420507764e-05,
+      "loss": 1.6087,
+      "step": 4045
+    },
+    {
+      "epoch": 1.7849272807404142,
+      "grad_norm": 0.24172711814982975,
+      "learning_rate": 8.437174914841261e-05,
+      "loss": 1.6365,
+      "step": 4050
+    },
+    {
+      "epoch": 1.7871308946672544,
+      "grad_norm": 0.3240513311621846,
+      "learning_rate": 8.411853684507744e-05,
+      "loss": 1.6818,
+      "step": 4055
+    },
+    {
+      "epoch": 1.7893345085940942,
+      "grad_norm": 0.26629546007810934,
+      "learning_rate": 8.38654289599038e-05,
+      "loss": 1.6165,
+      "step": 4060
+    },
+    {
+      "epoch": 1.7915381225209344,
+      "grad_norm": 0.2718697282057903,
+      "learning_rate": 8.36124271570369e-05,
+      "loss": 1.7767,
+      "step": 4065
+    },
+    {
+      "epoch": 1.7937417364477743,
+      "grad_norm": 0.2991997948062704,
+      "learning_rate": 8.335953309992442e-05,
+      "loss": 1.6968,
+      "step": 4070
+    },
+    {
+      "epoch": 1.7959453503746143,
+      "grad_norm": 0.33154914217795595,
+      "learning_rate": 8.310674845130563e-05,
+      "loss": 1.8523,
+      "step": 4075
+    },
+    {
+      "epoch": 1.7981489643014545,
+      "grad_norm": 0.29895299240181794,
+      "learning_rate": 8.285407487320042e-05,
+      "loss": 1.5945,
+      "step": 4080
+    },
+    {
+      "epoch": 1.8003525782282943,
+      "grad_norm": 0.2947016318335874,
+      "learning_rate": 8.260151402689848e-05,
+      "loss": 1.625,
+      "step": 4085
+    },
+    {
+      "epoch": 1.8025561921551345,
+      "grad_norm": 0.3039906812402064,
+      "learning_rate": 8.234906757294829e-05,
+      "loss": 1.8956,
+      "step": 4090
+    },
+    {
+      "epoch": 1.8047598060819743,
+      "grad_norm": 0.311603713861004,
+      "learning_rate": 8.209673717114618e-05,
+      "loss": 1.5808,
+      "step": 4095
+    },
+    {
+      "epoch": 1.8069634200088145,
+      "grad_norm": 0.3208961582381735,
+      "learning_rate": 8.184452448052547e-05,
+      "loss": 1.4928,
+      "step": 4100
+    },
+    {
+      "epoch": 1.8091670339356545,
+      "grad_norm": 0.27150517803601,
+      "learning_rate": 8.15924311593456e-05,
+      "loss": 1.7155,
+      "step": 4105
+    },
+    {
+      "epoch": 1.8113706478624945,
+      "grad_norm": 0.31637619824367874,
+      "learning_rate": 8.134045886508108e-05,
+      "loss": 1.4761,
+      "step": 4110
+    },
+    {
+      "epoch": 1.8135742617893347,
+      "grad_norm": 0.2536681416637017,
+      "learning_rate": 8.108860925441076e-05,
+      "loss": 1.7682,
+      "step": 4115
+    },
+    {
+      "epoch": 1.8157778757161744,
+      "grad_norm": 0.2683657294785762,
+      "learning_rate": 8.083688398320681e-05,
+      "loss": 1.6091,
+      "step": 4120
+    },
+    {
+      "epoch": 1.8179814896430146,
+      "grad_norm": 0.3004847316685122,
+      "learning_rate": 8.058528470652396e-05,
+      "loss": 1.7524,
+      "step": 4125
+    },
+    {
+      "epoch": 1.8201851035698544,
+      "grad_norm": 0.3033198553489031,
+      "learning_rate": 8.03338130785885e-05,
+      "loss": 1.6975,
+      "step": 4130
+    },
+    {
+      "epoch": 1.8223887174966946,
+      "grad_norm": 0.2932969746872377,
+      "learning_rate": 8.008247075278742e-05,
+      "loss": 1.6345,
+      "step": 4135
+    },
+    {
+      "epoch": 1.8245923314235346,
+      "grad_norm": 0.29419726566032006,
+      "learning_rate": 7.983125938165758e-05,
+      "loss": 2.0007,
+      "step": 4140
+    },
+    {
+      "epoch": 1.8267959453503746,
+      "grad_norm": 0.36994662608619383,
+      "learning_rate": 7.958018061687494e-05,
+      "loss": 1.8041,
+      "step": 4145
+    },
+    {
+      "epoch": 1.8289995592772146,
+      "grad_norm": 0.3171521170329957,
+      "learning_rate": 7.932923610924343e-05,
+      "loss": 1.8268,
+      "step": 4150
+    },
+    {
+      "epoch": 1.8312031732040546,
+      "grad_norm": 0.29749910928069495,
+      "learning_rate": 7.907842750868441e-05,
+      "loss": 1.7521,
+      "step": 4155
+    },
+    {
+      "epoch": 1.8334067871308948,
+      "grad_norm": 0.28978378846287695,
+      "learning_rate": 7.882775646422547e-05,
+      "loss": 1.8141,
+      "step": 4160
+    },
+    {
+      "epoch": 1.8356104010577345,
+      "grad_norm": 0.32370603539130244,
+      "learning_rate": 7.857722462399009e-05,
+      "loss": 1.5852,
+      "step": 4165
+    },
+    {
+      "epoch": 1.8378140149845748,
+      "grad_norm": 0.3000864365709902,
+      "learning_rate": 7.832683363518621e-05,
+      "loss": 1.5174,
+      "step": 4170
+    },
+    {
+      "epoch": 1.8400176289114147,
+      "grad_norm": 0.30867413513998754,
+      "learning_rate": 7.807658514409587e-05,
+      "loss": 1.7091,
+      "step": 4175
+    },
+    {
+      "epoch": 1.8422212428382547,
+      "grad_norm": 0.2941777962805528,
+      "learning_rate": 7.782648079606412e-05,
+      "loss": 1.9314,
+      "step": 4180
+    },
+    {
+      "epoch": 1.8444248567650947,
+      "grad_norm": 0.26789438468880916,
+      "learning_rate": 7.757652223548836e-05,
+      "loss": 1.5959,
+      "step": 4185
+    },
+    {
+      "epoch": 1.8466284706919347,
+      "grad_norm": 0.2852613225045537,
+      "learning_rate": 7.732671110580746e-05,
+      "loss": 1.4776,
+      "step": 4190
+    },
+    {
+      "epoch": 1.848832084618775,
+      "grad_norm": 0.28204667104600045,
+      "learning_rate": 7.707704904949085e-05,
+      "loss": 1.6044,
+      "step": 4195
+    },
+    {
+      "epoch": 1.8510356985456147,
+      "grad_norm": 0.3433461472883869,
+      "learning_rate": 7.682753770802791e-05,
+      "loss": 1.8343,
+      "step": 4200
+    },
+    {
+      "epoch": 1.8532393124724549,
+      "grad_norm": 0.27355747329642977,
+      "learning_rate": 7.657817872191713e-05,
+      "loss": 1.6496,
+      "step": 4205
+    },
+    {
+      "epoch": 1.8554429263992949,
+      "grad_norm": 0.2860370187153179,
+      "learning_rate": 7.632897373065522e-05,
+      "loss": 1.5719,
+      "step": 4210
+    },
+    {
+      "epoch": 1.8576465403261349,
+      "grad_norm": 0.2618172945255669,
+      "learning_rate": 7.607992437272642e-05,
+      "loss": 1.6911,
+      "step": 4215
+    },
+    {
+      "epoch": 1.8598501542529748,
+      "grad_norm": 0.2570942578849571,
+      "learning_rate": 7.583103228559164e-05,
+      "loss": 1.7205,
+      "step": 4220
+    },
+    {
+      "epoch": 1.8620537681798148,
+      "grad_norm": 0.30089829678159197,
+      "learning_rate": 7.558229910567794e-05,
+      "loss": 1.7114,
+      "step": 4225
+    },
+    {
+      "epoch": 1.864257382106655,
+      "grad_norm": 0.3120407343609943,
+      "learning_rate": 7.533372646836736e-05,
+      "loss": 1.6438,
+      "step": 4230
+    },
+    {
+      "epoch": 1.8664609960334948,
+      "grad_norm": 0.29920830893784145,
+      "learning_rate": 7.508531600798657e-05,
+      "loss": 1.8773,
+      "step": 4235
+    },
+    {
+      "epoch": 1.868664609960335,
+      "grad_norm": 0.28932016728007587,
+      "learning_rate": 7.483706935779584e-05,
+      "loss": 1.5398,
+      "step": 4240
+    },
+    {
+      "epoch": 1.870868223887175,
+      "grad_norm": 0.3323660400642421,
+      "learning_rate": 7.458898814997852e-05,
+      "loss": 1.8427,
+      "step": 4245
+    },
+    {
+      "epoch": 1.873071837814015,
+      "grad_norm": 0.24905133812180258,
+      "learning_rate": 7.434107401563016e-05,
+      "loss": 1.5713,
+      "step": 4250
+    },
+    {
+      "epoch": 1.875275451740855,
+      "grad_norm": 0.3177914709078092,
+      "learning_rate": 7.409332858474772e-05,
+      "loss": 1.8163,
+      "step": 4255
+    },
+    {
+      "epoch": 1.877479065667695,
+      "grad_norm": 0.2731160161178577,
+      "learning_rate": 7.384575348621909e-05,
+      "loss": 1.502,
+      "step": 4260
+    },
+    {
+      "epoch": 1.8796826795945352,
+      "grad_norm": 0.318150552351912,
+      "learning_rate": 7.359835034781227e-05,
+      "loss": 1.709,
+      "step": 4265
+    },
+    {
+      "epoch": 1.881886293521375,
+      "grad_norm": 0.2992460408245423,
+      "learning_rate": 7.335112079616456e-05,
+      "loss": 1.6948,
+      "step": 4270
+    },
+    {
+      "epoch": 1.8840899074482151,
+      "grad_norm": 0.26242651145469986,
+      "learning_rate": 7.31040664567719e-05,
+      "loss": 1.5288,
+      "step": 4275
+    },
+    {
+      "epoch": 1.8862935213750551,
+      "grad_norm": 0.32400746012853343,
+      "learning_rate": 7.285718895397848e-05,
+      "loss": 1.7299,
+      "step": 4280
+    },
+    {
+      "epoch": 1.8884971353018951,
+      "grad_norm": 0.30579297267422695,
+      "learning_rate": 7.261048991096558e-05,
+      "loss": 1.8361,
+      "step": 4285
+    },
+    {
+      "epoch": 1.890700749228735,
+      "grad_norm": 0.28173595807802665,
+      "learning_rate": 7.236397094974119e-05,
+      "loss": 1.737,
+      "step": 4290
+    },
+    {
+      "epoch": 1.892904363155575,
+      "grad_norm": 0.3062487579152163,
+      "learning_rate": 7.211763369112934e-05,
+      "loss": 1.6612,
+      "step": 4295
+    },
+    {
+      "epoch": 1.8951079770824153,
+      "grad_norm": 0.29979139920156694,
+      "learning_rate": 7.18714797547594e-05,
+      "loss": 1.762,
+      "step": 4300
+    },
+    {
+      "epoch": 1.897311591009255,
+      "grad_norm": 0.3021721281737111,
+      "learning_rate": 7.162551075905538e-05,
+      "loss": 1.8317,
+      "step": 4305
+    },
+    {
+      "epoch": 1.8995152049360953,
+      "grad_norm": 0.2809347403792435,
+      "learning_rate": 7.137972832122532e-05,
+      "loss": 1.7406,
+      "step": 4310
+    },
+    {
+      "epoch": 1.9017188188629353,
+      "grad_norm": 0.31981826220168086,
+      "learning_rate": 7.113413405725069e-05,
+      "loss": 1.8273,
+      "step": 4315
+    },
+    {
+      "epoch": 1.9039224327897752,
+      "grad_norm": 0.27829653919403746,
+      "learning_rate": 7.088872958187578e-05,
+      "loss": 1.6196,
+      "step": 4320
+    },
+    {
+      "epoch": 1.9061260467166152,
+      "grad_norm": 0.2612389626891207,
+      "learning_rate": 7.064351650859704e-05,
+      "loss": 1.7173,
+      "step": 4325
+    },
+    {
+      "epoch": 1.9083296606434552,
+      "grad_norm": 0.3351698187645255,
+      "learning_rate": 7.039849644965246e-05,
+      "loss": 1.5561,
+      "step": 4330
+    },
+    {
+      "epoch": 1.9105332745702954,
+      "grad_norm": 0.276151473793176,
+      "learning_rate": 7.015367101601091e-05,
+      "loss": 1.5952,
+      "step": 4335
+    },
+    {
+      "epoch": 1.9127368884971352,
+      "grad_norm": 0.2866569066490694,
+      "learning_rate": 6.990904181736187e-05,
+      "loss": 1.7386,
+      "step": 4340
+    },
+    {
+      "epoch": 1.9149405024239754,
+      "grad_norm": 0.30680340268949835,
+      "learning_rate": 6.96646104621043e-05,
+      "loss": 1.793,
+      "step": 4345
+    },
+    {
+      "epoch": 1.9171441163508154,
+      "grad_norm": 0.2853878652168025,
+      "learning_rate": 6.942037855733661e-05,
+      "loss": 1.8032,
+      "step": 4350
+    },
+    {
+      "epoch": 1.9193477302776554,
+      "grad_norm": 0.3335656030865567,
+      "learning_rate": 6.917634770884571e-05,
+      "loss": 1.8019,
+      "step": 4355
+    },
+    {
+      "epoch": 1.9215513442044954,
+      "grad_norm": 0.26469328077404297,
+      "learning_rate": 6.893251952109668e-05,
+      "loss": 1.7769,
+      "step": 4360
+    },
+    {
+      "epoch": 1.9237549581313353,
+      "grad_norm": 0.3611778018097071,
+      "learning_rate": 6.868889559722213e-05,
+      "loss": 1.845,
+      "step": 4365
+    },
+    {
+      "epoch": 1.9259585720581756,
+      "grad_norm": 0.2607913972126291,
+      "learning_rate": 6.84454775390116e-05,
+      "loss": 1.6583,
+      "step": 4370
+    },
+    {
+      "epoch": 1.9281621859850153,
+      "grad_norm": 0.2952126078040435,
+      "learning_rate": 6.820226694690112e-05,
+      "loss": 1.4631,
+      "step": 4375
+    },
+    {
+      "epoch": 1.9303657999118555,
+      "grad_norm": 0.28021793360716346,
+      "learning_rate": 6.795926541996273e-05,
+      "loss": 1.7197,
+      "step": 4380
+    },
+    {
+      "epoch": 1.9325694138386955,
+      "grad_norm": 0.2431287772236666,
+      "learning_rate": 6.771647455589384e-05,
+      "loss": 1.7528,
+      "step": 4385
+    },
+    {
+      "epoch": 1.9347730277655355,
+      "grad_norm": 0.24277132942171645,
+      "learning_rate": 6.74738959510068e-05,
+      "loss": 1.52,
+      "step": 4390
+    },
