diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,16483 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 494.7368421052632,
+  "eval_steps": 500,
+  "global_step": 23500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.6672950983047485,
+      "learning_rate": 0.0001999999106418929,
+      "loss": 2.6441,
+      "step": 10
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.7060385346412659,
+      "learning_rate": 0.00019999964256773125,
+      "loss": 2.4667,
+      "step": 20
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 1.0146973133087158,
+      "learning_rate": 0.00019999919577799412,
+      "loss": 2.1191,
+      "step": 30
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.9969562292098999,
+      "learning_rate": 0.00019999857027348008,
+      "loss": 1.7304,
+      "step": 40
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 1.6712026596069336,
+      "learning_rate": 0.0001999977660553069,
+      "loss": 1.5827,
+      "step": 50
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 1.2836483716964722,
+      "learning_rate": 0.00019999678312491195,
+      "loss": 1.3176,
+      "step": 60
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 1.425754189491272,
+      "learning_rate": 0.00019999562148405184,
+      "loss": 1.1686,
+      "step": 70
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 1.4164149761199951,
+      "learning_rate": 0.00019999428113480258,
+      "loss": 0.9951,
+      "step": 80
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 1.2195743322372437,
+      "learning_rate": 0.00019999276207955963,
+      "loss": 0.858,
+      "step": 90
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 1.5156681537628174,
+      "learning_rate": 0.0001999910643210378,
+      "loss": 0.8601,
+      "step": 100
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 1.2401676177978516,
+      "learning_rate": 0.00019998918786227123,
+      "loss": 0.8665,
+      "step": 110
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 1.550382137298584,
+      "learning_rate": 0.0001999871327066135,
+      "loss": 0.7616,
+      "step": 120
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 1.254746437072754,
+      "learning_rate": 0.00019998489885773743,
+      "loss": 0.7674,
+      "step": 130
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 1.6343107223510742,
+      "learning_rate": 0.00019998248631963533,
+      "loss": 0.7868,
+      "step": 140
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 1.5750573873519897,
+      "learning_rate": 0.0001999798950966188,
+      "loss": 0.7374,
+      "step": 150
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 1.6087182760238647,
+      "learning_rate": 0.0001999771251933187,
+      "loss": 0.7236,
+      "step": 160
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 1.918428659439087,
+      "learning_rate": 0.0001999741766146854,
+      "loss": 0.711,
+      "step": 170
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 1.64993155002594,
+      "learning_rate": 0.00019997137013085418,
+      "loss": 0.6914,
+      "step": 180
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 2.8061554431915283,
+      "learning_rate": 0.00019996808208386932,
+      "loss": 0.6858,
+      "step": 190
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 1.626227617263794,
+      "learning_rate": 0.00019996461537771275,
+      "loss": 0.6806,
+      "step": 200
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 1.505091905593872,
+      "learning_rate": 0.00019996097001857995,
+      "loss": 0.6467,
+      "step": 210
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 2.134430170059204,
+      "learning_rate": 0.00019995714601298584,
+      "loss": 0.5943,
+      "step": 220
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 1.6179273128509521,
+      "learning_rate": 0.00019995314336776452,
+      "loss": 0.6368,
+      "step": 230
+    },
+    {
+      "epoch": 5.052631578947368,
+      "grad_norm": 1.7538886070251465,
+      "learning_rate": 0.00019994896209006932,
+      "loss": 0.5949,
+      "step": 240
+    },
+    {
+      "epoch": 5.2631578947368425,
+      "grad_norm": 1.7176129817962646,
+      "learning_rate": 0.00019994460218737293,
+      "loss": 0.5982,
+      "step": 250
+    },
+    {
+      "epoch": 5.473684210526316,
+      "grad_norm": 1.5106130838394165,
+      "learning_rate": 0.00019994006366746723,
+      "loss": 0.6008,
+      "step": 260
+    },
+    {
+      "epoch": 5.684210526315789,
+      "grad_norm": 1.4125791788101196,
+      "learning_rate": 0.00019993534653846317,
+      "loss": 0.5772,
+      "step": 270
+    },
+    {
+      "epoch": 5.894736842105263,
+      "grad_norm": 1.6951205730438232,
+      "learning_rate": 0.0001999304508087911,
+      "loss": 0.6063,
+      "step": 280
+    },
+    {
+      "epoch": 6.105263157894737,
+      "grad_norm": 2.1900129318237305,
+      "learning_rate": 0.0001999253764872005,
+      "loss": 0.5803,
+      "step": 290
+    },
+    {
+      "epoch": 6.315789473684211,
+      "grad_norm": 1.5468244552612305,
+      "learning_rate": 0.00019992012358276,
+      "loss": 0.5515,
+      "step": 300
+    },
+    {
+      "epoch": 6.526315789473684,
+      "grad_norm": 1.7162660360336304,
+      "learning_rate": 0.00019991469210485732,
+      "loss": 0.5634,
+      "step": 310
+    },
+    {
+      "epoch": 6.7368421052631575,
+      "grad_norm": 1.6360913515090942,
+      "learning_rate": 0.00019990908206319948,
+      "loss": 0.5589,
+      "step": 320
+    },
+    {
+      "epoch": 6.947368421052632,
+      "grad_norm": 2.227215051651001,
+      "learning_rate": 0.0001999032934678125,
+      "loss": 0.5383,
+      "step": 330
+    },
+    {
+      "epoch": 7.157894736842105,
+      "grad_norm": 2.0045015811920166,
+      "learning_rate": 0.0001998973263290415,
+      "loss": 0.5386,
+      "step": 340
+    },
+    {
+      "epoch": 7.368421052631579,
+      "grad_norm": 1.3294235467910767,
+      "learning_rate": 0.0001998911806575508,
+      "loss": 0.5038,
+      "step": 350
+    },
+    {
+      "epoch": 7.578947368421053,
+      "grad_norm": 1.959319829940796,
+      "learning_rate": 0.00019988485646432364,
+      "loss": 0.5292,
+      "step": 360
+    },
+    {
+      "epoch": 7.7894736842105265,
+      "grad_norm": 2.2458395957946777,
+      "learning_rate": 0.00019987835376066243,
+      "loss": 0.4982,
+      "step": 370
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 2.295234203338623,
+      "learning_rate": 0.00019987167255818854,
+      "loss": 0.5044,
+      "step": 380
+    },
+    {
+      "epoch": 8.210526315789474,
+      "grad_norm": 2.3475089073181152,
+      "learning_rate": 0.00019986481286884234,
+      "loss": 0.476,
+      "step": 390
+    },
+    {
+      "epoch": 8.421052631578947,
+      "grad_norm": 1.692319393157959,
+      "learning_rate": 0.00019985777470488326,
+      "loss": 0.5153,
+      "step": 400
+    },
+    {
+      "epoch": 8.631578947368421,
+      "grad_norm": 1.9400404691696167,
+      "learning_rate": 0.00019985055807888958,
+      "loss": 0.5034,
+      "step": 410
+    },
+    {
+      "epoch": 8.842105263157894,
+      "grad_norm": 2.033897876739502,
+      "learning_rate": 0.00019984316300375863,
+      "loss": 0.505,
+      "step": 420
+    },
+    {
+      "epoch": 9.052631578947368,
+      "grad_norm": 1.9000393152236938,
+      "learning_rate": 0.00019983558949270656,
+      "loss": 0.5238,
+      "step": 430
+    },
+    {
+      "epoch": 9.263157894736842,
+      "grad_norm": 2.177555561065674,
+      "learning_rate": 0.0001998278375592685,
+      "loss": 0.4655,
+      "step": 440
+    },
+    {
+      "epoch": 9.473684210526315,
+      "grad_norm": 1.841349720954895,
+      "learning_rate": 0.00019981990721729842,
+      "loss": 0.4829,
+      "step": 450
+    },
+    {
+      "epoch": 9.68421052631579,
+      "grad_norm": 2.5707108974456787,
+      "learning_rate": 0.0001998117984809691,
+      "loss": 0.4808,
+      "step": 460
+    },
+    {
+      "epoch": 9.894736842105264,
+      "grad_norm": 1.9689738750457764,
+      "learning_rate": 0.00019980351136477214,
+      "loss": 0.476,
+      "step": 470
+    },
+    {
+      "epoch": 10.105263157894736,
+      "grad_norm": 2.5027239322662354,
+      "learning_rate": 0.000199795045883518,
+      "loss": 0.4365,
+      "step": 480
+    },
+    {
+      "epoch": 10.31578947368421,
+      "grad_norm": 2.014554500579834,
+      "learning_rate": 0.0001997864020523359,
+      "loss": 0.431,
+      "step": 490
+    },
+    {
+      "epoch": 10.526315789473685,
+      "grad_norm": 1.5790047645568848,
+      "learning_rate": 0.0001997775798866737,
+      "loss": 0.4584,
+      "step": 500
+    },
+    {
+      "epoch": 10.736842105263158,
+      "grad_norm": 1.7680073976516724,
+      "learning_rate": 0.00019976857940229807,
+      "loss": 0.4736,
+      "step": 510
+    },
+    {
+      "epoch": 10.947368421052632,
+      "grad_norm": 2.524589776992798,
+      "learning_rate": 0.00019975940061529434,
+      "loss": 0.5041,
+      "step": 520
+    },
+    {
+      "epoch": 11.157894736842104,
+      "grad_norm": 2.058872699737549,
+      "learning_rate": 0.00019975004354206647,
+      "loss": 0.4339,
+      "step": 530
+    },
+    {
+      "epoch": 11.368421052631579,
+      "grad_norm": 2.1387035846710205,
+      "learning_rate": 0.00019974050819933709,
+      "loss": 0.4611,
+      "step": 540
+    },
+    {
+      "epoch": 11.578947368421053,
+      "grad_norm": 2.866647720336914,
+      "learning_rate": 0.00019973079460414744,
+      "loss": 0.4361,
+      "step": 550
+    },
+    {
+      "epoch": 11.789473684210526,
+      "grad_norm": 2.1138830184936523,
+      "learning_rate": 0.0001997209027738572,
+      "loss": 0.4475,
+      "step": 560
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 2.3789708614349365,
+      "learning_rate": 0.00019971083272614474,
+      "loss": 0.4515,
+      "step": 570
+    },
+    {
+      "epoch": 12.210526315789474,
+      "grad_norm": 1.8684824705123901,
+      "learning_rate": 0.00019970058447900684,
+      "loss": 0.3823,
+      "step": 580
+    },
+    {
+      "epoch": 12.421052631578947,
+      "grad_norm": 2.0576510429382324,
+      "learning_rate": 0.00019969015805075877,
+      "loss": 0.4218,
+      "step": 590
+    },
+    {
+      "epoch": 12.631578947368421,
+      "grad_norm": 2.553516387939453,
+      "learning_rate": 0.00019967955346003428,
+      "loss": 0.4429,
+      "step": 600
+    },
+    {
+      "epoch": 12.842105263157894,
+      "grad_norm": 2.1283926963806152,
+      "learning_rate": 0.00019966877072578548,
+      "loss": 0.4622,
+      "step": 610
+    },
+    {
+      "epoch": 13.052631578947368,
+      "grad_norm": 2.276144504547119,
+      "learning_rate": 0.00019965780986728286,
+      "loss": 0.4225,
+      "step": 620
+    },
+    {
+      "epoch": 13.263157894736842,
+      "grad_norm": 1.8524142503738403,
+      "learning_rate": 0.00019964667090411524,
+      "loss": 0.4001,
+      "step": 630
+    },
+    {
+      "epoch": 13.473684210526315,
+      "grad_norm": 1.7885457277297974,
+      "learning_rate": 0.0001996353538561898,
+      "loss": 0.3619,
+      "step": 640
+    },
+    {
+      "epoch": 13.68421052631579,
+      "grad_norm": 2.346691370010376,
+      "learning_rate": 0.0001996238587437319,
+      "loss": 0.4234,
+      "step": 650
+    },
+    {
+      "epoch": 13.894736842105264,
+      "grad_norm": 2.195350170135498,
+      "learning_rate": 0.00019961218558728515,
+      "loss": 0.4107,
+      "step": 660
+    },
+    {
+      "epoch": 14.105263157894736,
+      "grad_norm": 1.7246651649475098,
+      "learning_rate": 0.00019960033440771143,
+      "loss": 0.4447,
+      "step": 670
+    },
+    {
+      "epoch": 14.31578947368421,
+      "grad_norm": 1.8804007768630981,
+      "learning_rate": 0.00019958830522619065,
+      "loss": 0.3807,
+      "step": 680
+    },
+    {
+      "epoch": 14.526315789473685,
+      "grad_norm": 1.7152245044708252,
+      "learning_rate": 0.000199576098064221,
+      "loss": 0.3791,
+      "step": 690
+    },
+    {
+      "epoch": 14.736842105263158,
+      "grad_norm": 1.8070579767227173,
+      "learning_rate": 0.00019956371294361857,
+      "loss": 0.3976,
+      "step": 700
+    },
+    {
+      "epoch": 14.947368421052632,
+      "grad_norm": 1.8141262531280518,
+      "learning_rate": 0.00019955114988651765,
+      "loss": 0.4205,
+      "step": 710
+    },
+    {
+      "epoch": 15.157894736842104,
+      "grad_norm": 1.9717251062393188,
+      "learning_rate": 0.00019953840891537037,
+      "loss": 0.3989,
+      "step": 720
+    },
+    {
+      "epoch": 15.368421052631579,
+      "grad_norm": 2.3383822441101074,
+      "learning_rate": 0.000199525490052947,
+      "loss": 0.3747,
+      "step": 730
+    },
+    {
+      "epoch": 15.578947368421053,
+      "grad_norm": 2.188035726547241,
+      "learning_rate": 0.0001995123933223356,
+      "loss": 0.3734,
+      "step": 740
+    },
+    {
+      "epoch": 15.789473684210526,
+      "grad_norm": 2.967191457748413,
+      "learning_rate": 0.00019949911874694217,
+      "loss": 0.3857,
+      "step": 750
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 2.172882080078125,
+      "learning_rate": 0.0001994856663504905,
+      "loss": 0.417,
+      "step": 760
+    },
+    {
+      "epoch": 16.210526315789473,
+      "grad_norm": 2.480196952819824,
+      "learning_rate": 0.00019947203615702224,
+      "loss": 0.3707,
+      "step": 770
+    },
+    {
+      "epoch": 16.42105263157895,
+      "grad_norm": 1.7600008249282837,
+      "learning_rate": 0.0001994582281908967,
+      "loss": 0.3799,
+      "step": 780
+    },
+    {
+      "epoch": 16.63157894736842,
+      "grad_norm": 1.875954031944275,
+      "learning_rate": 0.00019944424247679102,
+      "loss": 0.365,
+      "step": 790
+    },
+    {
+      "epoch": 16.842105263157894,
+      "grad_norm": 1.9729390144348145,
+      "learning_rate": 0.0001994300790396999,
+      "loss": 0.3698,
+      "step": 800
+    },
+    {
+      "epoch": 17.05263157894737,
+      "grad_norm": 1.9465049505233765,
+      "learning_rate": 0.0001994157379049357,
+      "loss": 0.3945,
+      "step": 810
+    },
+    {
+      "epoch": 17.263157894736842,
+      "grad_norm": 2.3426170349121094,
+      "learning_rate": 0.00019940121909812838,
+      "loss": 0.3627,
+      "step": 820
+    },
+    {
+      "epoch": 17.473684210526315,
+      "grad_norm": 1.4503591060638428,
+      "learning_rate": 0.00019938652264522538,
+      "loss": 0.3709,
+      "step": 830
+    },
+    {
+      "epoch": 17.68421052631579,
+      "grad_norm": 2.0927889347076416,
+      "learning_rate": 0.00019937164857249164,
+      "loss": 0.358,
+      "step": 840
+    },
+    {
+      "epoch": 17.894736842105264,
+      "grad_norm": 2.4150617122650146,
+      "learning_rate": 0.00019935659690650955,
+      "loss": 0.3579,
+      "step": 850
+    },
+    {
+      "epoch": 18.105263157894736,
+      "grad_norm": 1.8455313444137573,
+      "learning_rate": 0.00019934136767417888,
+      "loss": 0.3693,
+      "step": 860
+    },
+    {
+      "epoch": 18.31578947368421,
+      "grad_norm": 2.0851595401763916,
+      "learning_rate": 0.00019932596090271672,
+      "loss": 0.3305,
+      "step": 870
+    },
+    {
+      "epoch": 18.526315789473685,
+      "grad_norm": 2.065988779067993,
+      "learning_rate": 0.00019931037661965749,
+      "loss": 0.362,
+      "step": 880
+    },
+    {
+      "epoch": 18.736842105263158,
+      "grad_norm": 2.07012939453125,
+      "learning_rate": 0.0001992946148528528,
+      "loss": 0.3819,
+      "step": 890
+    },
+    {
+      "epoch": 18.94736842105263,
+      "grad_norm": 1.9075250625610352,
+      "learning_rate": 0.00019927867563047152,
+      "loss": 0.3557,
+      "step": 900
+    },
+    {
+      "epoch": 19.157894736842106,
+      "grad_norm": 1.9468849897384644,
+      "learning_rate": 0.0001992625589809996,
+      "loss": 0.32,
+      "step": 910
+    },
+    {
+      "epoch": 19.36842105263158,
+      "grad_norm": 2.025853395462036,
+      "learning_rate": 0.00019924626493324015,
+      "loss": 0.3372,
+      "step": 920
+    },
+    {
+      "epoch": 19.57894736842105,
+      "grad_norm": 1.625706434249878,
+      "learning_rate": 0.00019922979351631322,
+      "loss": 0.3378,
+      "step": 930
+    },
+    {
+      "epoch": 19.789473684210527,
+      "grad_norm": 1.993627667427063,
+      "learning_rate": 0.0001992131447596559,
+      "loss": 0.3488,
+      "step": 940
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 1.914183497428894,
+      "learning_rate": 0.00019919631869302226,
+      "loss": 0.3836,
+      "step": 950
+    },
+    {
+      "epoch": 20.210526315789473,
+      "grad_norm": 1.5059126615524292,
+      "learning_rate": 0.0001991793153464832,
+      "loss": 0.331,
+      "step": 960
+    },
+    {
+      "epoch": 20.42105263157895,
+      "grad_norm": 1.9410295486450195,
+      "learning_rate": 0.00019916213475042644,
+      "loss": 0.3474,
+      "step": 970
+    },
+    {
+      "epoch": 20.63157894736842,
+      "grad_norm": 2.240558624267578,
+      "learning_rate": 0.00019914477693555647,
+      "loss": 0.3364,
+      "step": 980
+    },
+    {
+      "epoch": 20.842105263157894,
+      "grad_norm": 1.5858815908432007,
+      "learning_rate": 0.00019912724193289457,
+      "loss": 0.3417,
+      "step": 990
+    },
+    {
+      "epoch": 21.05263157894737,
+      "grad_norm": 2.031370162963867,
+      "learning_rate": 0.00019910952977377863,
+      "loss": 0.3281,
+      "step": 1000
+    },
+    {
+      "epoch": 21.263157894736842,
+      "grad_norm": 1.5948817729949951,
+      "learning_rate": 0.0001990916404898631,
+      "loss": 0.3147,
+      "step": 1010
+    },
+    {
+      "epoch": 21.473684210526315,
+      "grad_norm": 1.7134989500045776,
+      "learning_rate": 0.00019907357411311907,
+      "loss": 0.3325,
+      "step": 1020
+    },
+    {
+      "epoch": 21.68421052631579,
+      "grad_norm": 1.8072121143341064,
+      "learning_rate": 0.0001990553306758341,
+      "loss": 0.3453,
+      "step": 1030
+    },
+    {
+      "epoch": 21.894736842105264,
+      "grad_norm": 2.8281702995300293,
+      "learning_rate": 0.00019903691021061213,
+      "loss": 0.3514,
+      "step": 1040
+    },
+    {
+      "epoch": 22.105263157894736,
+      "grad_norm": 2.1076014041900635,
+      "learning_rate": 0.00019901831275037353,
+      "loss": 0.3241,
+      "step": 1050
+    },
+    {
+      "epoch": 22.31578947368421,
+      "grad_norm": 2.392726421356201,
+      "learning_rate": 0.00019899953832835498,
+      "loss": 0.3078,
+      "step": 1060
+    },
+    {
+      "epoch": 22.526315789473685,
+      "grad_norm": 2.0120134353637695,
+      "learning_rate": 0.00019898058697810945,
+      "loss": 0.3438,
+      "step": 1070
+    },
+    {
+      "epoch": 22.736842105263158,
+      "grad_norm": 1.7498539686203003,
+      "learning_rate": 0.000198961458733506,
+      "loss": 0.3152,
+      "step": 1080
+    },
+    {
+      "epoch": 22.94736842105263,
+      "grad_norm": 1.975158452987671,
+      "learning_rate": 0.00019894215362872996,
+      "loss": 0.3443,
+      "step": 1090
+    },
+    {
+      "epoch": 23.157894736842106,
+      "grad_norm": 1.8120285272598267,
+      "learning_rate": 0.0001989226716982827,
+      "loss": 0.3078,
+      "step": 1100
+    },
+    {
+      "epoch": 23.36842105263158,
+      "grad_norm": 2.1680185794830322,
+      "learning_rate": 0.0001989030129769815,
+      "loss": 0.2995,
+      "step": 1110
+    },
+    {
+      "epoch": 23.57894736842105,
+      "grad_norm": 1.811830997467041,
+      "learning_rate": 0.00019888317749995978,
+      "loss": 0.3232,
+      "step": 1120
+    },
+    {
+      "epoch": 23.789473684210527,
+      "grad_norm": 1.6087640523910522,
+      "learning_rate": 0.00019886316530266673,
+      "loss": 0.3307,
+      "step": 1130
+    },
+    {
+      "epoch": 24.0,
+      "grad_norm": 2.173571825027466,
+      "learning_rate": 0.00019884297642086736,
+      "loss": 0.3286,
+      "step": 1140
+    },
+    {
+      "epoch": 24.210526315789473,
+      "grad_norm": 1.9915390014648438,
+      "learning_rate": 0.0001988226108906425,
+      "loss": 0.287,
+      "step": 1150
+    },
+    {
+      "epoch": 24.42105263157895,
+      "grad_norm": 1.5517760515213013,
+      "learning_rate": 0.0001988020687483886,
+      "loss": 0.3169,
+      "step": 1160
+    },
+    {
+      "epoch": 24.63157894736842,
+      "grad_norm": 1.8992213010787964,
+      "learning_rate": 0.0001987813500308179,
+      "loss": 0.3393,
+      "step": 1170
+    },
+    {
+      "epoch": 24.842105263157894,
+      "grad_norm": 2.124034881591797,
+      "learning_rate": 0.00019876045477495804,
+      "loss": 0.3214,
+      "step": 1180
+    },
+    {
+      "epoch": 25.05263157894737,
+      "grad_norm": 1.5663875341415405,
+      "learning_rate": 0.00019873938301815224,
+      "loss": 0.3076,
+      "step": 1190
+    },
+    {
+      "epoch": 25.263157894736842,
+      "grad_norm": 2.0292165279388428,
+      "learning_rate": 0.00019871813479805915,
+      "loss": 0.3133,
+      "step": 1200
+    },
+    {
+      "epoch": 25.473684210526315,
+      "grad_norm": 1.7555932998657227,
+      "learning_rate": 0.00019869671015265277,
+      "loss": 0.3019,
+      "step": 1210
+    },
+    {
+      "epoch": 25.68421052631579,
+      "grad_norm": 2.3077430725097656,
+      "learning_rate": 0.00019867510912022245,
+      "loss": 0.3369,
+      "step": 1220
+    },
+    {
+      "epoch": 25.894736842105264,
+      "grad_norm": 2.0065627098083496,
+      "learning_rate": 0.0001986533317393727,
+      "loss": 0.3225,
+      "step": 1230
+    },
+    {
+      "epoch": 26.105263157894736,
+      "grad_norm": 1.531724214553833,
+      "learning_rate": 0.00019863137804902324,
+      "loss": 0.2914,
+      "step": 1240
+    },
+    {
+      "epoch": 26.31578947368421,
+      "grad_norm": 1.7669227123260498,
+      "learning_rate": 0.00019860924808840893,
+      "loss": 0.2988,
+      "step": 1250
+    },
+    {
+      "epoch": 26.526315789473685,
+      "grad_norm": 1.8059635162353516,
+      "learning_rate": 0.0001985869418970795,
+      "loss": 0.2805,
+      "step": 1260
+    },
+    {
+      "epoch": 26.736842105263158,
+      "grad_norm": 1.8464925289154053,
+      "learning_rate": 0.00019856445951489982,
+      "loss": 0.297,
+      "step": 1270
+    },
+    {
+      "epoch": 26.94736842105263,
+      "grad_norm": 2.014112710952759,
+      "learning_rate": 0.00019854180098204948,
+      "loss": 0.308,
+      "step": 1280
+    },
+    {
+      "epoch": 27.157894736842106,
+      "grad_norm": 1.5433177947998047,
+      "learning_rate": 0.000198518966339023,
+      "loss": 0.3105,
+      "step": 1290
+    },
+    {
+      "epoch": 27.36842105263158,
+      "grad_norm": 1.5101069211959839,
+      "learning_rate": 0.00019849595562662956,
+      "loss": 0.2919,
+      "step": 1300
+    },
+    {
+      "epoch": 27.57894736842105,
+      "grad_norm": 1.640114188194275,
+      "learning_rate": 0.00019847276888599306,
+      "loss": 0.278,
+      "step": 1310
+    },
+    {
+      "epoch": 27.789473684210527,
+      "grad_norm": 1.9816792011260986,
+      "learning_rate": 0.00019844940615855198,
+      "loss": 0.3087,
+      "step": 1320
+    },
+    {
+      "epoch": 28.0,
+      "grad_norm": 2.4635629653930664,
+      "learning_rate": 0.0001984258674860592,
+      "loss": 0.3174,
+      "step": 1330
+    },
+    {
+      "epoch": 28.210526315789473,
+      "grad_norm": 1.7017345428466797,
+      "learning_rate": 0.00019840215291058228,
+      "loss": 0.2855,
+      "step": 1340
+    },
+    {
+      "epoch": 28.42105263157895,
+      "grad_norm": 1.8059803247451782,
+      "learning_rate": 0.00019837826247450294,
+      "loss": 0.2753,
+      "step": 1350
+    },
+    {
+      "epoch": 28.63157894736842,
+      "grad_norm": 1.9220434427261353,
+      "learning_rate": 0.00019835419622051727,
+      "loss": 0.313,
+      "step": 1360
+    },
+    {
+      "epoch": 28.842105263157894,
+      "grad_norm": 1.790998101234436,
+      "learning_rate": 0.00019832995419163555,
+      "loss": 0.3051,
+      "step": 1370
+    },
+    {
+      "epoch": 29.05263157894737,
+      "grad_norm": 1.4004263877868652,
+      "learning_rate": 0.0001983055364311823,
+      "loss": 0.2949,
+      "step": 1380
+    },
+    {
+      "epoch": 29.263157894736842,
+      "grad_norm": 1.7643471956253052,
+      "learning_rate": 0.00019828094298279589,
+      "loss": 0.2933,
+      "step": 1390
+    },
+    {
+      "epoch": 29.473684210526315,
+      "grad_norm": 1.7527259588241577,
+      "learning_rate": 0.0001982561738904289,
+      "loss": 0.2849,
+      "step": 1400
+    },
+    {
+      "epoch": 29.68421052631579,
+      "grad_norm": 2.0102221965789795,
+      "learning_rate": 0.00019823122919834766,
+      "loss": 0.2916,
+      "step": 1410
+    },
+    {
+      "epoch": 29.894736842105264,
+      "grad_norm": 1.9086006879806519,
+      "learning_rate": 0.0001982061089511324,
+      "loss": 0.2833,
+      "step": 1420
+    },
+    {
+      "epoch": 30.105263157894736,
+      "grad_norm": 2.0103325843811035,
+      "learning_rate": 0.00019818081319367709,
+      "loss": 0.2753,
+      "step": 1430
+    },
+    {
+      "epoch": 30.31578947368421,
+      "grad_norm": 1.5917645692825317,
+      "learning_rate": 0.0001981553419711893,
+      "loss": 0.2705,
+      "step": 1440
+    },
+    {
+      "epoch": 30.526315789473685,
+      "grad_norm": 1.8320797681808472,
+      "learning_rate": 0.00019812969532919032,
+      "loss": 0.295,
+      "step": 1450
+    },
+    {
+      "epoch": 30.736842105263158,
+      "grad_norm": 2.35292911529541,
+      "learning_rate": 0.00019810387331351478,
+      "loss": 0.2913,
+      "step": 1460
+    },
+    {
+      "epoch": 30.94736842105263,
+      "grad_norm": 1.739706039428711,
+      "learning_rate": 0.00019807787597031084,
+      "loss": 0.2926,
+      "step": 1470
+    },
+    {
+      "epoch": 31.157894736842106,
+      "grad_norm": 1.9969830513000488,
+      "learning_rate": 0.00019805170334603997,
+      "loss": 0.2991,
+      "step": 1480
+    },
+    {
+      "epoch": 31.36842105263158,
+      "grad_norm": 1.7005611658096313,
+      "learning_rate": 0.00019802535548747686,
+      "loss": 0.2941,
+      "step": 1490
+    },
+    {
+      "epoch": 31.57894736842105,
+      "grad_norm": 1.9884356260299683,
+      "learning_rate": 0.00019799883244170946,
+      "loss": 0.2743,
+      "step": 1500
+    },
+    {
+      "epoch": 31.789473684210527,
+      "grad_norm": 2.2900278568267822,
+      "learning_rate": 0.00019797213425613869,
+      "loss": 0.299,
+      "step": 1510
+    },
+    {
+      "epoch": 32.0,
+      "grad_norm": 2.4475691318511963,
+      "learning_rate": 0.00019794526097847862,
+      "loss": 0.2805,
+      "step": 1520
+    },
+    {
+      "epoch": 32.21052631578947,
+      "grad_norm": 1.836012601852417,
+      "learning_rate": 0.00019791821265675603,
+      "loss": 0.2549,
+      "step": 1530
+    },
+    {
+      "epoch": 32.421052631578945,
+      "grad_norm": 2.088721752166748,
+      "learning_rate": 0.0001978909893393108,
+      "loss": 0.3096,
+      "step": 1540
+    },
+    {
+      "epoch": 32.63157894736842,
+      "grad_norm": 1.7685314416885376,
+      "learning_rate": 0.0001978635910747953,
+      "loss": 0.2725,
+      "step": 1550
+    },
+    {
+      "epoch": 32.8421052631579,
+      "grad_norm": 1.8119862079620361,
+      "learning_rate": 0.00019783601791217474,
+      "loss": 0.2915,
+      "step": 1560
+    },
+    {
+      "epoch": 33.05263157894737,
+      "grad_norm": 1.7417136430740356,
+      "learning_rate": 0.0001978082699007268,
+      "loss": 0.2695,
+      "step": 1570
+    },
+    {
+      "epoch": 33.26315789473684,
+      "grad_norm": 1.4141062498092651,
+      "learning_rate": 0.0001977803470900417,
+      "loss": 0.2647,
+      "step": 1580
+    },
+    {
+      "epoch": 33.473684210526315,
+      "grad_norm": 2.370269775390625,
+      "learning_rate": 0.000197752249530022,
+      "loss": 0.2753,
+      "step": 1590
+    },
+    {
+      "epoch": 33.68421052631579,
+      "grad_norm": 1.886096715927124,
+      "learning_rate": 0.00019772397727088262,
+      "loss": 0.283,
+      "step": 1600
+    },
+    {
+      "epoch": 33.89473684210526,
+      "grad_norm": 1.6945075988769531,
+      "learning_rate": 0.00019769553036315065,
+      "loss": 0.2774,
+      "step": 1610
+    },
+    {
+      "epoch": 34.10526315789474,
+      "grad_norm": 1.800110101699829,
+      "learning_rate": 0.00019766690885766533,
+      "loss": 0.2646,
+      "step": 1620
+    },
+    {
+      "epoch": 34.31578947368421,
+      "grad_norm": 1.670495867729187,
+      "learning_rate": 0.00019763811280557793,
+      "loss": 0.2409,
+      "step": 1630
+    },
+    {
+      "epoch": 34.526315789473685,
+      "grad_norm": 2.0900466442108154,
+      "learning_rate": 0.0001976091422583517,
+      "loss": 0.2994,
+      "step": 1640
+    },
+    {
+      "epoch": 34.73684210526316,
+      "grad_norm": 2.043393611907959,
+      "learning_rate": 0.00019757999726776167,
+      "loss": 0.2862,
+      "step": 1650
+    },
+    {
+      "epoch": 34.94736842105263,
+      "grad_norm": 2.35016131401062,
+      "learning_rate": 0.00019755067788589467,
+      "loss": 0.2816,
+      "step": 1660
+    },
+    {
+      "epoch": 35.1578947368421,
+      "grad_norm": 1.4299836158752441,
+      "learning_rate": 0.00019752118416514915,
+      "loss": 0.255,
+      "step": 1670
+    },
+    {
+      "epoch": 35.36842105263158,
+      "grad_norm": 1.9748225212097168,
+      "learning_rate": 0.00019749151615823526,
+      "loss": 0.2635,
+      "step": 1680
+    },
+    {
+      "epoch": 35.578947368421055,
+      "grad_norm": 1.6960207223892212,
+      "learning_rate": 0.00019746167391817448,
+      "loss": 0.2682,
+      "step": 1690
+    },
+    {
+      "epoch": 35.78947368421053,
+      "grad_norm": 2.131016969680786,
+      "learning_rate": 0.00019743165749829973,
+      "loss": 0.2812,
+      "step": 1700
+    },
+    {
+      "epoch": 36.0,
+      "grad_norm": 1.7239797115325928,
+      "learning_rate": 0.00019740146695225525,
+      "loss": 0.2858,
+      "step": 1710
+    },
+    {
+      "epoch": 36.21052631578947,
+      "grad_norm": 1.5549391508102417,
+      "learning_rate": 0.0001973711023339964,
+      "loss": 0.2619,
+      "step": 1720
+    },
+    {
+      "epoch": 36.421052631578945,
+      "grad_norm": 1.949060320854187,
+      "learning_rate": 0.00019734056369778973,
+      "loss": 0.2489,
+      "step": 1730
+    },
+    {
+      "epoch": 36.63157894736842,
+      "grad_norm": 1.5633726119995117,
+      "learning_rate": 0.00019730985109821266,
+      "loss": 0.2791,
+      "step": 1740
+    },
+    {
+      "epoch": 36.8421052631579,
+      "grad_norm": 1.7164374589920044,
+      "learning_rate": 0.00019727896459015366,
+      "loss": 0.2764,
+      "step": 1750
+    },
+    {
+      "epoch": 37.05263157894737,
+      "grad_norm": 1.5717780590057373,
+      "learning_rate": 0.00019724790422881187,
+      "loss": 0.2574,
+      "step": 1760
+    },
+    {
+      "epoch": 37.26315789473684,
+      "grad_norm": 1.7981038093566895,
+      "learning_rate": 0.00019721667006969727,
+      "loss": 0.2558,
+      "step": 1770
+    },
+    {
+      "epoch": 37.473684210526315,
+      "grad_norm": 1.596404790878296,
+      "learning_rate": 0.00019718526216863026,
+      "loss": 0.2742,
+      "step": 1780
+    },
+    {
+      "epoch": 37.68421052631579,
+      "grad_norm": 1.3674874305725098,
+      "learning_rate": 0.00019715368058174194,
+      "loss": 0.2583,
+      "step": 1790
+    },
+    {
+      "epoch": 37.89473684210526,
+      "grad_norm": 1.6358920335769653,
+      "learning_rate": 0.00019712192536547364,
+      "loss": 0.2728,
+      "step": 1800
+    },
+    {
+      "epoch": 38.10526315789474,
+      "grad_norm": 1.784155011177063,
+      "learning_rate": 0.0001970899965765772,
+      "loss": 0.2706,
+      "step": 1810
+    },
+    {
+      "epoch": 38.31578947368421,
+      "grad_norm": 1.6264429092407227,
+      "learning_rate": 0.00019705789427211444,