+    {
+      "epoch": 1.9369766416923755,
+      "grad_norm": 0.2616438959393551,
+      "learning_rate": 6.723153120021833e-05,
+      "loss": 1.6386,
+      "step": 4395
+    },
+    {
+      "epoch": 1.9391802556192155,
+      "grad_norm": 0.3044048108764881,
+      "learning_rate": 6.698938189703918e-05,
+      "loss": 1.653,
+      "step": 4400
+    },
+    {
+      "epoch": 1.9413838695460557,
+      "grad_norm": 0.2952474841434869,
+      "learning_rate": 6.674744963356357e-05,
+      "loss": 1.6325,
+      "step": 4405
+    },
+    {
+      "epoch": 1.9435874834728954,
+      "grad_norm": 0.3104804592935538,
+      "learning_rate": 6.65057360004586e-05,
+      "loss": 1.7827,
+      "step": 4410
+    },
+    {
+      "epoch": 1.9457910973997357,
+      "grad_norm": 0.27421700814052996,
+      "learning_rate": 6.626424258695403e-05,
+      "loss": 1.6614,
+      "step": 4415
+    },
+    {
+      "epoch": 1.9479947113265754,
+      "grad_norm": 0.28977404672357854,
+      "learning_rate": 6.60229709808317e-05,
+      "loss": 1.8225,
+      "step": 4420
+    },
+    {
+      "epoch": 1.9501983252534156,
+      "grad_norm": 0.30875392634058607,
+      "learning_rate": 6.578192276841501e-05,
+      "loss": 1.7437,
+      "step": 4425
+    },
+    {
+      "epoch": 1.9524019391802556,
+      "grad_norm": 0.307292209389746,
+      "learning_rate": 6.554109953455864e-05,
+      "loss": 1.7637,
+      "step": 4430
+    },
+    {
+      "epoch": 1.9546055531070956,
+      "grad_norm": 0.34674701377289235,
+      "learning_rate": 6.53005028626381e-05,
+      "loss": 1.6782,
+      "step": 4435
+    },
+    {
+      "epoch": 1.9568091670339358,
+      "grad_norm": 0.2997355914742966,
+      "learning_rate": 6.506013433453926e-05,
+      "loss": 1.7479,
+      "step": 4440
+    },
+    {
+      "epoch": 1.9590127809607756,
+      "grad_norm": 0.33789074614600445,
+      "learning_rate": 6.4819995530648e-05,
+      "loss": 1.6811,
+      "step": 4445
+    },
+    {
+      "epoch": 1.9612163948876158,
+      "grad_norm": 0.3008656237866022,
+      "learning_rate": 6.45800880298397e-05,
+      "loss": 1.6704,
+      "step": 4450
+    },
+    {
+      "epoch": 1.9634200088144556,
+      "grad_norm": 0.31248757394845805,
+      "learning_rate": 6.434041340946909e-05,
+      "loss": 1.6695,
+      "step": 4455
+    },
+    {
+      "epoch": 1.9656236227412958,
+      "grad_norm": 0.3086687965739077,
+      "learning_rate": 6.41009732453597e-05,
+      "loss": 1.5949,
+      "step": 4460
+    },
+    {
+      "epoch": 1.9678272366681357,
+      "grad_norm": 0.2536971137620243,
+      "learning_rate": 6.386176911179353e-05,
+      "loss": 1.6463,
+      "step": 4465
+    },
+    {
+      "epoch": 1.9700308505949757,
+      "grad_norm": 0.3287438241265034,
+      "learning_rate": 6.362280258150074e-05,
+      "loss": 1.6429,
+      "step": 4470
+    },
+    {
+      "epoch": 1.972234464521816,
+      "grad_norm": 0.2565655853570647,
+      "learning_rate": 6.33840752256492e-05,
+      "loss": 1.5546,
+      "step": 4475
+    },
+    {
+      "epoch": 1.9744380784486557,
+      "grad_norm": 0.3055708289742853,
+      "learning_rate": 6.314558861383442e-05,
+      "loss": 1.7164,
+      "step": 4480
+    },
+    {
+      "epoch": 1.976641692375496,
+      "grad_norm": 0.26286511562137227,
+      "learning_rate": 6.29073443140689e-05,
+      "loss": 1.7841,
+      "step": 4485
+    },
+    {
+      "epoch": 1.9788453063023357,
+      "grad_norm": 0.2842174046049182,
+      "learning_rate": 6.266934389277204e-05,
+      "loss": 1.7053,
+      "step": 4490
+    },
+    {
+      "epoch": 1.981048920229176,
+      "grad_norm": 0.34820565260475556,
+      "learning_rate": 6.24315889147597e-05,
+      "loss": 1.7649,
+      "step": 4495
+    },
+    {
+      "epoch": 1.9832525341560159,
+      "grad_norm": 0.3290871426082691,
+      "learning_rate": 6.219408094323415e-05,
+      "loss": 1.6402,
+      "step": 4500
+    },
+    {
+      "epoch": 1.9854561480828559,
+      "grad_norm": 0.25604694542787715,
+      "learning_rate": 6.195682153977351e-05,
+      "loss": 1.5192,
+      "step": 4505
+    },
+    {
+      "epoch": 1.9876597620096959,
+      "grad_norm": 0.28478382829773047,
+      "learning_rate": 6.17198122643216e-05,
+      "loss": 1.649,
+      "step": 4510
+    },
+    {
+      "epoch": 1.9898633759365358,
+      "grad_norm": 0.2776833563766362,
+      "learning_rate": 6.148305467517768e-05,
+      "loss": 1.7351,
+      "step": 4515
+    },
+    {
+      "epoch": 1.992066989863376,
+      "grad_norm": 0.32063858764742265,
+      "learning_rate": 6.124655032898631e-05,
+      "loss": 1.8315,
+      "step": 4520
+    },
+    {
+      "epoch": 1.9942706037902158,
+      "grad_norm": 0.29572004437320715,
+      "learning_rate": 6.1010300780726925e-05,
+      "loss": 1.7337,
+      "step": 4525
+    },
+    {
+      "epoch": 1.996474217717056,
+      "grad_norm": 0.3096259639483799,
+      "learning_rate": 6.077430758370376e-05,
+      "loss": 1.759,
+      "step": 4530
+    },
+    {
+      "epoch": 1.998677831643896,
+      "grad_norm": 0.31442409226477874,
+      "learning_rate": 6.053857228953546e-05,
+      "loss": 1.7822,
+      "step": 4535
+    },
+    {
+      "epoch": 2.000881445570736,
+      "grad_norm": 0.2377883462279695,
+      "learning_rate": 6.03030964481452e-05,
+      "loss": 1.4966,
+      "step": 4540
+    },
+    {
+      "epoch": 2.003085059497576,
+      "grad_norm": 0.2900063226146335,
+      "learning_rate": 6.0067881607750134e-05,
+      "loss": 1.7189,
+      "step": 4545
+    },
+    {
+      "epoch": 2.005288673424416,
+      "grad_norm": 0.30629874279137115,
+      "learning_rate": 5.983292931485142e-05,
+      "loss": 1.5509,
+      "step": 4550
+    },
+    {
+      "epoch": 2.007492287351256,
+      "grad_norm": 0.2682166094278232,
+      "learning_rate": 5.9598241114223986e-05,
+      "loss": 1.5487,
+      "step": 4555
+    },
+    {
+      "epoch": 2.009695901278096,
+      "grad_norm": 0.3321377020430881,
+      "learning_rate": 5.936381854890646e-05,
+      "loss": 1.8033,
+      "step": 4560
+    },
+    {
+      "epoch": 2.011899515204936,
+      "grad_norm": 0.31810161747738896,
+      "learning_rate": 5.912966316019093e-05,
+      "loss": 1.621,
+      "step": 4565
+    },
+    {
+      "epoch": 2.014103129131776,
+      "grad_norm": 0.24729310525461057,
+      "learning_rate": 5.8895776487612765e-05,
+      "loss": 1.5993,
+      "step": 4570
+    },
+    {
+      "epoch": 2.016306743058616,
+      "grad_norm": 0.30363634336485235,
+      "learning_rate": 5.8662160068940655e-05,
+      "loss": 1.6749,
+      "step": 4575
+    },
+    {
+      "epoch": 2.0185103569854563,
+      "grad_norm": 0.2851113038101266,
+      "learning_rate": 5.84288154401664e-05,
+      "loss": 1.6956,
+      "step": 4580
+    },
+    {
+      "epoch": 2.020713970912296,
+      "grad_norm": 0.345208802987356,
+      "learning_rate": 5.81957441354948e-05,
+      "loss": 1.5851,
+      "step": 4585
+    },
+    {
+      "epoch": 2.0229175848391363,
+      "grad_norm": 0.33771374182942,
+      "learning_rate": 5.796294768733362e-05,
+      "loss": 1.5908,
+      "step": 4590
+    },
+    {
+      "epoch": 2.025121198765976,
+      "grad_norm": 0.29219932563018164,
+      "learning_rate": 5.773042762628342e-05,
+      "loss": 1.677,
+      "step": 4595
+    },
+    {
+      "epoch": 2.0273248126928163,
+      "grad_norm": 0.26977700042523883,
+      "learning_rate": 5.749818548112762e-05,
+      "loss": 1.6073,
+      "step": 4600
+    },
+    {
+      "epoch": 2.029528426619656,
+      "grad_norm": 0.30654875158910516,
+      "learning_rate": 5.726622277882243e-05,
+      "loss": 1.7622,
+      "step": 4605
+    },
+    {
+      "epoch": 2.0317320405464963,
+      "grad_norm": 0.3217700066222481,
+      "learning_rate": 5.703454104448665e-05,
+      "loss": 1.804,
+      "step": 4610
+    },
+    {
+      "epoch": 2.0339356544733365,
+      "grad_norm": 0.30632620282091305,
+      "learning_rate": 5.680314180139178e-05,
+      "loss": 1.7833,
+      "step": 4615
+    },
+    {
+      "epoch": 2.0361392684001762,
+      "grad_norm": 0.3023215763344122,
+      "learning_rate": 5.657202657095206e-05,
+      "loss": 1.7969,
+      "step": 4620
+    },
+    {
+      "epoch": 2.0383428823270164,
+      "grad_norm": 0.3078282400161021,
+      "learning_rate": 5.6341196872714394e-05,
+      "loss": 1.6958,
+      "step": 4625
+    },
+    {
+      "epoch": 2.040546496253856,
+      "grad_norm": 0.27870283271610047,
+      "learning_rate": 5.611065422434828e-05,
+      "loss": 1.5725,
+      "step": 4630
+    },
+    {
+      "epoch": 2.0427501101806964,
+      "grad_norm": 0.3114689828172716,
+      "learning_rate": 5.588040014163585e-05,
+      "loss": 1.562,
+      "step": 4635
+    },
+    {
+      "epoch": 2.044953724107536,
+      "grad_norm": 0.2912835457860766,
+      "learning_rate": 5.565043613846219e-05,
+      "loss": 1.7486,
+      "step": 4640
+    },
+    {
+      "epoch": 2.0471573380343764,
+      "grad_norm": 0.29628653391558857,
+      "learning_rate": 5.542076372680498e-05,
+      "loss": 1.7084,
+      "step": 4645
+    },
+    {
+      "epoch": 2.0493609519612166,
+      "grad_norm": 0.29948576765849355,
+      "learning_rate": 5.519138441672471e-05,
+      "loss": 1.6903,
+      "step": 4650
+    },
+    {
+      "epoch": 2.0515645658880564,
+      "grad_norm": 0.26477213023267704,
+      "learning_rate": 5.496229971635487e-05,
+      "loss": 1.6743,
+      "step": 4655
+    },
+    {
+      "epoch": 2.0537681798148966,
+      "grad_norm": 0.36060338821204513,
+      "learning_rate": 5.473351113189194e-05,
+      "loss": 1.8093,
+      "step": 4660
+    },
+    {
+      "epoch": 2.0559717937417363,
+      "grad_norm": 0.31416116481556966,
+      "learning_rate": 5.4505020167585396e-05,
+      "loss": 1.6409,
+      "step": 4665
+    },
+    {
+      "epoch": 2.0581754076685765,
+      "grad_norm": 0.23960293640658495,
+      "learning_rate": 5.4276828325727934e-05,
+      "loss": 1.5688,
+      "step": 4670
+    },
+    {
+      "epoch": 2.0603790215954163,
+      "grad_norm": 0.33607407453371085,
+      "learning_rate": 5.4048937106645613e-05,
+      "loss": 1.5812,
+      "step": 4675
+    },
+    {
+      "epoch": 2.0625826355222565,
+      "grad_norm": 0.3362219928372315,
+      "learning_rate": 5.3821348008687967e-05,
+      "loss": 1.7184,
+      "step": 4680
+    },
+    {
+      "epoch": 2.0647862494490967,
+      "grad_norm": 0.31128427440833256,
+      "learning_rate": 5.3594062528218025e-05,
+      "loss": 1.6606,
+      "step": 4685
+    },
+    {
+      "epoch": 2.0669898633759365,
+      "grad_norm": 0.28720777906212147,
+      "learning_rate": 5.336708215960258e-05,
+      "loss": 1.6505,
+      "step": 4690
+    },
+    {
+      "epoch": 2.0691934773027767,
+      "grad_norm": 0.35035039806258184,
+      "learning_rate": 5.314040839520253e-05,
+      "loss": 1.7716,
+      "step": 4695
+    },
+    {
+      "epoch": 2.0713970912296165,
+      "grad_norm": 0.2931752659797425,
+      "learning_rate": 5.291404272536275e-05,
+      "loss": 1.6877,
+      "step": 4700
+    },
+    {
+      "epoch": 2.0736007051564567,
+      "grad_norm": 0.2873712201718596,
+      "learning_rate": 5.268798663840243e-05,
+      "loss": 1.6062,
+      "step": 4705
+    },
+    {
+      "epoch": 2.0758043190832964,
+      "grad_norm": 0.26676705722923527,
+      "learning_rate": 5.2462241620605366e-05,
+      "loss": 1.6592,
+      "step": 4710
+    },
+    {
+      "epoch": 2.0780079330101366,
+      "grad_norm": 0.3509576273254501,
+      "learning_rate": 5.223680915621014e-05,
+      "loss": 1.671,
+      "step": 4715
+    },
+    {
+      "epoch": 2.080211546936977,
+      "grad_norm": 0.39209953129493824,
+      "learning_rate": 5.2011690727400285e-05,
+      "loss": 1.6385,
+      "step": 4720
+    },
+    {
+      "epoch": 2.0824151608638166,
+      "grad_norm": 0.3402308135526598,
+      "learning_rate": 5.178688781429455e-05,
+      "loss": 1.6095,
+      "step": 4725
+    },
+    {
+      "epoch": 2.084618774790657,
+      "grad_norm": 0.29891446961728113,
+      "learning_rate": 5.1562401894937365e-05,
+      "loss": 1.6653,
+      "step": 4730
+    },
+    {
+      "epoch": 2.0868223887174966,
+      "grad_norm": 0.2466567525935987,
+      "learning_rate": 5.133823444528889e-05,
+      "loss": 1.6558,
+      "step": 4735
+    },
+    {
+      "epoch": 2.089026002644337,
+      "grad_norm": 0.3676958248067488,