+      "loss": 0.2566,
+      "step": 1820
+    },
+    {
+      "epoch": 38.526315789473685,
+      "grad_norm": 1.6494293212890625,
+      "learning_rate": 0.00019702561850945744,
+      "loss": 0.2691,
+      "step": 1830
+    },
+    {
+      "epoch": 38.73684210526316,
+      "grad_norm": 2.1020774841308594,
+      "learning_rate": 0.00019699316934628818,
+      "loss": 0.2508,
+      "step": 1840
+    },
+    {
+      "epoch": 38.94736842105263,
+      "grad_norm": 1.4856878519058228,
+      "learning_rate": 0.0001969605468405986,
+      "loss": 0.261,
+      "step": 1850
+    },
+    {
+      "epoch": 39.1578947368421,
+      "grad_norm": 2.1468899250030518,
+      "learning_rate": 0.00019692775105069042,
+      "loss": 0.261,
+      "step": 1860
+    },
+    {
+      "epoch": 39.36842105263158,
+      "grad_norm": 1.5866405963897705,
+      "learning_rate": 0.00019689478203517498,
+      "loss": 0.2684,
+      "step": 1870
+    },
+    {
+      "epoch": 39.578947368421055,
+      "grad_norm": 1.5345929861068726,
+      "learning_rate": 0.0001968616398529733,
+      "loss": 0.2641,
+      "step": 1880
+    },
+    {
+      "epoch": 39.78947368421053,
+      "grad_norm": 1.664137601852417,
+      "learning_rate": 0.0001968283245633159,
+      "loss": 0.2619,
+      "step": 1890
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 1.8158512115478516,
+      "learning_rate": 0.00019679483622574246,
+      "loss": 0.2493,
+      "step": 1900
+    },
+    {
+      "epoch": 40.21052631578947,
+      "grad_norm": 1.8335703611373901,
+      "learning_rate": 0.00019676117490010215,
+      "loss": 0.2377,
+      "step": 1910
+    },
+    {
+      "epoch": 40.421052631578945,
+      "grad_norm": 1.6368952989578247,
+      "learning_rate": 0.00019672734064655326,
+      "loss": 0.238,
+      "step": 1920
+    },
+    {
+      "epoch": 40.63157894736842,
+      "grad_norm": 1.600355863571167,
+      "learning_rate": 0.00019669333352556297,
+      "loss": 0.2727,
+      "step": 1930
+    },
+    {
+      "epoch": 40.8421052631579,
+      "grad_norm": 1.7188122272491455,
+      "learning_rate": 0.0001966591535979076,
+      "loss": 0.2563,
+      "step": 1940
+    },
+    {
+      "epoch": 41.05263157894737,
+      "grad_norm": 1.2512385845184326,
+      "learning_rate": 0.0001966248009246722,
+      "loss": 0.2662,
+      "step": 1950
+    },
+    {
+      "epoch": 41.26315789473684,
+      "grad_norm": 1.8242894411087036,
+      "learning_rate": 0.0001965902755672506,
+      "loss": 0.2451,
+      "step": 1960
+    },
+    {
+      "epoch": 41.473684210526315,
+      "grad_norm": 1.5848056077957153,
+      "learning_rate": 0.00019655557758734517,
+      "loss": 0.2509,
+      "step": 1970
+    },
+    {
+      "epoch": 41.68421052631579,
+      "grad_norm": 2.00626277923584,
+      "learning_rate": 0.00019652070704696687,
+      "loss": 0.2662,
+      "step": 1980
+    },
+    {
+      "epoch": 41.89473684210526,
+      "grad_norm": 1.7578226327896118,
+      "learning_rate": 0.00019648566400843492,
+      "loss": 0.2553,
+      "step": 1990
+    },
+    {
+      "epoch": 42.10526315789474,
+      "grad_norm": 1.1532166004180908,
+      "learning_rate": 0.00019645044853437704,
+      "loss": 0.2525,
+      "step": 2000
+    },
+    {
+      "epoch": 42.31578947368421,
+      "grad_norm": 1.355414628982544,
+      "learning_rate": 0.00019641506068772887,
+      "loss": 0.2409,
+      "step": 2010
+    },
+    {
+      "epoch": 42.526315789473685,
+      "grad_norm": 1.8452808856964111,
+      "learning_rate": 0.00019637950053173433,
+      "loss": 0.2716,
+      "step": 2020
+    },
+    {
+      "epoch": 42.73684210526316,
+      "grad_norm": 2.008028745651245,
+      "learning_rate": 0.0001963437681299451,
+      "loss": 0.2482,
+      "step": 2030
+    },
+    {
+      "epoch": 42.94736842105263,
+      "grad_norm": 1.8328056335449219,
+      "learning_rate": 0.00019630786354622086,
+      "loss": 0.26,
+      "step": 2040
+    },
+    {
+      "epoch": 43.1578947368421,
+      "grad_norm": 1.332192301750183,
+      "learning_rate": 0.00019627178684472884,
+      "loss": 0.2586,
+      "step": 2050
+    },
+    {
+      "epoch": 43.36842105263158,
+      "grad_norm": 1.4395008087158203,
+      "learning_rate": 0.00019623553808994403,
+      "loss": 0.2461,
+      "step": 2060
+    },
+    {
+      "epoch": 43.578947368421055,
+      "grad_norm": 1.8931459188461304,
+      "learning_rate": 0.00019619911734664877,
+      "loss": 0.2476,
+      "step": 2070
+    },
+    {
+      "epoch": 43.78947368421053,
+      "grad_norm": 1.6473420858383179,
+      "learning_rate": 0.00019616252467993283,
+      "loss": 0.2497,
+      "step": 2080
+    },
+    {
+      "epoch": 44.0,
+      "grad_norm": 1.7745323181152344,
+      "learning_rate": 0.0001961257601551933,
+      "loss": 0.255,
+      "step": 2090
+    },
+    {
+      "epoch": 44.21052631578947,
+      "grad_norm": 1.6052919626235962,
+      "learning_rate": 0.00019608882383813435,
+      "loss": 0.2411,
+      "step": 2100
+    },
+    {
+      "epoch": 44.421052631578945,
+      "grad_norm": 1.1463373899459839,
+      "learning_rate": 0.00019605171579476708,
+      "loss": 0.2393,
+      "step": 2110
+    },
+    {
+      "epoch": 44.63157894736842,
+      "grad_norm": 1.6653571128845215,
+      "learning_rate": 0.00019601443609140967,
+      "loss": 0.2608,
+      "step": 2120
+    },
+    {
+      "epoch": 44.8421052631579,
+      "grad_norm": 2.1662609577178955,
+      "learning_rate": 0.0001959769847946869,
+      "loss": 0.252,
+      "step": 2130
+    },
+    {
+      "epoch": 45.05263157894737,
+      "grad_norm": 1.2638843059539795,
+      "learning_rate": 0.0001959393619715304,
+      "loss": 0.2529,
+      "step": 2140
+    },
+    {
+      "epoch": 45.26315789473684,
+      "grad_norm": 1.7441433668136597,
+      "learning_rate": 0.00019590156768917822,
+      "loss": 0.2544,
+      "step": 2150
+    },
+    {
+      "epoch": 45.473684210526315,
+      "grad_norm": 1.5743330717086792,
+      "learning_rate": 0.0001958636020151749,
+      "loss": 0.2504,
+      "step": 2160
+    },
+    {
+      "epoch": 45.68421052631579,
+      "grad_norm": 1.2976667881011963,
+      "learning_rate": 0.0001958254650173712,
+      "loss": 0.2479,
+      "step": 2170
+    },
+    {
+      "epoch": 45.89473684210526,
+      "grad_norm": 1.9019654989242554,
+      "learning_rate": 0.00019578715676392414,
+      "loss": 0.2612,
+      "step": 2180
+    },
+    {
+      "epoch": 46.10526315789474,
+      "grad_norm": 1.24739670753479,
+      "learning_rate": 0.0001957486773232968,
+      "loss": 0.2225,
+      "step": 2190
+    },
+    {
+      "epoch": 46.31578947368421,
+      "grad_norm": 1.4010628461837769,
+      "learning_rate": 0.00019571002676425816,
+      "loss": 0.2322,
+      "step": 2200
+    },
+    {
+      "epoch": 46.526315789473685,
+      "grad_norm": 1.3741133213043213,
+      "learning_rate": 0.00019567120515588308,
+      "loss": 0.2537,
+      "step": 2210
+    },
+    {
+      "epoch": 46.73684210526316,
+      "grad_norm": 1.5921913385391235,
+      "learning_rate": 0.00019563221256755197,
+      "loss": 0.2566,
+      "step": 2220
+    },
+    {
+      "epoch": 46.94736842105263,
+      "grad_norm": 1.7186962366104126,
+      "learning_rate": 0.000195593049068951,
+      "loss": 0.2549,
+      "step": 2230
+    },
+    {
+      "epoch": 47.1578947368421,
+      "grad_norm": 1.5977070331573486,
+      "learning_rate": 0.00019555371473007168,
+      "loss": 0.2303,
+      "step": 2240
+    },
+    {
+      "epoch": 47.36842105263158,
+      "grad_norm": 1.6327919960021973,
+      "learning_rate": 0.00019551420962121078,
+      "loss": 0.2463,
+      "step": 2250
+    },
+    {
+      "epoch": 47.578947368421055,
+      "grad_norm": 1.2250909805297852,
+      "learning_rate": 0.00019547453381297042,
+      "loss": 0.247,
+      "step": 2260
+    },
+    {
+      "epoch": 47.78947368421053,
+      "grad_norm": 1.8708374500274658,
+      "learning_rate": 0.00019543468737625766,
+      "loss": 0.2478,
+      "step": 2270
+    },
+    {
+      "epoch": 48.0,
+      "grad_norm": 1.1603832244873047,
+      "learning_rate": 0.00019539467038228456,
+      "loss": 0.2447,
+      "step": 2280
+    },
+    {
+      "epoch": 48.21052631578947,
+      "grad_norm": 1.687369704246521,
+      "learning_rate": 0.00019535448290256796,
+      "loss": 0.2228,
+      "step": 2290
+    },
+    {
+      "epoch": 48.421052631578945,
+      "grad_norm": 1.7475471496582031,
+      "learning_rate": 0.00019531412500892943,
+      "loss": 0.2424,
+      "step": 2300
+    },
+    {
+      "epoch": 48.63157894736842,
+      "grad_norm": 1.5487775802612305,
+      "learning_rate": 0.00019527359677349505,
+      "loss": 0.2447,
+      "step": 2310
+    },
+    {
+      "epoch": 48.8421052631579,
+      "grad_norm": 1.3614391088485718,
+      "learning_rate": 0.00019523289826869534,
+      "loss": 0.2548,
+      "step": 2320
+    },
+    {
+      "epoch": 49.05263157894737,
+      "grad_norm": 1.19809091091156,
+      "learning_rate": 0.00019519202956726512,
+      "loss": 0.2374,
+      "step": 2330
+    },
+    {
+      "epoch": 49.26315789473684,
+      "grad_norm": 1.5943597555160522,
+      "learning_rate": 0.00019515099074224343,
+      "loss": 0.2411,
+      "step": 2340
+    },
+    {
+      "epoch": 49.473684210526315,
+      "grad_norm": 1.6664204597473145,
+      "learning_rate": 0.0001951097818669733,
+      "loss": 0.2429,
+      "step": 2350
+    },
+    {
+      "epoch": 49.68421052631579,
+      "grad_norm": 1.5013232231140137,
+      "learning_rate": 0.00019506840301510158,
+      "loss": 0.2406,
+      "step": 2360
+    },
+    {
+      "epoch": 49.89473684210526,
+      "grad_norm": 1.2071912288665771,
+      "learning_rate": 0.0001950268542605791,
+      "loss": 0.251,
+      "step": 2370
+    },
+    {
+      "epoch": 50.10526315789474,
+      "grad_norm": 1.1436349153518677,
+      "learning_rate": 0.00019498513567766016,
+      "loss": 0.2405,
+      "step": 2380
+    },
+    {
+      "epoch": 50.31578947368421,
+      "grad_norm": 1.7050838470458984,
+      "learning_rate": 0.00019494324734090266,
+      "loss": 0.229,
+      "step": 2390
+    },
+    {
+      "epoch": 50.526315789473685,
+      "grad_norm": 1.4147546291351318,
+      "learning_rate": 0.00019490118932516786,
+      "loss": 0.2456,
+      "step": 2400
+    },
+    {
+      "epoch": 50.73684210526316,
+      "grad_norm": 1.4983457326889038,
+      "learning_rate": 0.00019485896170562018,
+      "loss": 0.2424,
+      "step": 2410
+    },
+    {
+      "epoch": 50.94736842105263,
+      "grad_norm": 1.6988271474838257,
+      "learning_rate": 0.00019481656455772734,
+      "loss": 0.2476,
+      "step": 2420
+    },
+    {
+      "epoch": 51.1578947368421,
+      "grad_norm": 1.4890905618667603,
+      "learning_rate": 0.00019477399795725983,
+      "loss": 0.2414,
+      "step": 2430
+    },
+    {
+      "epoch": 51.36842105263158,
+      "grad_norm": 1.7915074825286865,
+      "learning_rate": 0.00019473126198029105,
+      "loss": 0.2407,
+      "step": 2440
+    },
+    {
+      "epoch": 51.578947368421055,
+      "grad_norm": 1.7061022520065308,
+      "learning_rate": 0.0001946883567031972,
+      "loss": 0.2362,
+      "step": 2450
+    },
+    {
+      "epoch": 51.78947368421053,
+      "grad_norm": 2.6335580348968506,
+      "learning_rate": 0.00019464528220265693,
+      "loss": 0.2481,
+      "step": 2460
+    },
+    {
+      "epoch": 52.0,
+      "grad_norm": 1.3824809789657593,
+      "learning_rate": 0.00019460203855565134,
+      "loss": 0.2303,
+      "step": 2470
+    },
+    {
+      "epoch": 52.21052631578947,
+      "grad_norm": 1.4830249547958374,
+      "learning_rate": 0.00019455862583946386,
+      "loss": 0.216,
+      "step": 2480
+    },
+    {
+      "epoch": 52.421052631578945,
+      "grad_norm": 1.3665930032730103,
+      "learning_rate": 0.00019451504413168003,
+      "loss": 0.2372,
+      "step": 2490
+    },
+    {
+      "epoch": 52.63157894736842,
+      "grad_norm": 1.4474061727523804,
+      "learning_rate": 0.00019447129351018744,
+      "loss": 0.2359,
+      "step": 2500
+    },
+    {
+      "epoch": 52.8421052631579,
+      "grad_norm": 1.353022813796997,
+      "learning_rate": 0.00019442737405317556,
+      "loss": 0.2523,
+      "step": 2510
+    },
+    {
+      "epoch": 53.05263157894737,
+      "grad_norm": 1.4885058403015137,
+      "learning_rate": 0.00019438328583913558,
+      "loss": 0.2276,
+      "step": 2520
+    },
+    {
+      "epoch": 53.26315789473684,
+      "grad_norm": 1.1576405763626099,
+      "learning_rate": 0.00019433902894686026,
+      "loss": 0.2277,
+      "step": 2530
+    },
+    {
+      "epoch": 53.473684210526315,
+      "grad_norm": 1.4939807653427124,
+      "learning_rate": 0.0001942946034554439,
+      "loss": 0.2312,
+      "step": 2540
+    },
+    {
+      "epoch": 53.68421052631579,
+      "grad_norm": 1.3100343942642212,
+      "learning_rate": 0.00019425000944428198,
+      "loss": 0.242,
+      "step": 2550
+    },
+    {
+      "epoch": 53.89473684210526,
+      "grad_norm": 1.3382755517959595,
+      "learning_rate": 0.00019420524699307126,
+      "loss": 0.2554,
+      "step": 2560
+    },
+    {
+      "epoch": 54.10526315789474,
+      "grad_norm": 1.488303542137146,
+      "learning_rate": 0.0001941603161818095,
+      "loss": 0.2217,
+      "step": 2570
+    },
+    {
+      "epoch": 54.31578947368421,
+      "grad_norm": 1.3764780759811401,
+      "learning_rate": 0.00019411521709079534,
+      "loss": 0.2352,
+      "step": 2580
+    },
+    {
+      "epoch": 54.526315789473685,
+      "grad_norm": 1.777108073234558,
+      "learning_rate": 0.0001940699498006282,
+      "loss": 0.2366,
+      "step": 2590
+    },
+    {
+      "epoch": 54.73684210526316,
+      "grad_norm": 1.7822355031967163,
+      "learning_rate": 0.00019402451439220803,
+      "loss": 0.2348,
+      "step": 2600
+    },
+    {
+      "epoch": 54.94736842105263,
+      "grad_norm": 1.7189970016479492,
+      "learning_rate": 0.00019397891094673529,
+      "loss": 0.2439,
+      "step": 2610
+    },
+    {
+      "epoch": 55.1578947368421,
+      "grad_norm": 1.3337494134902954,
+      "learning_rate": 0.00019393313954571074,
+      "loss": 0.2216,
+      "step": 2620
+    },
+    {
+      "epoch": 55.36842105263158,
+      "grad_norm": 1.533355951309204,
+      "learning_rate": 0.00019388720027093523,
+      "loss": 0.2194,
+      "step": 2630
+    },
+    {
+      "epoch": 55.578947368421055,
+      "grad_norm": 1.2764497995376587,
+      "learning_rate": 0.00019384109320450977,
+      "loss": 0.2517,
+      "step": 2640
+    },
+    {
+      "epoch": 55.78947368421053,
+      "grad_norm": 1.2663401365280151,
+      "learning_rate": 0.00019379481842883518,
+      "loss": 0.2331,
+      "step": 2650
+    },
+    {
+      "epoch": 56.0,
+      "grad_norm": 1.443335771560669,
+      "learning_rate": 0.00019374837602661188,
+      "loss": 0.2451,
+      "step": 2660
+    },
+    {
+      "epoch": 56.21052631578947,
+      "grad_norm": 1.2762813568115234,
+      "learning_rate": 0.00019370176608084008,
+      "loss": 0.2138,
+      "step": 2670
+    },
+    {
+      "epoch": 56.421052631578945,
+      "grad_norm": 1.145351767539978,
+      "learning_rate": 0.00019365498867481923,
+      "loss": 0.2269,
+      "step": 2680
+    },
+    {
+      "epoch": 56.63157894736842,
+      "grad_norm": 1.6088831424713135,
+      "learning_rate": 0.00019360804389214822,
+      "loss": 0.2383,
+      "step": 2690
+    },
+    {
+      "epoch": 56.8421052631579,
+      "grad_norm": 1.5034024715423584,
+      "learning_rate": 0.0001935609318167249,
+      "loss": 0.2433,
+      "step": 2700
+    },
+    {
+      "epoch": 57.05263157894737,
+      "grad_norm": 1.2446277141571045,
+      "learning_rate": 0.00019351365253274626,
+      "loss": 0.2329,
+      "step": 2710
+    },
+    {
+      "epoch": 57.26315789473684,
+      "grad_norm": 1.3016337156295776,
+      "learning_rate": 0.00019346620612470803,
+      "loss": 0.2334,
+      "step": 2720
+    },
+    {
+      "epoch": 57.473684210526315,
+      "grad_norm": 1.444482684135437,
+      "learning_rate": 0.0001934185926774046,
+      "loss": 0.2056,
+      "step": 2730
+    },
+    {
+      "epoch": 57.68421052631579,
+      "grad_norm": 1.6010749340057373,
+      "learning_rate": 0.00019337081227592897,
+      "loss": 0.2506,
+      "step": 2740
+    },
+    {
+      "epoch": 57.89473684210526,
+      "grad_norm": 2.1765823364257812,
+      "learning_rate": 0.0001933228650056724,
+      "loss": 0.2421,
+      "step": 2750
+    },
+    {
+      "epoch": 58.10526315789474,
+      "grad_norm": 1.2616775035858154,
+      "learning_rate": 0.0001932747509523245,
+      "loss": 0.2308,
+      "step": 2760
+    },
+    {
+      "epoch": 58.31578947368421,
+      "grad_norm": 1.6567051410675049,
+      "learning_rate": 0.00019322647020187286,
+      "loss": 0.2325,
+      "step": 2770
+    },
+    {
+      "epoch": 58.526315789473685,
+      "grad_norm": 1.4858084917068481,
+      "learning_rate": 0.000193178022840603,
+      "loss": 0.2286,
+      "step": 2780
+    },
+    {
+      "epoch": 58.73684210526316,
+      "grad_norm": 1.2015066146850586,
+      "learning_rate": 0.00019312940895509822,
+      "loss": 0.2314,
+      "step": 2790
+    },
+    {
+      "epoch": 58.94736842105263,
+      "grad_norm": 1.4122145175933838,
+      "learning_rate": 0.00019308062863223943,
+      "loss": 0.235,
+      "step": 2800
+    },
+    {
+      "epoch": 59.1578947368421,
+      "grad_norm": 1.3632580041885376,
+      "learning_rate": 0.00019303168195920493,
+      "loss": 0.2159,
+      "step": 2810
+    },
+    {
+      "epoch": 59.36842105263158,
+      "grad_norm": 1.4167340993881226,
+      "learning_rate": 0.00019298256902347042,
+      "loss": 0.2385,
+      "step": 2820
+    },
+    {
+      "epoch": 59.578947368421055,
+      "grad_norm": 1.299408197402954,
+      "learning_rate": 0.00019293328991280863,
+      "loss": 0.2438,
+      "step": 2830
+    },
+    {
+      "epoch": 59.78947368421053,
+      "grad_norm": 1.329671025276184,
+      "learning_rate": 0.00019288384471528936,
+      "loss": 0.2228,
+      "step": 2840
+    },
+    {
+      "epoch": 60.0,
+      "grad_norm": 1.7981964349746704,
+      "learning_rate": 0.00019283423351927918,
+      "loss": 0.2182,
+      "step": 2850
+    },
+    {
+      "epoch": 60.21052631578947,
+      "grad_norm": 1.8409744501113892,
+      "learning_rate": 0.00019278445641344135,
+      "loss": 0.2194,
+      "step": 2860
+    },
+    {
+      "epoch": 60.421052631578945,
+      "grad_norm": 1.713848352432251,
+      "learning_rate": 0.0001927345134867356,
+      "loss": 0.2307,
+      "step": 2870
+    },
+    {
+      "epoch": 60.63157894736842,
+      "grad_norm": 1.1938320398330688,
+      "learning_rate": 0.00019268440482841804,
+      "loss": 0.2272,
+      "step": 2880
+    },
+    {
+      "epoch": 60.8421052631579,
+      "grad_norm": 1.3815594911575317,
+      "learning_rate": 0.000192634130528041,
+      "loss": 0.2379,
+      "step": 2890
+    },
+    {
+      "epoch": 61.05263157894737,
+      "grad_norm": 1.0334292650222778,
+      "learning_rate": 0.00019258369067545278,
+      "loss": 0.2196,
+      "step": 2900
+    },
+    {
+      "epoch": 61.26315789473684,
+      "grad_norm": 1.033290147781372,
+      "learning_rate": 0.0001925330853607976,
+      "loss": 0.2215,
+      "step": 2910
+    },
+    {
+      "epoch": 61.473684210526315,
+      "grad_norm": 1.375057339668274,
+      "learning_rate": 0.00019248231467451534,
+      "loss": 0.2373,
+      "step": 2920
+    },
+    {
+      "epoch": 61.68421052631579,
+      "grad_norm": 1.3225274085998535,
+      "learning_rate": 0.00019243137870734146,
+      "loss": 0.2256,
+      "step": 2930
+    },
+    {
+      "epoch": 61.89473684210526,
+      "grad_norm": 1.3929612636566162,
+      "learning_rate": 0.0001923802775503068,
+      "loss": 0.233,
+      "step": 2940
+    },
+    {
+      "epoch": 62.10526315789474,
+      "grad_norm": 1.6330879926681519,
+      "learning_rate": 0.00019232901129473734,
+      "loss": 0.2181,
+      "step": 2950
+    },
+    {
+      "epoch": 62.31578947368421,
+      "grad_norm": 1.1357747316360474,
+      "learning_rate": 0.0001922775800322543,
+      "loss": 0.2196,
+      "step": 2960
+    },
+    {
+      "epoch": 62.526315789473685,
+      "grad_norm": 1.415575623512268,
+      "learning_rate": 0.00019222598385477366,
+      "loss": 0.2299,
+      "step": 2970
+    },
+    {
+      "epoch": 62.73684210526316,
+      "grad_norm": 1.2679227590560913,
+      "learning_rate": 0.00019217422285450607,
+      "loss": 0.2175,
+      "step": 2980
+    },
+    {
+      "epoch": 62.94736842105263,
+      "grad_norm": 1.4680553674697876,
+      "learning_rate": 0.00019212229712395695,
+      "loss": 0.2357,
+      "step": 2990
+    },
+    {
+      "epoch": 63.1578947368421,
+      "grad_norm": 1.255954623222351,
+      "learning_rate": 0.00019207020675592593,
+      "loss": 0.2195,
+      "step": 3000
+    },
+    {
+      "epoch": 63.36842105263158,
+      "grad_norm": 1.352297306060791,
+      "learning_rate": 0.00019201795184350693,
+      "loss": 0.2203,
+      "step": 3010
+    },
+    {
+      "epoch": 63.578947368421055,
+      "grad_norm": 1.590889573097229,
+      "learning_rate": 0.000191965532480088,
+      "loss": 0.2299,
+      "step": 3020
+    },
+    {
+      "epoch": 63.78947368421053,
+      "grad_norm": 1.265857458114624,
+      "learning_rate": 0.00019191294875935103,
+      "loss": 0.2343,
+      "step": 3030
+    },
+    {
+      "epoch": 64.0,
+      "grad_norm": 1.2470191717147827,
+      "learning_rate": 0.00019186020077527162,
+      "loss": 0.2238,
+      "step": 3040
+    },
+    {
+      "epoch": 64.21052631578948,
+      "grad_norm": 1.4418771266937256,
+      "learning_rate": 0.000191807288622119,
+      "loss": 0.2242,
+      "step": 3050
+    },
+    {
+      "epoch": 64.42105263157895,
+      "grad_norm": 1.957590103149414,
+      "learning_rate": 0.00019175421239445576,
+      "loss": 0.2332,
+      "step": 3060
+    },
+    {
+      "epoch": 64.63157894736842,
+      "grad_norm": 1.2326531410217285,
+      "learning_rate": 0.00019170097218713773,
+      "loss": 0.2238,
+      "step": 3070
+    },
+    {
+      "epoch": 64.84210526315789,
+      "grad_norm": 1.2268998622894287,
+      "learning_rate": 0.0001916475680953138,
+      "loss": 0.2154,
+      "step": 3080
+    },
+    {
+      "epoch": 65.05263157894737,
+      "grad_norm": 1.0560153722763062,
+      "learning_rate": 0.0001915940002144257,
+      "loss": 0.2253,
+      "step": 3090
+    },
+    {
+      "epoch": 65.26315789473684,
+      "grad_norm": 1.13203763961792,
+      "learning_rate": 0.00019154026864020798,
+      "loss": 0.2109,
+      "step": 3100
+    },
+    {
+      "epoch": 65.47368421052632,
+      "grad_norm": 1.2309409379959106,
+      "learning_rate": 0.00019148637346868763,
+      "loss": 0.2148,
+      "step": 3110
+    },
+    {
+      "epoch": 65.6842105263158,
+      "grad_norm": 1.1464983224868774,
+      "learning_rate": 0.00019143231479618405,
+      "loss": 0.224,
+      "step": 3120
+    },
+    {
+      "epoch": 65.89473684210526,
+      "grad_norm": 0.9732810258865356,
+      "learning_rate": 0.00019137809271930893,
+      "loss": 0.2306,
+      "step": 3130
+    },
+    {
+      "epoch": 66.10526315789474,
+      "grad_norm": 1.1428371667861938,
+      "learning_rate": 0.00019132370733496582,
+      "loss": 0.2163,
+      "step": 3140
+    },
+    {
+      "epoch": 66.3157894736842,
+      "grad_norm": 1.4518449306488037,
+      "learning_rate": 0.0001912691587403503,
+      "loss": 0.2374,
+      "step": 3150
+    },
+    {
+      "epoch": 66.52631578947368,
+      "grad_norm": 1.1604626178741455,
+      "learning_rate": 0.0001912144470329495,
+      "loss": 0.2398,
+      "step": 3160
+    },
+    {
+      "epoch": 66.73684210526316,
+      "grad_norm": 1.2528883218765259,
+      "learning_rate": 0.0001911595723105421,
+      "loss": 0.2231,
+      "step": 3170
+    },
+    {
+      "epoch": 66.94736842105263,
+      "grad_norm": 1.35756516456604,
+      "learning_rate": 0.00019110453467119815,
+      "loss": 0.2136,
+      "step": 3180
+    },
+    {
+      "epoch": 67.15789473684211,
+      "grad_norm": 1.3628764152526855,
+      "learning_rate": 0.00019104933421327887,
+      "loss": 0.2084,
+      "step": 3190
+    },
+    {
+      "epoch": 67.36842105263158,
+      "grad_norm": 0.9904689192771912,
+      "learning_rate": 0.0001909939710354364,
+      "loss": 0.2278,
+      "step": 3200
+    },
+    {
+      "epoch": 67.57894736842105,
+      "grad_norm": 1.4833720922470093,
+      "learning_rate": 0.00019093844523661367,
+      "loss": 0.2396,
+      "step": 3210
+    },
+    {
+      "epoch": 67.78947368421052,
+      "grad_norm": 2.6204235553741455,
+      "learning_rate": 0.00019088275691604435,
+      "loss": 0.2349,
+      "step": 3220
+    },
+    {
+      "epoch": 68.0,
+      "grad_norm": 1.6207849979400635,
+      "learning_rate": 0.00019082690617325246,
+      "loss": 0.2195,
+      "step": 3230
+    },
+    {
+      "epoch": 68.21052631578948,
+      "grad_norm": 1.9536974430084229,
+      "learning_rate": 0.00019077089310805238,
+      "loss": 0.2093,
+      "step": 3240
+    },
+    {
+      "epoch": 68.42105263157895,
+      "grad_norm": 1.4202009439468384,
+      "learning_rate": 0.00019071471782054853,
+      "loss": 0.2127,
+      "step": 3250
+    },
+    {
+      "epoch": 68.63157894736842,
+      "grad_norm": 1.3091250658035278,
+      "learning_rate": 0.00019065838041113517,
+      "loss": 0.223,
+      "step": 3260
+    },
+    {
+      "epoch": 68.84210526315789,
+      "grad_norm": 1.2577513456344604,
+      "learning_rate": 0.0001906018809804965,
+      "loss": 0.2396,
+      "step": 3270
+    },
+    {
+      "epoch": 69.05263157894737,
+      "grad_norm": 0.9996993541717529,
+      "learning_rate": 0.0001905452196296061,
+      "loss": 0.2333,
+      "step": 3280
+    },
+    {
+      "epoch": 69.26315789473684,
+      "grad_norm": 1.3234609365463257,
+      "learning_rate": 0.000190488396459727,
+      "loss": 0.2289,
+      "step": 3290
+    },
+    {
+      "epoch": 69.47368421052632,
+      "grad_norm": 1.3428514003753662,
+      "learning_rate": 0.00019043141157241143,
+      "loss": 0.2127,
+      "step": 3300
+    },
+    {
+      "epoch": 69.6842105263158,
+      "grad_norm": 1.731479287147522,
+      "learning_rate": 0.0001903742650695006,
+      "loss": 0.2224,
+      "step": 3310
+    },
+    {
+      "epoch": 69.89473684210526,
+      "grad_norm": 0.9988399147987366,
+      "learning_rate": 0.0001903169570531246,
+      "loss": 0.216,
+      "step": 3320
+    },
+    {
+      "epoch": 70.10526315789474,
+      "grad_norm": 0.9177957773208618,
+      "learning_rate": 0.00019025948762570216,
+      "loss": 0.2024,
+      "step": 3330
+    },
+    {
+      "epoch": 70.3157894736842,
+      "grad_norm": 1.2759298086166382,
+      "learning_rate": 0.00019020185688994046,
+      "loss": 0.2181,
+      "step": 3340
+    },
+    {
+      "epoch": 70.52631578947368,
+      "grad_norm": 0.9687845706939697,
+      "learning_rate": 0.0001901440649488349,
+      "loss": 0.2183,
+      "step": 3350
+    },
+    {
+      "epoch": 70.73684210526316,
+      "grad_norm": 1.1401742696762085,
+      "learning_rate": 0.00019008611190566918,
+      "loss": 0.2339,
+      "step": 3360
+    },
+    {
+      "epoch": 70.94736842105263,
+      "grad_norm": 1.353368878364563,
+      "learning_rate": 0.00019002799786401462,
+      "loss": 0.2372,
+      "step": 3370
+    },
+    {
+      "epoch": 71.15789473684211,
+      "grad_norm": 1.0063115358352661,
+      "learning_rate": 0.00018996972292773057,
+      "loss": 0.2327,
+      "step": 3380
+    },
+    {
+      "epoch": 71.36842105263158,
+      "grad_norm": 1.4881367683410645,
+      "learning_rate": 0.00018991128720096377,
+      "loss": 0.212,
+      "step": 3390
+    },
+    {
+      "epoch": 71.57894736842105,
+      "grad_norm": 1.4579118490219116,
+      "learning_rate": 0.00018985269078814827,
+      "loss": 0.2224,
+      "step": 3400
+    },
+    {
+      "epoch": 71.78947368421052,
+      "grad_norm": 0.9688888192176819,
+      "learning_rate": 0.00018979393379400542,
+      "loss": 0.2216,
+      "step": 3410
+    },
+    {
+      "epoch": 72.0,
+      "grad_norm": 1.7144838571548462,
+      "learning_rate": 0.0001897350163235435,
+      "loss": 0.2312,
+      "step": 3420
+    },
+    {
+      "epoch": 72.21052631578948,
+      "grad_norm": 0.9466618299484253,
+      "learning_rate": 0.00018967593848205754,
+      "loss": 0.21,
+      "step": 3430
+    },
+    {
+      "epoch": 72.42105263157895,
+      "grad_norm": 1.3977335691452026,
+      "learning_rate": 0.00018961670037512924,
+      "loss": 0.2285,
+      "step": 3440
+    },
+    {
+      "epoch": 72.63157894736842,
+      "grad_norm": 1.2872693538665771,
+      "learning_rate": 0.0001895573021086267,
+      "loss": 0.2152,
+      "step": 3450
+    },
+    {
+      "epoch": 72.84210526315789,
+      "grad_norm": 0.9000837802886963,
+      "learning_rate": 0.00018949774378870427,
+      "loss": 0.2184,
+      "step": 3460
+    },
+    {
+      "epoch": 73.05263157894737,
+      "grad_norm": 1.1097438335418701,
+      "learning_rate": 0.0001894380255218023,
+      "loss": 0.2178,
+      "step": 3470
+    },
+    {
+      "epoch": 73.26315789473684,
+      "grad_norm": 1.0013376474380493,
+      "learning_rate": 0.0001893781474146471,
+      "loss": 0.2052,
+      "step": 3480
+    },
+    {
+      "epoch": 73.47368421052632,
+      "grad_norm": 0.9624537229537964,
+      "learning_rate": 0.0001893181095742504,
+      "loss": 0.2118,
+      "step": 3490
+    },
+    {
+      "epoch": 73.6842105263158,
+      "grad_norm": 0.8211550712585449,
+      "learning_rate": 0.00018925791210790968,
+      "loss": 0.2262,
+      "step": 3500
+    },
+    {
+      "epoch": 73.89473684210526,
+      "grad_norm": 0.9911038279533386,
+      "learning_rate": 0.00018919755512320752,
+      "loss": 0.2336,
+      "step": 3510
+    },
+    {
+      "epoch": 74.10526315789474,
+      "grad_norm": 1.0063726902008057,
+      "learning_rate": 0.00018913703872801166,
+      "loss": 0.2104,
+      "step": 3520
+    },
+    {
+      "epoch": 74.3157894736842,
+      "grad_norm": 1.1858199834823608,
+      "learning_rate": 0.00018907636303047468,
+      "loss": 0.2155,
+      "step": 3530
+    },
+    {
+      "epoch": 74.52631578947368,
+      "grad_norm": 0.9891308546066284,
+      "learning_rate": 0.00018901552813903395,
+      "loss": 0.2059,
+      "step": 3540
+    },
+    {
+      "epoch": 74.73684210526316,
+      "grad_norm": 1.5848833322525024,
+      "learning_rate": 0.0001889545341624112,
+      "loss": 0.2288,
+      "step": 3550
+    },
+    {
+      "epoch": 74.94736842105263,
+      "grad_norm": 0.734965980052948,
+      "learning_rate": 0.00018889338120961266,
+      "loss": 0.2202,
+      "step": 3560
+    },
+    {
+      "epoch": 75.15789473684211,
+      "grad_norm": 0.9111061692237854,
+      "learning_rate": 0.0001888320693899285,
+      "loss": 0.2135,
+      "step": 3570
+    },
+    {
+      "epoch": 75.36842105263158,
+      "grad_norm": 0.923260509967804,
+      "learning_rate": 0.00018877059881293288,
+      "loss": 0.2089,
+      "step": 3580
+    },
+    {
+      "epoch": 75.57894736842105,
+      "grad_norm": 1.0902652740478516,
+      "learning_rate": 0.00018870896958848372,
+      "loss": 0.226,
+      "step": 3590
+    },
+    {
+      "epoch": 75.78947368421052,
+      "grad_norm": 1.3361833095550537,
+      "learning_rate": 0.00018864718182672242,
+      "loss": 0.2156,
+      "step": 3600
+    },
+    {
+      "epoch": 76.0,
+      "grad_norm": 0.9090056419372559,
+      "learning_rate": 0.00018858523563807373,
+      "loss": 0.2344,
+      "step": 3610
+    },
+    {
+      "epoch": 76.21052631578948,
+      "grad_norm": 1.0681171417236328,
+      "learning_rate": 0.00018852313113324552,
+      "loss": 0.2057,
+      "step": 3620
+    },
+    {
+      "epoch": 76.42105263157895,
+      "grad_norm": 1.2302827835083008,
+      "learning_rate": 0.00018846086842322864,
+      "loss": 0.2168,
+      "step": 3630
+    },
+    {
+      "epoch": 76.63157894736842,
+      "grad_norm": 1.1854231357574463,
+      "learning_rate": 0.00018839844761929663,
+      "loss": 0.2093,
+      "step": 3640
+    },
+    {
+      "epoch": 76.84210526315789,
+      "grad_norm": 1.017804741859436,
+      "learning_rate": 0.0001883358688330056,
+      "loss": 0.2088,
+      "step": 3650
+    },
+    {
+      "epoch": 77.05263157894737,
+      "grad_norm": 0.9746867418289185,
+      "learning_rate": 0.00018827313217619399,
+      "loss": 0.2297,
+      "step": 3660
+    },
+    {
+      "epoch": 77.26315789473684,
+      "grad_norm": 1.3612266778945923,
+      "learning_rate": 0.00018821023776098233,
+      "loss": 0.2159,
+      "step": 3670
+    },
+    {
+      "epoch": 77.47368421052632,
+      "grad_norm": 1.2489289045333862,
+      "learning_rate": 0.0001881471856997732,
+      "loss": 0.2214,
+      "step": 3680
+    },
+    {
+      "epoch": 77.6842105263158,
+      "grad_norm": 0.9268271923065186,
+      "learning_rate": 0.00018808397610525085,
+      "loss": 0.2199,
+      "step": 3690
+    },
+    {
+      "epoch": 77.89473684210526,
+      "grad_norm": 1.429926872253418,
+      "learning_rate": 0.00018802060909038103,
+      "loss": 0.2153,
+      "step": 3700
+    },
+    {
+      "epoch": 78.10526315789474,
+      "grad_norm": 1.1016651391983032,
+      "learning_rate": 0.0001879570847684109,
+      "loss": 0.2119,
+      "step": 3710
+    },
+    {
+      "epoch": 78.3157894736842,
+      "grad_norm": 0.8588700294494629,
+      "learning_rate": 0.00018789340325286872,
+      "loss": 0.2107,
+      "step": 3720
+    },
+    {
+      "epoch": 78.52631578947368,
+      "grad_norm": 1.3524593114852905,
+      "learning_rate": 0.00018782956465756366,
+      "loss": 0.2245,
+      "step": 3730
+    },
+    {
+      "epoch": 78.73684210526316,
+      "grad_norm": 1.5214595794677734,
+      "learning_rate": 0.0001877655690965857,
+      "loss": 0.2131,
+      "step": 3740
+    },
+    {
+      "epoch": 78.94736842105263,
+      "grad_norm": 1.0766842365264893,
+      "learning_rate": 0.00018770141668430522,
+      "loss": 0.2288,
+      "step": 3750
+    },
+    {
+      "epoch": 79.15789473684211,
+      "grad_norm": 1.0418442487716675,
+      "learning_rate": 0.00018763710753537301,
+      "loss": 0.22,
+      "step": 3760
+    },
+    {
+      "epoch": 79.36842105263158,
+      "grad_norm": 1.1370205879211426,
+      "learning_rate": 0.00018757264176471998,
+      "loss": 0.2124,
+      "step": 3770
+    },
+    {
+      "epoch": 79.57894736842105,
+      "grad_norm": 0.9709566235542297,
+      "learning_rate": 0.00018750801948755685,
+      "loss": 0.2032,
+      "step": 3780
+    },
+    {
+      "epoch": 79.78947368421052,
+      "grad_norm": 1.1694402694702148,
+      "learning_rate": 0.00018744324081937415,
+      "loss": 0.2244,
+      "step": 3790
+    },
+    {
+      "epoch": 80.0,
+      "grad_norm": 1.1555752754211426,
+      "learning_rate": 0.00018737830587594185,
+      "loss": 0.2204,
+      "step": 3800
+    },
+    {
+      "epoch": 80.21052631578948,
+      "grad_norm": 0.89194655418396,
+      "learning_rate": 0.0001873132147733092,
+      "loss": 0.2024,
+      "step": 3810
+    },
+    {
+      "epoch": 80.42105263157895,
+      "grad_norm": 3.491398811340332,
+      "learning_rate": 0.00018724796762780463,
+      "loss": 0.2363,
+      "step": 3820
+    },
+    {
+      "epoch": 80.63157894736842,
+      "grad_norm": 0.9344122409820557,
+      "learning_rate": 0.00018718256455603526,
+      "loss": 0.2222,
+      "step": 3830
+    },
+    {
+      "epoch": 80.84210526315789,
+      "grad_norm": 1.2333403825759888,
+      "learning_rate": 0.0001871170056748871,
+      "loss": 0.2085,
+      "step": 3840
+    },
+    {
+      "epoch": 81.05263157894737,
+      "grad_norm": 0.8288215398788452,
+      "learning_rate": 0.0001870512911015244,
+      "loss": 0.2197,
+      "step": 3850
+    },
+    {
+      "epoch": 81.26315789473684,
+      "grad_norm": 1.2045210599899292,
+      "learning_rate": 0.00018698542095338982,
+      "loss": 0.2155,
+      "step": 3860
+    },
+    {
+      "epoch": 81.47368421052632,
+      "grad_norm": 1.1142665147781372,
+      "learning_rate": 0.00018691939534820398,
+      "loss": 0.2156,
+      "step": 3870
+    },
+    {
+      "epoch": 81.6842105263158,
+      "grad_norm": 0.994482159614563,
+      "learning_rate": 0.0001868532144039653,
+      "loss": 0.2075,
+      "step": 3880
+    },
+    {
+      "epoch": 81.89473684210526,
+      "grad_norm": 1.2781574726104736,
+      "learning_rate": 0.00018678687823894992,
+      "loss": 0.2203,
+      "step": 3890
+    },
+    {
+      "epoch": 82.10526315789474,
+      "grad_norm": 1.054432988166809,
+      "learning_rate": 0.0001867203869717113,
+      "loss": 0.2248,
+      "step": 3900
+    },
+    {
+      "epoch": 82.3157894736842,
+      "grad_norm": 1.141908049583435,
+      "learning_rate": 0.00018665374072108012,
+      "loss": 0.2061,
+      "step": 3910
+    },
+    {
+      "epoch": 82.52631578947368,
+      "grad_norm": 1.210333228111267,
+      "learning_rate": 0.000186586939606164,
+      "loss": 0.2082,
+      "step": 3920
+    },
+    {
+      "epoch": 82.73684210526316,
+      "grad_norm": 1.2293156385421753,
+      "learning_rate": 0.00018651998374634744,
+      "loss": 0.2153,
+      "step": 3930
+    },
+    {
+      "epoch": 82.94736842105263,
+      "grad_norm": 0.8742419481277466,
+      "learning_rate": 0.0001864528732612913,
+      "loss": 0.2146,
+      "step": 3940
+    },
+    {
+      "epoch": 83.15789473684211,
+      "grad_norm": 0.9708311557769775,
+      "learning_rate": 0.00018638560827093304,
+      "loss": 0.2171,
+      "step": 3950
+    },
+    {
+      "epoch": 83.36842105263158,
+      "grad_norm": 1.1058762073516846,
+      "learning_rate": 0.000186318188895486,
+      "loss": 0.2209,
+      "step": 3960
+    },
+    {
+      "epoch": 83.57894736842105,
+      "grad_norm": 1.1592589616775513,
+      "learning_rate": 0.00018625061525543956,
+      "loss": 0.2162,
+      "step": 3970
+    },
+    {
+      "epoch": 83.78947368421052,
+      "grad_norm": 1.1639076471328735,
+      "learning_rate": 0.00018618288747155882,
+      "loss": 0.2217,
+      "step": 3980
+    },
+    {
+      "epoch": 84.0,
+      "grad_norm": 0.9892878532409668,
+      "learning_rate": 0.00018611500566488421,
+      "loss": 0.2165,
+      "step": 3990
+    },
+    {
+      "epoch": 84.21052631578948,
+      "grad_norm": 1.1064993143081665,
+      "learning_rate": 0.00018604696995673162,
+      "loss": 0.2023,
+      "step": 4000
+    },
+    {
+      "epoch": 84.42105263157895,
+      "grad_norm": 1.0864388942718506,
+      "learning_rate": 0.00018597878046869187,
+      "loss": 0.2118,
+      "step": 4010
+    },
+    {
+      "epoch": 84.63157894736842,
+      "grad_norm": 1.1595569849014282,
+      "learning_rate": 0.0001859104373226306,
+      "loss": 0.215,
+      "step": 4020
+    },
+    {
+      "epoch": 84.84210526315789,
+      "grad_norm": 0.945059061050415,
+      "learning_rate": 0.00018584194064068813,
+      "loss": 0.2203,
+      "step": 4030
+    },
+    {
+      "epoch": 85.05263157894737,
+      "grad_norm": 0.8729904890060425,
+      "learning_rate": 0.0001857732905452791,
+      "loss": 0.2077,
+      "step": 4040
+    },
+    {
+      "epoch": 85.26315789473684,
+      "grad_norm": 1.315699815750122,
+      "learning_rate": 0.0001857044871590924,
+      "loss": 0.2098,
+      "step": 4050
+    },
+    {
+      "epoch": 85.47368421052632,
+      "grad_norm": 0.8844873309135437,
+      "learning_rate": 0.0001856355306050908,
+      "loss": 0.2146,
+      "step": 4060
+    },
+    {
+      "epoch": 85.6842105263158,
+      "grad_norm": 1.0418226718902588,
+      "learning_rate": 0.00018556642100651087,
+      "loss": 0.2182,
+      "step": 4070
+    },
+    {
+      "epoch": 85.89473684210526,
+      "grad_norm": 0.7228378653526306,
+      "learning_rate": 0.00018549715848686267,
+      "loss": 0.2144,
+      "step": 4080
+    },
+    {
+      "epoch": 86.10526315789474,
+      "grad_norm": 0.8830803632736206,
+      "learning_rate": 0.0001854277431699295,
+      "loss": 0.1954,
+      "step": 4090
+    },
+    {
+      "epoch": 86.3157894736842,
+      "grad_norm": 0.7646278738975525,
+      "learning_rate": 0.0001853581751797679,
+      "loss": 0.1989,
+      "step": 4100
+    },
+    {
+      "epoch": 86.52631578947368,
+      "grad_norm": 1.2258400917053223,
+      "learning_rate": 0.00018528845464070703,
+      "loss": 0.2189,
+      "step": 4110
+    },
+    {
+      "epoch": 86.73684210526316,
+      "grad_norm": 0.8631286025047302,
+      "learning_rate": 0.00018521858167734885,
+      "loss": 0.2215,
+      "step": 4120
+    },
+    {
+      "epoch": 86.94736842105263,
+      "grad_norm": 1.0356979370117188,
+      "learning_rate": 0.00018514855641456768,
+      "loss": 0.2294,
+      "step": 4130
+    },
+    {
+      "epoch": 87.15789473684211,
+      "grad_norm": 1.2814905643463135,
+      "learning_rate": 0.00018507837897751,
+      "loss": 0.2127,
+      "step": 4140
+    },
+    {
+      "epoch": 87.36842105263158,
+      "grad_norm": 1.0343648195266724,
+      "learning_rate": 0.00018500804949159427,
+      "loss": 0.2026,
+      "step": 4150
+    },
+    {
+      "epoch": 87.57894736842105,
+      "grad_norm": 1.256469488143921,
+      "learning_rate": 0.00018493756808251073,
+      "loss": 0.2157,
+      "step": 4160
+    },
+    {
+      "epoch": 87.78947368421052,
+      "grad_norm": 1.116888165473938,
+      "learning_rate": 0.000184866934876221,
+      "loss": 0.2184,
+      "step": 4170
+    },
+    {
+      "epoch": 88.0,
+      "grad_norm": 1.3331102132797241,
+      "learning_rate": 0.00018479614999895814,
+      "loss": 0.2189,
+      "step": 4180
+    },
+    {
+      "epoch": 88.21052631578948,
+      "grad_norm": 1.0215224027633667,
+      "learning_rate": 0.00018472521357722622,
+      "loss": 0.2124,
+      "step": 4190
+    },
+    {
+      "epoch": 88.42105263157895,
+      "grad_norm": 1.1717097759246826,
+      "learning_rate": 0.00018465412573780003,
+      "loss": 0.2028,
+      "step": 4200
+    },
+    {
+      "epoch": 88.63157894736842,
+      "grad_norm": 1.110374093055725,
+      "learning_rate": 0.00018458288660772515,
+      "loss": 0.2056,
+      "step": 4210
+    },
+    {
+      "epoch": 88.84210526315789,
+      "grad_norm": 1.1572178602218628,
+      "learning_rate": 0.00018451149631431744,
+      "loss": 0.2213,
+      "step": 4220
+    },
+    {
+      "epoch": 89.05263157894737,
+      "grad_norm": 0.8311715722084045,
+      "learning_rate": 0.00018443995498516294,
+      "loss": 0.2155,
+      "step": 4230
+    },
+    {
+      "epoch": 89.26315789473684,
+      "grad_norm": 0.6976020932197571,
+      "learning_rate": 0.00018436826274811753,
+      "loss": 0.193,
+      "step": 4240
+    },
+    {
+      "epoch": 89.47368421052632,
+      "grad_norm": 1.1248303651809692,
+      "learning_rate": 0.00018429641973130697,
+      "loss": 0.1934,
+      "step": 4250
+    },
+    {
+      "epoch": 89.6842105263158,
+      "grad_norm": 1.1168967485427856,
+      "learning_rate": 0.00018422442606312633,
+      "loss": 0.2176,
+      "step": 4260
+    },
+    {
+      "epoch": 89.89473684210526,
+      "grad_norm": 0.9903426766395569,
+      "learning_rate": 0.00018415228187223997,
+      "loss": 0.2214,
+      "step": 4270
+    },
+    {
+      "epoch": 90.10526315789474,
+      "grad_norm": 0.7373839020729065,
+      "learning_rate": 0.00018407998728758122,
+      "loss": 0.2047,
+      "step": 4280
+    },
+    {
+      "epoch": 90.3157894736842,
+      "grad_norm": 0.8506777882575989,
+      "learning_rate": 0.00018400754243835227,
+      "loss": 0.2025,
+      "step": 4290
+    },
+    {
+      "epoch": 90.52631578947368,
+      "grad_norm": 1.0671459436416626,
+      "learning_rate": 0.0001839349474540238,
+      "loss": 0.2118,
+      "step": 4300
+    },
+    {
+      "epoch": 90.73684210526316,
+      "grad_norm": 0.9546430706977844,
+      "learning_rate": 0.00018386220246433484,
+      "loss": 0.2131,
+      "step": 4310
+    },
+    {
+      "epoch": 90.94736842105263,
+      "grad_norm": 0.998571515083313,
+      "learning_rate": 0.00018378930759929246,
+      "loss": 0.2234,
+      "step": 4320
+    },
+    {
+      "epoch": 91.15789473684211,
+      "grad_norm": 1.444380521774292,
+      "learning_rate": 0.00018371626298917156,
+      "loss": 0.2124,
+      "step": 4330
+    },
+    {
+      "epoch": 91.36842105263158,
+      "grad_norm": 0.929400622844696,
+      "learning_rate": 0.0001836430687645148,
+      "loss": 0.2091,
+      "step": 4340
+    },
+    {
+      "epoch": 91.57894736842105,
+      "grad_norm": 0.7143007516860962,
+      "learning_rate": 0.00018356972505613204,
+      "loss": 0.1907,
+      "step": 4350
+    },
+    {
+      "epoch": 91.78947368421052,
+      "grad_norm": 1.2031328678131104,
+      "learning_rate": 0.0001834962319951004,
+      "loss": 0.2276,
+      "step": 4360
+    },
+    {
+      "epoch": 92.0,
+      "grad_norm": 1.0419529676437378,
+      "learning_rate": 0.00018342258971276395,
+      "loss": 0.2175,
+      "step": 4370
+    },
+    {
+      "epoch": 92.21052631578948,
+      "grad_norm": 0.9380586743354797,
+      "learning_rate": 0.00018334879834073332,
+      "loss": 0.1941,
+      "step": 4380
+    },
+    {
+      "epoch": 92.42105263157895,
+      "grad_norm": 1.299815058708191,
+      "learning_rate": 0.0001832748580108857,
+      "loss": 0.2108,
+      "step": 4390
+    },
+    {
+      "epoch": 92.63157894736842,
+      "grad_norm": 0.9901102185249329,
+      "learning_rate": 0.00018320076885536445,
+      "loss": 0.2002,
+      "step": 4400
+    },
+    {
+      "epoch": 92.84210526315789,
+      "grad_norm": 1.0252271890640259,
+      "learning_rate": 0.00018312653100657883,
+      "loss": 0.2323,
+      "step": 4410
+    },
+    {
+      "epoch": 93.05263157894737,
+      "grad_norm": 0.9015369415283203,
+      "learning_rate": 0.00018305214459720398,
+      "loss": 0.2066,
+      "step": 4420
+    },
+    {
+      "epoch": 93.26315789473684,
+      "grad_norm": 0.7637150287628174,
+      "learning_rate": 0.00018297760976018052,
+      "loss": 0.209,
+      "step": 4430
+    },
+    {
+      "epoch": 93.47368421052632,
+      "grad_norm": 0.784934937953949,
+      "learning_rate": 0.00018290292662871417,
+      "loss": 0.2161,
+      "step": 4440
+    },
+    {
+      "epoch": 93.6842105263158,
+      "grad_norm": 0.7980606555938721,
+      "learning_rate": 0.0001828280953362759,
+      "loss": 0.2133,
+      "step": 4450
+    },
+    {
+      "epoch": 93.89473684210526,
+      "grad_norm": 1.3009899854660034,
+      "learning_rate": 0.0001827531160166013,
+      "loss": 0.2143,
+      "step": 4460
+    },
+    {
+      "epoch": 94.10526315789474,
+      "grad_norm": 0.9536803364753723,
+      "learning_rate": 0.0001826779888036906,
+      "loss": 0.21,
+      "step": 4470
+    },
+    {
+      "epoch": 94.3157894736842,
+      "grad_norm": 1.1637252569198608,
+      "learning_rate": 0.0001826027138318083,
+      "loss": 0.2102,
+      "step": 4480
+    },
+    {
+      "epoch": 94.52631578947368,
+      "grad_norm": 1.1899417638778687,
+      "learning_rate": 0.00018252729123548295,
+      "loss": 0.2033,
+      "step": 4490
+    },
+    {
+      "epoch": 94.73684210526316,
+      "grad_norm": 0.8081888556480408,
+      "learning_rate": 0.00018245172114950703,
+      "loss": 0.2071,
+      "step": 4500
+    },
+    {
+      "epoch": 94.94736842105263,
+      "grad_norm": 0.962952196598053,
+      "learning_rate": 0.0001823760037089365,
+      "loss": 0.2121,
+      "step": 4510
+    },
+    {
+      "epoch": 95.15789473684211,
+      "grad_norm": 0.9023415446281433,
+      "learning_rate": 0.0001823001390490907,
+      "loss": 0.2044,
+      "step": 4520
+    },
+    {
+      "epoch": 95.36842105263158,
+      "grad_norm": 0.9460718035697937,
+      "learning_rate": 0.00018222412730555207,
+      "loss": 0.2085,
+      "step": 4530
+    },
+    {
+      "epoch": 95.57894736842105,
+      "grad_norm": 0.9123182892799377,
+      "learning_rate": 0.00018214796861416594,
+      "loss": 0.2036,
+      "step": 4540
+    },
+    {
+      "epoch": 95.78947368421052,
+      "grad_norm": 0.9567075967788696,
+      "learning_rate": 0.00018207166311104024,
+      "loss": 0.2142,
+      "step": 4550
+    },
+    {
+      "epoch": 96.0,
+      "grad_norm": 1.246997594833374,
+      "learning_rate": 0.00018199521093254523,
+      "loss": 0.2081,
+      "step": 4560
+    },
+    {
+      "epoch": 96.21052631578948,
+      "grad_norm": 0.8137680292129517,
+      "learning_rate": 0.0001819186122153134,
+      "loss": 0.2071,
+      "step": 4570
+    },
+    {
+      "epoch": 96.42105263157895,
+      "grad_norm": 0.901633620262146,
+      "learning_rate": 0.0001818418670962391,
+      "loss": 0.2161,
+      "step": 4580
+    },
+    {
+      "epoch": 96.63157894736842,
+      "grad_norm": 0.9326063990592957,
+      "learning_rate": 0.00018176497571247824,
+      "loss": 0.2102,
+      "step": 4590
+    },
+    {
+      "epoch": 96.84210526315789,
+      "grad_norm": 1.112606406211853,
+      "learning_rate": 0.0001816879382014482,
+      "loss": 0.1983,
+      "step": 4600
+    },
+    {
+      "epoch": 97.05263157894737,
+      "grad_norm": 0.6728799939155579,
+      "learning_rate": 0.00018161075470082754,
+      "loss": 0.2174,
+      "step": 4610
+    },
+    {
+      "epoch": 97.26315789473684,
+      "grad_norm": 1.1379280090332031,
+      "learning_rate": 0.00018153342534855566,
+      "loss": 0.2151,
+      "step": 4620
+    },
+    {
+      "epoch": 97.47368421052632,
+      "grad_norm": 1.089913010597229,
+      "learning_rate": 0.00018145595028283267,
+      "loss": 0.2061,
+      "step": 4630
+    },
+    {
+      "epoch": 97.6842105263158,
+      "grad_norm": 0.8472727537155151,
+      "learning_rate": 0.00018137832964211905,
+      "loss": 0.1973,
+      "step": 4640
+    },
+    {
+      "epoch": 97.89473684210526,
+      "grad_norm": 1.0479404926300049,
+      "learning_rate": 0.0001813005635651355,
+      "loss": 0.2134,
+      "step": 4650
+    },
+    {
+      "epoch": 98.10526315789474,
+      "grad_norm": 0.8933761715888977,
+      "learning_rate": 0.00018122265219086258,
+      "loss": 0.2074,
+      "step": 4660
+    },
+    {
+      "epoch": 98.3157894736842,
+      "grad_norm": 0.8513138294219971,
+      "learning_rate": 0.00018114459565854056,
+      "loss": 0.1951,
+      "step": 4670
+    },
+    {
+      "epoch": 98.52631578947368,
+      "grad_norm": 1.1558623313903809,
+      "learning_rate": 0.00018106639410766912,
+      "loss": 0.2212,
+      "step": 4680
+    },
+    {
+      "epoch": 98.73684210526316,
+      "grad_norm": 1.7782244682312012,
+      "learning_rate": 0.00018098804767800711,
+      "loss": 0.2202,
+      "step": 4690
+    },
+    {
+      "epoch": 98.94736842105263,
+      "grad_norm": 0.9603464603424072,
+      "learning_rate": 0.00018090955650957232,
+      "loss": 0.1953,
+      "step": 4700
+    },
+    {
+      "epoch": 99.15789473684211,
+      "grad_norm": 0.7804552316665649,
+      "learning_rate": 0.0001808309207426412,
+      "loss": 0.1969,
+      "step": 4710
+    },
+    {
+      "epoch": 99.36842105263158,
+      "grad_norm": 1.0236091613769531,
+      "learning_rate": 0.00018075214051774857,
+      "loss": 0.2108,
+      "step": 4720
+    },
+    {
+      "epoch": 99.57894736842105,
+      "grad_norm": 0.8507013320922852,
+      "learning_rate": 0.00018067321597568746,
+      "loss": 0.2114,
+      "step": 4730
+    },
+    {
+      "epoch": 99.78947368421052,
+      "grad_norm": 1.1381756067276,
+      "learning_rate": 0.0001805941472575089,
+      "loss": 0.2161,
+      "step": 4740
+    },
+    {
+      "epoch": 100.0,
+      "grad_norm": 1.0529285669326782,
+      "learning_rate": 0.00018051493450452148,
+      "loss": 0.2094,
+      "step": 4750
+    },
+    {
+      "epoch": 100.21052631578948,
+      "grad_norm": 0.7255027294158936,
+      "learning_rate": 0.0001804355778582912,
+      "loss": 0.1996,
+      "step": 4760
+    },
+    {
+      "epoch": 100.42105263157895,
+      "grad_norm": 0.8361448049545288,
+      "learning_rate": 0.00018035607746064126,
+      "loss": 0.1991,
+      "step": 4770
+    },
+    {
+      "epoch": 100.63157894736842,
+      "grad_norm": 1.0485684871673584,
+      "learning_rate": 0.00018027643345365176,
+      "loss": 0.2255,
+      "step": 4780
+    },
+    {
+      "epoch": 100.84210526315789,
+      "grad_norm": 0.8241237998008728,
+      "learning_rate": 0.00018019664597965947,
+      "loss": 0.2072,
+      "step": 4790
+    },
+    {
+      "epoch": 101.05263157894737,
+      "grad_norm": 0.5731729865074158,
+      "learning_rate": 0.00018011671518125758,
+      "loss": 0.208,
+      "step": 4800
+    },
+    {
+      "epoch": 101.26315789473684,
+      "grad_norm": 1.0061657428741455,
+      "learning_rate": 0.00018003664120129533,
+      "loss": 0.1995,
+      "step": 4810
+    },
+    {
+      "epoch": 101.47368421052632,
+      "grad_norm": 1.0121266841888428,
+      "learning_rate": 0.00017995642418287792,
+      "loss": 0.2157,
+      "step": 4820
+    },
+    {
+      "epoch": 101.6842105263158,
+      "grad_norm": 0.7068321108818054,
+      "learning_rate": 0.00017987606426936615,
+      "loss": 0.2062,
+      "step": 4830
+    },
+    {
+      "epoch": 101.89473684210526,
+      "grad_norm": 1.0422128438949585,
+      "learning_rate": 0.00017979556160437627,
+      "loss": 0.2076,
+      "step": 4840
+    },
+    {
+      "epoch": 102.10526315789474,
+      "grad_norm": 0.9450851082801819,
+      "learning_rate": 0.00017971491633177956,
+      "loss": 0.2104,
+      "step": 4850
+    },
+    {
+      "epoch": 102.3157894736842,
+      "grad_norm": 0.9264403581619263,
+      "learning_rate": 0.0001796341285957022,
+      "loss": 0.1923,
+      "step": 4860
+    },
+    {
+      "epoch": 102.52631578947368,
+      "grad_norm": 0.8550981283187866,
+      "learning_rate": 0.00017955319854052494,
+      "loss": 0.223,
+      "step": 4870
+    },
+    {
+      "epoch": 102.73684210526316,
+      "grad_norm": 0.9619629979133606,
+      "learning_rate": 0.00017947212631088297,
+      "loss": 0.1947,
+      "step": 4880
+    },
+    {
+      "epoch": 102.94736842105263,
+      "grad_norm": 0.872445285320282,
+      "learning_rate": 0.00017939091205166548,
+      "loss": 0.2202,
+      "step": 4890
+    },
+    {
+      "epoch": 103.15789473684211,
+      "grad_norm": 0.8909777402877808,
+      "learning_rate": 0.00017930955590801553,
+      "loss": 0.2104,
+      "step": 4900
+    },
+    {
+      "epoch": 103.36842105263158,
+      "grad_norm": 1.1058716773986816,
+      "learning_rate": 0.00017922805802532974,
+      "loss": 0.1991,
+      "step": 4910
+    },
+    {
+      "epoch": 103.57894736842105,
+      "grad_norm": 0.9872148633003235,
+      "learning_rate": 0.000179146418549258,
+      "loss": 0.1916,
+      "step": 4920
+    },
+    {
+      "epoch": 103.78947368421052,
+      "grad_norm": 0.8365228176116943,
+      "learning_rate": 0.00017906463762570337,
+      "loss": 0.223,
+      "step": 4930
+    },
+    {
+      "epoch": 104.0,
+      "grad_norm": 0.7262590527534485,
+      "learning_rate": 0.00017898271540082154,
+      "loss": 0.2089,
+      "step": 4940
+    },
+    {
+      "epoch": 104.21052631578948,
+      "grad_norm": 0.753351628780365,
+      "learning_rate": 0.00017890065202102085,
+      "loss": 0.1939,
+      "step": 4950
+    },
+    {
+      "epoch": 104.42105263157895,
+      "grad_norm": 0.7613105773925781,
+      "learning_rate": 0.00017881844763296186,
+      "loss": 0.2005,
+      "step": 4960
+    },
+    {
+      "epoch": 104.63157894736842,
+      "grad_norm": 0.79398512840271,
+      "learning_rate": 0.00017873610238355715,
+      "loss": 0.2098,
+      "step": 4970
+    },
+    {
+      "epoch": 104.84210526315789,
+      "grad_norm": 0.7749453186988831,
+      "learning_rate": 0.00017865361641997103,
+      "loss": 0.2094,
+      "step": 4980
+    },
+    {
+      "epoch": 105.05263157894737,
+      "grad_norm": 0.8748012185096741,
+      "learning_rate": 0.0001785709898896193,
+      "loss": 0.2133,
+      "step": 4990
+    },
+    {
+      "epoch": 105.26315789473684,
+      "grad_norm": 0.9628874659538269,
+      "learning_rate": 0.0001784882229401689,
+      "loss": 0.1999,
+      "step": 5000
+    },
+    {
+      "epoch": 105.47368421052632,
+      "grad_norm": 1.3108863830566406,
+      "learning_rate": 0.00017840531571953786,
+      "loss": 0.2095,
+      "step": 5010
+    },
+    {
+      "epoch": 105.6842105263158,
+      "grad_norm": 0.9370902180671692,
+      "learning_rate": 0.0001783222683758948,
+      "loss": 0.2298,
+      "step": 5020
+    },
+    {
+      "epoch": 105.89473684210526,
+      "grad_norm": 0.8537701964378357,
+      "learning_rate": 0.0001782390810576588,
+      "loss": 0.1997,
+      "step": 5030
+    },
+    {
+      "epoch": 106.10526315789474,
+      "grad_norm": 0.572172224521637,
+      "learning_rate": 0.0001781557539134991,
+      "loss": 0.194,
+      "step": 5040
+    },
+    {
+      "epoch": 106.3157894736842,
+      "grad_norm": 0.7815021276473999,
+      "learning_rate": 0.00017807228709233478,
+      "loss": 0.2138,
+      "step": 5050
+    },
+    {
+      "epoch": 106.52631578947368,
+      "grad_norm": 0.8081856966018677,
+      "learning_rate": 0.00017798868074333463,
+      "loss": 0.1877,
+      "step": 5060
+    },
+    {
+      "epoch": 106.73684210526316,
+      "grad_norm": 0.891789436340332,
+      "learning_rate": 0.00017790493501591668,
+      "loss": 0.2164,
+      "step": 5070
+    },
+    {
+      "epoch": 106.94736842105263,
+      "grad_norm": 0.8877257108688354,
+      "learning_rate": 0.0001778210500597482,
+      "loss": 0.2139,
+      "step": 5080
+    },
+    {
+      "epoch": 107.15789473684211,
+      "grad_norm": 0.8057231903076172,
+      "learning_rate": 0.00017773702602474515,
+      "loss": 0.1985,
+      "step": 5090
+    },
+    {
+      "epoch": 107.36842105263158,
+      "grad_norm": 0.9601947069168091,
+      "learning_rate": 0.00017765286306107214,
+      "loss": 0.2082,
+      "step": 5100
+    },
+    {
+      "epoch": 107.57894736842105,
+      "grad_norm": 1.2363498210906982,
+      "learning_rate": 0.000177568561319142,
+      "loss": 0.1961,
+      "step": 5110
+    },
+    {
+      "epoch": 107.78947368421052,
+      "grad_norm": 0.8430647850036621,
+      "learning_rate": 0.00017748412094961566,
+      "loss": 0.2112,
+      "step": 5120
+    },
+    {
+      "epoch": 108.0,
+      "grad_norm": 1.2295119762420654,
+      "learning_rate": 0.00017739954210340173,
+      "loss": 0.2104,
+      "step": 5130
+    },
+    {
+      "epoch": 108.21052631578948,
+      "grad_norm": 0.6926615238189697,
+      "learning_rate": 0.0001773148249316563,
+      "loss": 0.2053,
+      "step": 5140
+    },
+    {
+      "epoch": 108.42105263157895,
+      "grad_norm": 0.6129204034805298,
+      "learning_rate": 0.0001772299695857827,
+      "loss": 0.1957,
+      "step": 5150
+    },
+    {
+      "epoch": 108.63157894736842,
+      "grad_norm": 0.7398890256881714,
+      "learning_rate": 0.00017714497621743123,
+      "loss": 0.2114,
+      "step": 5160
+    },
+    {
+      "epoch": 108.84210526315789,
+      "grad_norm": 0.9888297915458679,
+      "learning_rate": 0.00017705984497849874,
+      "loss": 0.2087,
+      "step": 5170
+    },
+    {
+      "epoch": 109.05263157894737,
+      "grad_norm": 0.8859359622001648,
+      "learning_rate": 0.00017697457602112863,
+      "loss": 0.2095,
+      "step": 5180
+    },
+    {
+      "epoch": 109.26315789473684,
+      "grad_norm": 0.80406254529953,
+      "learning_rate": 0.00017688916949771036,
+      "loss": 0.2149,
+      "step": 5190
+    },
+    {
+      "epoch": 109.47368421052632,
+      "grad_norm": 0.9358986020088196,
+      "learning_rate": 0.0001768036255608792,
+      "loss": 0.193,
+      "step": 5200
+    },
+    {
+      "epoch": 109.6842105263158,
+      "grad_norm": 0.9183883666992188,
+      "learning_rate": 0.000176717944363516,
+      "loss": 0.2081,
+      "step": 5210
+    },
+    {
+      "epoch": 109.89473684210526,
+      "grad_norm": 0.8253007531166077,
+      "learning_rate": 0.00017663212605874704,
+      "loss": 0.2275,
+      "step": 5220
+    },
+    {
+      "epoch": 110.10526315789474,
+      "grad_norm": 0.8964976668357849,
+      "learning_rate": 0.00017654617079994347,
+      "loss": 0.2045,
+      "step": 5230
+    },
+    {
+      "epoch": 110.3157894736842,
+      "grad_norm": 0.922153115272522,
+      "learning_rate": 0.0001764600787407213,
+      "loss": 0.205,
+      "step": 5240
+    },
+    {
+      "epoch": 110.52631578947368,
+      "grad_norm": 0.8380544185638428,
+      "learning_rate": 0.00017637385003494102,
+      "loss": 0.1918,
+      "step": 5250
+    },
+    {
+      "epoch": 110.73684210526316,
+      "grad_norm": 0.7515467405319214,
+      "learning_rate": 0.00017628748483670728,
+      "loss": 0.21,
+      "step": 5260
+    },
+    {
+      "epoch": 110.94736842105263,
+      "grad_norm": 0.9965086579322815,
+      "learning_rate": 0.00017620098330036873,
+      "loss": 0.2108,
+      "step": 5270
+    },
+    {
+      "epoch": 111.15789473684211,
+      "grad_norm": 0.7745735049247742,
+      "learning_rate": 0.00017611434558051757,
+      "loss": 0.21,
+      "step": 5280
+    },
+    {
+      "epoch": 111.36842105263158,
+      "grad_norm": 0.8300684690475464,
+      "learning_rate": 0.00017602757183198952,
+      "loss": 0.1987,
+      "step": 5290
+    },
+    {
+      "epoch": 111.57894736842105,
+      "grad_norm": 0.9019736051559448,
+      "learning_rate": 0.00017594066220986333,
+      "loss": 0.1983,
+      "step": 5300
+    },
+    {
+      "epoch": 111.78947368421052,
+      "grad_norm": 0.7582821846008301,
+      "learning_rate": 0.00017585361686946055,
+      "loss": 0.2161,
+      "step": 5310
+    },
+    {
+      "epoch": 112.0,
+      "grad_norm": 0.7558669447898865,
+      "learning_rate": 0.00017576643596634538,
+      "loss": 0.208,
+      "step": 5320
+    },
+    {
+      "epoch": 112.21052631578948,
+      "grad_norm": 0.9202759265899658,
+      "learning_rate": 0.00017567911965632414,
+      "loss": 0.2082,
+      "step": 5330
+    },
+    {
+      "epoch": 112.42105263157895,
+      "grad_norm": 0.8538267016410828,
+      "learning_rate": 0.0001755916680954453,
+      "loss": 0.2044,
+      "step": 5340
+    },
+    {
+      "epoch": 112.63157894736842,
+      "grad_norm": 0.6622471213340759,
+      "learning_rate": 0.00017550408143999894,
+      "loss": 0.1959,
+      "step": 5350
+    },
+    {
+      "epoch": 112.84210526315789,
+      "grad_norm": 1.5098806619644165,
+      "learning_rate": 0.00017541635984651667,
+      "loss": 0.2011,
+      "step": 5360
+    },
+    {
+      "epoch": 113.05263157894737,
+      "grad_norm": 1.0928170680999756,
+      "learning_rate": 0.00017532850347177118,
+      "loss": 0.2145,
+      "step": 5370
+    },
+    {
+      "epoch": 113.26315789473684,
+      "grad_norm": 0.7540849447250366,
+      "learning_rate": 0.00017524051247277603,
+      "loss": 0.1989,
+      "step": 5380
+    },
+    {
+      "epoch": 113.47368421052632,
+      "grad_norm": 0.7334789633750916,
+      "learning_rate": 0.00017515238700678538,
+      "loss": 0.1966,
+      "step": 5390
+    },
+    {
+      "epoch": 113.6842105263158,
+      "grad_norm": 0.948559045791626,
+      "learning_rate": 0.0001750641272312938,
+      "loss": 0.2164,
+      "step": 5400
+    },
+    {
+      "epoch": 113.89473684210526,
+      "grad_norm": 1.1310551166534424,
+      "learning_rate": 0.00017497573330403578,
+      "loss": 0.2174,
+      "step": 5410
+    },
+    {
+      "epoch": 114.10526315789474,
+      "grad_norm": 0.6442914605140686,
+      "learning_rate": 0.00017488720538298558,
+      "loss": 0.1954,