+      "learning_rate": 5.111438693921536e-05,
+      "loss": 1.6279,
+      "step": 4740
+    },
+    {
+      "epoch": 2.0912296165711766,
+      "grad_norm": 0.343788030408807,
+      "learning_rate": 5.089086084847954e-05,
+      "loss": 1.6951,
+      "step": 4745
+    },
+    {
+      "epoch": 2.0934332304980168,
+      "grad_norm": 0.3319476261884821,
+      "learning_rate": 5.066765764273078e-05,
+      "loss": 1.617,
+      "step": 4750
+    },
+    {
+      "epoch": 2.095636844424857,
+      "grad_norm": 0.33586220021618074,
+      "learning_rate": 5.044477878949571e-05,
+      "loss": 1.6601,
+      "step": 4755
+    },
+    {
+      "epoch": 2.0978404583516967,
+      "grad_norm": 0.27982003995931076,
+      "learning_rate": 5.0222225754168175e-05,
+      "loss": 1.6063,
+      "step": 4760
+    },
+    {
+      "epoch": 2.100044072278537,
+      "grad_norm": 0.3330827204915635,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 1.7513,
+      "step": 4765
+    },
+    {
+      "epoch": 2.1022476862053767,
+      "grad_norm": 0.34805614239612326,
+      "learning_rate": 4.97781029880911e-05,
+      "loss": 1.5524,
+      "step": 4770
+    },
+    {
+      "epoch": 2.104451300132217,
+      "grad_norm": 0.287355630557104,
+      "learning_rate": 4.955653617737995e-05,
+      "loss": 1.6138,
+      "step": 4775
+    },
+    {
+      "epoch": 2.1066549140590567,
+      "grad_norm": 0.29898227350779466,
+      "learning_rate": 4.9335301024634094e-05,
+      "loss": 1.6648,
+      "step": 4780
+    },
+    {
+      "epoch": 2.108858527985897,
+      "grad_norm": 0.26174467755819714,
+      "learning_rate": 4.911439898444036e-05,
+      "loss": 1.594,
+      "step": 4785
+    },
+    {
+      "epoch": 2.111062141912737,
+      "grad_norm": 0.35692436121383353,
+      "learning_rate": 4.889383150919543e-05,
+      "loss": 1.5403,
+      "step": 4790
+    },
+    {
+      "epoch": 2.113265755839577,
+      "grad_norm": 0.2926942813060104,
+      "learning_rate": 4.867360004909635e-05,
+      "loss": 1.754,
+      "step": 4795
+    },
+    {
+      "epoch": 2.115469369766417,
+      "grad_norm": 0.3290438654965286,
+      "learning_rate": 4.845370605213091e-05,
+      "loss": 1.5578,
+      "step": 4800
+    },
+    {
+      "epoch": 2.117672983693257,
+      "grad_norm": 0.3373040566376742,
+      "learning_rate": 4.823415096406806e-05,
+      "loss": 1.6939,
+      "step": 4805
+    },
+    {
+      "epoch": 2.119876597620097,
+      "grad_norm": 0.34391563105196393,
+      "learning_rate": 4.801493622844847e-05,
+      "loss": 1.7067,
+      "step": 4810
+    },
+    {
+      "epoch": 2.122080211546937,
+      "grad_norm": 0.3002969326243971,
+      "learning_rate": 4.779606328657513e-05,
+      "loss": 1.716,
+      "step": 4815
+    },
+    {
+      "epoch": 2.124283825473777,
+      "grad_norm": 0.3388857319061881,
+      "learning_rate": 4.75775335775038e-05,
+      "loss": 1.843,
+      "step": 4820
+    },
+    {
+      "epoch": 2.1264874394006172,
+      "grad_norm": 0.4156596099940998,
+      "learning_rate": 4.735934853803339e-05,
+      "loss": 1.7106,
+      "step": 4825
+    },
+    {
+      "epoch": 2.128691053327457,
+      "grad_norm": 0.2940085091207592,
+      "learning_rate": 4.71415096026968e-05,
+      "loss": 1.6581,
+      "step": 4830
+    },
+    {
+      "epoch": 2.130894667254297,
+      "grad_norm": 0.306112268047742,
+      "learning_rate": 4.692401820375134e-05,
+      "loss": 1.7315,
+      "step": 4835
+    },
+    {
+      "epoch": 2.133098281181137,
+      "grad_norm": 0.29996735672531655,
+      "learning_rate": 4.6706875771169265e-05,
+      "loss": 1.649,
+      "step": 4840
+    },
+    {
+      "epoch": 2.135301895107977,
+      "grad_norm": 0.27463359410451954,
+      "learning_rate": 4.64900837326284e-05,
+      "loss": 1.6608,
+      "step": 4845
+    },
+    {
+      "epoch": 2.137505509034817,
+      "grad_norm": 0.3165826117114464,
+      "learning_rate": 4.627364351350288e-05,
+      "loss": 1.6793,
+      "step": 4850
+    },
+    {
+      "epoch": 2.139709122961657,
+      "grad_norm": 0.3469777212157657,
+      "learning_rate": 4.605755653685366e-05,
+      "loss": 1.7027,
+      "step": 4855
+    },
+    {
+      "epoch": 2.1419127368884974,
+      "grad_norm": 0.28698621030093346,
+      "learning_rate": 4.584182422341915e-05,
+      "loss": 1.6516,
+      "step": 4860
+    },
+    {
+      "epoch": 2.144116350815337,
+      "grad_norm": 0.3294649962724426,
+      "learning_rate": 4.562644799160585e-05,
+      "loss": 1.7214,
+      "step": 4865
+    },
+    {
+      "epoch": 2.1463199647421773,
+      "grad_norm": 0.3185351974138239,
+      "learning_rate": 4.541142925747919e-05,
+      "loss": 1.6362,
+      "step": 4870
+    },
+    {
+      "epoch": 2.148523578669017,
+      "grad_norm": 0.3126109058989699,
+      "learning_rate": 4.519676943475408e-05,
+      "loss": 1.7064,
+      "step": 4875
+    },
+    {
+      "epoch": 2.1507271925958573,
+      "grad_norm": 0.5380540251828319,
+      "learning_rate": 4.4982469934785574e-05,
+      "loss": 1.6943,
+      "step": 4880
+    },
+    {
+      "epoch": 2.152930806522697,
+      "grad_norm": 0.3435754281589444,
+      "learning_rate": 4.4768532166559763e-05,
+      "loss": 1.715,
+      "step": 4885
+    },
+    {
+      "epoch": 2.1551344204495373,
+      "grad_norm": 0.28522234926734513,
+      "learning_rate": 4.455495753668428e-05,
+      "loss": 1.3429,
+      "step": 4890
+    },
+    {
+      "epoch": 2.157338034376377,
+      "grad_norm": 0.29145572840002815,
+      "learning_rate": 4.4341747449379335e-05,
+      "loss": 1.5995,
+      "step": 4895
+    },
+    {
+      "epoch": 2.1595416483032173,
+      "grad_norm": 0.47296843361442037,
+      "learning_rate": 4.412890330646815e-05,
+      "loss": 1.8911,
+      "step": 4900
+    },
+    {
+      "epoch": 2.1617452622300575,
+      "grad_norm": 0.3101777212408002,
+      "learning_rate": 4.391642650736811e-05,
+      "loss": 1.5388,
+      "step": 4905
+    },
+    {
+      "epoch": 2.1639488761568972,
+      "grad_norm": 0.27664700170021256,
+      "learning_rate": 4.370431844908119e-05,
+      "loss": 1.5866,
+      "step": 4910
+    },
+    {
+      "epoch": 2.1661524900837374,
+      "grad_norm": 0.29560765708014347,
+      "learning_rate": 4.349258052618509e-05,
+      "loss": 1.7198,
+      "step": 4915
+    },
+    {
+      "epoch": 2.168356104010577,
+      "grad_norm": 0.3363838001346494,
+      "learning_rate": 4.328121413082388e-05,
+      "loss": 1.6872,
+      "step": 4920
+    },
+    {
+      "epoch": 2.1705597179374174,
+      "grad_norm": 0.2850203542162651,
+      "learning_rate": 4.307022065269887e-05,
+      "loss": 1.6207,
+      "step": 4925
+    },
+    {
+      "epoch": 2.172763331864257,
+      "grad_norm": 0.32285248455004484,
+      "learning_rate": 4.285960147905946e-05,
+      "loss": 1.6117,
+      "step": 4930
+    },
+    {
+      "epoch": 2.1749669457910974,
+      "grad_norm": 0.27430213811378995,
+      "learning_rate": 4.264935799469417e-05,
+      "loss": 1.6949,
+      "step": 4935
+    },
+    {
+      "epoch": 2.1771705597179376,
+      "grad_norm": 0.46396848778851696,
+      "learning_rate": 4.2439491581921373e-05,
+      "loss": 1.6883,
+      "step": 4940
+    },
+    {
+      "epoch": 2.1793741736447774,
+      "grad_norm": 0.31449151532095687,
+      "learning_rate": 4.223000362058023e-05,
+      "loss": 1.6213,
+      "step": 4945
+    },
+    {
+      "epoch": 2.1815777875716176,
+      "grad_norm": 0.27602615724586665,
+      "learning_rate": 4.202089548802157e-05,
+      "loss": 1.6365,
+      "step": 4950
+    },
+    {
+      "epoch": 2.1837814014984573,
+      "grad_norm": 0.2989766105095882,
+      "learning_rate": 4.181216855909913e-05,
+      "loss": 1.6936,
+      "step": 4955
+    },
+    {
+      "epoch": 2.1859850154252976,
+      "grad_norm": 0.29722828765908,
+      "learning_rate": 4.16038242061601e-05,
+      "loss": 1.5737,
+      "step": 4960
+    },
+    {
+      "epoch": 2.1881886293521373,
+      "grad_norm": 0.2573098385617781,
+      "learning_rate": 4.139586379903629e-05,
+      "loss": 1.6852,
+      "step": 4965
+    },
+    {
+      "epoch": 2.1903922432789775,
+      "grad_norm": 0.2951816124750441,
+      "learning_rate": 4.1188288705035226e-05,
+      "loss": 1.645,
+      "step": 4970
+    },
+    {
+      "epoch": 2.1925958572058177,
+      "grad_norm": 0.33288427202987686,
+      "learning_rate": 4.098110028893105e-05,
+      "loss": 1.5257,
+      "step": 4975
+    },
+    {
+      "epoch": 2.1947994711326575,
+      "grad_norm": 0.3489398250177744,
+      "learning_rate": 4.077429991295549e-05,
+      "loss": 1.6671,
+      "step": 4980
+    },
+    {
+      "epoch": 2.1970030850594977,
+      "grad_norm": 0.2899374350308383,
+      "learning_rate": 4.056788893678898e-05,
+      "loss": 1.6132,
+      "step": 4985
+    },
+    {
+      "epoch": 2.1992066989863375,
+      "grad_norm": 0.3500814252732699,
+      "learning_rate": 4.036186871755173e-05,
+      "loss": 1.5695,
+      "step": 4990
+    },
+    {
+      "epoch": 2.2014103129131777,
+      "grad_norm": 0.3687163067932294,
+      "learning_rate": 4.015624060979486e-05,
+      "loss": 1.5143,
+      "step": 4995
+    },
+    {
+      "epoch": 2.2036139268400174,
+      "grad_norm": 0.33423795880289425,
+      "learning_rate": 3.995100596549128e-05,
+      "loss": 1.6156,
+      "step": 5000
+    },
+    {
+      "epoch": 2.2058175407668577,
+      "grad_norm": 0.2868033623703877,
+      "learning_rate": 3.9746166134026995e-05,
+      "loss": 1.5561,
+      "step": 5005
+    },
+    {
+      "epoch": 2.208021154693698,
+      "grad_norm": 0.34323096599301134,
+      "learning_rate": 3.9541722462192196e-05,
+      "loss": 1.4618,
+      "step": 5010
+    },
+    {
+      "epoch": 2.2102247686205376,
+      "grad_norm": 0.27958346490895863,
+      "learning_rate": 3.9337676294172424e-05,
+      "loss": 1.7082,
+      "step": 5015
+    },
+    {
+      "epoch": 2.212428382547378,
+      "grad_norm": 0.3257321749343037,
+      "learning_rate": 3.913402897153957e-05,
+      "loss": 1.6946,
+      "step": 5020
+    },
+    {
+      "epoch": 2.2146319964742176,
+      "grad_norm": 0.3365177337266649,
+      "learning_rate": 3.893078183324329e-05,
+      "loss": 1.6428,
+      "step": 5025
+    },
+    {
+      "epoch": 2.216835610401058,
+      "grad_norm": 0.34401308894946486,
+      "learning_rate": 3.8727936215602077e-05,
+      "loss": 1.5488,
+      "step": 5030
+    },
+    {
+      "epoch": 2.2190392243278976,
+      "grad_norm": 0.2845220746832339,
+      "learning_rate": 3.852549345229445e-05,
+      "loss": 1.6519,
+      "step": 5035
+    },
+    {
+      "epoch": 2.221242838254738,
+      "grad_norm": 0.25529471083550676,
+      "learning_rate": 3.832345487435019e-05,
+      "loss": 1.8166,
+      "step": 5040
+    },
+    {
+      "epoch": 2.223446452181578,
+      "grad_norm": 0.3591343853259783,
+      "learning_rate": 3.812182181014169e-05,
+      "loss": 1.7223,
+      "step": 5045
+    },
+    {
+      "epoch": 2.2256500661084178,
+      "grad_norm": 0.3054726611280714,
+      "learning_rate": 3.792059558537518e-05,
+      "loss": 1.8144,
+      "step": 5050
+    },
+    {
+      "epoch": 2.227853680035258,
+      "grad_norm": 0.30167111375727146,
+      "learning_rate": 3.7719777523081864e-05,
+      "loss": 1.5961,
+      "step": 5055
+    },
+    {
+      "epoch": 2.2300572939620977,
+      "grad_norm": 0.2916710789608964,
+      "learning_rate": 3.751936894360949e-05,
+      "loss": 1.7809,
+      "step": 5060
+    },
+    {
+      "epoch": 2.232260907888938,
+      "grad_norm": 0.3732355683913339,
+      "learning_rate": 3.731937116461336e-05,
+      "loss": 1.6552,
+      "step": 5065
+    },
+    {
+      "epoch": 2.2344645218157777,
+      "grad_norm": 0.35016354658091353,
+      "learning_rate": 3.7119785501047977e-05,
+      "loss": 1.649,
+      "step": 5070
+    },
+    {
+      "epoch": 2.236668135742618,
+      "grad_norm": 0.34880893384696754,
+      "learning_rate": 3.6920613265158124e-05,
+      "loss": 1.5914,
+      "step": 5075
+    },
+    {
+      "epoch": 2.2388717496694577,
+      "grad_norm": 0.3026190037167945,
+      "learning_rate": 3.672185576647047e-05,
+      "loss": 1.5736,
+      "step": 5080
+    },
+    {
+      "epoch": 2.241075363596298,
+      "grad_norm": 0.3484793101368692,
+      "learning_rate": 3.652351431178473e-05,
+      "loss": 1.7296,