+      "step": 5420
+    },
+    {
+      "epoch": 114.3157894736842,
+      "grad_norm": 0.6499135494232178,
+      "learning_rate": 0.000174798543626357,
+      "loss": 0.2052,
+      "step": 5430
+    },
+    {
+      "epoch": 114.52631578947368,
+      "grad_norm": 0.9780040979385376,
+      "learning_rate": 0.00017470974819260292,
+      "loss": 0.1946,
+      "step": 5440
+    },
+    {
+      "epoch": 114.73684210526316,
+      "grad_norm": 0.8085854053497314,
+      "learning_rate": 0.00017462081924041523,
+      "loss": 0.2252,
+      "step": 5450
+    },
+    {
+      "epoch": 114.94736842105263,
+      "grad_norm": 1.080230712890625,
+      "learning_rate": 0.00017453175692872436,
+      "loss": 0.2177,
+      "step": 5460
+    },
+    {
+      "epoch": 115.15789473684211,
+      "grad_norm": 0.6058759093284607,
+      "learning_rate": 0.00017444256141669907,
+      "loss": 0.1986,
+      "step": 5470
+    },
+    {
+      "epoch": 115.36842105263158,
+      "grad_norm": 0.9204846024513245,
+      "learning_rate": 0.00017435323286374627,
+      "loss": 0.1958,
+      "step": 5480
+    },
+    {
+      "epoch": 115.57894736842105,
+      "grad_norm": 0.781603991985321,
+      "learning_rate": 0.00017426377142951052,
+      "loss": 0.2055,
+      "step": 5490
+    },
+    {
+      "epoch": 115.78947368421052,
+      "grad_norm": 0.9358359575271606,
+      "learning_rate": 0.00017417417727387394,
+      "loss": 0.2253,
+      "step": 5500
+    },
+    {
+      "epoch": 116.0,
+      "grad_norm": 0.8498882055282593,
+      "learning_rate": 0.00017408445055695578,
+      "loss": 0.1943,
+      "step": 5510
+    },
+    {
+      "epoch": 116.21052631578948,
+      "grad_norm": 0.8480699062347412,
+      "learning_rate": 0.00017399459143911225,
+      "loss": 0.2052,
+      "step": 5520
+    },
+    {
+      "epoch": 116.42105263157895,
+      "grad_norm": 1.2214350700378418,
+      "learning_rate": 0.00017390460008093618,
+      "loss": 0.2145,
+      "step": 5530
+    },
+    {
+      "epoch": 116.63157894736842,
+      "grad_norm": 0.7576544880867004,
+      "learning_rate": 0.00017381447664325666,
+      "loss": 0.2106,
+      "step": 5540
+    },
+    {
+      "epoch": 116.84210526315789,
+      "grad_norm": 0.7696592211723328,
+      "learning_rate": 0.00017372422128713891,
+      "loss": 0.1956,
+      "step": 5550
+    },
+    {
+      "epoch": 117.05263157894737,
+      "grad_norm": 0.787528395652771,
+      "learning_rate": 0.00017364287880967868,
+      "loss": 0.2027,
+      "step": 5560
+    },
+    {
+      "epoch": 117.26315789473684,
+      "grad_norm": 0.7447987198829651,
+      "learning_rate": 0.00017355237325310713,
+      "loss": 0.1893,
+      "step": 5570
+    },
+    {
+      "epoch": 117.47368421052632,
+      "grad_norm": 0.8453590273857117,
+      "learning_rate": 0.00017346173624651856,
+      "loss": 0.2099,
+      "step": 5580
+    },
+    {
+      "epoch": 117.6842105263158,
+      "grad_norm": 0.7510943412780762,
+      "learning_rate": 0.0001733709679518961,
+      "loss": 0.2025,
+      "step": 5590
+    },
+    {
+      "epoch": 117.89473684210526,
+      "grad_norm": 0.8954883813858032,
+      "learning_rate": 0.00017328006853145739,
+      "loss": 0.2046,
+      "step": 5600
+    },
+    {
+      "epoch": 118.10526315789474,
+      "grad_norm": 0.9460748434066772,
+      "learning_rate": 0.00017318903814765436,
+      "loss": 0.215,
+      "step": 5610
+    },
+    {
+      "epoch": 118.3157894736842,
+      "grad_norm": 0.9654355645179749,
+      "learning_rate": 0.00017309787696317315,
+      "loss": 0.2006,
+      "step": 5620
+    },
+    {
+      "epoch": 118.52631578947368,
+      "grad_norm": 0.7358613014221191,
+      "learning_rate": 0.00017300658514093353,
+      "loss": 0.204,
+      "step": 5630
+    },
+    {
+      "epoch": 118.73684210526316,
+      "grad_norm": 0.6271918416023254,
+      "learning_rate": 0.00017291516284408882,
+      "loss": 0.2058,
+      "step": 5640
+    },
+    {
+      "epoch": 118.94736842105263,
+      "grad_norm": 0.7335479259490967,
+      "learning_rate": 0.00017282361023602546,
+      "loss": 0.2023,
+      "step": 5650
+    },
+    {
+      "epoch": 119.15789473684211,
+      "grad_norm": 0.9404072761535645,
+      "learning_rate": 0.0001727319274803628,
+      "loss": 0.1997,
+      "step": 5660
+    },
+    {
+      "epoch": 119.36842105263158,
+      "grad_norm": 0.7514150738716125,
+      "learning_rate": 0.00017264011474095282,
+      "loss": 0.1994,
+      "step": 5670
+    },
+    {
+      "epoch": 119.57894736842105,
+      "grad_norm": 0.9548558592796326,
+      "learning_rate": 0.00017254817218187977,
+      "loss": 0.2048,
+      "step": 5680
+    },
+    {
+      "epoch": 119.78947368421052,
+      "grad_norm": 1.1437008380889893,
+      "learning_rate": 0.00017245609996745985,
+      "loss": 0.2062,
+      "step": 5690
+    },
+    {
+      "epoch": 120.0,
+      "grad_norm": 0.9660143256187439,
+      "learning_rate": 0.00017236389826224115,
+      "loss": 0.2125,
+      "step": 5700
+    },
+    {
+      "epoch": 120.21052631578948,
+      "grad_norm": 0.6402197480201721,
+      "learning_rate": 0.000172271567231003,
+      "loss": 0.1845,
+      "step": 5710
+    },
+    {
+      "epoch": 120.42105263157895,
+      "grad_norm": 0.9553530216217041,
+      "learning_rate": 0.00017217910703875588,
+      "loss": 0.2018,
+      "step": 5720
+    },
+    {
+      "epoch": 120.63157894736842,
+      "grad_norm": 1.160772681236267,
+      "learning_rate": 0.00017208651785074122,
+      "loss": 0.2186,
+      "step": 5730
+    },
+    {
+      "epoch": 120.84210526315789,
+      "grad_norm": 0.9803076982498169,
+      "learning_rate": 0.00017199379983243087,
+      "loss": 0.2133,
+      "step": 5740
+    },
+    {
+      "epoch": 121.05263157894737,
+      "grad_norm": 0.7129189968109131,
+      "learning_rate": 0.00017190095314952697,
+      "loss": 0.2165,
+      "step": 5750
+    },
+    {
+      "epoch": 121.26315789473684,
+      "grad_norm": 0.8149412870407104,
+      "learning_rate": 0.0001718079779679616,
+      "loss": 0.2117,
+      "step": 5760
+    },
+    {
+      "epoch": 121.47368421052632,
+      "grad_norm": 0.7495967149734497,
+      "learning_rate": 0.0001717148744538965,
+      "loss": 0.1908,
+      "step": 5770
+    },
+    {
+      "epoch": 121.6842105263158,
+      "grad_norm": 1.0567039251327515,
+      "learning_rate": 0.00017162164277372273,
+      "loss": 0.2077,
+      "step": 5780
+    },
+    {
+      "epoch": 121.89473684210526,
+      "grad_norm": 1.0251121520996094,
+      "learning_rate": 0.0001715282830940604,
+      "loss": 0.207,
+      "step": 5790
+    },
+    {
+      "epoch": 122.10526315789474,
+      "grad_norm": 0.8159797191619873,
+      "learning_rate": 0.00017143479558175844,
+      "loss": 0.2052,
+      "step": 5800
+    },
+    {
+      "epoch": 122.3157894736842,
+      "grad_norm": 1.1662620306015015,
+      "learning_rate": 0.00017134118040389415,
+      "loss": 0.2021,
+      "step": 5810
+    },
+    {
+      "epoch": 122.52631578947368,
+      "grad_norm": 0.9236595630645752,
+      "learning_rate": 0.00017124743772777308,
+      "loss": 0.1977,
+      "step": 5820
+    },
+    {
+      "epoch": 122.73684210526316,
+      "grad_norm": 0.8414731621742249,
+      "learning_rate": 0.00017115356772092857,
+      "loss": 0.2003,
+      "step": 5830
+    },
+    {
+      "epoch": 122.94736842105263,
+      "grad_norm": 0.7867684364318848,
+      "learning_rate": 0.0001710595705511215,
+      "loss": 0.2051,
+      "step": 5840
+    },
+    {
+      "epoch": 123.15789473684211,
+      "grad_norm": 0.6422597169876099,
+      "learning_rate": 0.00017096544638634008,
+      "loss": 0.1987,
+      "step": 5850
+    },
+    {
+      "epoch": 123.36842105263158,
+      "grad_norm": 0.8452603220939636,
+      "learning_rate": 0.00017087119539479947,
+      "loss": 0.2116,
+      "step": 5860
+    },
+    {
+      "epoch": 123.57894736842105,
+      "grad_norm": 0.7509437203407288,
+      "learning_rate": 0.0001707768177449415,
+      "loss": 0.2036,
+      "step": 5870
+    },
+    {
+      "epoch": 123.78947368421052,
+      "grad_norm": 0.964089572429657,
+      "learning_rate": 0.00017068231360543425,
+      "loss": 0.2029,
+      "step": 5880
+    },
+    {
+      "epoch": 124.0,
+      "grad_norm": 1.1077382564544678,
+      "learning_rate": 0.00017058768314517203,
+      "loss": 0.2193,
+      "step": 5890
+    },
+    {
+      "epoch": 124.21052631578948,
+      "grad_norm": 1.1068215370178223,
+      "learning_rate": 0.00017049292653327478,
+      "loss": 0.1957,
+      "step": 5900
+    },
+    {
+      "epoch": 124.42105263157895,
+      "grad_norm": 0.6014849543571472,
+      "learning_rate": 0.0001703980439390879,
+      "loss": 0.1914,
+      "step": 5910
+    },
+    {
+      "epoch": 124.63157894736842,
+      "grad_norm": 0.895735502243042,
+      "learning_rate": 0.000170303035532182,
+      "loss": 0.2051,
+      "step": 5920
+    },
+    {
+      "epoch": 124.84210526315789,
+      "grad_norm": 0.758565366268158,
+      "learning_rate": 0.00017020790148235252,
+      "loss": 0.2218,
+      "step": 5930
+    },
+    {
+      "epoch": 125.05263157894737,
+      "grad_norm": 1.467057228088379,
+      "learning_rate": 0.00017011264195961937,
+      "loss": 0.2013,
+      "step": 5940
+    },
+    {
+      "epoch": 125.26315789473684,
+      "grad_norm": 0.9302055239677429,
+      "learning_rate": 0.00017001725713422684,
+      "loss": 0.2045,
+      "step": 5950
+    },
+    {
+      "epoch": 125.47368421052632,
+      "grad_norm": 0.963485598564148,
+      "learning_rate": 0.00016992174717664305,
+      "loss": 0.1982,
+      "step": 5960
+    },
+    {
+      "epoch": 125.6842105263158,
+      "grad_norm": 0.7401504516601562,
+      "learning_rate": 0.00016982611225755978,
+      "loss": 0.2028,
+      "step": 5970
+    },
+    {
+      "epoch": 125.89473684210526,
+      "grad_norm": 0.938339352607727,
+      "learning_rate": 0.00016973035254789213,
+      "loss": 0.2072,
+      "step": 5980
+    },
+    {
+      "epoch": 126.10526315789474,
+      "grad_norm": 0.8344370722770691,
+      "learning_rate": 0.00016963446821877825,
+      "loss": 0.2026,
+      "step": 5990
+    },
+    {
+      "epoch": 126.3157894736842,
+      "grad_norm": 0.695802628993988,
+      "learning_rate": 0.00016953845944157894,
+      "loss": 0.1893,
+      "step": 6000
+    },
+    {
+      "epoch": 126.52631578947368,
+      "grad_norm": 0.7851212620735168,
+      "learning_rate": 0.00016944232638787748,
+      "loss": 0.1912,
+      "step": 6010
+    },
+    {
+      "epoch": 126.73684210526316,
+      "grad_norm": 0.762319803237915,
+      "learning_rate": 0.00016934606922947923,
+      "loss": 0.2068,
+      "step": 6020
+    },
+    {
+      "epoch": 126.94736842105263,
+      "grad_norm": 0.9079723358154297,
+      "learning_rate": 0.0001692496881384113,
+      "loss": 0.2096,
+      "step": 6030
+    },
+    {
+      "epoch": 127.15789473684211,
+      "grad_norm": 0.7185937166213989,
+      "learning_rate": 0.00016915318328692243,
+      "loss": 0.2039,
+      "step": 6040
+    },
+    {
+      "epoch": 127.36842105263158,
+      "grad_norm": 0.9591910243034363,
+      "learning_rate": 0.0001690565548474823,
+      "loss": 0.2109,
+      "step": 6050
+    },
+    {
+      "epoch": 127.57894736842105,
+      "grad_norm": 0.8179908394813538,
+      "learning_rate": 0.0001689598029927817,
+      "loss": 0.1946,
+      "step": 6060
+    },
+    {
+      "epoch": 127.78947368421052,
+      "grad_norm": 0.9175289273262024,
+      "learning_rate": 0.00016886292789573183,
+      "loss": 0.2043,
+      "step": 6070
+    },
+    {
+      "epoch": 128.0,
+      "grad_norm": 0.7145066857337952,
+      "learning_rate": 0.0001687659297294642,
+      "loss": 0.2177,
+      "step": 6080
+    },
+    {
+      "epoch": 128.21052631578948,
+      "grad_norm": 0.6293530464172363,
+      "learning_rate": 0.0001686688086673303,
+      "loss": 0.2011,
+      "step": 6090
+    },
+    {
+      "epoch": 128.42105263157896,
+      "grad_norm": 0.8945363759994507,
+      "learning_rate": 0.0001685715648829012,
+      "loss": 0.2044,
+      "step": 6100
+    },
+    {
+      "epoch": 128.6315789473684,
+      "grad_norm": 0.9259580373764038,
+      "learning_rate": 0.00016847419854996724,
+      "loss": 0.2128,
+      "step": 6110
+    },
+    {
+      "epoch": 128.8421052631579,
+      "grad_norm": 1.0482829809188843,
+      "learning_rate": 0.00016837670984253794,
+      "loss": 0.2029,
+      "step": 6120
+    },
+    {
+      "epoch": 129.05263157894737,
+      "grad_norm": 0.7181470990180969,
+      "learning_rate": 0.0001682790989348414,
+      "loss": 0.2024,
+      "step": 6130
+    },
+    {
+      "epoch": 129.26315789473685,
+      "grad_norm": 0.9605810642242432,
+      "learning_rate": 0.00016818136600132416,
+      "loss": 0.2086,
+      "step": 6140
+    },
+    {
+      "epoch": 129.47368421052633,
+      "grad_norm": 0.651897668838501,
+      "learning_rate": 0.00016808351121665071,
+      "loss": 0.1877,
+      "step": 6150
+    },
+    {
+      "epoch": 129.68421052631578,
+      "grad_norm": 0.7484719753265381,
+      "learning_rate": 0.00016798553475570356,
+      "loss": 0.2019,
+      "step": 6160
+    },
+    {
+      "epoch": 129.89473684210526,
+      "grad_norm": 0.7095634937286377,
+      "learning_rate": 0.0001678874367935824,
+      "loss": 0.2328,
+      "step": 6170
+    },
+    {
+      "epoch": 130.10526315789474,
+      "grad_norm": 1.094728946685791,
+      "learning_rate": 0.0001677892175056043,
+      "loss": 0.2077,
+      "step": 6180
+    },
+    {
+      "epoch": 130.31578947368422,
+      "grad_norm": 0.7821156978607178,
+      "learning_rate": 0.00016769087706730302,
+      "loss": 0.2012,
+      "step": 6190
+    },
+    {
+      "epoch": 130.52631578947367,
+      "grad_norm": 0.6422997117042542,
+      "learning_rate": 0.00016759241565442884,
+      "loss": 0.2012,
+      "step": 6200
+    },
+    {
+      "epoch": 130.73684210526315,
+      "grad_norm": 0.7488658428192139,
+      "learning_rate": 0.00016749383344294834,
+      "loss": 0.199,
+      "step": 6210
+    },
+    {
+      "epoch": 130.94736842105263,
+      "grad_norm": 0.7575473785400391,
+      "learning_rate": 0.00016739513060904382,
+      "loss": 0.1988,
+      "step": 6220
+    },
+    {
+      "epoch": 131.1578947368421,
+      "grad_norm": 0.6518003344535828,
+      "learning_rate": 0.0001672963073291133,
+      "loss": 0.1945,
+      "step": 6230
+    },
+    {
+      "epoch": 131.3684210526316,
+      "grad_norm": 0.6180778741836548,
+      "learning_rate": 0.00016719736377977,
+      "loss": 0.1934,
+      "step": 6240
+    },
+    {
+      "epoch": 131.57894736842104,
+      "grad_norm": 0.5813746452331543,
+      "learning_rate": 0.00016709830013784212,
+      "loss": 0.2006,
+      "step": 6250
+    },
+    {
+      "epoch": 131.78947368421052,
+      "grad_norm": 0.880317211151123,
+      "learning_rate": 0.00016699911658037237,
+      "loss": 0.2136,
+      "step": 6260
+    },
+    {
+      "epoch": 132.0,
+      "grad_norm": 0.804076075553894,
+      "learning_rate": 0.00016689981328461793,
+      "loss": 0.2092,
+      "step": 6270
+    },
+    {
+      "epoch": 132.21052631578948,
+      "grad_norm": 0.7459381818771362,
+      "learning_rate": 0.00016680039042804982,
+      "loss": 0.194,
+      "step": 6280
+    },
+    {
+      "epoch": 132.42105263157896,
+      "grad_norm": 0.6983547806739807,
+      "learning_rate": 0.00016670084818835287,
+      "loss": 0.1926,
+      "step": 6290
+    },
+    {
+      "epoch": 132.6315789473684,
+      "grad_norm": 0.7421784400939941,
+      "learning_rate": 0.00016660118674342517,
+      "loss": 0.1958,
+      "step": 6300
+    },
+    {
+      "epoch": 132.8421052631579,
+      "grad_norm": 1.2000913619995117,
+      "learning_rate": 0.0001665014062713779,
+      "loss": 0.2105,
+      "step": 6310
+    },
+    {
+      "epoch": 133.05263157894737,
+      "grad_norm": 0.6287597417831421,
+      "learning_rate": 0.0001664015069505349,
+      "loss": 0.2012,
+      "step": 6320
+    },
+    {
+      "epoch": 133.26315789473685,
+      "grad_norm": 0.8468930125236511,
+      "learning_rate": 0.0001663014889594325,
+      "loss": 0.203,
+      "step": 6330
+    },
+    {
+      "epoch": 133.47368421052633,
+      "grad_norm": 0.9103506207466125,
+      "learning_rate": 0.00016620135247681902,
+      "loss": 0.1935,
+      "step": 6340
+    },
+    {
+      "epoch": 133.68421052631578,
+      "grad_norm": 0.619668185710907,
+      "learning_rate": 0.00016610109768165464,
+      "loss": 0.2187,
+      "step": 6350
+    },
+    {
+      "epoch": 133.89473684210526,
+      "grad_norm": 0.816078782081604,
+      "learning_rate": 0.00016600072475311096,
+      "loss": 0.1962,
+      "step": 6360
+    },
+    {
+      "epoch": 134.10526315789474,
+      "grad_norm": 0.721704363822937,
+      "learning_rate": 0.00016590023387057055,
+      "loss": 0.2024,
+      "step": 6370
+    },
+    {
+      "epoch": 134.31578947368422,
+      "grad_norm": 0.913533091545105,
+      "learning_rate": 0.00016579962521362708,
+      "loss": 0.1888,
+      "step": 6380
+    },
+    {
+      "epoch": 134.52631578947367,
+      "grad_norm": 0.8451491594314575,
+      "learning_rate": 0.00016569889896208436,
+      "loss": 0.2001,
+      "step": 6390
+    },
+    {
+      "epoch": 134.73684210526315,
+      "grad_norm": 0.7206440567970276,
+      "learning_rate": 0.00016559805529595668,
+      "loss": 0.2135,
+      "step": 6400
+    },
+    {
+      "epoch": 134.94736842105263,
+      "grad_norm": 0.5913563966751099,
+      "learning_rate": 0.00016549709439546794,
+      "loss": 0.2041,
+      "step": 6410
+    },
+    {
+      "epoch": 135.1578947368421,
+      "grad_norm": 1.0242068767547607,
+      "learning_rate": 0.00016539601644105167,
+      "loss": 0.2064,
+      "step": 6420
+    },
+    {
+      "epoch": 135.3684210526316,
+      "grad_norm": 0.7297083735466003,
+      "learning_rate": 0.00016529482161335054,
+      "loss": 0.1989,
+      "step": 6430
+    },
+    {
+      "epoch": 135.57894736842104,
+      "grad_norm": 0.9045604467391968,
+      "learning_rate": 0.00016519351009321612,
+      "loss": 0.1974,
+      "step": 6440
+    },
+    {
+      "epoch": 135.78947368421052,
+      "grad_norm": 0.6610522866249084,
+      "learning_rate": 0.00016509208206170857,
+      "loss": 0.1924,
+      "step": 6450
+    },
+    {
+      "epoch": 136.0,
+      "grad_norm": 0.9059966802597046,
+      "learning_rate": 0.00016499053770009618,
+      "loss": 0.2102,
+      "step": 6460
+    },
+    {
+      "epoch": 136.21052631578948,
+      "grad_norm": 0.7158136367797852,
+      "learning_rate": 0.0001648888771898552,
+      "loss": 0.1978,
+      "step": 6470
+    },
+    {
+      "epoch": 136.42105263157896,
+      "grad_norm": 0.8900578618049622,
+      "learning_rate": 0.00016478710071266944,
+      "loss": 0.2019,
+      "step": 6480
+    },
+    {
+      "epoch": 136.6315789473684,
+      "grad_norm": 0.7282655239105225,
+      "learning_rate": 0.00016468520845042996,
+      "loss": 0.2081,
+      "step": 6490
+    },
+    {
+      "epoch": 136.8421052631579,
+      "grad_norm": 0.6675572395324707,
+      "learning_rate": 0.0001645832005852348,
+      "loss": 0.1943,
+      "step": 6500
+    },
+    {
+      "epoch": 137.05263157894737,
+      "grad_norm": 0.6044570803642273,
+      "learning_rate": 0.0001644810772993885,
+      "loss": 0.1966,
+      "step": 6510
+    },
+    {
+      "epoch": 137.26315789473685,
+      "grad_norm": 0.7982662320137024,
+      "learning_rate": 0.00016437883877540194,
+      "loss": 0.204,
+      "step": 6520
+    },
+    {
+      "epoch": 137.47368421052633,
+      "grad_norm": 0.6893635392189026,
+      "learning_rate": 0.00016427648519599196,
+      "loss": 0.1854,
+      "step": 6530
+    },
+    {
+      "epoch": 137.68421052631578,
+      "grad_norm": 0.7114787101745605,
+      "learning_rate": 0.000164174016744081,
+      "loss": 0.2067,
+      "step": 6540
+    },
+    {
+      "epoch": 137.89473684210526,
+      "grad_norm": 0.8275425434112549,
+      "learning_rate": 0.00016407143360279682,
+      "loss": 0.2046,
+      "step": 6550
+    },
+    {
+      "epoch": 138.10526315789474,
+      "grad_norm": 0.6906023025512695,
+      "learning_rate": 0.00016396873595547206,
+      "loss": 0.2047,
+      "step": 6560
+    },
+    {
+      "epoch": 138.31578947368422,
+      "grad_norm": 0.7150043845176697,
+      "learning_rate": 0.00016386592398564412,
+      "loss": 0.1906,
+      "step": 6570
+    },
+    {
+      "epoch": 138.52631578947367,
+      "grad_norm": 0.6663026213645935,
+      "learning_rate": 0.00016376299787705464,
+      "loss": 0.2021,
+      "step": 6580
+    },
+    {
+      "epoch": 138.73684210526315,
+      "grad_norm": 0.9189106822013855,
+      "learning_rate": 0.00016365995781364925,
+      "loss": 0.2195,
+      "step": 6590
+    },
+    {
+      "epoch": 138.94736842105263,
+      "grad_norm": 0.7078124284744263,
+      "learning_rate": 0.0001635568039795773,
+      "loss": 0.1993,
+      "step": 6600
+    },
+    {
+      "epoch": 139.1578947368421,
+      "grad_norm": 0.9718939065933228,
+      "learning_rate": 0.00016345353655919137,
+      "loss": 0.1884,
+      "step": 6610
+    },
+    {
+      "epoch": 139.3684210526316,
+      "grad_norm": 1.3638027906417847,
+      "learning_rate": 0.0001633501557370471,
+      "loss": 0.1878,
+      "step": 6620
+    },
+    {
+      "epoch": 139.57894736842104,
+      "grad_norm": 0.8305347561836243,
+      "learning_rate": 0.00016324666169790283,
+      "loss": 0.1939,
+      "step": 6630
+    },
+    {
+      "epoch": 139.78947368421052,
+      "grad_norm": 1.2636075019836426,
+      "learning_rate": 0.0001631430546267191,
+      "loss": 0.2179,
+      "step": 6640
+    },
+    {
+      "epoch": 140.0,
+      "grad_norm": 1.4744796752929688,
+      "learning_rate": 0.0001630393347086586,
+      "loss": 0.2319,
+      "step": 6650
+    },
+    {
+      "epoch": 140.21052631578948,
+      "grad_norm": 0.9998826384544373,
+      "learning_rate": 0.0001629355021290856,
+      "loss": 0.1919,
+      "step": 6660
+    },
+    {
+      "epoch": 140.42105263157896,
+      "grad_norm": 0.6231628060340881,
+      "learning_rate": 0.0001628315570735658,
+      "loss": 0.1977,
+      "step": 6670
+    },
+    {
+      "epoch": 140.6315789473684,
+      "grad_norm": 0.6121691465377808,
+      "learning_rate": 0.00016272749972786587,
+      "loss": 0.1989,
+      "step": 6680
+    },
+    {
+      "epoch": 140.8421052631579,
+      "grad_norm": 0.7017862200737,
+      "learning_rate": 0.00016262333027795313,
+      "loss": 0.2033,
+      "step": 6690
+    },
+    {
+      "epoch": 141.05263157894737,
+      "grad_norm": 0.6834351420402527,
+      "learning_rate": 0.0001625190489099953,
+      "loss": 0.2034,
+      "step": 6700
+    },
+    {
+      "epoch": 141.26315789473685,
+      "grad_norm": 0.8252644538879395,
+      "learning_rate": 0.00016241465581036009,
+      "loss": 0.1984,
+      "step": 6710
+    },
+    {
+      "epoch": 141.47368421052633,
+      "grad_norm": 0.830450713634491,
+      "learning_rate": 0.00016231015116561487,
+      "loss": 0.2015,
+      "step": 6720
+    },
+    {
+      "epoch": 141.68421052631578,
+      "grad_norm": 0.644212007522583,
+      "learning_rate": 0.0001622055351625264,
+      "loss": 0.1998,
+      "step": 6730
+    },
+    {
+      "epoch": 141.89473684210526,
+      "grad_norm": 0.8591092824935913,
+      "learning_rate": 0.00016210080798806042,
+      "loss": 0.2094,
+      "step": 6740
+    },
+    {
+      "epoch": 142.10526315789474,
+      "grad_norm": 0.7406374216079712,
+      "learning_rate": 0.00016199596982938142,
+      "loss": 0.2068,
+      "step": 6750
+    },
+    {
+      "epoch": 142.31578947368422,
+      "grad_norm": 0.7162677645683289,
+      "learning_rate": 0.00016189102087385218,
+      "loss": 0.2031,
+      "step": 6760
+    },
+    {
+      "epoch": 142.52631578947367,
+      "grad_norm": 0.7859647870063782,
+      "learning_rate": 0.00016178596130903344,
+      "loss": 0.1974,
+      "step": 6770
+    },
+    {
+      "epoch": 142.73684210526315,
+      "grad_norm": 0.6114941239356995,
+      "learning_rate": 0.00016168079132268374,
+      "loss": 0.1965,
+      "step": 6780
+    },
+    {
+      "epoch": 142.94736842105263,
+      "grad_norm": 0.6999493837356567,
+      "learning_rate": 0.00016157551110275887,
+      "loss": 0.2135,
+      "step": 6790
+    },
+    {
+      "epoch": 143.1578947368421,
+      "grad_norm": 0.6540025472640991,
+      "learning_rate": 0.00016147012083741168,
+      "loss": 0.1862,
+      "step": 6800
+    },
+    {
+      "epoch": 143.3684210526316,
+      "grad_norm": 0.7767214775085449,
+      "learning_rate": 0.0001613646207149916,
+      "loss": 0.2028,
+      "step": 6810
+    },
+    {
+      "epoch": 143.57894736842104,
+      "grad_norm": 0.8844462633132935,
+      "learning_rate": 0.00016125901092404457,
+      "loss": 0.2026,
+      "step": 6820
+    },
+    {
+      "epoch": 143.78947368421052,
+      "grad_norm": 0.6872785687446594,
+      "learning_rate": 0.00016115329165331227,
+      "loss": 0.206,
+      "step": 6830
+    },
+    {
+      "epoch": 144.0,
+      "grad_norm": 1.2136003971099854,
+      "learning_rate": 0.0001610474630917323,
+      "loss": 0.2072,
+      "step": 6840
+    },
+    {
+      "epoch": 144.21052631578948,
+      "grad_norm": 0.7713127136230469,
+      "learning_rate": 0.00016094152542843733,
+      "loss": 0.1923,
+      "step": 6850
+    },
+    {
+      "epoch": 144.42105263157896,
+      "grad_norm": 0.7771678566932678,
+      "learning_rate": 0.0001608354788527553,
+      "loss": 0.2069,
+      "step": 6860
+    },
+    {
+      "epoch": 144.6315789473684,
+      "grad_norm": 0.7056291699409485,
+      "learning_rate": 0.0001607293235542085,
+      "loss": 0.1953,
+      "step": 6870
+    },
+    {
+      "epoch": 144.8421052631579,
+      "grad_norm": 0.9615933299064636,
+      "learning_rate": 0.00016062305972251373,
+      "loss": 0.2048,
+      "step": 6880
+    },
+    {
+      "epoch": 145.05263157894737,
+      "grad_norm": 0.7184662222862244,
+      "learning_rate": 0.00016051668754758167,
+      "loss": 0.2147,
+      "step": 6890
+    },
+    {
+      "epoch": 145.26315789473685,
+      "grad_norm": 0.7650150656700134,
+      "learning_rate": 0.00016041020721951666,
+      "loss": 0.1993,
+      "step": 6900
+    },
+    {
+      "epoch": 145.47368421052633,
+      "grad_norm": 0.6379842162132263,
+      "learning_rate": 0.00016030361892861622,
+      "loss": 0.2053,
+      "step": 6910
+    },
+    {
+      "epoch": 145.68421052631578,
+      "grad_norm": 0.7708711624145508,
+      "learning_rate": 0.00016019692286537107,
+      "loss": 0.1985,
+      "step": 6920
+    },
+    {
+      "epoch": 145.89473684210526,
+      "grad_norm": 0.7167929410934448,
+      "learning_rate": 0.00016009011922046425,
+      "loss": 0.1991,
+      "step": 6930
+    },
+    {
+      "epoch": 146.10526315789474,
+      "grad_norm": 0.7796617746353149,
+      "learning_rate": 0.00015998320818477125,
+      "loss": 0.2039,
+      "step": 6940
+    },
+    {
+      "epoch": 146.31578947368422,
+      "grad_norm": 0.6694478988647461,
+      "learning_rate": 0.00015987618994935937,
+      "loss": 0.1863,
+      "step": 6950
+    },
+    {
+      "epoch": 146.52631578947367,
+      "grad_norm": 0.7410951256752014,
+      "learning_rate": 0.0001597690647054876,
+      "loss": 0.2019,
+      "step": 6960
+    },
+    {
+      "epoch": 146.73684210526315,
+      "grad_norm": 0.6894869804382324,
+      "learning_rate": 0.0001596618326446061,
+      "loss": 0.187,
+      "step": 6970
+    },
+    {
+      "epoch": 146.94736842105263,
+      "grad_norm": 0.6591702699661255,
+      "learning_rate": 0.00015955449395835597,
+      "loss": 0.208,
+      "step": 6980
+    },
+    {
+      "epoch": 147.1578947368421,
+      "grad_norm": 0.5875535607337952,
+      "learning_rate": 0.00015944704883856883,
+      "loss": 0.2059,
+      "step": 6990
+    },
+    {
+      "epoch": 147.3684210526316,
+      "grad_norm": 0.7606227993965149,
+      "learning_rate": 0.00015933949747726653,
+      "loss": 0.2009,
+      "step": 7000
+    },
+    {
+      "epoch": 147.57894736842104,
+      "grad_norm": 0.8052495718002319,
+      "learning_rate": 0.00015923184006666076,
+      "loss": 0.2118,
+      "step": 7010
+    },
+    {
+      "epoch": 147.78947368421052,
+      "grad_norm": 0.6055042743682861,
+      "learning_rate": 0.00015912407679915283,
+      "loss": 0.1936,
+      "step": 7020
+    },
+    {
+      "epoch": 148.0,
+      "grad_norm": 0.6587730050086975,
+      "learning_rate": 0.00015901620786733312,
+      "loss": 0.1942,
+      "step": 7030
+    },
+    {
+      "epoch": 148.21052631578948,
+      "grad_norm": 0.6731119751930237,
+      "learning_rate": 0.00015890823346398095,
+      "loss": 0.1803,
+      "step": 7040
+    },
+    {
+      "epoch": 148.42105263157896,
+      "grad_norm": 0.7457992434501648,
+      "learning_rate": 0.00015880015378206408,
+      "loss": 0.2079,
+      "step": 7050
+    },
+    {
+      "epoch": 148.6315789473684,
+      "grad_norm": 0.7178442478179932,
+      "learning_rate": 0.00015869196901473838,
+      "loss": 0.1932,
+      "step": 7060
+    },
+    {
+      "epoch": 148.8421052631579,
+      "grad_norm": 1.3482714891433716,
+      "learning_rate": 0.00015858367935534754,
+      "loss": 0.2095,
+      "step": 7070
+    },
+    {
+      "epoch": 149.05263157894737,
+      "grad_norm": 0.669923722743988,
+      "learning_rate": 0.00015847528499742287,
+      "loss": 0.2085,
+      "step": 7080
+    },
+    {
+      "epoch": 149.26315789473685,
+      "grad_norm": 0.6748956441879272,
+      "learning_rate": 0.00015836678613468256,
+      "loss": 0.1856,
+      "step": 7090
+    },
+    {
+      "epoch": 149.47368421052633,
+      "grad_norm": 0.7660018801689148,
+      "learning_rate": 0.0001582581829610317,
+      "loss": 0.1958,
+      "step": 7100
+    },
+    {
+      "epoch": 149.68421052631578,
+      "grad_norm": 0.8634538650512695,
+      "learning_rate": 0.00015814947567056178,
+      "loss": 0.2109,
+      "step": 7110
+    },
+    {
+      "epoch": 149.89473684210526,
+      "grad_norm": 0.713979959487915,
+      "learning_rate": 0.0001580406644575503,
+      "loss": 0.2067,
+      "step": 7120
+    },
+    {
+      "epoch": 150.10526315789474,
+      "grad_norm": 0.5765550136566162,
+      "learning_rate": 0.00015793174951646063,
+      "loss": 0.1914,
+      "step": 7130
+    },
+    {
+      "epoch": 150.31578947368422,
+      "grad_norm": 0.8116117119789124,
+      "learning_rate": 0.00015782273104194137,
+      "loss": 0.189,
+      "step": 7140
+    },
+    {
+      "epoch": 150.52631578947367,
+      "grad_norm": 0.7519946694374084,
+      "learning_rate": 0.00015771360922882624,
+      "loss": 0.202,
+      "step": 7150
+    },
+    {
+      "epoch": 150.73684210526315,
+      "grad_norm": 0.7752880454063416,
+      "learning_rate": 0.0001576043842721336,
+      "loss": 0.2098,
+      "step": 7160
+    },
+    {
+      "epoch": 150.94736842105263,
+      "grad_norm": 0.726283073425293,
+      "learning_rate": 0.0001574950563670661,
+      "loss": 0.1998,
+      "step": 7170
+    },
+    {
+      "epoch": 151.1578947368421,
+      "grad_norm": 0.7491242289543152,
+      "learning_rate": 0.00015738562570901055,
+      "loss": 0.2044,
+      "step": 7180
+    },
+    {
+      "epoch": 151.3684210526316,
+      "grad_norm": 0.9558908343315125,
+      "learning_rate": 0.00015727609249353722,
+      "loss": 0.2053,
+      "step": 7190
+    },
+    {
+      "epoch": 151.57894736842104,
+      "grad_norm": 0.6602309346199036,
+      "learning_rate": 0.00015716645691639966,
+      "loss": 0.1879,
+      "step": 7200
+    },
+    {
+      "epoch": 151.78947368421052,
+      "grad_norm": 0.6388182640075684,
+      "learning_rate": 0.00015705671917353456,
+      "loss": 0.2012,
+      "step": 7210
+    },
+    {
+      "epoch": 152.0,
+      "grad_norm": 0.653289258480072,
+      "learning_rate": 0.00015694687946106093,
+      "loss": 0.1997,
+      "step": 7220
+    },
+    {
+      "epoch": 152.21052631578948,
+      "grad_norm": 0.693202555179596,
+      "learning_rate": 0.00015683693797528022,
+      "loss": 0.1867,
+      "step": 7230
+    },
+    {
+      "epoch": 152.42105263157896,
+      "grad_norm": 0.8159785866737366,
+      "learning_rate": 0.00015672689491267567,
+      "loss": 0.1981,
+      "step": 7240
+    },
+    {
+      "epoch": 152.6315789473684,
+      "grad_norm": 0.604954183101654,
+      "learning_rate": 0.00015661675046991206,
+      "loss": 0.2078,
+      "step": 7250
+    },
+    {
+      "epoch": 152.8421052631579,
+      "grad_norm": 0.6773616075515747,
+      "learning_rate": 0.0001565065048438354,
+      "loss": 0.2006,
+      "step": 7260
+    },
+    {
+      "epoch": 153.05263157894737,
+      "grad_norm": 0.6346316933631897,
+      "learning_rate": 0.0001563961582314725,
+      "loss": 0.1864,
+      "step": 7270
+    },
+    {
+      "epoch": 153.26315789473685,
+      "grad_norm": 0.6382675170898438,
+      "learning_rate": 0.00015628571083003062,
+      "loss": 0.2017,
+      "step": 7280
+    },
+    {
+      "epoch": 153.47368421052633,
+      "grad_norm": 0.7643656134605408,
+      "learning_rate": 0.00015617516283689722,
+      "loss": 0.1972,
+      "step": 7290
+    },
+    {
+      "epoch": 153.68421052631578,
+      "grad_norm": 0.7513641715049744,
+      "learning_rate": 0.0001560645144496394,
+      "loss": 0.2053,
+      "step": 7300
+    },
+    {
+      "epoch": 153.89473684210526,
+      "grad_norm": 0.5776196122169495,
+      "learning_rate": 0.00015595376586600388,
+      "loss": 0.1991,
+      "step": 7310
+    },
+    {
+      "epoch": 154.10526315789474,
+      "grad_norm": 0.743189811706543,
+      "learning_rate": 0.00015584291728391625,
+      "loss": 0.1923,
+      "step": 7320
+    },
+    {
+      "epoch": 154.31578947368422,
+      "grad_norm": 0.6481560468673706,
+      "learning_rate": 0.00015573196890148093,
+      "loss": 0.201,
+      "step": 7330
+    },
+    {
+      "epoch": 154.52631578947367,
+      "grad_norm": 0.5811365842819214,
+      "learning_rate": 0.00015562092091698067,
+      "loss": 0.1926,
+      "step": 7340
+    },
+    {
+      "epoch": 154.73684210526315,
+      "grad_norm": 0.8171225190162659,
+      "learning_rate": 0.00015550977352887622,
+      "loss": 0.2118,
+      "step": 7350
+    },
+    {
+      "epoch": 154.94736842105263,
+      "grad_norm": 0.6557311415672302,
+      "learning_rate": 0.00015539852693580603,
+      "loss": 0.1983,
+      "step": 7360
+    },
+    {
+      "epoch": 155.1578947368421,
+      "grad_norm": 0.6950130462646484,
+      "learning_rate": 0.00015528718133658571,
+      "loss": 0.1867,
+      "step": 7370
+    },
+    {
+      "epoch": 155.3684210526316,
+      "grad_norm": 0.525143027305603,
+      "learning_rate": 0.00015517573693020798,
+      "loss": 0.1907,
+      "step": 7380
+    },
+    {
+      "epoch": 155.57894736842104,
+      "grad_norm": 0.6554726958274841,
+      "learning_rate": 0.00015506419391584202,
+      "loss": 0.2006,
+      "step": 7390
+    },
+    {
+      "epoch": 155.78947368421052,
+      "grad_norm": 0.8342046141624451,
+      "learning_rate": 0.00015495255249283328,
+      "loss": 0.2063,
+      "step": 7400
+    },
+    {
+      "epoch": 156.0,
+      "grad_norm": 0.7520895004272461,
+      "learning_rate": 0.00015484081286070312,
+      "loss": 0.1961,
+      "step": 7410
+    },
+    {
+      "epoch": 156.21052631578948,
+      "grad_norm": 0.7809695601463318,
+      "learning_rate": 0.00015472897521914836,
+      "loss": 0.2021,
+      "step": 7420
+    },
+    {
+      "epoch": 156.42105263157896,
+      "grad_norm": 0.6753067374229431,
+      "learning_rate": 0.00015461703976804095,
+      "loss": 0.1963,
+      "step": 7430
+    },
+    {
+      "epoch": 156.6315789473684,
+      "grad_norm": 0.9035070538520813,
+      "learning_rate": 0.0001545050067074278,
+      "loss": 0.2009,
+      "step": 7440
+    },
+    {
+      "epoch": 156.8421052631579,
+      "grad_norm": 0.6189687252044678,
+      "learning_rate": 0.00015439287623753007,
+      "loss": 0.1901,
+      "step": 7450
+    },
+    {
+      "epoch": 157.05263157894737,
+      "grad_norm": 0.6014378666877747,
+      "learning_rate": 0.00015428064855874308,
+      "loss": 0.206,
+      "step": 7460
+    },
+    {
+      "epoch": 157.26315789473685,
+      "grad_norm": 0.73002028465271,
+      "learning_rate": 0.00015416832387163596,
+      "loss": 0.2043,
+      "step": 7470
+    },
+    {
+      "epoch": 157.47368421052633,
+      "grad_norm": 0.6451747417449951,
+      "learning_rate": 0.0001540559023769511,
+      "loss": 0.1884,
+      "step": 7480
+    },
+    {
+      "epoch": 157.68421052631578,
+      "grad_norm": 0.7301135659217834,
+      "learning_rate": 0.00015394338427560396,
+      "loss": 0.1977,
+      "step": 7490
+    },
+    {
+      "epoch": 157.89473684210526,
+      "grad_norm": 0.6060409545898438,
+      "learning_rate": 0.0001538307697686826,
+      "loss": 0.2155,
+      "step": 7500
+    },
+    {
+      "epoch": 158.10526315789474,
+      "grad_norm": 0.571553647518158,
+      "learning_rate": 0.00015371805905744736,
+      "loss": 0.1992,
+      "step": 7510
+    },
+    {
+      "epoch": 158.31578947368422,
+      "grad_norm": 0.5945022106170654,
+      "learning_rate": 0.00015360525234333066,
+      "loss": 0.1823,
+      "step": 7520
+    },
+    {
+      "epoch": 158.52631578947367,
+      "grad_norm": 0.7339075803756714,
+      "learning_rate": 0.00015349234982793634,
+      "loss": 0.2059,
+      "step": 7530
+    },
+    {
+      "epoch": 158.73684210526315,
+      "grad_norm": 0.633240282535553,
+      "learning_rate": 0.00015337935171303948,
+      "loss": 0.1908,
+      "step": 7540
+    },
+    {
+      "epoch": 158.94736842105263,
+      "grad_norm": 0.6861400008201599,
+      "learning_rate": 0.00015326625820058612,
+      "loss": 0.2089,
+      "step": 7550
+    },
+    {
+      "epoch": 159.1578947368421,
+      "grad_norm": 0.46820777654647827,
+      "learning_rate": 0.00015315306949269255,
+      "loss": 0.1958,
+      "step": 7560
+    },
+    {
+      "epoch": 159.3684210526316,
+      "grad_norm": 0.7215929627418518,
+      "learning_rate": 0.00015303978579164545,
+      "loss": 0.1922,
+      "step": 7570
+    },
+    {
+      "epoch": 159.57894736842104,
+      "grad_norm": 0.5514410138130188,
+      "learning_rate": 0.00015292640729990117,
+      "loss": 0.1902,
+      "step": 7580
+    },
+    {
+      "epoch": 159.78947368421052,
+      "grad_norm": 0.807755172252655,
+      "learning_rate": 0.00015281293422008543,
+      "loss": 0.2058,
+      "step": 7590
+    },
+    {
+      "epoch": 160.0,
+      "grad_norm": 0.8933700323104858,
+      "learning_rate": 0.00015269936675499306,
+      "loss": 0.2119,
+      "step": 7600
+    },
+    {
+      "epoch": 160.21052631578948,
+      "grad_norm": 0.7081916332244873,
+      "learning_rate": 0.00015258570510758745,
+      "loss": 0.195,
+      "step": 7610
+    },
+    {
+      "epoch": 160.42105263157896,
+      "grad_norm": 0.6982260942459106,
+      "learning_rate": 0.00015247194948100047,
+      "loss": 0.1978,
+      "step": 7620
+    },
+    {
+      "epoch": 160.6315789473684,
+      "grad_norm": 0.7810565233230591,
+      "learning_rate": 0.00015235810007853179,
+      "loss": 0.1982,
+      "step": 7630
+    },
+    {
+      "epoch": 160.8421052631579,
+      "grad_norm": 0.679720938205719,
+      "learning_rate": 0.00015224415710364883,
+      "loss": 0.2144,
+      "step": 7640
+    },
+    {
+      "epoch": 161.05263157894737,
+      "grad_norm": 0.6578332185745239,
+      "learning_rate": 0.00015213012075998615,
+      "loss": 0.1944,
+      "step": 7650
+    },
+    {
+      "epoch": 161.26315789473685,
+      "grad_norm": 0.7118348479270935,
+      "learning_rate": 0.00015201599125134517,
+      "loss": 0.2015,
+      "step": 7660
+    },
+    {
+      "epoch": 161.47368421052633,
+      "grad_norm": 0.5404331684112549,
+      "learning_rate": 0.00015190176878169384,
+      "loss": 0.1995,
+      "step": 7670
+    },
+    {
+      "epoch": 161.68421052631578,
+      "grad_norm": 0.685939371585846,
+      "learning_rate": 0.0001517874535551662,
+      "loss": 0.2089,
+      "step": 7680
+    },
+    {
+      "epoch": 161.89473684210526,
+      "grad_norm": 0.6666051745414734,
+      "learning_rate": 0.0001516730457760621,
+      "loss": 0.1867,
+      "step": 7690
+    },
+    {
+      "epoch": 162.10526315789474,
+      "grad_norm": 0.7303677797317505,
+      "learning_rate": 0.0001515585456488468,
+      "loss": 0.1986,
+      "step": 7700
+    },
+    {
+      "epoch": 162.31578947368422,
+      "grad_norm": 0.7837722301483154,
+      "learning_rate": 0.00015144395337815064,
+      "loss": 0.1879,
+      "step": 7710
+    },
+    {
+      "epoch": 162.52631578947367,
+      "grad_norm": 0.6972552537918091,
+      "learning_rate": 0.00015132926916876856,
+      "loss": 0.187,
+      "step": 7720
+    },
+    {
+      "epoch": 162.73684210526315,
+      "grad_norm": 0.7570910453796387,
+      "learning_rate": 0.0001512144932256598,
+      "loss": 0.216,
+      "step": 7730
+    },
+    {
+      "epoch": 162.94736842105263,
+      "grad_norm": 0.5790467858314514,
+      "learning_rate": 0.0001510996257539476,
+      "loss": 0.2098,
+      "step": 7740
+    },
+    {
+      "epoch": 163.1578947368421,
+      "grad_norm": 0.6272661089897156,
+      "learning_rate": 0.0001509846669589188,
+      "loss": 0.1912,
+      "step": 7750
+    },
+    {
+      "epoch": 163.3684210526316,
+      "grad_norm": 0.9367342591285706,
+      "learning_rate": 0.0001508696170460233,
+      "loss": 0.1912,
+      "step": 7760
+    },
+    {
+      "epoch": 163.57894736842104,
+      "grad_norm": 0.5576528310775757,
+      "learning_rate": 0.00015075447622087408,
+      "loss": 0.194,
+      "step": 7770
+    },
+    {
+      "epoch": 163.78947368421052,
+      "grad_norm": 0.7530799508094788,
+      "learning_rate": 0.0001506392446892464,
+      "loss": 0.2107,
+      "step": 7780
+    },
+    {
+      "epoch": 164.0,
+      "grad_norm": 0.7183924913406372,
+      "learning_rate": 0.00015052392265707767,
+      "loss": 0.1968,
+      "step": 7790
+    },
+    {
+      "epoch": 164.21052631578948,
+      "grad_norm": 0.5782009959220886,
+      "learning_rate": 0.0001504085103304671,
+      "loss": 0.1872,
+      "step": 7800
+    },
+    {
+      "epoch": 164.42105263157896,
+      "grad_norm": 0.7883465886116028,
+      "learning_rate": 0.0001502930079156752,
+      "loss": 0.1902,
+      "step": 7810
+    },
+    {
+      "epoch": 164.6315789473684,
+      "grad_norm": 0.6550580263137817,
+      "learning_rate": 0.00015017741561912352,
+      "loss": 0.198,
+      "step": 7820
+    },
+    {
+      "epoch": 164.8421052631579,
+      "grad_norm": 0.6090496182441711,
+      "learning_rate": 0.00015006173364739427,
+      "loss": 0.1965,
+      "step": 7830
+    },
+    {
+      "epoch": 165.05263157894737,
+      "grad_norm": 0.6356409192085266,
+      "learning_rate": 0.00014994596220722987,
+      "loss": 0.2125,
+      "step": 7840
+    },
+    {
+      "epoch": 165.26315789473685,
+      "grad_norm": 0.5657672882080078,
+      "learning_rate": 0.00014983010150553262,
+      "loss": 0.2024,
+      "step": 7850
+    },
+    {
+      "epoch": 165.47368421052633,
+      "grad_norm": 0.7943452596664429,
+      "learning_rate": 0.00014971415174936444,
+      "loss": 0.1978,
+      "step": 7860
+    },
+    {
+      "epoch": 165.68421052631578,
+      "grad_norm": 0.5848918557167053,
+      "learning_rate": 0.00014959811314594628,
+      "loss": 0.1951,
+      "step": 7870
+    },
+    {
+      "epoch": 165.89473684210526,
+      "grad_norm": 0.6190322041511536,
+      "learning_rate": 0.000149481985902658,
+      "loss": 0.1859,
+      "step": 7880
+    },
+    {
+      "epoch": 166.10526315789474,
+      "grad_norm": 0.6574214100837708,
+      "learning_rate": 0.00014936577022703777,
+      "loss": 0.2012,
+      "step": 7890
+    },
+    {
+      "epoch": 166.31578947368422,
+      "grad_norm": 0.6191821694374084,
+      "learning_rate": 0.00014924946632678186,
+      "loss": 0.1964,
+      "step": 7900
+    },
+    {
+      "epoch": 166.52631578947367,
+      "grad_norm": 0.5656545162200928,
+      "learning_rate": 0.0001491330744097442,
+      "loss": 0.1917,
+      "step": 7910
+    },
+    {
+      "epoch": 166.73684210526315,
+      "grad_norm": 0.6235769987106323,
+      "learning_rate": 0.00014901659468393602,
+      "loss": 0.1846,
+      "step": 7920
+    },
+    {
+      "epoch": 166.94736842105263,
+      "grad_norm": 0.753472626209259,
+      "learning_rate": 0.00014890002735752547,
+      "loss": 0.2266,
+      "step": 7930
+    },
+    {
+      "epoch": 167.1578947368421,
+      "grad_norm": 0.6449782252311707,
+      "learning_rate": 0.00014878337263883728,
+      "loss": 0.1922,
+      "step": 7940
+    },
+    {
+      "epoch": 167.3684210526316,
+      "grad_norm": 0.7990699410438538,
+      "learning_rate": 0.00014866663073635232,
+      "loss": 0.1932,
+      "step": 7950
+    },
+    {
+      "epoch": 167.57894736842104,
+      "grad_norm": 0.7350342273712158,
+      "learning_rate": 0.00014854980185870733,
+      "loss": 0.2028,
+      "step": 7960
+    },
+    {
+      "epoch": 167.78947368421052,
+      "grad_norm": 0.5282038450241089,
+      "learning_rate": 0.00014843288621469442,
+      "loss": 0.2125,
+      "step": 7970
+    },
+    {
+      "epoch": 168.0,
+      "grad_norm": 0.6253044605255127,
+      "learning_rate": 0.00014831588401326083,
+      "loss": 0.1925,
+      "step": 7980
+    },
+    {
+      "epoch": 168.21052631578948,
+      "grad_norm": 0.8048043847084045,
+      "learning_rate": 0.00014819879546350842,
+      "loss": 0.2032,
+      "step": 7990
+    },
+    {
+      "epoch": 168.42105263157896,
+      "grad_norm": 0.6136367917060852,
+      "learning_rate": 0.00014808162077469347,
+      "loss": 0.1854,
+      "step": 8000
+    },
+    {
+      "epoch": 168.6315789473684,
+      "grad_norm": 0.6089524626731873,
+      "learning_rate": 0.00014796436015622618,
+      "loss": 0.1951,
+      "step": 8010
+    },
+    {
+      "epoch": 168.8421052631579,
+      "grad_norm": 0.6197385787963867,
+      "learning_rate": 0.00014784701381767018,
+      "loss": 0.2084,
+      "step": 8020
+    },
+    {
+      "epoch": 169.05263157894737,
+      "grad_norm": 0.5597293376922607,
+      "learning_rate": 0.00014772958196874246,
+      "loss": 0.1921,
+      "step": 8030
+    },
+    {
+      "epoch": 169.26315789473685,
+      "grad_norm": 0.5773986577987671,
+      "learning_rate": 0.00014761206481931282,
+      "loss": 0.208,
+      "step": 8040
+    },
+    {
+      "epoch": 169.47368421052633,
+      "grad_norm": 0.6238757967948914,
+      "learning_rate": 0.00014749446257940335,
+      "loss": 0.1902,
+      "step": 8050
+    },
+    {
+      "epoch": 169.68421052631578,
+      "grad_norm": 0.6683410406112671,
+      "learning_rate": 0.00014737677545918843,
+      "loss": 0.187,
+      "step": 8060
+    },
+    {
+      "epoch": 169.89473684210526,
+      "grad_norm": 0.8527116179466248,
+      "learning_rate": 0.000147259003668994,
+      "loss": 0.2193,
+      "step": 8070
+    },
+    {
+      "epoch": 170.10526315789474,
+      "grad_norm": 0.6407440900802612,
+      "learning_rate": 0.00014714114741929728,
+      "loss": 0.1941,
+      "step": 8080
+    },
+    {
+      "epoch": 170.31578947368422,
+      "grad_norm": 0.6082753539085388,
+      "learning_rate": 0.00014702320692072657,
+      "loss": 0.2025,
+      "step": 8090
+    },
+    {
+      "epoch": 170.52631578947367,
+      "grad_norm": 0.7939004302024841,
+      "learning_rate": 0.00014690518238406064,
+      "loss": 0.1986,
+      "step": 8100
+    },
+    {
+      "epoch": 170.73684210526315,
+      "grad_norm": 0.5582263469696045,
+      "learning_rate": 0.00014678707402022845,
+      "loss": 0.1946,
+      "step": 8110
+    },
+    {
+      "epoch": 170.94736842105263,
+      "grad_norm": 0.5746979117393494,
+      "learning_rate": 0.00014666888204030885,
+      "loss": 0.2034,
+      "step": 8120
+    },
+    {
+      "epoch": 171.1578947368421,
+      "grad_norm": 0.825444757938385,
+      "learning_rate": 0.00014655060665553005,
+      "loss": 0.1941,
+      "step": 8130
+    },
+    {
+      "epoch": 171.3684210526316,
+      "grad_norm": 0.9377308487892151,
+      "learning_rate": 0.0001464322480772693,
+      "loss": 0.2003,
+      "step": 8140
+    },
+    {
+      "epoch": 171.57894736842104,
+      "grad_norm": 0.6175430417060852,
+      "learning_rate": 0.0001463138065170526,
+      "loss": 0.1862,
+      "step": 8150
+    },
+    {
+      "epoch": 171.78947368421052,
+      "grad_norm": 0.6299969553947449,
+      "learning_rate": 0.00014619528218655424,
+      "loss": 0.2055,
+      "step": 8160
+    },
+    {
+      "epoch": 172.0,
+      "grad_norm": 0.8263299465179443,
+      "learning_rate": 0.00014607667529759635,
+      "loss": 0.2038,
+      "step": 8170
+    },
+    {
+      "epoch": 172.21052631578948,
+      "grad_norm": 0.522394061088562,
+      "learning_rate": 0.00014595798606214882,
+      "loss": 0.1864,
+      "step": 8180
+    },
+    {
+      "epoch": 172.42105263157896,
+      "grad_norm": 0.8208839893341064,
+      "learning_rate": 0.00014583921469232838,
+      "loss": 0.2064,
+      "step": 8190
+    },
+    {
+      "epoch": 172.6315789473684,
+      "grad_norm": 0.6659351587295532,
+      "learning_rate": 0.00014572036140039885,
+      "loss": 0.2048,
+      "step": 8200
+    },
+    {
+      "epoch": 172.8421052631579,
+      "grad_norm": 0.7370153665542603,
+      "learning_rate": 0.00014560142639877025,
+      "loss": 0.1909,
+      "step": 8210
+    },
+    {
+      "epoch": 173.05263157894737,
+      "grad_norm": 0.6763059496879578,
+      "learning_rate": 0.0001454824098999988,
+      "loss": 0.1938,
+      "step": 8220
+    },
+    {
+      "epoch": 173.26315789473685,
+      "grad_norm": 0.8497671484947205,
+      "learning_rate": 0.0001453633121167862,
+      "loss": 0.1956,
+      "step": 8230
+    },
+    {
+      "epoch": 173.47368421052633,
+      "grad_norm": 0.599168598651886,
+      "learning_rate": 0.00014524413326197952,
+      "loss": 0.1873,
+      "step": 8240
+    },
+    {
+      "epoch": 173.68421052631578,
+      "grad_norm": 0.6262931823730469,
+      "learning_rate": 0.00014512487354857075,
+      "loss": 0.2077,
+      "step": 8250
+    },
+    {
+      "epoch": 173.89473684210526,
+      "grad_norm": 0.6336191296577454,
+      "learning_rate": 0.00014500553318969628,
+      "loss": 0.2007,
+      "step": 8260
+    },
+    {
+      "epoch": 174.10526315789474,
+      "grad_norm": 0.6542587876319885,
+      "learning_rate": 0.00014488611239863667,
+      "loss": 0.1922,
+      "step": 8270
+    },
+    {
+      "epoch": 174.31578947368422,
+      "grad_norm": 0.5030838251113892,
+      "learning_rate": 0.00014476661138881629,
+      "loss": 0.1903,
+      "step": 8280
+    },
+    {
+      "epoch": 174.52631578947367,
+      "grad_norm": 0.5871732831001282,
+      "learning_rate": 0.00014464703037380278,
+      "loss": 0.1861,
+      "step": 8290
+    },
+    {
+      "epoch": 174.73684210526315,
+      "grad_norm": 0.9264737963676453,
+      "learning_rate": 0.00014452736956730683,
+      "loss": 0.2027,
+      "step": 8300
+    },
+    {
+      "epoch": 174.94736842105263,
+      "grad_norm": 0.5170317888259888,
+      "learning_rate": 0.0001444076291831817,
+      "loss": 0.2027,
+      "step": 8310
+    },
+    {
+      "epoch": 175.1578947368421,
+      "grad_norm": 0.5175766348838806,
+      "learning_rate": 0.00014428780943542285,
+      "loss": 0.2065,
+      "step": 8320
+    },
+    {
+      "epoch": 175.3684210526316,
+      "grad_norm": 0.7319501042366028,
+      "learning_rate": 0.0001441679105381676,
+      "loss": 0.1838,
+      "step": 8330
+    },
+    {
+      "epoch": 175.57894736842104,
+      "grad_norm": 0.7036318182945251,
+      "learning_rate": 0.00014404793270569475,
+      "loss": 0.1948,
+      "step": 8340
+    },
+    {
+      "epoch": 175.78947368421052,
+      "grad_norm": 0.7479885220527649,
+      "learning_rate": 0.0001439278761524241,
+      "loss": 0.2003,
+      "step": 8350
+    },
+    {
+      "epoch": 176.0,
+      "grad_norm": 0.7207411527633667,
+      "learning_rate": 0.0001438077410929162,
+      "loss": 0.2014,
+      "step": 8360
+    },
+    {
+      "epoch": 176.21052631578948,
+      "grad_norm": 0.7896419167518616,
+      "learning_rate": 0.00014368752774187186,
+      "loss": 0.1973,
+      "step": 8370
+    },
+    {
+      "epoch": 176.42105263157896,
+      "grad_norm": 0.7755358815193176,
+      "learning_rate": 0.00014356723631413188,
+      "loss": 0.1972,
+      "step": 8380
+    },
+    {
+      "epoch": 176.6315789473684,
+      "grad_norm": 0.855574905872345,
+      "learning_rate": 0.00014344686702467648,
+      "loss": 0.2078,
+      "step": 8390
+    },
+    {
+      "epoch": 176.8421052631579,
+      "grad_norm": 0.753102719783783,
+      "learning_rate": 0.00014332642008862514,
+      "loss": 0.1881,
+      "step": 8400
+    },
+    {
+      "epoch": 177.05263157894737,
+      "grad_norm": 0.5475196838378906,
+      "learning_rate": 0.00014320589572123607,
+      "loss": 0.1862,
+      "step": 8410
+    },
+    {
+      "epoch": 177.26315789473685,
+      "grad_norm": 0.7424550652503967,
+      "learning_rate": 0.0001430852941379058,
+      "loss": 0.1923,
+      "step": 8420
+    },
+    {
+      "epoch": 177.47368421052633,
+      "grad_norm": 0.5902268290519714,
+      "learning_rate": 0.000142964615554169,
+      "loss": 0.1925,
+      "step": 8430
+    },
+    {
+      "epoch": 177.68421052631578,
+      "grad_norm": 0.6847654581069946,
+      "learning_rate": 0.0001428438601856978,
+      "loss": 0.2016,
+      "step": 8440
+    },
+    {
+      "epoch": 177.89473684210526,
+      "grad_norm": 0.81497722864151,
+      "learning_rate": 0.00014272302824830166,
+      "loss": 0.1913,
+      "step": 8450
+    },
+    {
+      "epoch": 178.10526315789474,
+      "grad_norm": 0.6223244071006775,
+      "learning_rate": 0.00014260211995792679,
+      "loss": 0.1937,
+      "step": 8460
+    },
+    {
+      "epoch": 178.31578947368422,
+      "grad_norm": 0.8048047423362732,
+      "learning_rate": 0.00014248113553065597,
+      "loss": 0.1931,
+      "step": 8470
+    },
+    {
+      "epoch": 178.52631578947367,
+      "grad_norm": 0.606454074382782,
+      "learning_rate": 0.00014236007518270797,
+      "loss": 0.2105,
+      "step": 8480
+    },
+    {
+      "epoch": 178.73684210526315,
+      "grad_norm": 0.6434724926948547,
+      "learning_rate": 0.00014223893913043725,
+      "loss": 0.1907,
+      "step": 8490
+    },
+    {
+      "epoch": 178.94736842105263,
+      "grad_norm": 0.6487394571304321,
+      "learning_rate": 0.00014211772759033359,
+      "loss": 0.1984,
+      "step": 8500
+    },
+    {
+      "epoch": 179.1578947368421,
+      "grad_norm": 0.7166820168495178,
+      "learning_rate": 0.00014199644077902165,
+      "loss": 0.2023,
+      "step": 8510
+    },
+    {
+      "epoch": 179.3684210526316,
+      "grad_norm": 0.684614360332489,
+      "learning_rate": 0.00014187507891326063,
+      "loss": 0.1902,
+      "step": 8520
+    },
+    {
+      "epoch": 179.57894736842104,
+      "grad_norm": 0.5545143485069275,
+      "learning_rate": 0.00014175364220994388,
+      "loss": 0.1909,
+      "step": 8530
+    },
+    {
+      "epoch": 179.78947368421052,
+      "grad_norm": 0.7007152438163757,
+      "learning_rate": 0.00014163213088609847,
+      "loss": 0.2038,
+      "step": 8540
+    },
+    {
+      "epoch": 180.0,
+      "grad_norm": 0.8921439051628113,
+      "learning_rate": 0.00014151054515888482,
+      "loss": 0.2096,
+      "step": 8550
+    },
+    {
+      "epoch": 180.21052631578948,
+      "grad_norm": 0.5631104707717896,
+      "learning_rate": 0.00014138888524559636,
+      "loss": 0.1777,
+      "step": 8560
+    },
+    {
+      "epoch": 180.42105263157896,
+      "grad_norm": 0.6226398348808289,
+      "learning_rate": 0.0001412671513636591,
+      "loss": 0.2028,
+      "step": 8570
+    },
+    {
+      "epoch": 180.6315789473684,
+      "grad_norm": 0.5310755968093872,
+      "learning_rate": 0.00014114534373063113,
+      "loss": 0.2058,
+      "step": 8580
+    },
+    {
+      "epoch": 180.8421052631579,
+      "grad_norm": 0.7497243285179138,
+      "learning_rate": 0.00014102346256420257,
+      "loss": 0.197,
+      "step": 8590
+    },
+    {
+      "epoch": 181.05263157894737,
+      "grad_norm": 0.6481735706329346,
+      "learning_rate": 0.00014090150808219475,
+      "loss": 0.2011,
+      "step": 8600
+    },
+    {
+      "epoch": 181.26315789473685,
+      "grad_norm": 0.4959108829498291,
+      "learning_rate": 0.0001407794805025601,
+      "loss": 0.1873,
+      "step": 8610
+    },
+    {
+      "epoch": 181.47368421052633,
+      "grad_norm": 0.6530917882919312,
+      "learning_rate": 0.00014065738004338175,
+      "loss": 0.1928,
+      "step": 8620
+    },
+    {
+      "epoch": 181.68421052631578,
+      "grad_norm": 0.819771945476532,
+      "learning_rate": 0.00014053520692287297,
+      "loss": 0.1971,
+      "step": 8630
+    },
+    {
+      "epoch": 181.89473684210526,
+      "grad_norm": 0.8146491050720215,
+      "learning_rate": 0.00014041296135937692,
+      "loss": 0.2056,
+      "step": 8640
+    },
+    {
+      "epoch": 182.10526315789474,
+      "grad_norm": 0.6754285097122192,
+      "learning_rate": 0.00014029064357136628,
+      "loss": 0.1985,
+      "step": 8650
+    },
+    {
+      "epoch": 182.31578947368422,
+      "grad_norm": 0.651917576789856,
+      "learning_rate": 0.00014016825377744275,
+      "loss": 0.1852,
+      "step": 8660
+    },
+    {
+      "epoch": 182.52631578947367,
+      "grad_norm": 0.5566385984420776,
+      "learning_rate": 0.0001400457921963368,
+      "loss": 0.1838,
+      "step": 8670
+    },
+    {
+      "epoch": 182.73684210526315,
+      "grad_norm": 0.6756837368011475,
+      "learning_rate": 0.00013992325904690697,
+      "loss": 0.2143,
+      "step": 8680
+    },
+    {
+      "epoch": 182.94736842105263,
+      "grad_norm": 0.5130738019943237,
+      "learning_rate": 0.00013980065454814004,
+      "loss": 0.1997,
+      "step": 8690
+    },
+    {
+      "epoch": 183.1578947368421,
+      "grad_norm": 0.69854336977005,
+      "learning_rate": 0.00013967797891915003,
+      "loss": 0.1889,
+      "step": 8700
+    },
+    {
+      "epoch": 183.3684210526316,
+      "grad_norm": 0.8691510558128357,
+      "learning_rate": 0.00013955523237917824,
+      "loss": 0.2069,
+      "step": 8710
+    },
+    {
+      "epoch": 183.57894736842104,
+      "grad_norm": 0.5549122095108032,
+      "learning_rate": 0.00013943241514759262,
+      "loss": 0.1862,
+      "step": 8720
+    },
+    {
+      "epoch": 183.78947368421052,
+      "grad_norm": 0.6922882199287415,
+      "learning_rate": 0.00013930952744388743,
+      "loss": 0.2016,
+      "step": 8730
+    },
+    {
+      "epoch": 184.0,
+      "grad_norm": 1.2639508247375488,
+      "learning_rate": 0.000139186569487683,
+      "loss": 0.2071,
+      "step": 8740
+    },
+    {
+      "epoch": 184.21052631578948,
+      "grad_norm": 0.5515004396438599,
+      "learning_rate": 0.00013906354149872504,
+      "loss": 0.1907,
+      "step": 8750
+    },
+    {
+      "epoch": 184.42105263157896,
+      "grad_norm": 0.571441113948822,
+      "learning_rate": 0.00013894044369688462,
+      "loss": 0.1918,
+      "step": 8760
+    },
+    {
+      "epoch": 184.6315789473684,
+      "grad_norm": 0.7092531323432922,
+      "learning_rate": 0.00013881727630215738,
+      "loss": 0.1966,
+      "step": 8770
+    },
+    {
+      "epoch": 184.8421052631579,
+      "grad_norm": 0.8415676951408386,
+      "learning_rate": 0.00013869403953466346,
+      "loss": 0.2004,
+      "step": 8780
+    },
+    {
+      "epoch": 185.05263157894737,
+      "grad_norm": 1.0705996751785278,
+      "learning_rate": 0.00013857073361464697,
+      "loss": 0.209,
+      "step": 8790
+    },
+    {
+      "epoch": 185.26315789473685,
+      "grad_norm": 0.530075192451477,
+      "learning_rate": 0.00013844735876247558,
+      "loss": 0.1937,
+      "step": 8800
+    },
+    {
+      "epoch": 185.47368421052633,
+      "grad_norm": 0.5449802875518799,
+      "learning_rate": 0.00013832391519864008,
+      "loss": 0.1881,
+      "step": 8810
+    },
+    {
+      "epoch": 185.68421052631578,
+      "grad_norm": 0.7269825339317322,
+      "learning_rate": 0.00013820040314375422,
+      "loss": 0.1986,
+      "step": 8820
+    },
+    {
+      "epoch": 185.89473684210526,
+      "grad_norm": 1.3125274181365967,
+      "learning_rate": 0.00013807682281855404,
+      "loss": 0.203,
+      "step": 8830
+    },
+    {
+      "epoch": 186.10526315789474,
+      "grad_norm": 0.5790199041366577,
+      "learning_rate": 0.00013795317444389763,
+      "loss": 0.1958,
+      "step": 8840
+    },
+    {
+      "epoch": 186.31578947368422,
+      "grad_norm": 0.637947142124176,
+      "learning_rate": 0.00013782945824076465,
+      "loss": 0.2031,
+      "step": 8850
+    },
+    {
+      "epoch": 186.52631578947367,
+      "grad_norm": 0.561091959476471,
+      "learning_rate": 0.00013770567443025606,
+      "loss": 0.1828,
+      "step": 8860
+    },
+    {
+      "epoch": 186.73684210526315,
+      "grad_norm": 0.48705366253852844,
+      "learning_rate": 0.00013758182323359355,
+      "loss": 0.1902,
+      "step": 8870
+    },
+    {
+      "epoch": 186.94736842105263,
+      "grad_norm": 0.4845261573791504,
+      "learning_rate": 0.0001374579048721193,
+      "loss": 0.2079,
+      "step": 8880
+    },
+    {
+      "epoch": 187.1578947368421,
+      "grad_norm": 0.6765108108520508,
+      "learning_rate": 0.00013733391956729555,
+      "loss": 0.1987,
+      "step": 8890
+    },
+    {
+      "epoch": 187.3684210526316,
+      "grad_norm": 0.6771337389945984,
+      "learning_rate": 0.00013720986754070413,
+      "loss": 0.2101,
+      "step": 8900
+    },
+    {
+      "epoch": 187.57894736842104,
+      "grad_norm": 0.8644657731056213,
+      "learning_rate": 0.00013708574901404613,
+      "loss": 0.1934,
+      "step": 8910
+    },
+    {
+      "epoch": 187.78947368421052,
+      "grad_norm": 0.733113169670105,
+      "learning_rate": 0.00013696156420914146,
+      "loss": 0.2098,
+      "step": 8920
+    },
+    {
+      "epoch": 188.0,
+      "grad_norm": 0.7218254208564758,
+      "learning_rate": 0.0001368373133479285,
+      "loss": 0.1852,
+      "step": 8930
+    },
+    {
+      "epoch": 188.21052631578948,
+      "grad_norm": 0.7735620141029358,
+      "learning_rate": 0.0001367129966524637,
+      "loss": 0.1996,
+      "step": 8940
+    },
+    {
+      "epoch": 188.42105263157896,
+      "grad_norm": 0.6765386462211609,
+      "learning_rate": 0.00013658861434492117,
+      "loss": 0.1981,
+      "step": 8950
+    },
+    {
+      "epoch": 188.6315789473684,
+      "grad_norm": 0.7172536849975586,
+      "learning_rate": 0.00013646416664759222,
+      "loss": 0.1918,
+      "step": 8960
+    },
+    {
+      "epoch": 188.8421052631579,
+      "grad_norm": 0.6445688605308533,
+      "learning_rate": 0.00013633965378288509,
+      "loss": 0.2076,
+      "step": 8970
+    },
+    {
+      "epoch": 189.05263157894737,
+      "grad_norm": 0.6303783655166626,
+      "learning_rate": 0.00013621507597332447,
+      "loss": 0.1827,
+      "step": 8980
+    },
+    {
+      "epoch": 189.26315789473685,
+      "grad_norm": 0.6926207542419434,
+      "learning_rate": 0.00013609043344155108,
+      "loss": 0.1913,
+      "step": 8990
+    },
+    {
+      "epoch": 189.47368421052633,
+      "grad_norm": 0.4824252426624298,
+      "learning_rate": 0.00013596572641032132,
+      "loss": 0.1955,
+      "step": 9000
+    },
+    {
+      "epoch": 189.68421052631578,
+      "grad_norm": 0.5803983807563782,