+      "step": 5085
+    },
+    {
+      "epoch": 2.243278977523138,
+      "grad_norm": 0.2587698411979413,
+      "learning_rate": 3.6325590205165314e-05,
+      "loss": 1.7112,
+      "step": 5090
+    },
+    {
+      "epoch": 2.245482591449978,
+      "grad_norm": 0.25882732375987005,
+      "learning_rate": 3.612808474793261e-05,
+      "loss": 1.7223,
+      "step": 5095
+    },
+    {
+      "epoch": 2.247686205376818,
+      "grad_norm": 0.27481092620324143,
+      "learning_rate": 3.593099923865438e-05,
+      "loss": 1.5473,
+      "step": 5100
+    },
+    {
+      "epoch": 2.249889819303658,
+      "grad_norm": 0.3411799767876264,
+      "learning_rate": 3.573433497313731e-05,
+      "loss": 1.6459,
+      "step": 5105
+    },
+    {
+      "epoch": 2.252093433230498,
+      "grad_norm": 0.2887170405054378,
+      "learning_rate": 3.5538093244418525e-05,
+      "loss": 1.6195,
+      "step": 5110
+    },
+    {
+      "epoch": 2.254297047157338,
+      "grad_norm": 0.3884613083691619,
+      "learning_rate": 3.5342275342757046e-05,
+      "loss": 1.8638,
+      "step": 5115
+    },
+    {
+      "epoch": 2.256500661084178,
+      "grad_norm": 0.28793406477711025,
+      "learning_rate": 3.5146882555625226e-05,
+      "loss": 1.5124,
+      "step": 5120
+    },
+    {
+      "epoch": 2.2587042750110182,
+      "grad_norm": 0.30122690590343093,
+      "learning_rate": 3.495191616770034e-05,
+      "loss": 1.7147,
+      "step": 5125
+    },
+    {
+      "epoch": 2.260907888937858,
+      "grad_norm": 0.27840602962105204,
+      "learning_rate": 3.475737746085631e-05,
+      "loss": 1.5467,
+      "step": 5130
+    },
+    {
+      "epoch": 2.263111502864698,
+      "grad_norm": 0.3165143945785378,
+      "learning_rate": 3.456326771415498e-05,
+      "loss": 1.6215,
+      "step": 5135
+    },
+    {
+      "epoch": 2.265315116791538,
+      "grad_norm": 0.315000953366212,
+      "learning_rate": 3.436958820383783e-05,
+      "loss": 1.5548,
+      "step": 5140
+    },
+    {
+      "epoch": 2.267518730718378,
+      "grad_norm": 0.29091408470720853,
+      "learning_rate": 3.417634020331769e-05,
+      "loss": 1.786,
+      "step": 5145
+    },
+    {
+      "epoch": 2.269722344645218,
+      "grad_norm": 0.335006684656636,
+      "learning_rate": 3.398352498317029e-05,
+      "loss": 1.6015,
+      "step": 5150
+    },
+    {
+      "epoch": 2.271925958572058,
+      "grad_norm": 0.3244552469471718,
+      "learning_rate": 3.379114381112581e-05,
+      "loss": 1.653,
+      "step": 5155
+    },
+    {
+      "epoch": 2.2741295724988984,
+      "grad_norm": 0.32402505918566016,
+      "learning_rate": 3.359919795206065e-05,
+      "loss": 1.5578,
+      "step": 5160
+    },
+    {
+      "epoch": 2.276333186425738,
+      "grad_norm": 0.3417399295841799,
+      "learning_rate": 3.3407688667989124e-05,
+      "loss": 1.8143,
+      "step": 5165
+    },
+    {
+      "epoch": 2.2785368003525783,
+      "grad_norm": 0.3293628623372523,
+      "learning_rate": 3.321661721805519e-05,
+      "loss": 1.62,
+      "step": 5170
+    },
+    {
+      "epoch": 2.280740414279418,
+      "grad_norm": 0.3594982936202251,
+      "learning_rate": 3.302598485852401e-05,
+      "loss": 1.5937,
+      "step": 5175
+    },
+    {
+      "epoch": 2.2829440282062583,
+      "grad_norm": 0.2986899594367847,
+      "learning_rate": 3.283579284277378e-05,
+      "loss": 1.5761,
+      "step": 5180
+    },
+    {
+      "epoch": 2.285147642133098,
+      "grad_norm": 0.2633442789610725,
+      "learning_rate": 3.2646042421287625e-05,
+      "loss": 1.7272,
+      "step": 5185
+    },
+    {
+      "epoch": 2.2873512560599383,
+      "grad_norm": 0.3664081223341677,
+      "learning_rate": 3.245673484164521e-05,
+      "loss": 1.6607,
+      "step": 5190
+    },
+    {
+      "epoch": 2.2895548699867785,
+      "grad_norm": 0.36095917628329555,
+      "learning_rate": 3.2267871348514475e-05,
+      "loss": 1.6644,
+      "step": 5195
+    },
+    {
+      "epoch": 2.2917584839136182,
+      "grad_norm": 0.3155817104820564,
+      "learning_rate": 3.207945318364376e-05,
+      "loss": 1.7833,
+      "step": 5200
+    },
+    {
+      "epoch": 2.2939620978404585,
+      "grad_norm": 0.5525894084347752,
+      "learning_rate": 3.1891481585853224e-05,
+      "loss": 1.7846,
+      "step": 5205
+    },
+    {
+      "epoch": 2.2961657117672982,
+      "grad_norm": 0.3183611553729405,
+      "learning_rate": 3.1703957791027104e-05,
+      "loss": 1.8015,
+      "step": 5210
+    },
+    {
+      "epoch": 2.2983693256941384,
+      "grad_norm": 0.3341156230287187,
+      "learning_rate": 3.151688303210525e-05,
+      "loss": 1.4901,
+      "step": 5215
+    },
+    {
+      "epoch": 2.300572939620978,
+      "grad_norm": 0.29108487309353126,
+      "learning_rate": 3.133025853907531e-05,
+      "loss": 1.6021,
+      "step": 5220
+    },
+    {
+      "epoch": 2.3027765535478184,
+      "grad_norm": 0.31003713411448125,
+      "learning_rate": 3.114408553896437e-05,
+      "loss": 1.6835,
+      "step": 5225
+    },
+    {
+      "epoch": 2.3049801674746586,
+      "grad_norm": 0.31639971696298724,
+      "learning_rate": 3.09583652558311e-05,
+      "loss": 1.7131,
+      "step": 5230
+    },
+    {
+      "epoch": 2.3071837814014984,
+      "grad_norm": 0.320459066996447,
+      "learning_rate": 3.077309891075766e-05,
+      "loss": 1.7207,
+      "step": 5235
+    },
+    {
+      "epoch": 2.3093873953283386,
+      "grad_norm": 0.3658345444037042,
+      "learning_rate": 3.058828772184155e-05,
+      "loss": 1.637,
+      "step": 5240
+    },
+    {
+      "epoch": 2.3115910092551784,
+      "grad_norm": 0.3119356619189155,
+      "learning_rate": 3.0403932904187694e-05,
+      "loss": 1.7374,
+      "step": 5245
+    },
+    {
+      "epoch": 2.3137946231820186,
+      "grad_norm": 0.34062681606017264,
+      "learning_rate": 3.0220035669900493e-05,
+      "loss": 1.3662,
+      "step": 5250
+    },
+    {
+      "epoch": 2.3159982371088583,
+      "grad_norm": 0.3231089104173347,
+      "learning_rate": 3.0036597228075847e-05,
+      "loss": 1.7862,
+      "step": 5255
+    },
+    {
+      "epoch": 2.3182018510356985,
+      "grad_norm": 0.2845067164587982,
+      "learning_rate": 2.985361878479307e-05,
+      "loss": 1.6374,
+      "step": 5260
+    },
+    {
+      "epoch": 2.3204054649625387,
+      "grad_norm": 0.3158916323796938,
+      "learning_rate": 2.9671101543107037e-05,
+      "loss": 1.7791,
+      "step": 5265
+    },
+    {
+      "epoch": 2.3226090788893785,
+      "grad_norm": 0.3473567438345015,
+      "learning_rate": 2.9489046703040478e-05,
+      "loss": 1.6438,
+      "step": 5270
+    },
+    {
+      "epoch": 2.3248126928162187,
+      "grad_norm": 0.32956604729846517,
+      "learning_rate": 2.9307455461575728e-05,
+      "loss": 1.5174,
+      "step": 5275
+    },
+    {
+      "epoch": 2.3270163067430585,
+      "grad_norm": 0.31139915651994976,
+      "learning_rate": 2.9126329012647048e-05,
+      "loss": 1.6661,
+      "step": 5280
+    },
+    {
+      "epoch": 2.3292199206698987,
+      "grad_norm": 0.3916001778917859,
+      "learning_rate": 2.894566854713283e-05,
+      "loss": 1.7324,
+      "step": 5285
+    },
+    {
+      "epoch": 2.3314235345967385,
+      "grad_norm": 0.33538535373990935,
+      "learning_rate": 2.8765475252847696e-05,
+      "loss": 1.8397,
+      "step": 5290
+    },
+    {
+      "epoch": 2.3336271485235787,
+      "grad_norm": 0.4045876334606375,
+      "learning_rate": 2.8585750314534633e-05,
+      "loss": 1.761,
+      "step": 5295
+    },
+    {
+      "epoch": 2.335830762450419,
+      "grad_norm": 0.34476500843118907,
+      "learning_rate": 2.8406494913857264e-05,
+      "loss": 1.7239,
+      "step": 5300
+    },
+    {
+      "epoch": 2.3380343763772586,
+      "grad_norm": 0.3377854881404091,
+      "learning_rate": 2.8227710229392102e-05,
+      "loss": 1.6767,
+      "step": 5305
+    },
+    {
+      "epoch": 2.340237990304099,
+      "grad_norm": 0.2943468051471504,
+      "learning_rate": 2.8049397436620817e-05,
+      "loss": 1.7027,
+      "step": 5310
+    },
+    {
+      "epoch": 2.3424416042309386,
+      "grad_norm": 0.3325646272609782,
+      "learning_rate": 2.7871557707922356e-05,
+      "loss": 1.7092,
+      "step": 5315
+    },
+    {
+      "epoch": 2.344645218157779,
+      "grad_norm": 0.3194408987322575,
+      "learning_rate": 2.769419221256546e-05,
+      "loss": 1.5551,
+      "step": 5320
+    },
+    {
+      "epoch": 2.3468488320846186,
+      "grad_norm": 0.3154791851100986,
+      "learning_rate": 2.751730211670075e-05,
+      "loss": 1.5952,
+      "step": 5325
+    },
+    {
+      "epoch": 2.349052446011459,
+      "grad_norm": 0.37152524303673207,
+      "learning_rate": 2.7340888583353263e-05,
+      "loss": 1.6328,
+      "step": 5330
+    },
+    {
+      "epoch": 2.351256059938299,
+      "grad_norm": 0.364737719498444,
+      "learning_rate": 2.716495277241463e-05,
+      "loss": 1.5125,
+      "step": 5335
+    },
+    {
+      "epoch": 2.3534596738651388,
+      "grad_norm": 0.3524349639748066,
+      "learning_rate": 2.6989495840635615e-05,
+      "loss": 1.6589,
+      "step": 5340
+    },
+    {
+      "epoch": 2.355663287791979,
+      "grad_norm": 0.27640673317944237,
+      "learning_rate": 2.6814518941618326e-05,
+      "loss": 1.5661,
+      "step": 5345
+    },
+    {
+      "epoch": 2.3578669017188187,
+      "grad_norm": 0.31837314971643976,
+      "learning_rate": 2.6640023225808852e-05,
+      "loss": 1.7214,
+      "step": 5350
+    },
+    {
+      "epoch": 2.360070515645659,
+      "grad_norm": 0.29946292415188364,
+      "learning_rate": 2.6466009840489436e-05,
+      "loss": 1.4745,
+      "step": 5355
+    },
+    {
+      "epoch": 2.3622741295724987,
+      "grad_norm": 0.3368911389398718,
+      "learning_rate": 2.629247992977122e-05,
+      "loss": 1.6371,
+      "step": 5360
+    },
+    {
+      "epoch": 2.364477743499339,
+      "grad_norm": 0.3078861715002911,
+      "learning_rate": 2.6119434634586427e-05,
+      "loss": 1.6562,
+      "step": 5365
+    },
+    {
+      "epoch": 2.366681357426179,
+      "grad_norm": 0.39044087680489714,
+      "learning_rate": 2.5946875092681134e-05,
+      "loss": 1.7854,
+      "step": 5370
+    },
+    {
+      "epoch": 2.368884971353019,
+      "grad_norm": 0.3905475645799491,
+      "learning_rate": 2.5774802438607627e-05,
+      "loss": 1.7027,
+      "step": 5375
+    },
+    {
+      "epoch": 2.371088585279859,
+      "grad_norm": 0.3344158542218199,
+      "learning_rate": 2.5603217803716938e-05,
+      "loss": 1.6856,
+      "step": 5380
+    },
+    {
+      "epoch": 2.373292199206699,
+      "grad_norm": 0.3171235587081351,
+      "learning_rate": 2.5432122316151463e-05,
+      "loss": 1.6338,
+      "step": 5385
+    },
+    {
+      "epoch": 2.375495813133539,
+      "grad_norm": 0.31814287079371045,
+      "learning_rate": 2.5261517100837563e-05,
+      "loss": 1.6072,
+      "step": 5390
+    },
+    {
+      "epoch": 2.377699427060379,
+      "grad_norm": 0.3715801248464969,
+      "learning_rate": 2.509140327947814e-05,
+      "loss": 1.7025,
+      "step": 5395
+    },
+    {
+      "epoch": 2.379903040987219,
+      "grad_norm": 0.30689235196019754,
+      "learning_rate": 2.4921781970545178e-05,
+      "loss": 1.704,
+      "step": 5400
+    },
+    {
+      "epoch": 2.3821066549140593,
+      "grad_norm": 0.2783210103898173,
+      "learning_rate": 2.4752654289272568e-05,
+      "loss": 1.8138,
+      "step": 5405
+    },
+    {
+      "epoch": 2.384310268840899,
+      "grad_norm": 0.4115102536349836,
+      "learning_rate": 2.4584021347648645e-05,
+      "loss": 1.7562,
+      "step": 5410
+    },
+    {
+      "epoch": 2.3865138827677392,
+      "grad_norm": 0.320397115810562,
+      "learning_rate": 2.441588425440886e-05,
+      "loss": 1.7002,
+      "step": 5415
+    },
+    {
+      "epoch": 2.388717496694579,
+      "grad_norm": 0.32551710279990625,
+      "learning_rate": 2.424824411502856e-05,
+      "loss": 1.6053,
+      "step": 5420
+    },
+    {
+      "epoch": 2.390921110621419,
+      "grad_norm": 0.37060732842102345,
+      "learning_rate": 2.408110203171572e-05,
+      "loss": 1.6564,
+      "step": 5425
+    },
+    {
+      "epoch": 2.393124724548259,
+      "grad_norm": 0.33585884001747096,
+      "learning_rate": 2.3914459103403696e-05,
+      "loss": 1.7012,
+      "step": 5430
+    },
+    {