+      "learning_rate": 0.00013584095510250693,
+      "loss": 0.1982,
+      "step": 9010
+    },
+    {
+      "epoch": 189.89473684210526,
+      "grad_norm": 0.6127357482910156,
+      "learning_rate": 0.0001357161197410944,
+      "loss": 0.2044,
+      "step": 9020
+    },
+    {
+      "epoch": 190.10526315789474,
+      "grad_norm": 0.5252726674079895,
+      "learning_rate": 0.00013559122054918483,
+      "loss": 0.1955,
+      "step": 9030
+    },
+    {
+      "epoch": 190.31578947368422,
+      "grad_norm": 0.6263585090637207,
+      "learning_rate": 0.00013546625774999327,
+      "loss": 0.1971,
+      "step": 9040
+    },
+    {
+      "epoch": 190.52631578947367,
+      "grad_norm": 0.5231578946113586,
+      "learning_rate": 0.00013534123156684852,
+      "loss": 0.2063,
+      "step": 9050
+    },
+    {
+      "epoch": 190.73684210526315,
+      "grad_norm": 0.6962978839874268,
+      "learning_rate": 0.00013521614222319268,
+      "loss": 0.1995,
+      "step": 9060
+    },
+    {
+      "epoch": 190.94736842105263,
+      "grad_norm": 0.5984175801277161,
+      "learning_rate": 0.00013509098994258064,
+      "loss": 0.1859,
+      "step": 9070
+    },
+    {
+      "epoch": 191.1578947368421,
+      "grad_norm": 0.534826397895813,
+      "learning_rate": 0.00013496577494867985,
+      "loss": 0.1959,
+      "step": 9080
+    },
+    {
+      "epoch": 191.3684210526316,
+      "grad_norm": 0.579361617565155,
+      "learning_rate": 0.00013484049746526977,
+      "loss": 0.1814,
+      "step": 9090
+    },
+    {
+      "epoch": 191.57894736842104,
+      "grad_norm": 0.7044987082481384,
+      "learning_rate": 0.0001347151577162416,
+      "loss": 0.1996,
+      "step": 9100
+    },
+    {
+      "epoch": 191.78947368421052,
+      "grad_norm": 0.6362590193748474,
+      "learning_rate": 0.00013458975592559781,
+      "loss": 0.1943,
+      "step": 9110
+    },
+    {
+      "epoch": 192.0,
+      "grad_norm": 0.7269014120101929,
+      "learning_rate": 0.0001344642923174517,
+      "loss": 0.2107,
+      "step": 9120
+    },
+    {
+      "epoch": 192.21052631578948,
+      "grad_norm": 0.6897448897361755,
+      "learning_rate": 0.00013433876711602713,
+      "loss": 0.184,
+      "step": 9130
+    },
+    {
+      "epoch": 192.42105263157896,
+      "grad_norm": 0.6716914176940918,
+      "learning_rate": 0.00013421318054565793,
+      "loss": 0.1982,
+      "step": 9140
+    },
+    {
+      "epoch": 192.6315789473684,
+      "grad_norm": 0.6334236264228821,
+      "learning_rate": 0.00013408753283078768,
+      "loss": 0.2077,
+      "step": 9150
+    },
+    {
+      "epoch": 192.8421052631579,
+      "grad_norm": 0.6236383318901062,
+      "learning_rate": 0.00013396182419596925,
+      "loss": 0.1933,
+      "step": 9160
+    },
+    {
+      "epoch": 193.05263157894737,
+      "grad_norm": 0.7885491251945496,
+      "learning_rate": 0.00013383605486586432,
+      "loss": 0.1919,
+      "step": 9170
+    },
+    {
+      "epoch": 193.26315789473685,
+      "grad_norm": 0.6639363765716553,
+      "learning_rate": 0.0001337102250652431,
+      "loss": 0.1943,
+      "step": 9180
+    },
+    {
+      "epoch": 193.47368421052633,
+      "grad_norm": 0.8824483752250671,
+      "learning_rate": 0.00013358433501898381,
+      "loss": 0.2155,
+      "step": 9190
+    },
+    {
+      "epoch": 193.68421052631578,
+      "grad_norm": 0.6293418407440186,
+      "learning_rate": 0.0001334583849520724,
+      "loss": 0.1853,
+      "step": 9200
+    },
+    {
+      "epoch": 193.89473684210526,
+      "grad_norm": 0.5062781572341919,
+      "learning_rate": 0.00013333237508960207,
+      "loss": 0.197,
+      "step": 9210
+    },
+    {
+      "epoch": 194.10526315789474,
+      "grad_norm": 0.5617866516113281,
+      "learning_rate": 0.00013320630565677287,
+      "loss": 0.1886,
+      "step": 9220
+    },
+    {
+      "epoch": 194.31578947368422,
+      "grad_norm": 0.8578426241874695,
+      "learning_rate": 0.0001330801768788913,
+      "loss": 0.2,
+      "step": 9230
+    },
+    {
+      "epoch": 194.52631578947367,
+      "grad_norm": 0.6378150582313538,
+      "learning_rate": 0.00013295398898136994,
+      "loss": 0.193,
+      "step": 9240
+    },
+    {
+      "epoch": 194.73684210526315,
+      "grad_norm": 0.6234794855117798,
+      "learning_rate": 0.00013282774218972707,
+      "loss": 0.189,
+      "step": 9250
+    },
+    {
+      "epoch": 194.94736842105263,
+      "grad_norm": 0.548239529132843,
+      "learning_rate": 0.0001327014367295861,
+      "loss": 0.2063,
+      "step": 9260
+    },
+    {
+      "epoch": 195.1578947368421,
+      "grad_norm": 0.6770426034927368,
+      "learning_rate": 0.00013257507282667542,
+      "loss": 0.2092,
+      "step": 9270
+    },
+    {
+      "epoch": 195.3684210526316,
+      "grad_norm": 0.7459677457809448,
+      "learning_rate": 0.00013244865070682785,
+      "loss": 0.1917,
+      "step": 9280
+    },
+    {
+      "epoch": 195.57894736842104,
+      "grad_norm": 0.6438161134719849,
+      "learning_rate": 0.0001323221705959801,
+      "loss": 0.1967,
+      "step": 9290
+    },
+    {
+      "epoch": 195.78947368421052,
+      "grad_norm": 0.6017579436302185,
+      "learning_rate": 0.00013219563272017271,
+      "loss": 0.197,
+      "step": 9300
+    },
+    {
+      "epoch": 196.0,
+      "grad_norm": 0.6799076199531555,
+      "learning_rate": 0.00013206903730554937,
+      "loss": 0.1958,
+      "step": 9310
+    },
+    {
+      "epoch": 196.21052631578948,
+      "grad_norm": 0.5263960361480713,
+      "learning_rate": 0.00013194238457835665,
+      "loss": 0.1785,
+      "step": 9320
+    },
+    {
+      "epoch": 196.42105263157896,
+      "grad_norm": 0.5635556578636169,
+      "learning_rate": 0.00013181567476494346,
+      "loss": 0.1882,
+      "step": 9330
+    },
+    {
+      "epoch": 196.6315789473684,
+      "grad_norm": 0.6033175587654114,
+      "learning_rate": 0.00013168890809176075,
+      "loss": 0.1971,
+      "step": 9340
+    },
+    {
+      "epoch": 196.8421052631579,
+      "grad_norm": 0.916451632976532,
+      "learning_rate": 0.00013156208478536124,
+      "loss": 0.2021,
+      "step": 9350
+    },
+    {
+      "epoch": 197.05263157894737,
+      "grad_norm": 0.6587386131286621,
+      "learning_rate": 0.0001314352050723986,
+      "loss": 0.2214,
+      "step": 9360
+    },
+    {
+      "epoch": 197.26315789473685,
+      "grad_norm": 0.6655704975128174,
+      "learning_rate": 0.00013130826917962755,
+      "loss": 0.1986,
+      "step": 9370
+    },
+    {
+      "epoch": 197.47368421052633,
+      "grad_norm": 0.6699230074882507,
+      "learning_rate": 0.0001311812773339031,
+      "loss": 0.1883,
+      "step": 9380
+    },
+    {
+      "epoch": 197.68421052631578,
+      "grad_norm": 0.515881359577179,
+      "learning_rate": 0.0001310542297621802,
+      "loss": 0.1909,
+      "step": 9390
+    },
+    {
+      "epoch": 197.89473684210526,
+      "grad_norm": 0.4768686890602112,
+      "learning_rate": 0.00013092712669151356,
+      "loss": 0.1918,
+      "step": 9400
+    },
+    {
+      "epoch": 198.10526315789474,
+      "grad_norm": 0.6306703090667725,
+      "learning_rate": 0.00013079996834905688,
+      "loss": 0.1894,
+      "step": 9410
+    },
+    {
+      "epoch": 198.31578947368422,
+      "grad_norm": 0.5824154019355774,
+      "learning_rate": 0.0001306727549620628,
+      "loss": 0.1845,
+      "step": 9420
+    },
+    {
+      "epoch": 198.52631578947367,
+      "grad_norm": 0.6079233288764954,
+      "learning_rate": 0.00013054548675788224,
+      "loss": 0.2015,
+      "step": 9430
+    },
+    {
+      "epoch": 198.73684210526315,
+      "grad_norm": 0.607089638710022,
+      "learning_rate": 0.00013041816396396416,
+      "loss": 0.1983,
+      "step": 9440
+    },
+    {
+      "epoch": 198.94736842105263,
+      "grad_norm": 0.642997682094574,
+      "learning_rate": 0.00013029078680785498,
+      "loss": 0.2135,
+      "step": 9450
+    },
+    {
+      "epoch": 199.1578947368421,
+      "grad_norm": 0.769798219203949,
+      "learning_rate": 0.00013016335551719837,
+      "loss": 0.2018,
+      "step": 9460
+    },
+    {
+      "epoch": 199.3684210526316,
+      "grad_norm": 0.72061687707901,
+      "learning_rate": 0.00013003587031973465,
+      "loss": 0.1863,
+      "step": 9470
+    },
+    {
+      "epoch": 199.57894736842104,
+      "grad_norm": 0.5181578993797302,
+      "learning_rate": 0.00012990833144330062,
+      "loss": 0.204,
+      "step": 9480
+    },
+    {
+      "epoch": 199.78947368421052,
+      "grad_norm": 0.6537469625473022,
+      "learning_rate": 0.00012978073911582886,
+      "loss": 0.1968,
+      "step": 9490
+    },
+    {
+      "epoch": 200.0,
+      "grad_norm": 0.5669968128204346,
+      "learning_rate": 0.00012965309356534764,
+      "loss": 0.1947,
+      "step": 9500
+    },
+    {
+      "epoch": 200.21052631578948,
+      "grad_norm": 0.5587459206581116,
+      "learning_rate": 0.00012952539501998012,
+      "loss": 0.1982,
+      "step": 9510
+    },
+    {
+      "epoch": 200.42105263157896,
+      "grad_norm": 0.6033953428268433,
+      "learning_rate": 0.00012939764370794446,
+      "loss": 0.2012,
+      "step": 9520
+    },
+    {
+      "epoch": 200.6315789473684,
+      "grad_norm": 0.6542585492134094,
+      "learning_rate": 0.00012926983985755283,
+      "loss": 0.194,
+      "step": 9530
+    },
+    {
+      "epoch": 200.8421052631579,
+      "grad_norm": 0.4944283664226532,
+      "learning_rate": 0.0001291419836972115,
+      "loss": 0.1766,
+      "step": 9540
+    },
+    {
+      "epoch": 201.05263157894737,
+      "grad_norm": 0.5921803116798401,
+      "learning_rate": 0.0001290140754554202,
+      "loss": 0.2082,
+      "step": 9550
+    },
+    {
+      "epoch": 201.26315789473685,
+      "grad_norm": 0.631533682346344,
+      "learning_rate": 0.0001288861153607716,
+      "loss": 0.1876,
+      "step": 9560
+    },
+    {
+      "epoch": 201.47368421052633,
+      "grad_norm": 0.6319711208343506,
+      "learning_rate": 0.00012875810364195123,
+      "loss": 0.1849,
+      "step": 9570
+    },
+    {
+      "epoch": 201.68421052631578,
+      "grad_norm": 0.8693262934684753,
+      "learning_rate": 0.0001286300405277367,
+      "loss": 0.195,
+      "step": 9580
+    },
+    {
+      "epoch": 201.89473684210526,
+      "grad_norm": 0.6572778820991516,
+      "learning_rate": 0.0001285019262469976,
+      "loss": 0.2099,
+      "step": 9590
+    },
+    {
+      "epoch": 202.10526315789474,
+      "grad_norm": 0.7793500423431396,
+      "learning_rate": 0.000128373761028695,
+      "loss": 0.1986,
+      "step": 9600
+    },
+    {
+      "epoch": 202.31578947368422,
+      "grad_norm": 0.6409270167350769,
+      "learning_rate": 0.0001282455451018808,
+      "loss": 0.1927,
+      "step": 9610
+    },
+    {
+      "epoch": 202.52631578947367,
+      "grad_norm": 0.6425662040710449,
+      "learning_rate": 0.0001281172786956977,
+      "loss": 0.1952,
+      "step": 9620
+    },
+    {
+      "epoch": 202.73684210526315,
+      "grad_norm": 0.5852437019348145,
+      "learning_rate": 0.00012798896203937855,
+      "loss": 0.1831,
+      "step": 9630
+    },
+    {
+      "epoch": 202.94736842105263,
+      "grad_norm": 0.768488347530365,
+      "learning_rate": 0.00012786059536224611,
+      "loss": 0.1994,
+      "step": 9640
+    },
+    {
+      "epoch": 203.1578947368421,
+      "grad_norm": 0.7588623762130737,
+      "learning_rate": 0.00012774502277463864,
+      "loss": 0.1998,
+      "step": 9650
+    },
+    {
+      "epoch": 203.3684210526316,
+      "grad_norm": 0.5900463461875916,
+      "learning_rate": 0.00012761656169006457,
+      "loss": 0.2037,
+      "step": 9660
+    },
+    {
+      "epoch": 203.57894736842104,
+      "grad_norm": 0.5838971734046936,
+      "learning_rate": 0.00012748805125021694,
+      "loss": 0.1923,
+      "step": 9670
+    },
+    {
+      "epoch": 203.78947368421052,
+      "grad_norm": 0.6288439035415649,
+      "learning_rate": 0.00012735949168476477,
+      "loss": 0.2109,
+      "step": 9680
+    },
+    {
+      "epoch": 204.0,
+      "grad_norm": 0.6366193890571594,
+      "learning_rate": 0.00012723088322346478,
+      "loss": 0.193,
+      "step": 9690
+    },
+    {
+      "epoch": 204.21052631578948,
+      "grad_norm": 0.49313342571258545,
+      "learning_rate": 0.00012710222609616125,
+      "loss": 0.1762,
+      "step": 9700
+    },
+    {
+      "epoch": 204.42105263157896,
+      "grad_norm": 0.6465559601783752,
+      "learning_rate": 0.0001269735205327852,
+      "loss": 0.1943,
+      "step": 9710
+    },
+    {
+      "epoch": 204.6315789473684,
+      "grad_norm": 0.6326197385787964,
+      "learning_rate": 0.00012684476676335445,
+      "loss": 0.1925,
+      "step": 9720
+    },
+    {
+      "epoch": 204.8421052631579,
+      "grad_norm": 0.8847769498825073,
+      "learning_rate": 0.00012671596501797282,
+      "loss": 0.2049,
+      "step": 9730
+    },
+    {
+      "epoch": 205.05263157894737,
+      "grad_norm": 0.5743268728256226,
+      "learning_rate": 0.00012658711552682988,
+      "loss": 0.1966,
+      "step": 9740
+    },
+    {
+      "epoch": 205.26315789473685,
+      "grad_norm": 0.5662633180618286,
+      "learning_rate": 0.00012645821852020066,
+      "loss": 0.1855,
+      "step": 9750
+    },
+    {
+      "epoch": 205.47368421052633,
+      "grad_norm": 0.6402904987335205,
+      "learning_rate": 0.0001263292742284449,
+      "loss": 0.1904,
+      "step": 9760
+    },
+    {
+      "epoch": 205.68421052631578,
+      "grad_norm": 0.5689384341239929,
+      "learning_rate": 0.000126200282882007,
+      "loss": 0.2074,
+      "step": 9770
+    },
+    {
+      "epoch": 205.89473684210526,
+      "grad_norm": 0.7823508977890015,
+      "learning_rate": 0.00012607124471141542,
+      "loss": 0.1976,
+      "step": 9780
+    },
+    {
+      "epoch": 206.10526315789474,
+      "grad_norm": 0.4707532227039337,
+      "learning_rate": 0.0001259421599472823,
+      "loss": 0.199,
+      "step": 9790
+    },
+    {
+      "epoch": 206.31578947368422,
+      "grad_norm": 0.6291844248771667,
+      "learning_rate": 0.000125813028820303,
+      "loss": 0.185,
+      "step": 9800
+    },
+    {
+      "epoch": 206.52631578947367,
+      "grad_norm": 0.6552050709724426,
+      "learning_rate": 0.00012568385156125586,
+      "loss": 0.1928,
+      "step": 9810
+    },
+    {
+      "epoch": 206.73684210526315,
+      "grad_norm": 0.5243370532989502,
+      "learning_rate": 0.0001255546284010015,
+      "loss": 0.1958,
+      "step": 9820
+    },
+    {
+      "epoch": 206.94736842105263,
+      "grad_norm": 0.5976846218109131,
+      "learning_rate": 0.0001254253595704827,
+      "loss": 0.1948,
+      "step": 9830
+    },
+    {
+      "epoch": 207.1578947368421,
+      "grad_norm": 0.473967969417572,
+      "learning_rate": 0.00012529604530072384,
+      "loss": 0.2067,
+      "step": 9840
+    },
+    {
+      "epoch": 207.3684210526316,
+      "grad_norm": 0.7041087746620178,
+      "learning_rate": 0.00012516668582283045,
+      "loss": 0.1843,
+      "step": 9850
+    },
+    {
+      "epoch": 207.57894736842104,
+      "grad_norm": 0.5606855154037476,
+      "learning_rate": 0.00012503728136798893,
+      "loss": 0.1965,
+      "step": 9860
+    },
+    {
+      "epoch": 207.78947368421052,
+      "grad_norm": 0.7192473411560059,
+      "learning_rate": 0.00012490783216746601,
+      "loss": 0.1948,
+      "step": 9870
+    },
+    {
+      "epoch": 208.0,
+      "grad_norm": 0.673022449016571,
+      "learning_rate": 0.00012477833845260836,
+      "loss": 0.2078,
+      "step": 9880
+    },
+    {
+      "epoch": 208.21052631578948,
+      "grad_norm": 0.7071841359138489,
+      "learning_rate": 0.0001246488004548423,
+      "loss": 0.1937,
+      "step": 9890
+    },
+    {
+      "epoch": 208.42105263157896,
+      "grad_norm": 0.518195390701294,
+      "learning_rate": 0.0001245192184056732,
+      "loss": 0.1941,
+      "step": 9900
+    },
+    {
+      "epoch": 208.6315789473684,
+      "grad_norm": 0.6964612603187561,
+      "learning_rate": 0.0001243895925366852,
+      "loss": 0.1954,
+      "step": 9910
+    },
+    {
+      "epoch": 208.8421052631579,
+      "grad_norm": 0.6792444586753845,
+      "learning_rate": 0.00012425992307954075,
+      "loss": 0.2075,
+      "step": 9920
+    },
+    {
+      "epoch": 209.05263157894737,
+      "grad_norm": 0.4383447468280792,
+      "learning_rate": 0.0001241302102659802,
+      "loss": 0.1745,
+      "step": 9930
+    },
+    {
+      "epoch": 209.26315789473685,
+      "grad_norm": 0.6733793616294861,
+      "learning_rate": 0.00012400045432782138,
+      "loss": 0.1972,
+      "step": 9940
+    },
+    {
+      "epoch": 209.47368421052633,
+      "grad_norm": 0.5500863194465637,
+      "learning_rate": 0.00012387065549695917,
+      "loss": 0.1809,
+      "step": 9950
+    },
+    {
+      "epoch": 209.68421052631578,
+      "grad_norm": 0.8505430221557617,
+      "learning_rate": 0.00012374081400536518,
+      "loss": 0.2014,
+      "step": 9960
+    },
+    {
+      "epoch": 209.89473684210526,
+      "grad_norm": 0.5805114507675171,
+      "learning_rate": 0.00012361093008508714,
+      "loss": 0.202,
+      "step": 9970
+    },
+    {
+      "epoch": 210.10526315789474,
+      "grad_norm": 0.5001348853111267,
+      "learning_rate": 0.0001234810039682487,
+      "loss": 0.1882,
+      "step": 9980
+    },
+    {
+      "epoch": 210.31578947368422,
+      "grad_norm": 0.6055691242218018,
+      "learning_rate": 0.00012335103588704895,
+      "loss": 0.201,
+      "step": 9990
+    },
+    {
+      "epoch": 210.52631578947367,
+      "grad_norm": 0.6668257713317871,
+      "learning_rate": 0.00012322102607376182,
+      "loss": 0.1899,
+      "step": 10000
+    },
+    {
+      "epoch": 210.73684210526315,
+      "grad_norm": 0.8748318552970886,
+      "learning_rate": 0.00012309097476073598,
+      "loss": 0.2055,
+      "step": 10010
+    },
+    {
+      "epoch": 210.94736842105263,
+      "grad_norm": 0.6310805678367615,
+      "learning_rate": 0.00012296088218039427,
+      "loss": 0.1911,
+      "step": 10020
+    },
+    {
+      "epoch": 211.1578947368421,
+      "grad_norm": 0.5111770033836365,
+      "learning_rate": 0.00012283074856523314,
+      "loss": 0.1892,
+      "step": 10030
+    },
+    {
+      "epoch": 211.3684210526316,
+      "grad_norm": 0.8765570521354675,
+      "learning_rate": 0.00012270057414782252,
+      "loss": 0.1764,
+      "step": 10040
+    },
+    {
+      "epoch": 211.57894736842104,
+      "grad_norm": 0.5945147275924683,
+      "learning_rate": 0.00012257035916080514,
+      "loss": 0.1962,
+      "step": 10050
+    },
+    {
+      "epoch": 211.78947368421052,
+      "grad_norm": 0.7749220728874207,
+      "learning_rate": 0.00012244010383689636,
+      "loss": 0.204,
+      "step": 10060
+    },
+    {
+      "epoch": 212.0,
+      "grad_norm": 0.741985023021698,
+      "learning_rate": 0.00012230980840888348,
+      "loss": 0.2083,
+      "step": 10070
+    },
+    {
+      "epoch": 212.21052631578948,
+      "grad_norm": 0.5785123705863953,
+      "learning_rate": 0.00012217947310962565,
+      "loss": 0.2033,
+      "step": 10080
+    },
+    {
+      "epoch": 212.42105263157896,
+      "grad_norm": 0.5759528875350952,
+      "learning_rate": 0.00012204909817205314,
+      "loss": 0.1963,
+      "step": 10090
+    },
+    {
+      "epoch": 212.6315789473684,
+      "grad_norm": 0.8312242031097412,
+      "learning_rate": 0.00012191868382916709,
+      "loss": 0.1874,
+      "step": 10100
+    },
+    {
+      "epoch": 212.8421052631579,
+      "grad_norm": 0.5142462849617004,
+      "learning_rate": 0.00012178823031403911,
+      "loss": 0.2005,
+      "step": 10110
+    },
+    {
+      "epoch": 213.05263157894737,
+      "grad_norm": 0.49217525124549866,
+      "learning_rate": 0.00012165773785981076,
+      "loss": 0.1855,
+      "step": 10120
+    },
+    {
+      "epoch": 213.26315789473685,
+      "grad_norm": 0.4872840940952301,
+      "learning_rate": 0.00012152720669969317,
+      "loss": 0.1837,
+      "step": 10130
+    },
+    {
+      "epoch": 213.47368421052633,
+      "grad_norm": 0.6701936721801758,
+      "learning_rate": 0.0001213966370669668,
+      "loss": 0.2015,
+      "step": 10140
+    },
+    {
+      "epoch": 213.68421052631578,
+      "grad_norm": 0.6903809905052185,
+      "learning_rate": 0.00012126602919498065,
+      "loss": 0.1927,
+      "step": 10150
+    },
+    {
+      "epoch": 213.89473684210526,
+      "grad_norm": 0.4880557656288147,
+      "learning_rate": 0.00012113538331715224,
+      "loss": 0.2053,
+      "step": 10160
+    },
+    {
+      "epoch": 214.10526315789474,
+      "grad_norm": 0.5100629329681396,
+      "learning_rate": 0.00012100469966696686,
+      "loss": 0.1871,
+      "step": 10170
+    },
+    {
+      "epoch": 214.31578947368422,
+      "grad_norm": 0.630207896232605,
+      "learning_rate": 0.00012087397847797743,
+      "loss": 0.2026,
+      "step": 10180
+    },
+    {
+      "epoch": 214.52631578947367,
+      "grad_norm": 0.5134378671646118,
+      "learning_rate": 0.00012074321998380391,
+      "loss": 0.1936,
+      "step": 10190
+    },
+    {
+      "epoch": 214.73684210526315,
+      "grad_norm": 0.5366536378860474,
+      "learning_rate": 0.00012061242441813294,
+      "loss": 0.195,
+      "step": 10200
+    },
+    {
+      "epoch": 214.94736842105263,
+      "grad_norm": 0.5395473837852478,
+      "learning_rate": 0.00012048159201471739,
+      "loss": 0.1924,
+      "step": 10210
+    },
+    {
+      "epoch": 215.1578947368421,
+      "grad_norm": 0.9155479669570923,
+      "learning_rate": 0.00012035072300737596,
+      "loss": 0.207,
+      "step": 10220
+    },
+    {
+      "epoch": 215.3684210526316,
+      "grad_norm": 0.6539714336395264,
+      "learning_rate": 0.00012021981762999279,
+      "loss": 0.1854,
+      "step": 10230
+    },
+    {
+      "epoch": 215.57894736842104,
+      "grad_norm": 0.8252039551734924,
+      "learning_rate": 0.00012008887611651704,
+      "loss": 0.1946,
+      "step": 10240
+    },
+    {
+      "epoch": 215.78947368421052,
+      "grad_norm": 0.6097246408462524,
+      "learning_rate": 0.00011995789870096241,
+      "loss": 0.1948,
+      "step": 10250
+    },
+    {
+      "epoch": 216.0,
+      "grad_norm": 0.5688139796257019,
+      "learning_rate": 0.0001198268856174068,
+      "loss": 0.1929,
+      "step": 10260
+    },
+    {
+      "epoch": 216.21052631578948,
+      "grad_norm": 0.7054363489151001,
+      "learning_rate": 0.0001196958370999918,
+      "loss": 0.2023,
+      "step": 10270
+    },
+    {
+      "epoch": 216.42105263157896,
+      "grad_norm": 0.4730806350708008,
+      "learning_rate": 0.00011956475338292237,
+      "loss": 0.1812,
+      "step": 10280
+    },
+    {
+      "epoch": 216.6315789473684,
+      "grad_norm": 0.7729452252388,
+      "learning_rate": 0.00011943363470046636,
+      "loss": 0.2017,
+      "step": 10290
+    },
+    {
+      "epoch": 216.8421052631579,
+      "grad_norm": 0.6470244526863098,
+      "learning_rate": 0.00011930248128695414,
+      "loss": 0.1964,
+      "step": 10300
+    },
+    {
+      "epoch": 217.05263157894737,
+      "grad_norm": 0.5927222371101379,
+      "learning_rate": 0.00011917129337677809,
+      "loss": 0.1905,
+      "step": 10310
+    },
+    {
+      "epoch": 217.26315789473685,
+      "grad_norm": 0.6557316184043884,
+      "learning_rate": 0.00011904007120439232,
+      "loss": 0.1878,
+      "step": 10320
+    },
+    {
+      "epoch": 217.47368421052633,
+      "grad_norm": 0.5663248896598816,
+      "learning_rate": 0.0001189088150043121,
+      "loss": 0.1854,
+      "step": 10330
+    },
+    {
+      "epoch": 217.68421052631578,
+      "grad_norm": 0.5649807453155518,
+      "learning_rate": 0.00011877752501111354,
+      "loss": 0.2107,
+      "step": 10340
+    },
+    {
+      "epoch": 217.89473684210526,
+      "grad_norm": 0.570824384689331,
+      "learning_rate": 0.00011864620145943315,
+      "loss": 0.1906,
+      "step": 10350
+    },
+    {
+      "epoch": 218.10526315789474,
+      "grad_norm": 0.5065656304359436,
+      "learning_rate": 0.0001185148445839674,
+      "loss": 0.1995,
+      "step": 10360
+    },
+    {
+      "epoch": 218.31578947368422,
+      "grad_norm": 0.7696800231933594,
+      "learning_rate": 0.00011838345461947235,
+      "loss": 0.1957,
+      "step": 10370
+    },
+    {
+      "epoch": 218.52631578947367,
+      "grad_norm": 0.4969731569290161,
+      "learning_rate": 0.00011825203180076319,
+      "loss": 0.1994,
+      "step": 10380
+    },
+    {
+      "epoch": 218.73684210526315,
+      "grad_norm": 0.8026688694953918,
+      "learning_rate": 0.00011812057636271374,
+      "loss": 0.2052,
+      "step": 10390
+    },
+    {
+      "epoch": 218.94736842105263,
+      "grad_norm": 0.609519362449646,
+      "learning_rate": 0.00011798908854025623,
+      "loss": 0.1834,
+      "step": 10400
+    },
+    {
+      "epoch": 219.1578947368421,
+      "grad_norm": 0.5675802826881409,
+      "learning_rate": 0.00011785756856838071,
+      "loss": 0.1994,
+      "step": 10410
+    },
+    {
+      "epoch": 219.3684210526316,
+      "grad_norm": 0.7236883640289307,
+      "learning_rate": 0.00011772601668213468,
+      "loss": 0.2032,
+      "step": 10420
+    },
+    {
+      "epoch": 219.57894736842104,
+      "grad_norm": 0.513433575630188,
+      "learning_rate": 0.0001175944331166227,
+      "loss": 0.1878,
+      "step": 10430
+    },
+    {
+      "epoch": 219.78947368421052,
+      "grad_norm": 0.6006221771240234,
+      "learning_rate": 0.00011746281810700592,
+      "loss": 0.1938,
+      "step": 10440
+    },
+    {
+      "epoch": 220.0,
+      "grad_norm": 0.5742442607879639,
+      "learning_rate": 0.00011733117188850178,
+      "loss": 0.1967,
+      "step": 10450
+    },
+    {
+      "epoch": 220.21052631578948,
+      "grad_norm": 0.6315630078315735,
+      "learning_rate": 0.00011719949469638329,
+      "loss": 0.1978,
+      "step": 10460
+    },
+    {
+      "epoch": 220.42105263157896,
+      "grad_norm": 0.5011500120162964,
+      "learning_rate": 0.00011706778676597905,
+      "loss": 0.1868,
+      "step": 10470
+    },
+    {
+      "epoch": 220.6315789473684,
+      "grad_norm": 0.6937258243560791,
+      "learning_rate": 0.00011693604833267242,
+      "loss": 0.2015,
+      "step": 10480
+    },
+    {
+      "epoch": 220.8421052631579,
+      "grad_norm": 0.6345796585083008,
+      "learning_rate": 0.00011680427963190139,
+      "loss": 0.1986,
+      "step": 10490
+    },
+    {
+      "epoch": 221.05263157894737,
+      "grad_norm": 0.5916712880134583,
+      "learning_rate": 0.00011667248089915799,
+      "loss": 0.1911,
+      "step": 10500
+    },
+    {
+      "epoch": 221.26315789473685,
+      "grad_norm": 0.5585072040557861,
+      "learning_rate": 0.00011654065236998786,
+      "loss": 0.1885,
+      "step": 10510
+    },
+    {
+      "epoch": 221.47368421052633,
+      "grad_norm": 0.7705309987068176,
+      "learning_rate": 0.00011640879427999003,
+      "loss": 0.1897,
+      "step": 10520
+    },
+    {
+      "epoch": 221.68421052631578,
+      "grad_norm": 0.5290719866752625,
+      "learning_rate": 0.00011627690686481627,
+      "loss": 0.1949,
+      "step": 10530
+    },
+    {
+      "epoch": 221.89473684210526,
+      "grad_norm": 0.8180526494979858,
+      "learning_rate": 0.00011614499036017075,
+      "loss": 0.2041,
+      "step": 10540
+    },
+    {
+      "epoch": 222.10526315789474,
+      "grad_norm": 0.5164480209350586,
+      "learning_rate": 0.00011601304500180967,
+      "loss": 0.1852,
+      "step": 10550
+    },
+    {
+      "epoch": 222.31578947368422,
+      "grad_norm": 0.5821452140808105,
+      "learning_rate": 0.00011588107102554078,
+      "loss": 0.1907,
+      "step": 10560
+    },
+    {
+      "epoch": 222.52631578947367,
+      "grad_norm": 0.6250927448272705,
+      "learning_rate": 0.000115749068667223,
+      "loss": 0.1949,
+      "step": 10570
+    },
+    {
+      "epoch": 222.73684210526315,
+      "grad_norm": 0.7137615084648132,
+      "learning_rate": 0.0001156170381627659,
+      "loss": 0.1976,
+      "step": 10580
+    },
+    {
+      "epoch": 222.94736842105263,
+      "grad_norm": 0.6757993102073669,
+      "learning_rate": 0.00011548497974812942,
+      "loss": 0.1982,
+      "step": 10590
+    },
+    {
+      "epoch": 223.1578947368421,
+      "grad_norm": 0.49481892585754395,
+      "learning_rate": 0.00011535289365932333,
+      "loss": 0.1832,
+      "step": 10600
+    },
+    {
+      "epoch": 223.3684210526316,
+      "grad_norm": 0.6089574694633484,
+      "learning_rate": 0.00011522078013240696,
+      "loss": 0.1833,
+      "step": 10610
+    },
+    {
+      "epoch": 223.57894736842104,
+      "grad_norm": 0.6293704509735107,
+      "learning_rate": 0.00011508863940348855,
+      "loss": 0.1994,
+      "step": 10620
+    },
+    {
+      "epoch": 223.78947368421052,
+      "grad_norm": 0.5520006418228149,
+      "learning_rate": 0.000114956471708725,
+      "loss": 0.2023,
+      "step": 10630
+    },
+    {
+      "epoch": 224.0,
+      "grad_norm": 0.566952645778656,
+      "learning_rate": 0.00011482427728432144,
+      "loss": 0.1956,
+      "step": 10640
+    },
+    {
+      "epoch": 224.21052631578948,
+      "grad_norm": 0.7031182050704956,
+      "learning_rate": 0.00011469205636653075,
+      "loss": 0.1958,
+      "step": 10650
+    },
+    {
+      "epoch": 224.42105263157896,
+      "grad_norm": 0.5254794955253601,
+      "learning_rate": 0.00011455980919165308,
+      "loss": 0.1892,
+      "step": 10660
+    },
+    {
+      "epoch": 224.6315789473684,
+      "grad_norm": 0.539675235748291,
+      "learning_rate": 0.00011442753599603566,
+      "loss": 0.179,
+      "step": 10670
+    },
+    {
+      "epoch": 224.8421052631579,
+      "grad_norm": 0.5456839203834534,
+      "learning_rate": 0.0001142952370160721,
+      "loss": 0.1983,
+      "step": 10680
+    },
+    {
+      "epoch": 225.05263157894737,
+      "grad_norm": 0.5140777826309204,
+      "learning_rate": 0.0001141629124882021,
+      "loss": 0.2006,
+      "step": 10690
+    },
+    {
+      "epoch": 225.26315789473685,
+      "grad_norm": 0.48138687014579773,
+      "learning_rate": 0.00011403056264891112,
+      "loss": 0.1995,
+      "step": 10700
+    },
+    {
+      "epoch": 225.47368421052633,
+      "grad_norm": 0.6670958995819092,
+      "learning_rate": 0.00011389818773472971,
+      "loss": 0.1954,
+      "step": 10710
+    },
+    {
+      "epoch": 225.68421052631578,
+      "grad_norm": 0.7487067580223083,
+      "learning_rate": 0.00011376578798223332,
+      "loss": 0.1916,
+      "step": 10720
+    },
+    {
+      "epoch": 225.89473684210526,
+      "grad_norm": 0.6026431918144226,
+      "learning_rate": 0.0001136333636280418,
+      "loss": 0.1847,
+      "step": 10730
+    },
+    {
+      "epoch": 226.10526315789474,
+      "grad_norm": 0.6704311966896057,
+      "learning_rate": 0.00011350091490881893,
+      "loss": 0.1861,
+      "step": 10740
+    },
+    {
+      "epoch": 226.31578947368422,
+      "grad_norm": 0.8869838714599609,
+      "learning_rate": 0.00011336844206127209,
+      "loss": 0.2022,
+      "step": 10750
+    },
+    {
+      "epoch": 226.52631578947367,
+      "grad_norm": 0.7338431477546692,
+      "learning_rate": 0.00011323594532215165,
+      "loss": 0.1899,
+      "step": 10760
+    },
+    {
+      "epoch": 226.73684210526315,
+      "grad_norm": 0.5879060626029968,
+      "learning_rate": 0.00011310342492825081,
+      "loss": 0.1827,
+      "step": 10770
+    },
+    {
+      "epoch": 226.94736842105263,
+      "grad_norm": 0.5951240062713623,
+      "learning_rate": 0.00011297088111640499,
+      "loss": 0.1964,
+      "step": 10780
+    },
+    {
+      "epoch": 227.1578947368421,
+      "grad_norm": 0.4901759922504425,
+      "learning_rate": 0.00011283831412349153,
+      "loss": 0.1882,
+      "step": 10790
+    },
+    {
+      "epoch": 227.3684210526316,
+      "grad_norm": 0.6785090565681458,
+      "learning_rate": 0.00011270572418642909,
+      "loss": 0.1897,
+      "step": 10800
+    },
+    {
+      "epoch": 227.57894736842104,
+      "grad_norm": 0.6172801852226257,
+      "learning_rate": 0.00011257311154217736,
+      "loss": 0.2049,
+      "step": 10810
+    },
+    {
+      "epoch": 227.78947368421052,
+      "grad_norm": 0.6416763663291931,
+      "learning_rate": 0.0001124404764277367,
+      "loss": 0.1885,
+      "step": 10820
+    },
+    {
+      "epoch": 228.0,
+      "grad_norm": 0.7574992775917053,
+      "learning_rate": 0.00011230781908014752,
+      "loss": 0.2072,
+      "step": 10830
+    },
+    {
+      "epoch": 228.21052631578948,
+      "grad_norm": 0.7357071042060852,
+      "learning_rate": 0.00011217513973649003,
+      "loss": 0.2017,
+      "step": 10840
+    },
+    {
+      "epoch": 228.42105263157896,
+      "grad_norm": 0.45522359013557434,
+      "learning_rate": 0.0001120424386338837,
+      "loss": 0.1815,
+      "step": 10850
+    },
+    {
+      "epoch": 228.6315789473684,
+      "grad_norm": 0.5420458316802979,
+      "learning_rate": 0.00011190971600948699,
+      "loss": 0.1809,
+      "step": 10860
+    },
+    {
+      "epoch": 228.8421052631579,
+      "grad_norm": 0.7300458550453186,
+      "learning_rate": 0.00011177697210049667,
+      "loss": 0.2028,
+      "step": 10870
+    },
+    {
+      "epoch": 229.05263157894737,
+      "grad_norm": 0.5525165796279907,
+      "learning_rate": 0.00011164420714414768,
+      "loss": 0.2037,
+      "step": 10880
+    },
+    {
+      "epoch": 229.26315789473685,
+      "grad_norm": 0.5621128082275391,
+      "learning_rate": 0.00011151142137771246,
+      "loss": 0.1981,
+      "step": 10890
+    },
+    {
+      "epoch": 229.47368421052633,
+      "grad_norm": 0.6248337626457214,
+      "learning_rate": 0.0001113786150385008,
+      "loss": 0.1974,
+      "step": 10900
+    },
+    {
+      "epoch": 229.68421052631578,
+      "grad_norm": 0.6925014853477478,
+      "learning_rate": 0.00011124578836385908,
+      "loss": 0.1905,
+      "step": 10910
+    },
+    {
+      "epoch": 229.89473684210526,
+      "grad_norm": 0.4596860110759735,
+      "learning_rate": 0.00011111294159117016,
+      "loss": 0.1923,
+      "step": 10920
+    },
+    {
+      "epoch": 230.10526315789474,
+      "grad_norm": 0.5915650129318237,
+      "learning_rate": 0.00011098007495785272,
+      "loss": 0.1887,
+      "step": 10930
+    },
+    {
+      "epoch": 230.31578947368422,
+      "grad_norm": 0.47109338641166687,
+      "learning_rate": 0.00011084718870136102,
+      "loss": 0.1875,
+      "step": 10940
+    },
+    {
+      "epoch": 230.52631578947367,
+      "grad_norm": 0.7140452861785889,
+      "learning_rate": 0.0001107142830591843,
+      "loss": 0.1857,
+      "step": 10950
+    },
+    {
+      "epoch": 230.73684210526315,
+      "grad_norm": 0.6751444935798645,
+      "learning_rate": 0.0001105813582688465,
+      "loss": 0.1947,
+      "step": 10960
+    },
+    {
+      "epoch": 230.94736842105263,
+      "grad_norm": 0.5488141775131226,
+      "learning_rate": 0.0001104484145679058,
+      "loss": 0.1902,
+      "step": 10970
+    },
+    {
+      "epoch": 231.1578947368421,
+      "grad_norm": 0.5548514127731323,
+      "learning_rate": 0.00011031545219395413,
+      "loss": 0.1883,
+      "step": 10980
+    },
+    {
+      "epoch": 231.3684210526316,
+      "grad_norm": 0.6914103031158447,
+      "learning_rate": 0.00011018247138461684,
+      "loss": 0.1943,
+      "step": 10990
+    },
+    {
+      "epoch": 231.57894736842104,
+      "grad_norm": 0.7115802764892578,
+      "learning_rate": 0.00011004947237755217,
+      "loss": 0.198,
+      "step": 11000
+    },
+    {
+      "epoch": 231.78947368421052,
+      "grad_norm": 0.5460284352302551,
+      "learning_rate": 0.0001099164554104509,
+      "loss": 0.1859,
+      "step": 11010
+    },
+    {
+      "epoch": 232.0,
+      "grad_norm": 0.7350656390190125,
+      "learning_rate": 0.00010978342072103593,
+      "loss": 0.206,
+      "step": 11020
+    },
+    {
+      "epoch": 232.21052631578948,
+      "grad_norm": 0.4807220995426178,
+      "learning_rate": 0.00010965036854706183,
+      "loss": 0.1907,
+      "step": 11030
+    },
+    {
+      "epoch": 232.42105263157896,
+      "grad_norm": 0.6180817484855652,
+      "learning_rate": 0.00010951729912631443,
+      "loss": 0.1993,
+      "step": 11040
+    },
+    {
+      "epoch": 232.6315789473684,
+      "grad_norm": 0.5458692908287048,
+      "learning_rate": 0.0001093842126966103,
+      "loss": 0.185,
+      "step": 11050
+    },
+    {
+      "epoch": 232.8421052631579,
+      "grad_norm": 0.6068777441978455,
+      "learning_rate": 0.00010925110949579653,
+      "loss": 0.1931,
+      "step": 11060
+    },
+    {
+      "epoch": 233.05263157894737,
+      "grad_norm": 0.604051411151886,
+      "learning_rate": 0.00010911798976175008,
+      "loss": 0.1988,
+      "step": 11070
+    },
+    {
+      "epoch": 233.26315789473685,
+      "grad_norm": 0.621242344379425,
+      "learning_rate": 0.00010898485373237748,
+      "loss": 0.1924,
+      "step": 11080
+    },
+    {
+      "epoch": 233.47368421052633,
+      "grad_norm": 0.5285283923149109,
+      "learning_rate": 0.00010885170164561449,
+      "loss": 0.1833,
+      "step": 11090
+    },
+    {
+      "epoch": 233.68421052631578,
+      "grad_norm": 0.512630045413971,
+      "learning_rate": 0.0001087185337394254,
+      "loss": 0.1929,
+      "step": 11100
+    },
+    {
+      "epoch": 233.89473684210526,
+      "grad_norm": 0.602725088596344,
+      "learning_rate": 0.0001085853502518029,
+      "loss": 0.1954,
+      "step": 11110
+    },
+    {
+      "epoch": 234.10526315789474,
+      "grad_norm": 0.41769760847091675,
+      "learning_rate": 0.00010845215142076742,
+      "loss": 0.191,
+      "step": 11120
+    },
+    {
+      "epoch": 234.31578947368422,
+      "grad_norm": 0.5544017553329468,
+      "learning_rate": 0.0001083189374843669,
+      "loss": 0.1834,
+      "step": 11130
+    },
+    {
+      "epoch": 234.52631578947367,
+      "grad_norm": 0.5937119126319885,
+      "learning_rate": 0.00010818570868067623,
+      "loss": 0.1933,
+      "step": 11140
+    },
+    {
+      "epoch": 234.73684210526315,
+      "grad_norm": 0.6727398633956909,
+      "learning_rate": 0.00010805246524779693,
+      "loss": 0.1862,
+      "step": 11150
+    },
+    {
+      "epoch": 234.94736842105263,
+      "grad_norm": 0.6366226673126221,
+      "learning_rate": 0.00010791920742385657,
+      "loss": 0.2097,
+      "step": 11160
+    },
+    {
+      "epoch": 235.1578947368421,
+      "grad_norm": 0.5684067010879517,
+      "learning_rate": 0.00010778593544700852,
+      "loss": 0.2046,
+      "step": 11170
+    },
+    {
+      "epoch": 235.3684210526316,
+      "grad_norm": 0.8542050123214722,
+      "learning_rate": 0.0001076526495554314,
+      "loss": 0.1991,
+      "step": 11180
+    },
+    {
+      "epoch": 235.57894736842104,
+      "grad_norm": 0.6223448514938354,
+      "learning_rate": 0.00010751934998732867,
+      "loss": 0.186,
+      "step": 11190
+    },
+    {
+      "epoch": 235.78947368421052,
+      "grad_norm": 0.47746938467025757,
+      "learning_rate": 0.00010738603698092831,
+      "loss": 0.1916,
+      "step": 11200
+    },
+    {
+      "epoch": 236.0,
+      "grad_norm": 0.6482349038124084,
+      "learning_rate": 0.00010725271077448232,
+      "loss": 0.1981,
+      "step": 11210
+    },
+    {
+      "epoch": 236.21052631578948,
+      "grad_norm": 0.5127678513526917,
+      "learning_rate": 0.00010711937160626617,
+      "loss": 0.1871,
+      "step": 11220
+    },
+    {
+      "epoch": 236.42105263157896,
+      "grad_norm": 0.7345888614654541,
+      "learning_rate": 0.00010698601971457862,
+      "loss": 0.199,
+      "step": 11230
+    },
+    {
+      "epoch": 236.6315789473684,
+      "grad_norm": 0.4777671992778778,
+      "learning_rate": 0.00010685265533774109,
+      "loss": 0.1944,
+      "step": 11240
+    },
+    {
+      "epoch": 236.8421052631579,
+      "grad_norm": 0.5794364809989929,
+      "learning_rate": 0.0001067192787140974,
+      "loss": 0.1884,
+      "step": 11250
+    },
+    {
+      "epoch": 237.05263157894737,
+      "grad_norm": 0.6226224303245544,
+      "learning_rate": 0.00010658589008201314,
+      "loss": 0.189,
+      "step": 11260
+    },
+    {
+      "epoch": 237.26315789473685,
+      "grad_norm": 0.5198787450790405,
+      "learning_rate": 0.00010645248967987544,
+      "loss": 0.1817,
+      "step": 11270
+    },
+    {
+      "epoch": 237.47368421052633,
+      "grad_norm": 0.5877044200897217,
+      "learning_rate": 0.0001063190777460925,
+      "loss": 0.1982,
+      "step": 11280
+    },
+    {
+      "epoch": 237.68421052631578,
+      "grad_norm": 0.5517503619194031,
+      "learning_rate": 0.00010618565451909302,
+      "loss": 0.1969,
+      "step": 11290
+    },
+    {
+      "epoch": 237.89473684210526,
+      "grad_norm": 0.5407507419586182,
+      "learning_rate": 0.00010605222023732596,
+      "loss": 0.2072,
+      "step": 11300
+    },
+    {
+      "epoch": 238.10526315789474,
+      "grad_norm": 0.583928644657135,
+      "learning_rate": 0.00010591877513926,
+      "loss": 0.1818,
+      "step": 11310
+    },
+    {
+      "epoch": 238.31578947368422,
+      "grad_norm": 0.620707631111145,
+      "learning_rate": 0.00010578531946338319,
+      "loss": 0.1861,
+      "step": 11320
+    },
+    {
+      "epoch": 238.52631578947367,
+      "grad_norm": 0.83612459897995,
+      "learning_rate": 0.00010565185344820247,
+      "loss": 0.1924,
+      "step": 11330
+    },
+    {
+      "epoch": 238.73684210526315,
+      "grad_norm": 0.6024158596992493,
+      "learning_rate": 0.00010551837733224321,
+      "loss": 0.1889,
+      "step": 11340
+    },
+    {
+      "epoch": 238.94736842105263,
+      "grad_norm": 0.5878002047538757,
+      "learning_rate": 0.00010538489135404893,
+      "loss": 0.199,
+      "step": 11350
+    },
+    {
+      "epoch": 239.1578947368421,
+      "grad_norm": 0.6979864239692688,
+      "learning_rate": 0.00010525139575218063,
+      "loss": 0.1959,
+      "step": 11360
+    },
+    {
+      "epoch": 239.3684210526316,
+      "grad_norm": 0.6338476538658142,
+      "learning_rate": 0.00010511789076521668,
+      "loss": 0.1866,
+      "step": 11370
+    },
+    {
+      "epoch": 239.57894736842104,
+      "grad_norm": 0.7045919299125671,
+      "learning_rate": 0.0001049843766317521,
+      "loss": 0.1935,
+      "step": 11380
+    },
+    {
+      "epoch": 239.78947368421052,
+      "grad_norm": 0.5786710977554321,
+      "learning_rate": 0.00010485085359039828,
+      "loss": 0.1875,
+      "step": 11390
+    },
+    {
+      "epoch": 240.0,
+      "grad_norm": 0.6468011736869812,
+      "learning_rate": 0.0001047173218797826,
+      "loss": 0.1982,
+      "step": 11400
+    },
+    {
+      "epoch": 240.21052631578948,
+      "grad_norm": 0.4997261166572571,
+      "learning_rate": 0.00010458378173854783,
+      "loss": 0.186,
+      "step": 11410
+    },
+    {
+      "epoch": 240.42105263157896,
+      "grad_norm": 0.44703060388565063,
+      "learning_rate": 0.00010445023340535185,
+      "loss": 0.1809,
+      "step": 11420
+    },
+    {
+      "epoch": 240.6315789473684,
+      "grad_norm": 0.8224093317985535,
+      "learning_rate": 0.00010431667711886721,
+      "loss": 0.1949,
+      "step": 11430
+    },
+    {
+      "epoch": 240.8421052631579,
+      "grad_norm": 0.6010963320732117,
+      "learning_rate": 0.00010418311311778066,
+      "loss": 0.1917,
+      "step": 11440
+    },
+    {
+      "epoch": 241.05263157894737,
+      "grad_norm": 0.5227510929107666,
+      "learning_rate": 0.0001040495416407927,
+      "loss": 0.2126,
+      "step": 11450
+    },
+    {
+      "epoch": 241.26315789473685,
+      "grad_norm": 0.4908906817436218,
+      "learning_rate": 0.00010391596292661722,
+      "loss": 0.1938,
+      "step": 11460
+    },
+    {
+      "epoch": 241.47368421052633,
+      "grad_norm": 0.6381858587265015,
+      "learning_rate": 0.00010378237721398106,
+      "loss": 0.1838,
+      "step": 11470
+    },
+    {
+      "epoch": 241.68421052631578,
+      "grad_norm": 0.6832618713378906,
+      "learning_rate": 0.00010364878474162354,
+      "loss": 0.1942,
+      "step": 11480
+    },
+    {
+      "epoch": 241.89473684210526,
+      "grad_norm": 0.7349768280982971,
+      "learning_rate": 0.00010351518574829602,
+      "loss": 0.2055,
+      "step": 11490
+    },
+    {
+      "epoch": 242.10526315789474,
+      "grad_norm": 0.7650448083877563,
+      "learning_rate": 0.00010338158047276165,
+      "loss": 0.1882,
+      "step": 11500
+    },
+    {
+      "epoch": 242.31578947368422,
+      "grad_norm": 0.5663439631462097,
+      "learning_rate": 0.00010324796915379466,
+      "loss": 0.1913,
+      "step": 11510
+    },
+    {
+      "epoch": 242.52631578947367,
+      "grad_norm": 0.5075819492340088,
+      "learning_rate": 0.00010311435203018018,
+      "loss": 0.1845,
+      "step": 11520
+    },
+    {
+      "epoch": 242.73684210526315,
+      "grad_norm": 0.5255244970321655,
+      "learning_rate": 0.00010298072934071363,
+      "loss": 0.1835,
+      "step": 11530
+    },
+    {
+      "epoch": 242.94736842105263,
+      "grad_norm": 0.5974047780036926,
+      "learning_rate": 0.00010284710132420045,
+      "loss": 0.195,
+      "step": 11540
+    },
+    {
+      "epoch": 243.1578947368421,
+      "grad_norm": 0.719183623790741,
+      "learning_rate": 0.00010271346821945558,
+      "loss": 0.2077,
+      "step": 11550
+    },
+    {
+      "epoch": 243.3684210526316,
+      "grad_norm": 0.5450050234794617,
+      "learning_rate": 0.00010257983026530302,
+      "loss": 0.1757,
+      "step": 11560
+    },
+    {
+      "epoch": 243.57894736842104,
+      "grad_norm": 0.8635089993476868,
+      "learning_rate": 0.0001024461877005755,
+      "loss": 0.2051,
+      "step": 11570
+    },
+    {
+      "epoch": 243.78947368421052,
+      "grad_norm": 0.6234176158905029,
+      "learning_rate": 0.0001023125407641139,
+      "loss": 0.2019,
+      "step": 11580
+    },
+    {
+      "epoch": 244.0,
+      "grad_norm": 0.5475245714187622,
+      "learning_rate": 0.00010217888969476699,
+      "loss": 0.1883,
+      "step": 11590
+    },
+    {
+      "epoch": 244.21052631578948,
+      "grad_norm": 0.5728968381881714,
+      "learning_rate": 0.00010204523473139094,
+      "loss": 0.1979,
+      "step": 11600
+    },
+    {
+      "epoch": 244.42105263157896,
+      "grad_norm": 0.6688836812973022,
+      "learning_rate": 0.00010191157611284876,
+      "loss": 0.1858,
+      "step": 11610
+    },
+    {
+      "epoch": 244.6315789473684,
+      "grad_norm": 0.5823827385902405,
+      "learning_rate": 0.00010177791407801017,
+      "loss": 0.1918,
+      "step": 11620
+    },
+    {
+      "epoch": 244.8421052631579,
+      "grad_norm": 0.46818673610687256,
+      "learning_rate": 0.0001016442488657508,
+      "loss": 0.1982,
+      "step": 11630
+    },
+    {
+      "epoch": 245.05263157894737,
+      "grad_norm": 0.6950215101242065,
+      "learning_rate": 0.00010151058071495211,
+      "loss": 0.1994,
+      "step": 11640
+    },
+    {
+      "epoch": 245.26315789473685,
+      "grad_norm": 0.6790728569030762,
+      "learning_rate": 0.00010137690986450079,
+      "loss": 0.1833,
+      "step": 11650
+    },
+    {
+      "epoch": 245.47368421052633,
+      "grad_norm": 0.5021740198135376,
+      "learning_rate": 0.00010124323655328826,
+      "loss": 0.1758,
+      "step": 11660
+    },
+    {
+      "epoch": 245.68421052631578,
+      "grad_norm": 0.6168434619903564,
+      "learning_rate": 0.00010110956102021043,
+      "loss": 0.2001,
+      "step": 11670
+    },
+    {
+      "epoch": 245.89473684210526,
+      "grad_norm": 0.6883941888809204,
+      "learning_rate": 0.00010097588350416715,
+      "loss": 0.2136,
+      "step": 11680
+    },
+    {
+      "epoch": 246.10526315789474,
+      "grad_norm": 0.703221321105957,
+      "learning_rate": 0.00010084220424406183,
+      "loss": 0.1941,
+      "step": 11690
+    },
+    {
+      "epoch": 246.31578947368422,
+      "grad_norm": 0.5424191951751709,
+      "learning_rate": 0.00010070852347880095,
+      "loss": 0.1827,
+      "step": 11700
+    },
+    {
+      "epoch": 246.52631578947367,
+      "grad_norm": 0.6174150705337524,
+      "learning_rate": 0.00010057484144729375,
+      "loss": 0.1899,
+      "step": 11710
+    },
+    {
+      "epoch": 246.73684210526315,
+      "grad_norm": 0.6344186663627625,
+      "learning_rate": 0.00010044115838845167,
+      "loss": 0.1847,
+      "step": 11720
+    },
+    {
+      "epoch": 246.94736842105263,
+      "grad_norm": 0.5950223207473755,
+      "learning_rate": 0.000100307474541188,
+      "loss": 0.1944,
+      "step": 11730
+    },
+    {
+      "epoch": 247.1578947368421,
+      "grad_norm": 0.6062918305397034,
+      "learning_rate": 0.0001001737901444175,
+      "loss": 0.1882,
+      "step": 11740
+    },
+    {
+      "epoch": 247.3684210526316,
+      "grad_norm": 0.6734747290611267,
+      "learning_rate": 0.00010004010543705583,
+      "loss": 0.1873,
+      "step": 11750
+    },
+    {
+      "epoch": 247.57894736842104,
+      "grad_norm": 0.5567269325256348,
+      "learning_rate": 9.990642065801922e-05,
+      "loss": 0.1905,
+      "step": 11760
+    },
+    {
+      "epoch": 247.78947368421052,
+      "grad_norm": 0.638371467590332,
+      "learning_rate": 9.977273604622408e-05,
+      "loss": 0.1911,
+      "step": 11770
+    },
+    {
+      "epoch": 248.0,
+      "grad_norm": 0.7093550562858582,
+      "learning_rate": 9.963905184058648e-05,
+      "loss": 0.2124,
+      "step": 11780
+    },
+    {
+      "epoch": 248.21052631578948,
+      "grad_norm": 0.5440819263458252,
+      "learning_rate": 9.950536828002174e-05,
+      "loss": 0.1972,
+      "step": 11790
+    },
+    {
+      "epoch": 248.42105263157896,
+      "grad_norm": 0.4559662640094757,
+      "learning_rate": 9.937168560344412e-05,
+      "loss": 0.1837,
+      "step": 11800
+    },
+    {
+      "epoch": 248.6315789473684,
+      "grad_norm": 0.6494418978691101,
+      "learning_rate": 9.923800404976619e-05,
+      "loss": 0.2035,
+      "step": 11810
+    },
+    {
+      "epoch": 248.8421052631579,
+      "grad_norm": 0.6026273369789124,
+      "learning_rate": 9.910432385789855e-05,
+      "loss": 0.1968,
+      "step": 11820
+    },
+    {
+      "epoch": 249.05263157894737,
+      "grad_norm": 0.5863096714019775,
+      "learning_rate": 9.897064526674944e-05,
+      "loss": 0.198,
+      "step": 11830
+    },
+    {
+      "epoch": 249.26315789473685,
+      "grad_norm": 0.6923006176948547,
+      "learning_rate": 9.883696851522412e-05,
+      "loss": 0.1906,
+      "step": 11840
+    },
+    {
+      "epoch": 249.47368421052633,
+      "grad_norm": 0.6653925180435181,
+      "learning_rate": 9.870329384222465e-05,
+      "loss": 0.1976,
+      "step": 11850
+    },
+    {
+      "epoch": 249.68421052631578,
+      "grad_norm": 0.6321898698806763,
+      "learning_rate": 9.85696214866493e-05,
+      "loss": 0.1967,
+      "step": 11860
+    },
+    {
+      "epoch": 249.89473684210526,
+      "grad_norm": 0.5583874583244324,
+      "learning_rate": 9.843595168739233e-05,
+      "loss": 0.1899,
+      "step": 11870
+    },
+    {
+      "epoch": 250.10526315789474,
+      "grad_norm": 0.5806682109832764,
+      "learning_rate": 9.830228468334329e-05,
+      "loss": 0.1894,
+      "step": 11880
+    },
+    {
+      "epoch": 250.31578947368422,
+      "grad_norm": 0.519559383392334,
+      "learning_rate": 9.816862071338675e-05,
+      "loss": 0.1802,
+      "step": 11890
+    },
+    {
+      "epoch": 250.52631578947367,
+      "grad_norm": 0.6656150817871094,
+      "learning_rate": 9.803496001640198e-05,
+      "loss": 0.189,
+      "step": 11900
+    },
+    {
+      "epoch": 250.73684210526315,
+      "grad_norm": 1.0183082818984985,
+      "learning_rate": 9.790130283126226e-05,
+      "loss": 0.2056,
+      "step": 11910
+    },
+    {
+      "epoch": 250.94736842105263,
+      "grad_norm": 0.5636556148529053,
+      "learning_rate": 9.776764939683463e-05,
+      "loss": 0.1879,
+      "step": 11920
+    },
+    {
+      "epoch": 251.1578947368421,
+      "grad_norm": 0.5543136596679688,
+      "learning_rate": 9.763399995197955e-05,
+      "loss": 0.2136,
+      "step": 11930
+    },
+    {
+      "epoch": 251.3684210526316,
+      "grad_norm": 0.48448771238327026,
+      "learning_rate": 9.750035473555016e-05,
+      "loss": 0.1811,
+      "step": 11940
+    },
+    {
+      "epoch": 251.57894736842104,
+      "grad_norm": 0.7867603302001953,
+      "learning_rate": 9.736671398639217e-05,
+      "loss": 0.1843,
+      "step": 11950
+    },
+    {
+      "epoch": 251.78947368421052,
+      "grad_norm": 0.6914984583854675,
+      "learning_rate": 9.723307794334322e-05,
+      "loss": 0.1912,
+      "step": 11960
+    },
+    {
+      "epoch": 252.0,
+      "grad_norm": NaN,
+      "learning_rate": 9.711280972571503e-05,
+      "loss": 0.2048,
+      "step": 11970
+    },
+    {
+      "epoch": 252.21052631578948,
+      "grad_norm": 0.4894615411758423,
+      "learning_rate": 9.697918328224071e-05,
+      "loss": 0.188,
+      "step": 11980
+    },
+    {
+      "epoch": 252.42105263157896,
+      "grad_norm": 0.7946804761886597,
+      "learning_rate": 9.684556223745563e-05,
+      "loss": 0.1935,
+      "step": 11990
+    },
+    {
+      "epoch": 252.6315789473684,
+      "grad_norm": 0.5945575833320618,
+      "learning_rate": 9.671194683016235e-05,
+      "loss": 0.1974,
+      "step": 12000
+    },
+    {
+      "epoch": 252.8421052631579,
+      "grad_norm": 0.5779606103897095,
+      "learning_rate": 9.65783372991532e-05,
+      "loss": 0.1951,
+      "step": 12010
+    },
+    {
+      "epoch": 253.05263157894737,
+      "grad_norm": 0.5105938911437988,
+      "learning_rate": 9.644473388321008e-05,
+      "loss": 0.1904,
+      "step": 12020
+    },
+    {
+      "epoch": 253.26315789473685,
+      "grad_norm": 0.6342964172363281,
+      "learning_rate": 9.631113682110396e-05,
+      "loss": 0.1927,
+      "step": 12030
+    },
+    {
+      "epoch": 253.47368421052633,
+      "grad_norm": 0.6746343970298767,
+      "learning_rate": 9.61775463515945e-05,
+      "loss": 0.1941,
+      "step": 12040
+    },
+    {
+      "epoch": 253.68421052631578,
+      "grad_norm": 0.6832113862037659,
+      "learning_rate": 9.604396271342943e-05,
+      "loss": 0.1879,
+      "step": 12050
+    },
+    {
+      "epoch": 253.89473684210526,
+      "grad_norm": 0.4744756817817688,
+      "learning_rate": 9.59103861453445e-05,
+      "loss": 0.1896,
+      "step": 12060
+    },
+    {
+      "epoch": 254.10526315789474,
+      "grad_norm": 0.6902132034301758,
+      "learning_rate": 9.577681688606262e-05,
+      "loss": 0.1884,
+      "step": 12070
+    },
+    {
+      "epoch": 254.31578947368422,
+      "grad_norm": 0.5863637924194336,
+      "learning_rate": 9.564325517429369e-05,
+      "loss": 0.2072,
+      "step": 12080
+    },
+    {
+      "epoch": 254.52631578947367,
+      "grad_norm": 0.5211676359176636,
+      "learning_rate": 9.550970124873417e-05,
+      "loss": 0.1781,
+      "step": 12090
+    },
+    {
+      "epoch": 254.73684210526315,
+      "grad_norm": 0.5536790490150452,
+      "learning_rate": 9.537615534806662e-05,
+      "loss": 0.1973,
+      "step": 12100
+    },
+    {
+      "epoch": 254.94736842105263,
+      "grad_norm": 0.5267168283462524,
+      "learning_rate": 9.524261771095919e-05,
+      "loss": 0.1871,
+      "step": 12110
+    },
+    {
+      "epoch": 255.1578947368421,
+      "grad_norm": 0.6351366639137268,
+      "learning_rate": 9.510908857606522e-05,
+      "loss": 0.1895,
+      "step": 12120
+    },
+    {
+      "epoch": 255.3684210526316,
+      "grad_norm": 0.5926997661590576,
+      "learning_rate": 9.497556818202306e-05,
+      "loss": 0.181,
+      "step": 12130
+    },
+    {
+      "epoch": 255.57894736842104,
+      "grad_norm": 0.7168594002723694,
+      "learning_rate": 9.48420567674552e-05,
+      "loss": 0.1904,
+      "step": 12140
+    },
+    {
+      "epoch": 255.78947368421052,
+      "grad_norm": 0.6090984344482422,
+      "learning_rate": 9.470855457096824e-05,
+      "loss": 0.1957,
+      "step": 12150
+    },
+    {
+      "epoch": 256.0,
+      "grad_norm": 0.5657641887664795,
+      "learning_rate": 9.457506183115217e-05,
+      "loss": 0.2069,
+      "step": 12160
+    },
+    {
+      "epoch": 256.2105263157895,
+      "grad_norm": 0.5447742342948914,
+      "learning_rate": 9.444157878658028e-05,
+      "loss": 0.1941,
+      "step": 12170
+    },
+    {
+      "epoch": 256.42105263157896,
+      "grad_norm": 0.6158611178398132,
+      "learning_rate": 9.430810567580836e-05,
+      "loss": 0.194,
+      "step": 12180
+    },
+    {
+      "epoch": 256.63157894736844,
+      "grad_norm": 0.5748459100723267,
+      "learning_rate": 9.417464273737444e-05,
+      "loss": 0.1948,
+      "step": 12190
+    },
+    {
+      "epoch": 256.8421052631579,
+      "grad_norm": 0.5924521088600159,
+      "learning_rate": 9.404119020979853e-05,
+      "loss": 0.1962,
+      "step": 12200
+    },
+    {
+      "epoch": 257.05263157894734,
+      "grad_norm": 0.6660053730010986,
+      "learning_rate": 9.390774833158186e-05,
+      "loss": 0.1842,
+      "step": 12210
+    },
+    {
+      "epoch": 257.2631578947368,
+      "grad_norm": 0.5104465484619141,
+      "learning_rate": 9.377431734120673e-05,
+      "loss": 0.1955,
+      "step": 12220
+    },
+    {
+      "epoch": 257.4736842105263,
+      "grad_norm": 0.4942176342010498,
+      "learning_rate": 9.364089747713599e-05,
+      "loss": 0.1897,
+      "step": 12230
+    },
+    {
+      "epoch": 257.6842105263158,
+      "grad_norm": 0.49282151460647583,
+      "learning_rate": 9.350748897781254e-05,
+      "loss": 0.1835,
+      "step": 12240
+    },
+    {
+      "epoch": 257.89473684210526,
+      "grad_norm": 0.5877977013587952,
+      "learning_rate": 9.337409208165898e-05,
+      "loss": 0.2036,
+      "step": 12250
+    },
+    {
+      "epoch": 258.10526315789474,
+      "grad_norm": 0.4874722361564636,
+      "learning_rate": 9.32407070270772e-05,
+      "loss": 0.1842,
+      "step": 12260
+    },
+    {
+      "epoch": 258.3157894736842,
+      "grad_norm": 0.6626269221305847,
+      "learning_rate": 9.310733405244795e-05,
+      "loss": 0.187,
+      "step": 12270
+    },
+    {
+      "epoch": 258.5263157894737,
+      "grad_norm": 0.5759614109992981,
+      "learning_rate": 9.297397339613035e-05,
+      "loss": 0.2007,
+      "step": 12280
+    },
+    {
+      "epoch": 258.7368421052632,
+      "grad_norm": 0.5541868805885315,
+      "learning_rate": 9.284062529646146e-05,
+      "loss": 0.1879,
+      "step": 12290
+    },
+    {
+      "epoch": 258.94736842105266,
+      "grad_norm": 0.5124220252037048,
+      "learning_rate": 9.270728999175605e-05,
+      "loss": 0.1843,
+      "step": 12300
+    },
+    {
+      "epoch": 259.1578947368421,
+      "grad_norm": 0.39318570494651794,
+      "learning_rate": 9.257396772030589e-05,
+      "loss": 0.1904,
+      "step": 12310
+    },
+    {
+      "epoch": 259.36842105263156,
+      "grad_norm": 0.5387923121452332,
+      "learning_rate": 9.244065872037946e-05,
+      "loss": 0.1837,
+      "step": 12320
+    },
+    {
+      "epoch": 259.57894736842104,
+      "grad_norm": 0.5390989780426025,
+      "learning_rate": 9.230736323022157e-05,
+      "loss": 0.1773,
+      "step": 12330
+    },
+    {
+      "epoch": 259.7894736842105,
+      "grad_norm": 0.7210453748703003,
+      "learning_rate": 9.217408148805292e-05,
+      "loss": 0.2003,
+      "step": 12340
+    },
+    {
+      "epoch": 260.0,
+      "grad_norm": 0.5788223743438721,
+      "learning_rate": 9.204081373206958e-05,
+      "loss": 0.1979,
+      "step": 12350
+    },
+    {
+      "epoch": 260.2105263157895,
+      "grad_norm": 0.6655392050743103,
+      "learning_rate": 9.190756020044257e-05,
+      "loss": 0.1754,
+      "step": 12360
+    },
+    {
+      "epoch": 260.42105263157896,
+      "grad_norm": 0.6042269468307495,
+      "learning_rate": 9.177432113131766e-05,
+      "loss": 0.192,
+      "step": 12370
+    },
+    {
+      "epoch": 260.63157894736844,
+      "grad_norm": 0.8559859395027161,
+      "learning_rate": 9.164109676281458e-05,
+      "loss": 0.2043,
+      "step": 12380
+    },
+    {
+      "epoch": 260.8421052631579,
+      "grad_norm": 0.5184524059295654,
+      "learning_rate": 9.150788733302691e-05,
+      "loss": 0.1964,
+      "step": 12390
+    },
+    {
+      "epoch": 261.05263157894734,
+      "grad_norm": 0.6655091643333435,
+      "learning_rate": 9.137469308002154e-05,
+      "loss": 0.211,
+      "step": 12400
+    },
+    {
+      "epoch": 261.2631578947368,
+      "grad_norm": 0.5456916689872742,
+      "learning_rate": 9.124151424183817e-05,
+      "loss": 0.1753,
+      "step": 12410
+    },
+    {
+      "epoch": 261.4736842105263,
+      "grad_norm": 0.6656398177146912,
+      "learning_rate": 9.110835105648898e-05,
+      "loss": 0.1864,
+      "step": 12420
+    },
+    {
+      "epoch": 261.6842105263158,
+      "grad_norm": 0.5527902245521545,
+      "learning_rate": 9.097520376195811e-05,
+      "loss": 0.202,
+      "step": 12430
+    },
+    {
+      "epoch": 261.89473684210526,
+      "grad_norm": 0.5181113481521606,
+      "learning_rate": 9.084207259620144e-05,
+      "loss": 0.1909,
+      "step": 12440
+    },
+    {
+      "epoch": 262.10526315789474,
+      "grad_norm": 0.4919548034667969,
+      "learning_rate": 9.070895779714597e-05,
+      "loss": 0.1996,
+      "step": 12450
+    },
+    {
+      "epoch": 262.3157894736842,
+      "grad_norm": 0.5647410154342651,
+      "learning_rate": 9.057585960268931e-05,
+      "loss": 0.1774,
+      "step": 12460
+    },
+    {
+      "epoch": 262.5263157894737,
+      "grad_norm": 0.5741182565689087,
+      "learning_rate": 9.044277825069967e-05,
+      "loss": 0.1846,
+      "step": 12470
+    },
+    {
+      "epoch": 262.7368421052632,
+      "grad_norm": 0.7225587964057922,
+      "learning_rate": 9.030971397901491e-05,
+      "loss": 0.1999,
+      "step": 12480
+    },
+    {
+      "epoch": 262.94736842105266,
+      "grad_norm": 0.4701984226703644,
+      "learning_rate": 9.017666702544245e-05,
+      "loss": 0.1977,
+      "step": 12490
+    },
+    {
+      "epoch": 263.1578947368421,
+      "grad_norm": 0.5436278581619263,
+      "learning_rate": 9.00436376277588e-05,
+      "loss": 0.1948,
+      "step": 12500
+    },
+    {
+      "epoch": 263.36842105263156,
+      "grad_norm": 0.5927190780639648,
+      "learning_rate": 8.991062602370907e-05,
+      "loss": 0.2018,
+      "step": 12510
+    },
+    {
+      "epoch": 263.57894736842104,
+      "grad_norm": 0.5915952324867249,
+      "learning_rate": 8.977763245100656e-05,
+      "loss": 0.1815,
+      "step": 12520
+    },
+    {
+      "epoch": 263.7894736842105,
+      "grad_norm": 0.6053709387779236,
+      "learning_rate": 8.964465714733229e-05,
+      "loss": 0.2057,
+      "step": 12530
+    },
+    {
+      "epoch": 264.0,
+      "grad_norm": 0.5991526246070862,
+      "learning_rate": 8.951170035033478e-05,
+      "loss": 0.1847,
+      "step": 12540
+    },
+    {
+      "epoch": 264.2105263157895,
+      "grad_norm": 0.43415001034736633,
+      "learning_rate": 8.937876229762933e-05,
+      "loss": 0.1909,
+      "step": 12550
+    },
+    {
+      "epoch": 264.42105263157896,
+      "grad_norm": 0.571382462978363,
+      "learning_rate": 8.92458432267978e-05,
+      "loss": 0.1816,
+      "step": 12560
+    },
+    {
+      "epoch": 264.63157894736844,
+      "grad_norm": 0.5127904415130615,
+      "learning_rate": 8.911294337538813e-05,
+      "loss": 0.1895,
+      "step": 12570
+    },
+    {
+      "epoch": 264.8421052631579,
+      "grad_norm": 0.5136194825172424,
+      "learning_rate": 8.898006298091392e-05,
+      "loss": 0.2023,
+      "step": 12580
+    },
+    {