+      "epoch": 2.395328338475099,
+      "grad_norm": 0.297871745937653,
+      "learning_rate": 2.374831642574392e-05,
+      "loss": 1.7399,
+      "step": 5435
+    },
+    {
+      "epoch": 2.3975319524019394,
+      "grad_norm": 0.28026347006814506,
+      "learning_rate": 2.3582675091098717e-05,
+      "loss": 1.6698,
+      "step": 5440
+    },
+    {
+      "epoch": 2.399735566328779,
+      "grad_norm": 0.3092926951118008,
+      "learning_rate": 2.3417536188534327e-05,
+      "loss": 1.7019,
+      "step": 5445
+    },
+    {
+      "epoch": 2.4019391802556194,
+      "grad_norm": 0.29956315407231204,
+      "learning_rate": 2.3252900803813415e-05,
+      "loss": 1.7835,
+      "step": 5450
+    },
+    {
+      "epoch": 2.404142794182459,
+      "grad_norm": 0.3472861996721381,
+      "learning_rate": 2.3088770019388116e-05,
+      "loss": 1.7523,
+      "step": 5455
+    },
+    {
+      "epoch": 2.4063464081092993,
+      "grad_norm": 0.3346967409970768,
+      "learning_rate": 2.292514491439297e-05,
+      "loss": 1.543,
+      "step": 5460
+    },
+    {
+      "epoch": 2.408550022036139,
+      "grad_norm": 0.30667047442127315,
+      "learning_rate": 2.2762026564637717e-05,
+      "loss": 1.7131,
+      "step": 5465
+    },
+    {
+      "epoch": 2.4107536359629793,
+      "grad_norm": 0.32172911896887835,
+      "learning_rate": 2.259941604260024e-05,
+      "loss": 1.4888,
+      "step": 5470
+    },
+    {
+      "epoch": 2.4129572498898195,
+      "grad_norm": 0.33661859838509534,
+      "learning_rate": 2.2437314417419518e-05,
+      "loss": 1.6434,
+      "step": 5475
+    },
+    {
+      "epoch": 2.4151608638166593,
+      "grad_norm": 0.2766235524609549,
+      "learning_rate": 2.2275722754888662e-05,
+      "loss": 1.4497,
+      "step": 5480
+    },
+    {
+      "epoch": 2.4173644777434995,
+      "grad_norm": 0.31531488102987404,
+      "learning_rate": 2.211464211744787e-05,
+      "loss": 1.7619,
+      "step": 5485
+    },
+    {
+      "epoch": 2.4195680916703393,
+      "grad_norm": 0.4045821730130733,
+      "learning_rate": 2.195407356417737e-05,
+      "loss": 1.5253,
+      "step": 5490
+    },
+    {
+      "epoch": 2.4217717055971795,
+      "grad_norm": 0.3077739063590653,
+      "learning_rate": 2.1794018150790507e-05,
+      "loss": 1.3737,
+      "step": 5495
+    },
+    {
+      "epoch": 2.4239753195240192,
+      "grad_norm": 0.26933598773812595,
+      "learning_rate": 2.1634476929626868e-05,
+      "loss": 1.5562,
+      "step": 5500
+    },
+    {
+      "epoch": 2.4261789334508594,
+      "grad_norm": 0.3180561976663089,
+      "learning_rate": 2.1475450949645325e-05,
+      "loss": 1.8415,
+      "step": 5505
+    },
+    {
+      "epoch": 2.4283825473776997,
+      "grad_norm": 0.326460873305707,
+      "learning_rate": 2.1316941256417024e-05,
+      "loss": 1.6886,
+      "step": 5510
+    },
+    {
+      "epoch": 2.4305861613045394,
+      "grad_norm": 0.2651305009582311,
+      "learning_rate": 2.115894889211869e-05,
+      "loss": 1.556,
+      "step": 5515
+    },
+    {
+      "epoch": 2.4327897752313796,
+      "grad_norm": 0.285018315967615,
+      "learning_rate": 2.100147489552562e-05,
+      "loss": 1.6264,
+      "step": 5520
+    },
+    {
+      "epoch": 2.4349933891582194,
+      "grad_norm": 0.3039400323881749,
+      "learning_rate": 2.084452030200502e-05,
+      "loss": 1.68,
+      "step": 5525
+    },
+    {
+      "epoch": 2.4371970030850596,
+      "grad_norm": 0.4530331306706194,
+      "learning_rate": 2.068808614350899e-05,
+      "loss": 1.8822,
+      "step": 5530
+    },
+    {
+      "epoch": 2.4394006170118994,
+      "grad_norm": 0.32124417856720616,
+      "learning_rate": 2.0532173448567936e-05,
+      "loss": 1.6896,
+      "step": 5535
+    },
+    {
+      "epoch": 2.4416042309387396,
+      "grad_norm": 0.3526640578616687,
+      "learning_rate": 2.037678324228366e-05,
+      "loss": 1.3874,
+      "step": 5540
+    },
+    {
+      "epoch": 2.44380784486558,
+      "grad_norm": 0.3025841716329574,
+      "learning_rate": 2.022191654632274e-05,
+      "loss": 1.5026,
+      "step": 5545
+    },
+    {
+      "epoch": 2.4460114587924195,
+      "grad_norm": 0.3451825131993311,
+      "learning_rate": 2.0067574378909726e-05,
+      "loss": 1.6466,
+      "step": 5550
+    },
+    {
+      "epoch": 2.4482150727192598,
+      "grad_norm": 0.2946116685019073,
+      "learning_rate": 1.9913757754820483e-05,
+      "loss": 1.7246,
+      "step": 5555
+    },
+    {
+      "epoch": 2.4504186866460995,
+      "grad_norm": 0.3604250683804044,
+      "learning_rate": 1.976046768537544e-05,
+      "loss": 1.5967,
+      "step": 5560
+    },
+    {
+      "epoch": 2.4526223005729397,
+      "grad_norm": 0.356529608888727,
+      "learning_rate": 1.9607705178433124e-05,
+      "loss": 1.8344,
+      "step": 5565
+    },
+    {
+      "epoch": 2.4548259144997795,
+      "grad_norm": 0.3062637993405146,
+      "learning_rate": 1.9455471238383394e-05,
+      "loss": 1.727,
+      "step": 5570
+    },
+    {
+      "epoch": 2.4570295284266197,
+      "grad_norm": 0.2913261589880019,
+      "learning_rate": 1.9303766866140794e-05,
+      "loss": 1.6422,
+      "step": 5575
+    },
+    {
+      "epoch": 2.45923314235346,
+      "grad_norm": 0.28539967100565233,
+      "learning_rate": 1.9152593059138036e-05,
+      "loss": 1.5191,
+      "step": 5580
+    },
+    {
+      "epoch": 2.4614367562802997,
+      "grad_norm": 0.38278133166473177,
+      "learning_rate": 1.9001950811319624e-05,
+      "loss": 1.747,
+      "step": 5585
+    },
+    {
+      "epoch": 2.46364037020714,
+      "grad_norm": 0.3213863995857726,
+      "learning_rate": 1.885184111313494e-05,
+      "loss": 1.6493,
+      "step": 5590
+    },
+    {
+      "epoch": 2.4658439841339796,
+      "grad_norm": 0.3515687604082846,
+      "learning_rate": 1.870226495153199e-05,
+      "loss": 1.6207,
+      "step": 5595
+    },
+    {
+      "epoch": 2.46804759806082,
+      "grad_norm": 0.2818899743965069,
+      "learning_rate": 1.8553223309950907e-05,
+      "loss": 1.5783,
+      "step": 5600
+    },
+    {
+      "epoch": 2.4702512119876596,
+      "grad_norm": 0.32339197111018464,
+      "learning_rate": 1.8404717168317444e-05,
+      "loss": 1.6023,
+      "step": 5605
+    },
+    {
+      "epoch": 2.4724548259145,
+      "grad_norm": 0.32235437648153625,
+      "learning_rate": 1.8256747503036465e-05,
+      "loss": 1.5901,
+      "step": 5610
+    },
+    {
+      "epoch": 2.47465843984134,
+      "grad_norm": 0.2996816683579224,
+      "learning_rate": 1.8109315286985575e-05,
+      "loss": 1.7065,
+      "step": 5615
+    },
+    {
+      "epoch": 2.47686205376818,
+      "grad_norm": 0.34309522162413447,
+      "learning_rate": 1.7962421489508797e-05,
+      "loss": 1.7226,
+      "step": 5620
+    },
+    {
+      "epoch": 2.47906566769502,
+      "grad_norm": 0.29849313259138993,
+      "learning_rate": 1.7816067076410138e-05,
+      "loss": 1.7579,
+      "step": 5625
+    },
+    {
+      "epoch": 2.48126928162186,
+      "grad_norm": 0.3368071136503966,
+      "learning_rate": 1.7670253009947146e-05,
+      "loss": 1.5962,
+      "step": 5630
+    },
+    {
+      "epoch": 2.4834728955487,
+      "grad_norm": 0.3175198598873115,
+      "learning_rate": 1.7524980248824806e-05,
+      "loss": 1.7556,
+      "step": 5635
+    },
+    {
+      "epoch": 2.4856765094755398,
+      "grad_norm": 0.30889458937514946,
+      "learning_rate": 1.738024974818896e-05,
+      "loss": 1.7268,
+      "step": 5640
+    },
+    {
+      "epoch": 2.48788012340238,
+      "grad_norm": 0.30454000273861387,
+      "learning_rate": 1.7236062459620306e-05,
+      "loss": 1.6084,
+      "step": 5645
+    },
+    {
+      "epoch": 2.49008373732922,
+      "grad_norm": 0.30973625136809374,
+      "learning_rate": 1.7092419331127894e-05,
+      "loss": 1.581,
+      "step": 5650
+    },
+    {
+      "epoch": 2.49228735125606,
+      "grad_norm": 0.3312428081002817,
+      "learning_rate": 1.6949321307143096e-05,
+      "loss": 1.6826,
+      "step": 5655
+    },
+    {
+      "epoch": 2.4944909651829,
+      "grad_norm": 0.29438754562772956,
+      "learning_rate": 1.6806769328513226e-05,
+      "loss": 1.6531,
+      "step": 5660
+    },
+    {
+      "epoch": 2.49669457910974,
+      "grad_norm": 0.3072379816792179,
+      "learning_rate": 1.666476433249552e-05,
+      "loss": 1.6036,
+      "step": 5665
+    },
+    {
+      "epoch": 2.49889819303658,
+      "grad_norm": 0.2653188527319854,
+      "learning_rate": 1.6523307252750787e-05,
+      "loss": 1.6677,
+      "step": 5670
+    },
+    {
+      "epoch": 2.50110180696342,
+      "grad_norm": 0.31853001670175285,
+      "learning_rate": 1.6382399019337493e-05,
+      "loss": 1.7512,
+      "step": 5675
+    },
+    {
+      "epoch": 2.50330542089026,
+      "grad_norm": 0.3541014940940505,
+      "learning_rate": 1.6242040558705386e-05,
+      "loss": 1.4784,
+      "step": 5680
+    },
+    {
+      "epoch": 2.5055090348171003,
+      "grad_norm": 0.3022737220008577,
+      "learning_rate": 1.6102232793689652e-05,
+      "loss": 1.552,
+      "step": 5685
+    },
+    {
+      "epoch": 2.50771264874394,
+      "grad_norm": 0.2865580516371412,
+      "learning_rate": 1.5962976643504734e-05,
+      "loss": 1.6162,
+      "step": 5690
+    },
+    {
+      "epoch": 2.50991626267078,
+      "grad_norm": 0.33229646654342737,
+      "learning_rate": 1.5824273023738223e-05,
+      "loss": 1.7025,
+      "step": 5695
+    },
+    {
+      "epoch": 2.51211987659762,
+      "grad_norm": 0.25315774233489335,
+      "learning_rate": 1.5686122846344932e-05,
+      "loss": 1.6556,
+      "step": 5700
+    },
+    {
+      "epoch": 2.5143234905244602,
+      "grad_norm": 0.32926703632524384,
+      "learning_rate": 1.55485270196409e-05,
+      "loss": 1.7055,
+      "step": 5705
+    },
+    {
+      "epoch": 2.5165271044513,
+      "grad_norm": 0.38254621782613324,
+      "learning_rate": 1.541148644829743e-05,
+      "loss": 1.8189,
+      "step": 5710
+    },
+    {
+      "epoch": 2.51873071837814,
+      "grad_norm": 0.3262138673835723,
+      "learning_rate": 1.5275002033335016e-05,
+      "loss": 1.6328,
+      "step": 5715
+    },
+    {
+      "epoch": 2.5209343323049804,
+      "grad_norm": 0.36898284620242594,
+      "learning_rate": 1.5139074672117514e-05,
+      "loss": 1.7229,
+      "step": 5720
+    },
+    {
+      "epoch": 2.52313794623182,
+      "grad_norm": 0.29383774220673775,
+      "learning_rate": 1.500370525834639e-05,
+      "loss": 1.7057,
+      "step": 5725
+    },
+    {
+      "epoch": 2.52534156015866,
+      "grad_norm": 0.35255942435868104,
+      "learning_rate": 1.4868894682054535e-05,
+      "loss": 1.703,
+      "step": 5730
+    },
+    {
+      "epoch": 2.5275451740855,
+      "grad_norm": 0.3224975998851544,
+      "learning_rate": 1.473464382960057e-05,
+      "loss": 1.6255,
+      "step": 5735
+    },
+    {
+      "epoch": 2.5297487880123404,
+      "grad_norm": 0.2728863606551014,
+      "learning_rate": 1.4600953583663114e-05,
+      "loss": 1.5348,
+      "step": 5740
+    },
+    {
+      "epoch": 2.53195240193918,
+      "grad_norm": 0.3292693963632781,
+      "learning_rate": 1.4467824823234843e-05,
+      "loss": 1.6536,
+      "step": 5745
+    },
+    {
+      "epoch": 2.5341560158660204,
+      "grad_norm": 0.3530164649944846,
+      "learning_rate": 1.4335258423616737e-05,
+      "loss": 1.631,
+      "step": 5750
+    },
+    {
+      "epoch": 2.5363596297928606,
+      "grad_norm": 0.3107181811257503,
+      "learning_rate": 1.4203255256412318e-05,
+      "loss": 1.5969,
+      "step": 5755
+    },
+    {
+      "epoch": 2.5385632437197003,
+      "grad_norm": 0.3700803583018722,
+      "learning_rate": 1.407181618952199e-05,
+      "loss": 1.7883,
+      "step": 5760
+    },
+    {
+      "epoch": 2.54076685764654,
+      "grad_norm": 0.30243816752307084,
+      "learning_rate": 1.394094208713732e-05,
+      "loss": 1.652,
+      "step": 5765
+    },
+    {
+      "epoch": 2.5429704715733803,
+      "grad_norm": 0.3413616757398393,
+      "learning_rate": 1.3810633809735196e-05,
+      "loss": 1.7507,
+      "step": 5770
+    },
+    {
+      "epoch": 2.5451740855002205,
+      "grad_norm": 0.3120087119800215,
+      "learning_rate": 1.3680892214072405e-05,
+      "loss": 1.7198,
+      "step": 5775
+    },
+    {
+      "epoch": 2.5473776994270603,
+      "grad_norm": 0.3518504932831635,
+      "learning_rate": 1.3551718153179871e-05,
+      "loss": 1.8579,
+      "step": 5780
+    },
+    {
+      "epoch": 2.5495813133539005,
+      "grad_norm": 0.34752155952619,
+      "learning_rate": 1.3423112476357036e-05,
+      "loss": 1.5468,
+      "step": 5785
+    },
+    {
+      "epoch": 2.5517849272807407,
+      "grad_norm": 0.3260308226099618,
+      "learning_rate": 1.3295076029166265e-05,
+      "loss": 1.5258,
+      "step": 5790
+    },
+    {
+      "epoch": 2.5539885412075805,
+      "grad_norm": 0.35584838362058835,
+      "learning_rate": 1.3167609653427426e-05,
+      "loss": 1.7373,
+      "step": 5795
+    },
+    {
+      "epoch": 2.55619215513442,
+      "grad_norm": 0.32184918024972864,
+      "learning_rate": 1.304071418721221e-05,
+      "loss": 1.6741,
+      "step": 5800
+    },
+    {
+      "epoch": 2.5583957690612604,
+      "grad_norm": 0.315994062331127,
+      "learning_rate": 1.2914390464838655e-05,
+      "loss": 1.7156,
+      "step": 5805
+    },
+    {
+      "epoch": 2.5605993829881006,
+      "grad_norm": 0.23924524979523612,
+      "learning_rate": 1.2788639316865635e-05,
+      "loss": 1.738,
+      "step": 5810
+    },
+    {
+      "epoch": 2.5628029969149404,
+      "grad_norm": 0.3098121698841723,
+      "learning_rate": 1.266346157008753e-05,
+      "loss": 1.5198,
+      "step": 5815
+    },
+    {
+      "epoch": 2.5650066108417806,
+      "grad_norm": 0.3262487762949137,
+      "learning_rate": 1.2538858047528646e-05,
+      "loss": 1.587,
+      "step": 5820
+    },
+    {
+      "epoch": 2.5672102247686204,
+      "grad_norm": 0.33166973277913664,
+      "learning_rate": 1.2414829568437825e-05,
+      "loss": 1.5043,
+      "step": 5825
+    },
+    {
+      "epoch": 2.5694138386954606,
+      "grad_norm": 0.3022890966007592,
+      "learning_rate": 1.2291376948283139e-05,
+      "loss": 1.605,
+      "step": 5830
+    },
+    {
+      "epoch": 2.5716174526223003,
+      "grad_norm": 0.3389630036906691,
+      "learning_rate": 1.2168500998746435e-05,
+      "loss": 1.6955,
+      "step": 5835
+    },
+    {
+      "epoch": 2.5738210665491406,
+      "grad_norm": 0.3796380833169501,
+      "learning_rate": 1.2046202527718076e-05,
+      "loss": 1.6275,
+      "step": 5840
+    },
+    {
+      "epoch": 2.5760246804759808,
+      "grad_norm": 0.3238245323234113,
+      "learning_rate": 1.1924482339291554e-05,
+      "loss": 1.7289,
+      "step": 5845
+    },
+    {
+      "epoch": 2.5782282944028205,
+      "grad_norm": 0.30218042381327415,
+      "learning_rate": 1.1803341233758291e-05,
+      "loss": 1.5412,
+      "step": 5850
+    },
+    {
+      "epoch": 2.5804319083296607,
+      "grad_norm": 0.33206646260037787,
+      "learning_rate": 1.1682780007602268e-05,
+      "loss": 1.8011,
+      "step": 5855
+    },
+    {
+      "epoch": 2.5826355222565005,
+      "grad_norm": 0.3538395424898534,
+      "learning_rate": 1.1562799453494899e-05,
+      "loss": 1.7862,
+      "step": 5860
+    },
+    {
+      "epoch": 2.5848391361833407,
+      "grad_norm": 0.35528233062631853,
+      "learning_rate": 1.144340036028978e-05,
+      "loss": 1.685,
+      "step": 5865
+    },
+    {
+      "epoch": 2.5870427501101805,
+      "grad_norm": 0.33412660972869573,
+      "learning_rate": 1.132458351301744e-05,
+      "loss": 1.7575,
+      "step": 5870
+    },
+    {
+      "epoch": 2.5892463640370207,
+      "grad_norm": 0.3175660326892495,
+      "learning_rate": 1.1206349692880236e-05,
+      "loss": 1.5658,
+      "step": 5875
+    },
+    {
+      "epoch": 2.591449977963861,
+      "grad_norm": 0.30007433696171515,
+      "learning_rate": 1.1088699677247238e-05,
+      "loss": 1.586,
+      "step": 5880
+    },
+    {
+      "epoch": 2.5936535918907007,
+      "grad_norm": 0.3180669505120074,
+      "learning_rate": 1.097163423964912e-05,
+      "loss": 1.7819,
+      "step": 5885
+    },
+    {
+      "epoch": 2.595857205817541,
+      "grad_norm": 0.25188231699393177,
+      "learning_rate": 1.0855154149772994e-05,
+      "loss": 1.512,
+      "step": 5890
+    },
+    {
+      "epoch": 2.5980608197443806,
+      "grad_norm": 0.2725825366458783,
+      "learning_rate": 1.0739260173457355e-05,
+      "loss": 1.7237,
+      "step": 5895
+    },
+    {
+      "epoch": 2.600264433671221,
+      "grad_norm": 0.4353012547028696,
+      "learning_rate": 1.0623953072687265e-05,
+      "loss": 1.5664,
+      "step": 5900
+    },
+    {
+      "epoch": 2.6024680475980606,
+      "grad_norm": 0.3208919789812601,
+      "learning_rate": 1.0509233605588997e-05,
+      "loss": 1.7184,
+      "step": 5905
+    },
+    {
+      "epoch": 2.604671661524901,
+      "grad_norm": 0.2811764278977997,
+      "learning_rate": 1.0395102526425282e-05,
+      "loss": 1.6933,
+      "step": 5910
+    },
+    {
+      "epoch": 2.606875275451741,
+      "grad_norm": 0.3283193021777617,
+      "learning_rate": 1.0281560585590311e-05,
+      "loss": 1.709,
+      "step": 5915
+    },
+    {
+      "epoch": 2.609078889378581,
+      "grad_norm": 0.31601773043700576,
+      "learning_rate": 1.0168608529604783e-05,
+      "loss": 1.5517,
+      "step": 5920
+    },
+    {
+      "epoch": 2.611282503305421,
+      "grad_norm": 0.26388766629051863,
+      "learning_rate": 1.0056247101110972e-05,
+      "loss": 1.5716,
+      "step": 5925
+    },
+    {
+      "epoch": 2.6134861172322608,
+      "grad_norm": 0.3170004934043696,
+      "learning_rate": 9.944477038867838e-06,
+      "loss": 1.3933,
+      "step": 5930
+    },
+    {
+      "epoch": 2.615689731159101,
+      "grad_norm": 0.3026100677449506,
+      "learning_rate": 9.833299077746261e-06,
+      "loss": 1.538,
+      "step": 5935
+    },
+    {
+      "epoch": 2.6178933450859407,
+      "grad_norm": 0.32599901382842245,
+      "learning_rate": 9.72271394872416e-06,
+      "loss": 1.722,
+      "step": 5940
+    },
+    {
+      "epoch": 2.620096959012781,
+      "grad_norm": 0.31252235277224677,
+      "learning_rate": 9.612722378881578e-06,
+      "loss": 1.5777,
+      "step": 5945
+    },
+    {
+      "epoch": 2.622300572939621,
+      "grad_norm": 0.27949230064797415,
+      "learning_rate": 9.503325091396098e-06,
+      "loss": 1.7781,
+      "step": 5950
+    },
+    {
+      "epoch": 2.624504186866461,
+      "grad_norm": 0.3031767326868999,
+      "learning_rate": 9.394522805537931e-06,
+      "loss": 1.6123,
+      "step": 5955
+    },
+    {
+      "epoch": 2.626707800793301,
+      "grad_norm": 0.3596489809565846,
+      "learning_rate": 9.286316236665271e-06,
+      "loss": 1.8234,
+      "step": 5960
+    },
+    {
+      "epoch": 2.628911414720141,
+      "grad_norm": 0.37067976899261396,
+      "learning_rate": 9.178706096219547e-06,
+      "loss": 1.5176,
+      "step": 5965
+    },
+    {
+      "epoch": 2.631115028646981,
+      "grad_norm": 0.30187017500054564,
+      "learning_rate": 9.0716930917208e-06,
+      "loss": 1.5401,
+      "step": 5970
+    },
+    {
+      "epoch": 2.633318642573821,
+      "grad_norm": 0.338495742270805,
+      "learning_rate": 8.965277926762916e-06,
+      "loss": 1.5802,
+      "step": 5975
+    },
+    {
+      "epoch": 2.635522256500661,
+      "grad_norm": 0.33527970219407616,
+      "learning_rate": 8.859461301009186e-06,
+      "loss": 1.6115,
+      "step": 5980
+    },
+    {
+      "epoch": 2.6377258704275013,
+      "grad_norm": 0.3284526374151953,
+      "learning_rate": 8.754243910187498e-06,
+      "loss": 1.7051,
+      "step": 5985
+    },
+    {
+      "epoch": 2.639929484354341,
+      "grad_norm": 0.2820385948307506,
+      "learning_rate": 8.649626446085945e-06,
+      "loss": 1.4949,
+      "step": 5990
+    },
+    {
+      "epoch": 2.6421330982811813,
+      "grad_norm": 0.3357888909663005,
+      "learning_rate": 8.545609596548121e-06,
+      "loss": 1.5265,
+      "step": 5995
+    },
+    {
+      "epoch": 2.644336712208021,
+      "grad_norm": 0.24829105724152342,
+      "learning_rate": 8.442194045468733e-06,
+      "loss": 1.4418,
+      "step": 6000
+    },
+    {
+      "epoch": 2.6465403261348612,
+      "grad_norm": 0.3445954692738449,
+      "learning_rate": 8.339380472789016e-06,
+      "loss": 1.471,
+      "step": 6005
+    },
+    {
+      "epoch": 2.648743940061701,
+      "grad_norm": 0.34193200011238856,
+      "learning_rate": 8.237169554492297e-06,
+      "loss": 1.4714,
+      "step": 6010
+    },
+    {
+      "epoch": 2.650947553988541,
+      "grad_norm": 0.2994838498414857,
+      "learning_rate": 8.135561962599514e-06,
+      "loss": 1.5747,
+      "step": 6015
+    },
+    {
+      "epoch": 2.6531511679153814,
+      "grad_norm": 0.2510845629901009,
+      "learning_rate": 8.034558365164868e-06,
+      "loss": 1.6476,
+      "step": 6020
+    },
+    {
+      "epoch": 2.655354781842221,
+      "grad_norm": 0.3375793364015892,
+      "learning_rate": 7.934159426271403e-06,
+      "loss": 1.6063,
+      "step": 6025
+    },
+    {
+      "epoch": 2.6575583957690614,
+      "grad_norm": 0.33573003764748477,
+      "learning_rate": 7.834365806026578e-06,
+      "loss": 1.5814,
+      "step": 6030
+    },
+    {
+      "epoch": 2.659762009695901,
+      "grad_norm": 0.38295818181236296,
+      "learning_rate": 7.735178160557943e-06,
+      "loss": 1.7642,
+      "step": 6035
+    },
+    {
+      "epoch": 2.6619656236227414,
+      "grad_norm": 0.28003247319118707,
+      "learning_rate": 7.636597142009017e-06,
+      "loss": 1.7946,
+      "step": 6040
+    },
+    {
+      "epoch": 2.664169237549581,
+      "grad_norm": 0.367119348898622,
+      "learning_rate": 7.538623398534661e-06,
+      "loss": 1.5553,
+      "step": 6045
+    },
+    {
+      "epoch": 2.6663728514764213,
+      "grad_norm": 0.30087617582351794,
+      "learning_rate": 7.441257574297089e-06,
+      "loss": 1.56,
+      "step": 6050
+    },
+    {
+      "epoch": 2.6685764654032615,
+      "grad_norm": 0.2679029740855203,
+      "learning_rate": 7.344500309461511e-06,
+      "loss": 1.5622,
+      "step": 6055
+    },
+    {
+      "epoch": 2.6707800793301013,
+      "grad_norm": 0.38643212881302264,
+      "learning_rate": 7.248352240192002e-06,
+      "loss": 1.6344,
+      "step": 6060
+    },
+    {
+      "epoch": 2.6729836932569415,
+      "grad_norm": 0.3036596598041859,
+      "learning_rate": 7.15281399864719e-06,
+      "loss": 1.6126,
+      "step": 6065
+    },
+    {
+      "epoch": 2.6751873071837813,
+      "grad_norm": 0.3409000984647797,
+      "learning_rate": 7.057886212976239e-06,
+      "loss": 1.7453,
+      "step": 6070
+    },
+    {
+      "epoch": 2.6773909211106215,
+      "grad_norm": 0.3123434626912612,
+      "learning_rate": 6.963569507314627e-06,
+      "loss": 1.6624,
+      "step": 6075
+    },
+    {
+      "epoch": 2.6795945350374613,
+      "grad_norm": 0.3272418793360404,
+      "learning_rate": 6.8698645017801325e-06,
+      "loss": 1.8614,
+      "step": 6080
+    },
+    {
+      "epoch": 2.6817981489643015,
+      "grad_norm": 0.2432770598979605,
+      "learning_rate": 6.776771812468618e-06,
+      "loss": 1.6761,
+      "step": 6085
+    },
+    {
+      "epoch": 2.6840017628911417,
+      "grad_norm": 0.29638907037530365,
+      "learning_rate": 6.684292051450147e-06,
+      "loss": 1.5734,
+      "step": 6090
+    },
+    {
+      "epoch": 2.6862053768179814,
+      "grad_norm": 0.2927874947198909,
+      "learning_rate": 6.592425826764781e-06,
+      "loss": 1.6527,
+      "step": 6095
+    },
+    {
+      "epoch": 2.6884089907448216,
+      "grad_norm": 0.3262506822619773,
+      "learning_rate": 6.501173742418753e-06,
+      "loss": 1.7488,
+      "step": 6100
+    },
+    {
+      "epoch": 2.6906126046716614,
+      "grad_norm": 0.29388683856815984,
+      "learning_rate": 6.410536398380385e-06,
+      "loss": 1.7391,
+      "step": 6105
+    },
+    {
+      "epoch": 2.6928162185985016,
+      "grad_norm": 0.42564932659573224,
+      "learning_rate": 6.320514390576193e-06,
+      "loss": 1.5618,
+      "step": 6110
+    },
+    {
+      "epoch": 2.6950198325253414,
+      "grad_norm": 0.33120895230745345,
+      "learning_rate": 6.231108310886924e-06,
+      "loss": 1.5172,
+      "step": 6115
+    },
+    {
+      "epoch": 2.6972234464521816,
+      "grad_norm": 0.2970159546645706,
+      "learning_rate": 6.142318747143716e-06,
+      "loss": 1.5319,
+      "step": 6120
+    },
+    {
+      "epoch": 2.699427060379022,
+      "grad_norm": 0.3125793002150719,
+      "learning_rate": 6.054146283124218e-06,
+      "loss": 1.6401,
+      "step": 6125
+    },
+    {
+      "epoch": 2.7016306743058616,
+      "grad_norm": 0.30146872233206856,
+      "learning_rate": 5.966591498548724e-06,
+      "loss": 1.7384,
+      "step": 6130
+    },
+    {
+      "epoch": 2.7038342882327018,
+      "grad_norm": 0.3887993733574576,
+      "learning_rate": 5.8796549690763645e-06,
+      "loss": 1.8019,
+      "step": 6135
+    },
+    {
+      "epoch": 2.7060379021595415,
+      "grad_norm": 0.3033022818548619,
+      "learning_rate": 5.79333726630138e-06,
+      "loss": 1.5844,
+      "step": 6140
+    },
+    {
+      "epoch": 2.7082415160863818,
+      "grad_norm": 0.368025241100897,
+      "learning_rate": 5.7076389577493175e-06,
+      "loss": 1.8454,
+      "step": 6145
+    },
+    {
+      "epoch": 2.7104451300132215,
+      "grad_norm": 0.298047160015892,
+      "learning_rate": 5.622560606873262e-06,
+      "loss": 1.6045,
+      "step": 6150
+    },
+    {
+      "epoch": 2.7126487439400617,
+      "grad_norm": 0.30605420571316905,
+      "learning_rate": 5.538102773050235e-06,
+      "loss": 1.696,
+      "step": 6155
+    },
+    {
+      "epoch": 2.714852357866902,
+      "grad_norm": 0.3454395184873584,
+      "learning_rate": 5.454266011577369e-06,
+      "loss": 1.6258,
+      "step": 6160
+    },
+    {
+      "epoch": 2.7170559717937417,
+      "grad_norm": 0.30863792610727064,
+      "learning_rate": 5.371050873668437e-06,
+      "loss": 1.5895,
+      "step": 6165
+    },
+    {
+      "epoch": 2.719259585720582,
+      "grad_norm": 0.32671612304128317,
+      "learning_rate": 5.2884579064500615e-06,
+      "loss": 1.751,
+      "step": 6170
+    },
+    {
+      "epoch": 2.7214631996474217,
+      "grad_norm": 0.24404249683796003,
+      "learning_rate": 5.206487652958214e-06,
+      "loss": 1.5318,
+      "step": 6175
+    },
+    {
+      "epoch": 2.723666813574262,
+      "grad_norm": 0.31345216172176077,
+      "learning_rate": 5.125140652134652e-06,
+      "loss": 1.6814,
+      "step": 6180
+    },
+    {
+      "epoch": 2.7258704275011016,
+      "grad_norm": 0.3301202829233535,
+      "learning_rate": 5.044417438823279e-06,
+      "loss": 1.6688,
+      "step": 6185
+    },
+    {
+      "epoch": 2.728074041427942,
+      "grad_norm": 0.3180153311679256,
+      "learning_rate": 4.964318543766733e-06,
+      "loss": 1.8152,
+      "step": 6190
+    },
+    {
+      "epoch": 2.730277655354782,
+      "grad_norm": 0.3477053424267945,
+      "learning_rate": 4.884844493602847e-06,
+      "loss": 1.6068,
+      "step": 6195
+    },
+    {
+      "epoch": 2.732481269281622,
+      "grad_norm": 0.3829450747175287,
+      "learning_rate": 4.805995810861219e-06,
+      "loss": 1.5436,
+      "step": 6200
+    },
+    {
+      "epoch": 2.734684883208462,
+      "grad_norm": 0.35144134365971347,
+      "learning_rate": 4.727773013959702e-06,
+      "loss": 1.7733,
+      "step": 6205
+    },
+    {
+      "epoch": 2.736888497135302,
+      "grad_norm": 0.3511048030633929,
+      "learning_rate": 4.650176617201074e-06,
+      "loss": 1.7483,
+      "step": 6210
+    },
+    {
+      "epoch": 2.739092111062142,
+      "grad_norm": 0.3146009823620477,
+      "learning_rate": 4.573207130769663e-06,
+      "loss": 1.6416,
+      "step": 6215
+    },
+    {
+      "epoch": 2.7412957249889818,
+      "grad_norm": 0.2961123189548949,
+      "learning_rate": 4.496865060727917e-06,
+      "loss": 1.5871,
+      "step": 6220
+    },
+    {
+      "epoch": 2.743499338915822,
+      "grad_norm": 0.2827248539547275,
+      "learning_rate": 4.421150909013094e-06,
+      "loss": 1.6537,
+      "step": 6225
+    },
+    {
+      "epoch": 2.745702952842662,
+      "grad_norm": 0.3230949462447901,
+      "learning_rate": 4.346065173434055e-06,
+      "loss": 1.5128,
+      "step": 6230
+    },
+    {
+      "epoch": 2.747906566769502,
+      "grad_norm": 0.27203611650594783,
+      "learning_rate": 4.271608347667888e-06,
+      "loss": 1.6916,
+      "step": 6235
+    },
+    {
+      "epoch": 2.750110180696342,
+      "grad_norm": 0.316718312569841,
+      "learning_rate": 4.197780921256678e-06,
+      "loss": 1.7967,
+      "step": 6240
+    },
+    {
+      "epoch": 2.752313794623182,
+      "grad_norm": 0.31564921665656825,
+      "learning_rate": 4.1245833796043184e-06,
+      "loss": 1.5092,
+      "step": 6245
+    },
+    {
+      "epoch": 2.754517408550022,
+      "grad_norm": 0.3023941557144956,
+      "learning_rate": 4.052016203973319e-06,
+      "loss": 1.6864,
+      "step": 6250
+    },
+    {
+      "epoch": 2.756721022476862,
+      "grad_norm": 0.31013607515444674,
+      "learning_rate": 3.9800798714816566e-06,
+      "loss": 1.7096,
+      "step": 6255
+    },
+    {
+      "epoch": 2.758924636403702,
+      "grad_norm": 0.34766875505420175,
+      "learning_rate": 3.908774855099529e-06,
+      "loss": 1.6837,
+      "step": 6260
+    },
+    {
+      "epoch": 2.7611282503305423,
+      "grad_norm": 0.2943152497920515,
+      "learning_rate": 3.838101623646429e-06,
+      "loss": 1.6478,
+      "step": 6265
+    },
+    {
+      "epoch": 2.763331864257382,
+      "grad_norm": 0.32690567367741863,
+      "learning_rate": 3.768060641787874e-06,
+      "loss": 1.8321,
+      "step": 6270
+    },
+    {
+      "epoch": 2.765535478184222,
+      "grad_norm": 0.35174727110880194,
+      "learning_rate": 3.698652370032496e-06,
+      "loss": 1.7583,
+      "step": 6275
+    },
+    {
+      "epoch": 2.767739092111062,
+      "grad_norm": 0.30775438684619605,
+      "learning_rate": 3.6298772647289204e-06,
+      "loss": 1.7887,
+      "step": 6280
+    },
+    {
+      "epoch": 2.7699427060379023,
+      "grad_norm": 0.2974347187890948,
+      "learning_rate": 3.561735778062847e-06,
+      "loss": 1.5669,
+      "step": 6285
+    },
+    {
+      "epoch": 2.772146319964742,
+      "grad_norm": 0.32631762135645986,
+      "learning_rate": 3.4942283580539747e-06,
+      "loss": 1.5496,
+      "step": 6290
+    },
+    {
+      "epoch": 2.7743499338915822,
+      "grad_norm": 0.3078445012000501,
+      "learning_rate": 3.427355448553149e-06,
+      "loss": 1.4473,
+      "step": 6295
+    },
+    {
+      "epoch": 2.7765535478184225,
+      "grad_norm": 0.31874633296757016,
+      "learning_rate": 3.3611174892393848e-06,
+      "loss": 1.7297,
+      "step": 6300
+    },
+    {
+      "epoch": 2.778757161745262,
+      "grad_norm": 0.3877251343286691,
+      "learning_rate": 3.2955149156170373e-06,
+      "loss": 1.7889,
+      "step": 6305
+    },
+    {
+      "epoch": 2.780960775672102,
+      "grad_norm": 0.3050428005282347,
+      "learning_rate": 3.230548159012836e-06,
+      "loss": 1.7297,
+      "step": 6310
+    },
+    {
+      "epoch": 2.783164389598942,
+      "grad_norm": 0.3297939270880444,
+      "learning_rate": 3.1662176465731776e-06,
+      "loss": 1.7542,
+      "step": 6315
+    },
+    {
+      "epoch": 2.7853680035257824,
+      "grad_norm": 0.35972869738440505,
+      "learning_rate": 3.1025238012612146e-06,
+      "loss": 1.6169,
+      "step": 6320
+    },
+    {
+      "epoch": 2.787571617452622,
+      "grad_norm": 0.3384226865100565,
+      "learning_rate": 3.039467041854105e-06,
+      "loss": 1.6362,
+      "step": 6325
+    },
+    {
+      "epoch": 2.7897752313794624,
+      "grad_norm": 0.3670760451609635,
+      "learning_rate": 2.97704778294029e-06,
+      "loss": 1.6398,
+      "step": 6330
+    },
+    {
+      "epoch": 2.7919788453063026,
+      "grad_norm": 0.2854714227749342,
+      "learning_rate": 2.9152664349167415e-06,
+      "loss": 1.4325,
+      "step": 6335
+    },
+    {
+      "epoch": 2.7941824592331423,
+      "grad_norm": 0.3042468904319354,
+      "learning_rate": 2.854123403986253e-06,
+      "loss": 1.6423,
+      "step": 6340
+    },
+    {
+      "epoch": 2.796386073159982,
+      "grad_norm": 0.33954395775276786,
+      "learning_rate": 2.793619092154787e-06,
+      "loss": 1.6785,
+      "step": 6345
+    },
+    {
+      "epoch": 2.7985896870868223,
+      "grad_norm": 0.41052304755954266,
+      "learning_rate": 2.7337538972287967e-06,
+      "loss": 1.7808,
+      "step": 6350
+    },
+    {
+      "epoch": 2.8007933010136625,
+      "grad_norm": 0.2735021459128279,
+      "learning_rate": 2.674528212812721e-06,
+      "loss": 1.579,
+      "step": 6355
+    },
+    {
+      "epoch": 2.8029969149405023,
+      "grad_norm": 0.42116340427498267,
+      "learning_rate": 2.6159424283062507e-06,
+      "loss": 1.665,
+      "step": 6360
+    },
+    {
+      "epoch": 2.8052005288673425,
+      "grad_norm": 0.40395544915333126,
+      "learning_rate": 2.557996928901829e-06,
+      "loss": 1.6685,
+      "step": 6365
+    },
+    {
+      "epoch": 2.8074041427941827,
+      "grad_norm": 0.41455342226161546,
+      "learning_rate": 2.5006920955821465e-06,
+      "loss": 1.7578,
+      "step": 6370
+    },
+    {
+      "epoch": 2.8096077567210225,
+      "grad_norm": 0.32321328330549354,
+      "learning_rate": 2.4440283051176405e-06,
+      "loss": 1.7026,
+      "step": 6375
+    },
+    {
+      "epoch": 2.8118113706478622,
+      "grad_norm": 0.2956835138059502,
+      "learning_rate": 2.388005930063941e-06,
+      "loss": 1.8632,
+      "step": 6380
+    },
+    {
+      "epoch": 2.8140149845747024,
+      "grad_norm": 0.3268794222372435,
+      "learning_rate": 2.3326253387594753e-06,
+      "loss": 1.6233,
+      "step": 6385
+    },
+    {
+      "epoch": 2.8162185985015427,
+      "grad_norm": 0.3626797930342101,
+      "learning_rate": 2.277886895323078e-06,
+      "loss": 1.74,
+      "step": 6390
+    },
+    {
+      "epoch": 2.8184222124283824,
+      "grad_norm": 0.33358450270104334,
+      "learning_rate": 2.2237909596515396e-06,
+      "loss": 1.4655,
+      "step": 6395
+    },
+    {
+      "epoch": 2.8206258263552226,
+      "grad_norm": 0.29838126441173135,
+      "learning_rate": 2.1703378874172507e-06,
+      "loss": 1.4969,
+      "step": 6400
+    },
+    {
+      "epoch": 2.822829440282063,
+      "grad_norm": 0.34601838203569607,
+      "learning_rate": 2.117528030065907e-06,
+      "loss": 1.6886,
+      "step": 6405
+    },
+    {
+      "epoch": 2.8250330542089026,
+      "grad_norm": 0.3110562652655748,
+      "learning_rate": 2.0653617348141084e-06,
+      "loss": 1.4905,
+      "step": 6410
+    },
+    {
+      "epoch": 2.8272366681357424,
+      "grad_norm": 0.3274079081069346,
+      "learning_rate": 2.013839344647217e-06,
+      "loss": 1.6808,
+      "step": 6415
+    },
+    {
+      "epoch": 2.8294402820625826,
+      "grad_norm": 0.34133335820544286,
+      "learning_rate": 1.962961198316937e-06,
+      "loss": 1.7414,
+      "step": 6420
+    },
+    {
+      "epoch": 2.831643895989423,
+      "grad_norm": 0.3234135423197038,
+      "learning_rate": 1.912727630339217e-06,
+      "loss": 1.4927,
+      "step": 6425
+    },
+    {
+      "epoch": 2.8338475099162626,
+      "grad_norm": 0.31099973650003665,
+      "learning_rate": 1.8631389709919843e-06,
+      "loss": 1.5605,
+      "step": 6430
+    },
+    {
+      "epoch": 2.8360511238431028,
+      "grad_norm": 0.37303627703284004,
+      "learning_rate": 1.8141955463129912e-06,
+      "loss": 1.6712,
+      "step": 6435
+    },
+    {
+      "epoch": 2.838254737769943,
+      "grad_norm": 0.3614598619091711,
+      "learning_rate": 1.7658976780976944e-06,
+      "loss": 1.7914,
+      "step": 6440
+    },
+    {
+      "epoch": 2.8404583516967827,
+      "grad_norm": 0.30969623455642703,
+      "learning_rate": 1.7182456838971016e-06,
+      "loss": 1.5793,
+      "step": 6445
+    },
+    {
+      "epoch": 2.8426619656236225,
+      "grad_norm": 0.26680083024728835,
+      "learning_rate": 1.6712398770156734e-06,
+      "loss": 1.5423,
+      "step": 6450
+    },
+    {
+      "epoch": 2.8448655795504627,
+      "grad_norm": 0.38620184585663864,
+      "learning_rate": 1.6248805665093348e-06,
+      "loss": 1.7361,
+      "step": 6455
+    },
+    {
+      "epoch": 2.847069193477303,
+      "grad_norm": 0.29245866381267194,
+      "learning_rate": 1.5791680571833667e-06,
+      "loss": 1.4591,
+      "step": 6460
+    },
+    {
+      "epoch": 2.8492728074041427,
+      "grad_norm": 0.36925991583350115,
+      "learning_rate": 1.5341026495904409e-06,
+      "loss": 1.5466,
+      "step": 6465
+    },
+    {
+      "epoch": 2.851476421330983,
+      "grad_norm": 0.31621129108601936,
+      "learning_rate": 1.4896846400286323e-06,
+      "loss": 1.5198,
+      "step": 6470
+    },
+    {
+      "epoch": 2.853680035257823,
+      "grad_norm": 0.3918256317744239,
+      "learning_rate": 1.4459143205394876e-06,
+      "loss": 1.8413,
+      "step": 6475
+    },
+    {
+      "epoch": 2.855883649184663,
+      "grad_norm": 0.3333872096143664,
+      "learning_rate": 1.4027919789060818e-06,
+      "loss": 1.6091,
+      "step": 6480
+    },
+    {
+      "epoch": 2.8580872631115026,
+      "grad_norm": 0.350383265332482,
+      "learning_rate": 1.36031789865112e-06,
+      "loss": 1.77,
+      "step": 6485
+    },
+    {
+      "epoch": 2.860290877038343,
+      "grad_norm": 0.32052781126282354,
+      "learning_rate": 1.3184923590351062e-06,
+      "loss": 1.6178,
+      "step": 6490
+    },
+    {
+      "epoch": 2.862494490965183,
+      "grad_norm": 0.33800257995746813,
+      "learning_rate": 1.27731563505451e-06,
+      "loss": 1.6759,
+      "step": 6495
+    },
+    {
+      "epoch": 2.864698104892023,
+      "grad_norm": 0.32607540150243636,
+      "learning_rate": 1.236787997439892e-06,
+      "loss": 1.5576,
+      "step": 6500
+    },
+    {
+      "epoch": 2.866901718818863,
+      "grad_norm": 0.32271754224482374,
+      "learning_rate": 1.196909712654204e-06,
+      "loss": 1.5769,
+      "step": 6505
+    },
+    {
+      "epoch": 2.8691053327457032,
+      "grad_norm": 0.32279106125213675,
+      "learning_rate": 1.1576810428910012e-06,
+      "loss": 1.4904,
+      "step": 6510
+    },
+    {
+      "epoch": 2.871308946672543,
+      "grad_norm": 0.33356280248337183,
+      "learning_rate": 1.1191022460727007e-06,
+      "loss": 1.5742,
+      "step": 6515
+    },
+    {
+      "epoch": 2.8735125605993828,
+      "grad_norm": 0.3151589561132343,
+      "learning_rate": 1.0811735758489372e-06,
+      "loss": 1.6439,
+      "step": 6520
+    },
+    {
+      "epoch": 2.875716174526223,
+      "grad_norm": 0.3317166675779296,
+      "learning_rate": 1.04389528159482e-06,
+      "loss": 1.4983,
+      "step": 6525
+    },
+    {
+      "epoch": 2.877919788453063,
+      "grad_norm": 0.3198594908928924,
+      "learning_rate": 1.0072676084093902e-06,
+      "loss": 1.6749,
+      "step": 6530
+    },
+    {
+      "epoch": 2.880123402379903,
+      "grad_norm": 0.2937946173545102,
+      "learning_rate": 9.712907971139218e-07,
+      "loss": 1.7593,
+      "step": 6535
+    },
+    {
+      "epoch": 2.882327016306743,
+      "grad_norm": 0.3493222702605276,
+      "learning_rate": 9.359650842503565e-07,
+      "loss": 1.737,
+      "step": 6540
+    },
+    {
+      "epoch": 2.884530630233583,
+      "grad_norm": 0.3038129813169421,
+      "learning_rate": 9.012907020798156e-07,
+      "loss": 1.6078,
+      "step": 6545
+    },
+    {
+      "epoch": 2.886734244160423,
+      "grad_norm": 0.3031344362125413,
+      "learning_rate": 8.672678785809796e-07,
+      "loss": 1.6788,
+      "step": 6550
+    },
+    {
+      "epoch": 2.888937858087263,
+      "grad_norm": 0.31295231167033,
+      "learning_rate": 8.338968374486555e-07,
+      "loss": 1.734,
+      "step": 6555
+    },
+    {
+      "epoch": 2.891141472014103,
+      "grad_norm": 0.30464503391760195,
+      "learning_rate": 8.011777980922564e-07,
+      "loss": 1.6216,
+      "step": 6560
+    },
+    {
+      "epoch": 2.8933450859409433,
+      "grad_norm": 0.31057203787361704,
+      "learning_rate": 7.691109756344128e-07,
+      "loss": 1.6683,
+      "step": 6565
+    },
+    {
+      "epoch": 2.895548699867783,
+      "grad_norm": 0.3406431085406218,
+      "learning_rate": 7.376965809095193e-07,
+      "loss": 1.7457,
+      "step": 6570
+    },
+    {
+      "epoch": 2.8977523137946233,
+      "grad_norm": 0.26328954802317356,
+      "learning_rate": 7.06934820462346e-07,
+      "loss": 1.6027,
+      "step": 6575
+    },
+    {
+      "epoch": 2.899955927721463,
+      "grad_norm": 0.3427927943612674,
+      "learning_rate": 6.768258965467289e-07,
+      "loss": 1.7368,
+      "step": 6580
+    },
+    {
+      "epoch": 2.9021595416483033,
+      "grad_norm": 0.3099689467854988,
+      "learning_rate": 6.473700071241484e-07,
+      "loss": 1.7899,
+      "step": 6585
+    },
+    {
+      "epoch": 2.904363155575143,
+      "grad_norm": 0.38502367800150844,
+      "learning_rate": 6.185673458625418e-07,
+      "loss": 1.732,
+      "step": 6590
+    },
+    {
+      "epoch": 2.9065667695019832,
+      "grad_norm": 0.3501339480817073,
+      "learning_rate": 5.904181021349375e-07,
+      "loss": 1.6615,
+      "step": 6595
+    },
+    {
+      "epoch": 2.9087703834288234,
+      "grad_norm": 0.28836814250088966,
+      "learning_rate": 5.629224610182671e-07,
+      "loss": 1.5576,
+      "step": 6600
+    },
+    {
+      "epoch": 2.910973997355663,
+      "grad_norm": 0.2999170211226527,
+      "learning_rate": 5.360806032920995e-07,
+      "loss": 1.6333,
+      "step": 6605
+    },
+    {
+      "epoch": 2.9131776112825034,
+      "grad_norm": 0.3369079119860257,
+      "learning_rate": 5.09892705437498e-07,
+      "loss": 1.4815,
+      "step": 6610
+    },
+    {
+      "epoch": 2.915381225209343,
+      "grad_norm": 0.30167843329688654,
+      "learning_rate": 4.843589396358427e-07,
+      "loss": 1.7719,
+      "step": 6615
+    },
+    {
+      "epoch": 2.9175848391361834,
+      "grad_norm": 0.3352146963211379,
+      "learning_rate": 4.5947947376767663e-07,
+      "loss": 1.8039,
+      "step": 6620
+    },
+    {
+      "epoch": 2.919788453063023,
+      "grad_norm": 0.31153697039563194,
+      "learning_rate": 4.3525447141165023e-07,
+      "loss": 1.4281,
+      "step": 6625
+    },
+    {
+      "epoch": 2.9219920669898634,
+      "grad_norm": 0.3021129933679619,
+      "learning_rate": 4.116840918434006e-07,
+      "loss": 1.7845,
+      "step": 6630
+    },
+    {
+      "epoch": 2.9241956809167036,
+      "grad_norm": 0.3038448101025625,
+      "learning_rate": 3.887684900345301e-07,
+      "loss": 1.6785,
+      "step": 6635
+    },
+    {
+      "epoch": 2.9263992948435433,
+      "grad_norm": 0.29571689111577726,
+      "learning_rate": 3.665078166515623e-07,
+      "loss": 1.5903,
+      "step": 6640
+    },
+    {
+      "epoch": 2.9286029087703835,
+      "grad_norm": 0.3420405614774203,
+      "learning_rate": 3.449022180549766e-07,
+      "loss": 1.7721,
+      "step": 6645
+    },
+    {
+      "epoch": 2.9308065226972233,
+      "grad_norm": 0.28585767479477975,
+      "learning_rate": 3.2395183629824186e-07,
+      "loss": 1.6843,
+      "step": 6650
+    },
+    {
+      "epoch": 2.9330101366240635,
+      "grad_norm": 0.2974354225013896,
+      "learning_rate": 3.0365680912688434e-07,
+      "loss": 1.557,
+      "step": 6655
+    },
+    {
+      "epoch": 2.9352137505509033,
+      "grad_norm": 0.3294613096996469,
+      "learning_rate": 2.840172699775656e-07,
+      "loss": 1.5854,
+      "step": 6660
+    },
+    {
+      "epoch": 2.9374173644777435,
+      "grad_norm": 0.2754088803888805,
+      "learning_rate": 2.650333479771949e-07,
+      "loss": 1.5721,
+      "step": 6665
+    },
+    {
+      "epoch": 2.9396209784045837,
+      "grad_norm": 0.2983838481391387,
+      "learning_rate": 2.467051679421406e-07,
+      "loss": 1.6993,
+      "step": 6670
+    },
+    {
+      "epoch": 2.9418245923314235,
+      "grad_norm": 0.31155141036980233,
+      "learning_rate": 2.290328503773309e-07,
+      "loss": 1.687,
+      "step": 6675
+    },
+    {
+      "epoch": 2.9440282062582637,
+      "grad_norm": 0.29451801544591716,
+      "learning_rate": 2.1201651147554347e-07,
+      "loss": 1.7157,
+      "step": 6680
+    },
+    {
+      "epoch": 2.9462318201851034,
+      "grad_norm": 0.3316660290657273,
+      "learning_rate": 1.956562631165504e-07,
+      "loss": 1.6958,
+      "step": 6685
+    },
+    {
+      "epoch": 2.9484354341119436,
+      "grad_norm": 0.37499704382786475,
+      "learning_rate": 1.7995221286645215e-07,
+      "loss": 1.7061,
+      "step": 6690
+    },
+    {
+      "epoch": 2.9506390480387834,
+      "grad_norm": 0.21043643461218126,
+      "learning_rate": 1.6490446397696702e-07,
+      "loss": 1.3901,
+      "step": 6695
+    },
+    {
+      "epoch": 2.9528426619656236,
+      "grad_norm": 0.3260933032444566,
+      "learning_rate": 1.5051311538469837e-07,
+      "loss": 1.6567,
+      "step": 6700
+    },
+    {
+      "epoch": 2.955046275892464,
+      "grad_norm": 0.3164050584909926,
+      "learning_rate": 1.367782617105351e-07,
+      "loss": 1.6661,
+      "step": 6705
+    },
+    {
+      "epoch": 2.9572498898193036,
+      "grad_norm": 0.3495591454896855,
+      "learning_rate": 1.2369999325901881e-07,
+      "loss": 1.6197,
+      "step": 6710
+    },
+    {
+      "epoch": 2.959453503746144,
+      "grad_norm": 0.3351019168716464,
+      "learning_rate": 1.1127839601774437e-07,
+      "loss": 1.6162,
+      "step": 6715
+    },
+    {
+      "epoch": 2.9616571176729836,
+      "grad_norm": 0.34641497402663024,
+      "learning_rate": 9.951355165678244e-08,
+      "loss": 1.7908,
+      "step": 6720
+    },
+    {
+      "epoch": 2.9638607315998238,
+      "grad_norm": 0.37716614486780947,
+      "learning_rate": 8.840553752815783e-08,
+      "loss": 1.6302,
+      "step": 6725
+    },
+    {
+      "epoch": 2.9660643455266635,
+      "grad_norm": 0.3039829114023131,
+      "learning_rate": 7.79544266653609e-08,
+      "loss": 1.7006,
+      "step": 6730
+    },
+    {
+      "epoch": 2.9682679594535037,
+      "grad_norm": 0.31609676417425303,
+      "learning_rate": 6.816028778281469e-08,
+      "loss": 1.6702,
+      "step": 6735
+    },
+    {
+      "epoch": 2.970471573380344,
+      "grad_norm": 0.37777667052819425,
+      "learning_rate": 5.902318527547523e-08,
+      "loss": 1.4444,
+      "step": 6740
+    },
+    {
+      "epoch": 2.9726751873071837,
+      "grad_norm": 0.36388069806115425,
+      "learning_rate": 5.0543179218365265e-08,
+      "loss": 1.5438,
+      "step": 6745
+    },
+    {
+      "epoch": 2.974878801234024,
+      "grad_norm": 0.362171321908763,
+      "learning_rate": 4.272032536621895e-08,
+      "loss": 1.7638,
+      "step": 6750
+    },
+    {
+      "epoch": 2.9770824151608637,
+      "grad_norm": 0.3560152063618098,
+      "learning_rate": 3.5554675153082195e-08,
+      "loss": 1.6422,
+      "step": 6755
+    },
+    {
+      "epoch": 2.979286029087704,
+      "grad_norm": 0.2934857454303967,
+      "learning_rate": 2.9046275692012904e-08,
+      "loss": 1.5529,
+      "step": 6760
+    },
+    {
+      "epoch": 2.9814896430145437,
+      "grad_norm": 0.31059161987377404,
+      "learning_rate": 2.3195169774714586e-08,
+      "loss": 1.5975,
+      "step": 6765
+    },
+    {
+      "epoch": 2.983693256941384,
+      "grad_norm": 0.4364323493515609,
+      "learning_rate": 1.8001395871303228e-08,
+      "loss": 1.7395,
+      "step": 6770
+    },
+    {
+      "epoch": 2.985896870868224,
+      "grad_norm": 0.3199314191111005,
+      "learning_rate": 1.3464988130051925e-08,
+      "loss": 1.5351,
+      "step": 6775
+    },
+    {
+      "epoch": 2.988100484795064,
+      "grad_norm": 0.2946914049828253,
+      "learning_rate": 9.585976377124439e-09,
+      "loss": 1.519,
+      "step": 6780
+    },
+    {
+      "epoch": 2.990304098721904,
+      "grad_norm": 0.2758574818813751,
+      "learning_rate": 6.364386116419762e-09,
+      "loss": 1.4758,
+      "step": 6785
+    },
+    {
+      "epoch": 2.992507712648744,
+      "grad_norm": 0.329124521705272,
+      "learning_rate": 3.800238529416688e-09,
+      "loss": 1.8557,
+      "step": 6790
+    },
+    {
+      "epoch": 2.994711326575584,
+      "grad_norm": 0.3394680059091016,
+      "learning_rate": 1.8935504749628684e-09,
+      "loss": 1.7226,
+      "step": 6795
+    },
+    {
+      "epoch": 2.996914940502424,
+      "grad_norm": 0.27640804395611795,
+      "learning_rate": 6.443344892637093e-10,
+      "loss": 1.6186,
+      "step": 6800
+    },
+    {
+      "epoch": 2.999118554429264,
+      "grad_norm": 0.3138822790633703,
+      "learning_rate": 5.259878569363608e-11,
+      "loss": 1.6958,
+      "step": 6805
+    },
+    {
+      "epoch": 3.0,
+      "step": 6807,
+      "total_flos": 1.620091928969216e+16,
+      "train_loss": 1.72514724640379,
+      "train_runtime": 38823.0414,
+      "train_samples_per_second": 0.701,
+      "train_steps_per_second": 0.175
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 6807,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.620091928969216e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}