+      "epoch": 265.05263157894734,
+      "grad_norm": 0.5507082343101501,
+      "learning_rate": 8.884720228085397e-05,
+      "loss": 0.1958,
+      "step": 12590
+    },
+    {
+      "epoch": 265.2631578947368,
+      "grad_norm": 0.441049188375473,
+      "learning_rate": 8.871436151265184e-05,
+      "loss": 0.1869,
+      "step": 12600
+    },
+    {
+      "epoch": 265.4736842105263,
+      "grad_norm": 0.5187159776687622,
+      "learning_rate": 8.85815409137156e-05,
+      "loss": 0.184,
+      "step": 12610
+    },
+    {
+      "epoch": 265.6842105263158,
+      "grad_norm": 0.6891997456550598,
+      "learning_rate": 8.844874072141715e-05,
+      "loss": 0.1951,
+      "step": 12620
+    },
+    {
+      "epoch": 265.89473684210526,
+      "grad_norm": 0.6691162586212158,
+      "learning_rate": 8.831596117309195e-05,
+      "loss": 0.1935,
+      "step": 12630
+    },
+    {
+      "epoch": 266.10526315789474,
+      "grad_norm": 0.8119604587554932,
+      "learning_rate": 8.818320250603866e-05,
+      "loss": 0.1909,
+      "step": 12640
+    },
+    {
+      "epoch": 266.3157894736842,
+      "grad_norm": 0.5154120326042175,
+      "learning_rate": 8.80504649575185e-05,
+      "loss": 0.1858,
+      "step": 12650
+    },
+    {
+      "epoch": 266.5263157894737,
+      "grad_norm": 0.5835145711898804,
+      "learning_rate": 8.7917748764755e-05,
+      "loss": 0.1811,
+      "step": 12660
+    },
+    {
+      "epoch": 266.7368421052632,
+      "grad_norm": 0.601320207118988,
+      "learning_rate": 8.778505416493343e-05,
+      "loss": 0.1929,
+      "step": 12670
+    },
+    {
+      "epoch": 266.94736842105266,
+      "grad_norm": 0.6263984441757202,
+      "learning_rate": 8.765238139520067e-05,
+      "loss": 0.2011,
+      "step": 12680
+    },
+    {
+      "epoch": 267.1578947368421,
+      "grad_norm": 0.6324906945228577,
+      "learning_rate": 8.751973069266444e-05,
+      "loss": 0.1813,
+      "step": 12690
+    },
+    {
+      "epoch": 267.36842105263156,
+      "grad_norm": 0.5795466303825378,
+      "learning_rate": 8.7387102294393e-05,
+      "loss": 0.1942,
+      "step": 12700
+    },
+    {
+      "epoch": 267.57894736842104,
+      "grad_norm": 0.4995231330394745,
+      "learning_rate": 8.725449643741487e-05,
+      "loss": 0.1838,
+      "step": 12710
+    },
+    {
+      "epoch": 267.7894736842105,
+      "grad_norm": 0.6501296162605286,
+      "learning_rate": 8.712191335871822e-05,
+      "loss": 0.1951,
+      "step": 12720
+    },
+    {
+      "epoch": 268.0,
+      "grad_norm": 0.7934373617172241,
+      "learning_rate": 8.698935329525043e-05,
+      "loss": 0.1987,
+      "step": 12730
+    },
+    {
+      "epoch": 268.2105263157895,
+      "grad_norm": 0.5606156587600708,
+      "learning_rate": 8.685681648391791e-05,
+      "loss": 0.1904,
+      "step": 12740
+    },
+    {
+      "epoch": 268.42105263157896,
+      "grad_norm": 0.5056310296058655,
+      "learning_rate": 8.672430316158541e-05,
+      "loss": 0.1855,
+      "step": 12750
+    },
+    {
+      "epoch": 268.63157894736844,
+      "grad_norm": 0.48892274498939514,
+      "learning_rate": 8.659181356507571e-05,
+      "loss": 0.1878,
+      "step": 12760
+    },
+    {
+      "epoch": 268.8421052631579,
+      "grad_norm": 0.4638197422027588,
+      "learning_rate": 8.645934793116917e-05,
+      "loss": 0.1937,
+      "step": 12770
+    },
+    {
+      "epoch": 269.05263157894734,
+      "grad_norm": 0.3581981956958771,
+      "learning_rate": 8.632690649660342e-05,
+      "loss": 0.2066,
+      "step": 12780
+    },
+    {
+      "epoch": 269.2631578947368,
+      "grad_norm": 0.4948786795139313,
+      "learning_rate": 8.619448949807274e-05,
+      "loss": 0.1829,
+      "step": 12790
+    },
+    {
+      "epoch": 269.4736842105263,
+      "grad_norm": 0.6615347862243652,
+      "learning_rate": 8.606209717222777e-05,
+      "loss": 0.1841,
+      "step": 12800
+    },
+    {
+      "epoch": 269.6842105263158,
+      "grad_norm": 0.5024154782295227,
+      "learning_rate": 8.59297297556751e-05,
+      "loss": 0.1941,
+      "step": 12810
+    },
+    {
+      "epoch": 269.89473684210526,
+      "grad_norm": 0.5818947553634644,
+      "learning_rate": 8.579738748497675e-05,
+      "loss": 0.1993,
+      "step": 12820
+    },
+    {
+      "epoch": 270.10526315789474,
+      "grad_norm": 0.44670799374580383,
+      "learning_rate": 8.566507059664981e-05,
+      "loss": 0.1799,
+      "step": 12830
+    },
+    {
+      "epoch": 270.3157894736842,
+      "grad_norm": 0.5685772895812988,
+      "learning_rate": 8.553277932716599e-05,
+      "loss": 0.2018,
+      "step": 12840
+    },
+    {
+      "epoch": 270.5263157894737,
+      "grad_norm": 0.7774202823638916,
+      "learning_rate": 8.540051391295125e-05,
+      "loss": 0.175,
+      "step": 12850
+    },
+    {
+      "epoch": 270.7368421052632,
+      "grad_norm": 0.6112064719200134,
+      "learning_rate": 8.52682745903854e-05,
+      "loss": 0.1916,
+      "step": 12860
+    },
+    {
+      "epoch": 270.94736842105266,
+      "grad_norm": 0.5548771619796753,
+      "learning_rate": 8.513606159580142e-05,
+      "loss": 0.195,
+      "step": 12870
+    },
+    {
+      "epoch": 271.1578947368421,
+      "grad_norm": 0.5652728080749512,
+      "learning_rate": 8.500387516548549e-05,
+      "loss": 0.185,
+      "step": 12880
+    },
+    {
+      "epoch": 271.36842105263156,
+      "grad_norm": 0.5679665803909302,
+      "learning_rate": 8.487171553567616e-05,
+      "loss": 0.1895,
+      "step": 12890
+    },
+    {
+      "epoch": 271.57894736842104,
+      "grad_norm": 0.6408627033233643,
+      "learning_rate": 8.473958294256406e-05,
+      "loss": 0.1915,
+      "step": 12900
+    },
+    {
+      "epoch": 271.7894736842105,
+      "grad_norm": 0.6951084733009338,
+      "learning_rate": 8.460747762229164e-05,
+      "loss": 0.202,
+      "step": 12910
+    },
+    {
+      "epoch": 272.0,
+      "grad_norm": 0.6909904479980469,
+      "learning_rate": 8.447539981095246e-05,
+      "loss": 0.1847,
+      "step": 12920
+    },
+    {
+      "epoch": 272.2105263157895,
+      "grad_norm": 0.47111183404922485,
+      "learning_rate": 8.434334974459104e-05,
+      "loss": 0.1627,
+      "step": 12930
+    },
+    {
+      "epoch": 272.42105263157896,
+      "grad_norm": 0.5050517320632935,
+      "learning_rate": 8.421132765920219e-05,
+      "loss": 0.1828,
+      "step": 12940
+    },
+    {
+      "epoch": 272.63157894736844,
+      "grad_norm": 0.8324809670448303,
+      "learning_rate": 8.407933379073086e-05,
+      "loss": 0.2063,
+      "step": 12950
+    },
+    {
+      "epoch": 272.8421052631579,
+      "grad_norm": 0.5706649422645569,
+      "learning_rate": 8.394736837507148e-05,
+      "loss": 0.1949,
+      "step": 12960
+    },
+    {
+      "epoch": 273.05263157894734,
+      "grad_norm": 0.5625925064086914,
+      "learning_rate": 8.381543164806756e-05,
+      "loss": 0.2075,
+      "step": 12970
+    },
+    {
+      "epoch": 273.2631578947368,
+      "grad_norm": 0.4872249662876129,
+      "learning_rate": 8.368352384551153e-05,
+      "loss": 0.1857,
+      "step": 12980
+    },
+    {
+      "epoch": 273.4736842105263,
+      "grad_norm": 0.5305653810501099,
+      "learning_rate": 8.3551645203144e-05,
+      "loss": 0.1987,
+      "step": 12990
+    },
+    {
+      "epoch": 273.6842105263158,
+      "grad_norm": 0.46276259422302246,
+      "learning_rate": 8.341979595665346e-05,
+      "loss": 0.1812,
+      "step": 13000
+    },
+    {
+      "epoch": 273.89473684210526,
+      "grad_norm": 0.6188380122184753,
+      "learning_rate": 8.328797634167586e-05,
+      "loss": 0.1948,
+      "step": 13010
+    },
+    {
+      "epoch": 274.10526315789474,
+      "grad_norm": 0.4250720143318176,
+      "learning_rate": 8.315618659379429e-05,
+      "loss": 0.1867,
+      "step": 13020
+    },
+    {
+      "epoch": 274.3157894736842,
+      "grad_norm": 0.48751601576805115,
+      "learning_rate": 8.302442694853838e-05,
+      "loss": 0.1806,
+      "step": 13030
+    },
+    {
+      "epoch": 274.5263157894737,
+      "grad_norm": 0.45683354139328003,
+      "learning_rate": 8.289269764138393e-05,
+      "loss": 0.1804,
+      "step": 13040
+    },
+    {
+      "epoch": 274.7368421052632,
+      "grad_norm": 0.6656365990638733,
+      "learning_rate": 8.276099890775266e-05,
+      "loss": 0.2056,
+      "step": 13050
+    },
+    {
+      "epoch": 274.94736842105266,
+      "grad_norm": 0.4965307414531708,
+      "learning_rate": 8.262933098301152e-05,
+      "loss": 0.1938,
+      "step": 13060
+    },
+    {
+      "epoch": 275.1578947368421,
+      "grad_norm": 0.5132878422737122,
+      "learning_rate": 8.249769410247239e-05,
+      "loss": 0.192,
+      "step": 13070
+    },
+    {
+      "epoch": 275.36842105263156,
+      "grad_norm": 0.5989632606506348,
+      "learning_rate": 8.23660885013918e-05,
+      "loss": 0.1897,
+      "step": 13080
+    },
+    {
+      "epoch": 275.57894736842104,
+      "grad_norm": 0.6420390009880066,
+      "learning_rate": 8.223451441497026e-05,
+      "loss": 0.1839,
+      "step": 13090
+    },
+    {
+      "epoch": 275.7894736842105,
+      "grad_norm": 0.6834862232208252,
+      "learning_rate": 8.2102972078352e-05,
+      "loss": 0.205,
+      "step": 13100
+    },
+    {
+      "epoch": 276.0,
+      "grad_norm": 0.5762868523597717,
+      "learning_rate": 8.197146172662447e-05,
+      "loss": 0.195,
+      "step": 13110
+    },
+    {
+      "epoch": 276.2105263157895,
+      "grad_norm": 0.4554460346698761,
+      "learning_rate": 8.183998359481806e-05,
+      "loss": 0.1782,
+      "step": 13120
+    },
+    {
+      "epoch": 276.42105263157896,
+      "grad_norm": 0.720638632774353,
+      "learning_rate": 8.170853791790547e-05,
+      "loss": 0.1919,
+      "step": 13130
+    },
+    {
+      "epoch": 276.63157894736844,
+      "grad_norm": 0.5189310908317566,
+      "learning_rate": 8.15771249308014e-05,
+      "loss": 0.1967,
+      "step": 13140
+    },
+    {
+      "epoch": 276.8421052631579,
+      "grad_norm": 0.8014587759971619,
+      "learning_rate": 8.144574486836228e-05,
+      "loss": 0.1936,
+      "step": 13150
+    },
+    {
+      "epoch": 277.05263157894734,
+      "grad_norm": 0.5890443921089172,
+      "learning_rate": 8.131439796538546e-05,
+      "loss": 0.1888,
+      "step": 13160
+    },
+    {
+      "epoch": 277.2631578947368,
+      "grad_norm": 0.5942811369895935,
+      "learning_rate": 8.118308445660923e-05,
+      "loss": 0.1883,
+      "step": 13170
+    },
+    {
+      "epoch": 277.4736842105263,
+      "grad_norm": 0.5413142442703247,
+      "learning_rate": 8.105180457671204e-05,
+      "loss": 0.1807,
+      "step": 13180
+    },
+    {
+      "epoch": 277.6842105263158,
+      "grad_norm": 0.5064405798912048,
+      "learning_rate": 8.092055856031244e-05,
+      "loss": 0.1935,
+      "step": 13190
+    },
+    {
+      "epoch": 277.89473684210526,
+      "grad_norm": 0.6160650253295898,
+      "learning_rate": 8.078934664196825e-05,
+      "loss": 0.1909,
+      "step": 13200
+    },
+    {
+      "epoch": 278.10526315789474,
+      "grad_norm": 0.7421651482582092,
+      "learning_rate": 8.065816905617647e-05,
+      "loss": 0.1884,
+      "step": 13210
+    },
+    {
+      "epoch": 278.3157894736842,
+      "grad_norm": 0.5469620227813721,
+      "learning_rate": 8.052702603737272e-05,
+      "loss": 0.2007,
+      "step": 13220
+    },
+    {
+      "epoch": 278.5263157894737,
+      "grad_norm": 0.6061626076698303,
+      "learning_rate": 8.039591781993086e-05,
+      "loss": 0.1947,
+      "step": 13230
+    },
+    {
+      "epoch": 278.7368421052632,
+      "grad_norm": 0.5570159554481506,
+      "learning_rate": 8.026484463816245e-05,
+      "loss": 0.1858,
+      "step": 13240
+    },
+    {
+      "epoch": 278.94736842105266,
+      "grad_norm": 0.593967080116272,
+      "learning_rate": 8.013380672631664e-05,
+      "loss": 0.2003,
+      "step": 13250
+    },
+    {
+      "epoch": 279.1578947368421,
+      "grad_norm": 0.6793128848075867,
+      "learning_rate": 8.000280431857933e-05,
+      "loss": 0.2016,
+      "step": 13260
+    },
+    {
+      "epoch": 279.36842105263156,
+      "grad_norm": 0.7546000480651855,
+      "learning_rate": 7.98718376490731e-05,
+      "loss": 0.1938,
+      "step": 13270
+    },
+    {
+      "epoch": 279.57894736842104,
+      "grad_norm": 0.5035498142242432,
+      "learning_rate": 7.97409069518566e-05,
+      "loss": 0.1867,
+      "step": 13280
+    },
+    {
+      "epoch": 279.7894736842105,
+      "grad_norm": 0.6848862767219543,
+      "learning_rate": 7.961001246092427e-05,
+      "loss": 0.1838,
+      "step": 13290
+    },
+    {
+      "epoch": 280.0,
+      "grad_norm": 0.607519805431366,
+      "learning_rate": 7.947915441020575e-05,
+      "loss": 0.1892,
+      "step": 13300
+    },
+    {
+      "epoch": 280.2105263157895,
+      "grad_norm": 0.6071075797080994,
+      "learning_rate": 7.934833303356556e-05,
+      "loss": 0.1899,
+      "step": 13310
+    },
+    {
+      "epoch": 280.42105263157896,
+      "grad_norm": 0.5092148780822754,
+      "learning_rate": 7.921754856480279e-05,
+      "loss": 0.1826,
+      "step": 13320
+    },
+    {
+      "epoch": 280.63157894736844,
+      "grad_norm": 0.5188595652580261,
+      "learning_rate": 7.908680123765043e-05,
+      "loss": 0.1899,
+      "step": 13330
+    },
+    {
+      "epoch": 280.8421052631579,
+      "grad_norm": 0.6495827436447144,
+      "learning_rate": 7.895609128577514e-05,
+      "loss": 0.1933,
+      "step": 13340
+    },
+    {
+      "epoch": 281.05263157894734,
+      "grad_norm": 0.5069420337677002,
+      "learning_rate": 7.882541894277689e-05,
+      "loss": 0.1933,
+      "step": 13350
+    },
+    {
+      "epoch": 281.2631578947368,
+      "grad_norm": 0.6902370452880859,
+      "learning_rate": 7.869478444218828e-05,
+      "loss": 0.189,
+      "step": 13360
+    },
+    {
+      "epoch": 281.4736842105263,
+      "grad_norm": 0.6552804112434387,
+      "learning_rate": 7.856418801747435e-05,
+      "loss": 0.1977,
+      "step": 13370
+    },
+    {
+      "epoch": 281.6842105263158,
+      "grad_norm": 0.4557342529296875,
+      "learning_rate": 7.843362990203205e-05,
+      "loss": 0.1981,
+      "step": 13380
+    },
+    {
+      "epoch": 281.89473684210526,
+      "grad_norm": 0.44700777530670166,
+      "learning_rate": 7.830311032918994e-05,
+      "loss": 0.1859,
+      "step": 13390
+    },
+    {
+      "epoch": 282.10526315789474,
+      "grad_norm": 0.46295836567878723,
+      "learning_rate": 7.817262953220769e-05,
+      "loss": 0.1845,
+      "step": 13400
+    },
+    {
+      "epoch": 282.3157894736842,
+      "grad_norm": 0.4293200373649597,
+      "learning_rate": 7.804218774427558e-05,
+      "loss": 0.1884,
+      "step": 13410
+    },
+    {
+      "epoch": 282.5263157894737,
+      "grad_norm": 0.5236184000968933,
+      "learning_rate": 7.791178519851427e-05,
+      "loss": 0.1806,
+      "step": 13420
+    },
+    {
+      "epoch": 282.7368421052632,
+      "grad_norm": 0.6687787175178528,
+      "learning_rate": 7.778142212797428e-05,
+      "loss": 0.1915,
+      "step": 13430
+    },
+    {
+      "epoch": 282.94736842105266,
+      "grad_norm": 0.5880486369132996,
+      "learning_rate": 7.765109876563547e-05,
+      "loss": 0.1934,
+      "step": 13440
+    },
+    {
+      "epoch": 283.1578947368421,
+      "grad_norm": 0.4489704966545105,
+      "learning_rate": 7.752081534440689e-05,
+      "loss": 0.182,
+      "step": 13450
+    },
+    {
+      "epoch": 283.36842105263156,
+      "grad_norm": 0.6451957821846008,
+      "learning_rate": 7.739057209712612e-05,
+      "loss": 0.1884,
+      "step": 13460
+    },
+    {
+      "epoch": 283.57894736842104,
+      "grad_norm": 0.7118418216705322,
+      "learning_rate": 7.726036925655897e-05,
+      "loss": 0.2025,
+      "step": 13470
+    },
+    {
+      "epoch": 283.7894736842105,
+      "grad_norm": 0.49697622656822205,
+      "learning_rate": 7.713020705539898e-05,
+      "loss": 0.1944,
+      "step": 13480
+    },
+    {
+      "epoch": 284.0,
+      "grad_norm": 0.6399528384208679,
+      "learning_rate": 7.700008572626718e-05,
+      "loss": 0.1996,
+      "step": 13490
+    },
+    {
+      "epoch": 284.2105263157895,
+      "grad_norm": 0.4944116771221161,
+      "learning_rate": 7.687000550171143e-05,
+      "loss": 0.1754,
+      "step": 13500
+    },
+    {
+      "epoch": 284.42105263157896,
+      "grad_norm": 0.44895878434181213,
+      "learning_rate": 7.67399666142062e-05,
+      "loss": 0.1752,
+      "step": 13510
+    },
+    {
+      "epoch": 284.63157894736844,
+      "grad_norm": 0.5446719527244568,
+      "learning_rate": 7.660996929615206e-05,
+      "loss": 0.1885,
+      "step": 13520
+    },
+    {
+      "epoch": 284.8421052631579,
+      "grad_norm": 0.5553343296051025,
+      "learning_rate": 7.648001377987533e-05,
+      "loss": 0.1954,
+      "step": 13530
+    },
+    {
+      "epoch": 285.05263157894734,
+      "grad_norm": 0.5816590785980225,
+      "learning_rate": 7.635010029762756e-05,
+      "loss": 0.2043,
+      "step": 13540
+    },
+    {
+      "epoch": 285.2631578947368,
+      "grad_norm": 0.6026872992515564,
+      "learning_rate": 7.622022908158518e-05,
+      "loss": 0.1834,
+      "step": 13550
+    },
+    {
+      "epoch": 285.4736842105263,
+      "grad_norm": 0.4454641044139862,
+      "learning_rate": 7.609040036384915e-05,
+      "loss": 0.1829,
+      "step": 13560
+    },
+    {
+      "epoch": 285.6842105263158,
+      "grad_norm": 0.4715535044670105,
+      "learning_rate": 7.596061437644444e-05,
+      "loss": 0.1975,
+      "step": 13570
+    },
+    {
+      "epoch": 285.89473684210526,
+      "grad_norm": 0.5461314916610718,
+      "learning_rate": 7.583087135131961e-05,
+      "loss": 0.1976,
+      "step": 13580
+    },
+    {
+      "epoch": 286.10526315789474,
+      "grad_norm": 0.6173868179321289,
+      "learning_rate": 7.570117152034655e-05,
+      "loss": 0.1919,
+      "step": 13590
+    },
+    {
+      "epoch": 286.3157894736842,
+      "grad_norm": 0.5752949118614197,
+      "learning_rate": 7.557151511531986e-05,
+      "loss": 0.1893,
+      "step": 13600
+    },
+    {
+      "epoch": 286.5263157894737,
+      "grad_norm": 0.5035438537597656,
+      "learning_rate": 7.544190236795655e-05,
+      "loss": 0.1973,
+      "step": 13610
+    },
+    {
+      "epoch": 286.7368421052632,
+      "grad_norm": 0.5442306399345398,
+      "learning_rate": 7.531233350989558e-05,
+      "loss": 0.2008,
+      "step": 13620
+    },
+    {
+      "epoch": 286.94736842105266,
+      "grad_norm": 0.6274810433387756,
+      "learning_rate": 7.518280877269755e-05,
+      "loss": 0.1792,
+      "step": 13630
+    },
+    {
+      "epoch": 287.1578947368421,
+      "grad_norm": 0.5484516620635986,
+      "learning_rate": 7.50533283878442e-05,
+      "loss": 0.1898,
+      "step": 13640
+    },
+    {
+      "epoch": 287.36842105263156,
+      "grad_norm": 0.5255657434463501,
+      "learning_rate": 7.492389258673787e-05,
+      "loss": 0.1921,
+      "step": 13650
+    },
+    {
+      "epoch": 287.57894736842104,
+      "grad_norm": 0.4797198176383972,
+      "learning_rate": 7.479450160070145e-05,
+      "loss": 0.1794,
+      "step": 13660
+    },
+    {
+      "epoch": 287.7894736842105,
+      "grad_norm": 0.6554931402206421,
+      "learning_rate": 7.466515566097753e-05,
+      "loss": 0.1912,
+      "step": 13670
+    },
+    {
+      "epoch": 288.0,
+      "grad_norm": 0.5043107271194458,
+      "learning_rate": 7.453585499872826e-05,
+      "loss": 0.1884,
+      "step": 13680
+    },
+    {
+      "epoch": 288.2105263157895,
+      "grad_norm": 0.6006696224212646,
+      "learning_rate": 7.440659984503495e-05,
+      "loss": 0.1793,
+      "step": 13690
+    },
+    {
+      "epoch": 288.42105263157896,
+      "grad_norm": 0.5427776575088501,
+      "learning_rate": 7.427739043089753e-05,
+      "loss": 0.1773,
+      "step": 13700
+    },
+    {
+      "epoch": 288.63157894736844,
+      "grad_norm": 0.5292754769325256,
+      "learning_rate": 7.41482269872341e-05,
+      "loss": 0.2064,
+      "step": 13710
+    },
+    {
+      "epoch": 288.8421052631579,
+      "grad_norm": 0.485466867685318,
+      "learning_rate": 7.401910974488069e-05,
+      "loss": 0.1936,
+      "step": 13720
+    },
+    {
+      "epoch": 289.05263157894734,
+      "grad_norm": 0.5587201714515686,
+      "learning_rate": 7.389003893459081e-05,
+      "loss": 0.1942,
+      "step": 13730
+    },
+    {
+      "epoch": 289.2631578947368,
+      "grad_norm": 0.5549042820930481,
+      "learning_rate": 7.376101478703485e-05,
+      "loss": 0.1901,
+      "step": 13740
+    },
+    {
+      "epoch": 289.4736842105263,
+      "grad_norm": 0.48474541306495667,
+      "learning_rate": 7.363203753279992e-05,
+      "loss": 0.1956,
+      "step": 13750
+    },
+    {
+      "epoch": 289.6842105263158,
+      "grad_norm": 0.44179201126098633,
+      "learning_rate": 7.35031074023893e-05,
+      "loss": 0.1898,
+      "step": 13760
+    },
+    {
+      "epoch": 289.89473684210526,
+      "grad_norm": 0.7074770927429199,
+      "learning_rate": 7.337422462622203e-05,
+      "loss": 0.1908,
+      "step": 13770
+    },
+    {
+      "epoch": 290.10526315789474,
+      "grad_norm": 0.6862475872039795,
+      "learning_rate": 7.324538943463251e-05,
+      "loss": 0.1944,
+      "step": 13780
+    },
+    {
+      "epoch": 290.3157894736842,
+      "grad_norm": 0.5927271842956543,
+      "learning_rate": 7.31166020578701e-05,
+      "loss": 0.1913,
+      "step": 13790
+    },
+    {
+      "epoch": 290.5263157894737,
+      "grad_norm": 0.5208861231803894,
+      "learning_rate": 7.298786272609878e-05,
+      "loss": 0.1887,
+      "step": 13800
+    },
+    {
+      "epoch": 290.7368421052632,
+      "grad_norm": 0.6315051317214966,
+      "learning_rate": 7.285917166939658e-05,
+      "loss": 0.183,
+      "step": 13810
+    },
+    {
+      "epoch": 290.94736842105266,
+      "grad_norm": 0.6303961277008057,
+      "learning_rate": 7.273052911775524e-05,
+      "loss": 0.1909,
+      "step": 13820
+    },
+    {
+      "epoch": 291.1578947368421,
+      "grad_norm": 0.5073893666267395,
+      "learning_rate": 7.260193530107994e-05,
+      "loss": 0.192,
+      "step": 13830
+    },
+    {
+      "epoch": 291.36842105263156,
+      "grad_norm": 0.501746654510498,
+      "learning_rate": 7.247339044918867e-05,
+      "loss": 0.1808,
+      "step": 13840
+    },
+    {
+      "epoch": 291.57894736842104,
+      "grad_norm": 0.6303046941757202,
+      "learning_rate": 7.234489479181185e-05,
+      "loss": 0.1782,
+      "step": 13850
+    },
+    {
+      "epoch": 291.7894736842105,
+      "grad_norm": 0.6213247179985046,
+      "learning_rate": 7.221644855859213e-05,
+      "loss": 0.1964,
+      "step": 13860
+    },
+    {
+      "epoch": 292.0,
+      "grad_norm": 0.5928416848182678,
+      "learning_rate": 7.208805197908372e-05,
+      "loss": 0.1919,
+      "step": 13870
+    },
+    {
+      "epoch": 292.2105263157895,
+      "grad_norm": 0.5291090607643127,
+      "learning_rate": 7.195970528275213e-05,
+      "loss": 0.1875,
+      "step": 13880
+    },
+    {
+      "epoch": 292.42105263157896,
+      "grad_norm": 0.5044816732406616,
+      "learning_rate": 7.18314086989737e-05,
+      "loss": 0.1867,
+      "step": 13890
+    },
+    {
+      "epoch": 292.63157894736844,
+      "grad_norm": 0.6854403018951416,
+      "learning_rate": 7.170316245703528e-05,
+      "loss": 0.1819,
+      "step": 13900
+    },
+    {
+      "epoch": 292.8421052631579,
+      "grad_norm": 0.4865442216396332,
+      "learning_rate": 7.157496678613367e-05,
+      "loss": 0.1896,
+      "step": 13910
+    },
+    {
+      "epoch": 293.05263157894734,
+      "grad_norm": 0.6360872387886047,
+      "learning_rate": 7.144682191537527e-05,
+      "loss": 0.1981,
+      "step": 13920
+    },
+    {
+      "epoch": 293.2631578947368,
+      "grad_norm": 0.45435404777526855,
+      "learning_rate": 7.131872807377581e-05,
+      "loss": 0.1923,
+      "step": 13930
+    },
+    {
+      "epoch": 293.4736842105263,
+      "grad_norm": 0.6130965352058411,
+      "learning_rate": 7.119068549025976e-05,
+      "loss": 0.1824,
+      "step": 13940
+    },
+    {
+      "epoch": 293.6842105263158,
+      "grad_norm": 0.6408197283744812,
+      "learning_rate": 7.106269439365993e-05,
+      "loss": 0.1987,
+      "step": 13950
+    },
+    {
+      "epoch": 293.89473684210526,
+      "grad_norm": 0.5750312209129333,
+      "learning_rate": 7.093475501271716e-05,
+      "loss": 0.1844,
+      "step": 13960
+    },
+    {
+      "epoch": 294.10526315789474,
+      "grad_norm": 0.4592019021511078,
+      "learning_rate": 7.08068675760799e-05,
+      "loss": 0.1886,
+      "step": 13970
+    },
+    {
+      "epoch": 294.3157894736842,
+      "grad_norm": 0.5203351378440857,
+      "learning_rate": 7.067903231230374e-05,
+      "loss": 0.2025,
+      "step": 13980
+    },
+    {
+      "epoch": 294.5263157894737,
+      "grad_norm": 0.9253612160682678,
+      "learning_rate": 7.055124944985096e-05,
+      "loss": 0.1964,
+      "step": 13990
+    },
+    {
+      "epoch": 294.7368421052632,
+      "grad_norm": 0.5490900874137878,
+      "learning_rate": 7.042351921709037e-05,
+      "loss": 0.1786,
+      "step": 14000
+    },
+    {
+      "epoch": 294.94736842105266,
+      "grad_norm": 0.5033005475997925,
+      "learning_rate": 7.029584184229653e-05,
+      "loss": 0.193,
+      "step": 14010
+    },
+    {
+      "epoch": 295.1578947368421,
+      "grad_norm": 0.5323711037635803,
+      "learning_rate": 7.016821755364957e-05,
+      "loss": 0.1834,
+      "step": 14020
+    },
+    {
+      "epoch": 295.36842105263156,
+      "grad_norm": 0.5181049108505249,
+      "learning_rate": 7.00406465792349e-05,
+      "loss": 0.1997,
+      "step": 14030
+    },
+    {
+      "epoch": 295.57894736842104,
+      "grad_norm": 0.5805120468139648,
+      "learning_rate": 6.991312914704242e-05,
+      "loss": 0.187,
+      "step": 14040
+    },
+    {
+      "epoch": 295.7894736842105,
+      "grad_norm": 0.521169900894165,
+      "learning_rate": 6.978566548496657e-05,
+      "loss": 0.1825,
+      "step": 14050
+    },
+    {
+      "epoch": 296.0,
+      "grad_norm": 0.47887876629829407,
+      "learning_rate": 6.965825582080545e-05,
+      "loss": 0.2017,
+      "step": 14060
+    },
+    {
+      "epoch": 296.2105263157895,
+      "grad_norm": 0.49226275086402893,
+      "learning_rate": 6.953090038226092e-05,
+      "loss": 0.1998,
+      "step": 14070
+    },
+    {
+      "epoch": 296.42105263157896,
+      "grad_norm": 0.5467112064361572,
+      "learning_rate": 6.940359939693772e-05,
+      "loss": 0.1936,
+      "step": 14080
+    },
+    {
+      "epoch": 296.63157894736844,
+      "grad_norm": 0.5136857628822327,
+      "learning_rate": 6.927635309234335e-05,
+      "loss": 0.1757,
+      "step": 14090
+    },
+    {
+      "epoch": 296.8421052631579,
+      "grad_norm": 0.47038963437080383,
+      "learning_rate": 6.916187835818779e-05,
+      "loss": 0.1933,
+      "step": 14100
+    },
+    {
+      "epoch": 297.05263157894734,
+      "grad_norm": 0.5792372822761536,
+      "learning_rate": 6.903473657341111e-05,
+      "loss": 0.1895,
+      "step": 14110
+    },
+    {
+      "epoch": 297.2631578947368,
+      "grad_norm": 0.6727595925331116,
+      "learning_rate": 6.890765012858093e-05,
+      "loss": 0.1902,
+      "step": 14120
+    },
+    {
+      "epoch": 297.4736842105263,
+      "grad_norm": 0.5258743166923523,
+      "learning_rate": 6.878061925082137e-05,
+      "loss": 0.1787,
+      "step": 14130
+    },
+    {
+      "epoch": 297.6842105263158,
+      "grad_norm": 0.5945499539375305,
+      "learning_rate": 6.86536441671572e-05,
+      "loss": 0.2014,
+      "step": 14140
+    },
+    {
+      "epoch": 297.89473684210526,
+      "grad_norm": 0.49428433179855347,
+      "learning_rate": 6.852672510451346e-05,
+      "loss": 0.2045,
+      "step": 14150
+    },
+    {
+      "epoch": 298.10526315789474,
+      "grad_norm": 0.5705098509788513,
+      "learning_rate": 6.839986228971512e-05,
+      "loss": 0.1776,
+      "step": 14160
+    },
+    {
+      "epoch": 298.3157894736842,
+      "grad_norm": 0.5622084140777588,
+      "learning_rate": 6.827305594948658e-05,
+      "loss": 0.1858,
+      "step": 14170
+    },
+    {
+      "epoch": 298.5263157894737,
+      "grad_norm": 0.4996286630630493,
+      "learning_rate": 6.814630631045136e-05,
+      "loss": 0.1949,
+      "step": 14180
+    },
+    {
+      "epoch": 298.7368421052632,
+      "grad_norm": 0.7660753130912781,
+      "learning_rate": 6.801961359913156e-05,
+      "loss": 0.1768,
+      "step": 14190
+    },
+    {
+      "epoch": 298.94736842105266,
+      "grad_norm": 0.9577729105949402,
+      "learning_rate": 6.789297804194766e-05,
+      "loss": 0.2199,
+      "step": 14200
+    },
+    {
+      "epoch": 299.1578947368421,
+      "grad_norm": 0.5109399557113647,
+      "learning_rate": 6.776639986521792e-05,
+      "loss": 0.1809,
+      "step": 14210
+    },
+    {
+      "epoch": 299.36842105263156,
+      "grad_norm": 0.6572682857513428,
+      "learning_rate": 6.7639879295158e-05,
+      "loss": 0.1806,
+      "step": 14220
+    },
+    {
+      "epoch": 299.57894736842104,
+      "grad_norm": 0.45336204767227173,
+      "learning_rate": 6.751341655788077e-05,
+      "loss": 0.1925,
+      "step": 14230
+    },
+    {
+      "epoch": 299.7894736842105,
+      "grad_norm": 0.5275882482528687,
+      "learning_rate": 6.73870118793956e-05,
+      "loss": 0.1968,
+      "step": 14240
+    },
+    {
+      "epoch": 300.0,
+      "grad_norm": 0.628964364528656,
+      "learning_rate": 6.726066548560817e-05,
+      "loss": 0.1994,
+      "step": 14250
+    },
+    {
+      "epoch": 300.2105263157895,
+      "grad_norm": 0.6299924254417419,
+      "learning_rate": 6.71343776023199e-05,
+      "loss": 0.1782,
+      "step": 14260
+    },
+    {
+      "epoch": 300.42105263157896,
+      "grad_norm": 0.698186457157135,
+      "learning_rate": 6.700814845522779e-05,
+      "loss": 0.1941,
+      "step": 14270
+    },
+    {
+      "epoch": 300.63157894736844,
+      "grad_norm": 1.0937559604644775,
+      "learning_rate": 6.688197826992375e-05,
+      "loss": 0.2051,
+      "step": 14280
+    },
+    {
+      "epoch": 300.8421052631579,
+      "grad_norm": 0.5907070636749268,
+      "learning_rate": 6.675586727189436e-05,
+      "loss": 0.1948,
+      "step": 14290
+    },
+    {
+      "epoch": 301.05263157894734,
+      "grad_norm": 0.44216281175613403,
+      "learning_rate": 6.662981568652049e-05,
+      "loss": 0.1878,
+      "step": 14300
+    },
+    {
+      "epoch": 301.2631578947368,
+      "grad_norm": 0.6885428428649902,
+      "learning_rate": 6.650382373907672e-05,
+      "loss": 0.1857,
+      "step": 14310
+    },
+    {
+      "epoch": 301.4736842105263,
+      "grad_norm": 0.5614356398582458,
+      "learning_rate": 6.637789165473101e-05,
+      "loss": 0.1982,
+      "step": 14320
+    },
+    {
+      "epoch": 301.6842105263158,
+      "grad_norm": 0.5463254451751709,
+      "learning_rate": 6.625201965854453e-05,
+      "loss": 0.1884,
+      "step": 14330
+    },
+    {
+      "epoch": 301.89473684210526,
+      "grad_norm": 0.6541967988014221,
+      "learning_rate": 6.612620797547087e-05,
+      "loss": 0.1901,
+      "step": 14340
+    },
+    {
+      "epoch": 302.10526315789474,
+      "grad_norm": 0.5899919271469116,
+      "learning_rate": 6.600045683035597e-05,
+      "loss": 0.1976,
+      "step": 14350
+    },
+    {
+      "epoch": 302.3157894736842,
+      "grad_norm": 0.5349833369255066,
+      "learning_rate": 6.587476644793742e-05,
+      "loss": 0.1729,
+      "step": 14360
+    },
+    {
+      "epoch": 302.5263157894737,
+      "grad_norm": 0.8576452732086182,
+      "learning_rate": 6.574913705284443e-05,
+      "loss": 0.2013,
+      "step": 14370
+    },
+    {
+      "epoch": 302.7368421052632,
+      "grad_norm": 0.5418473482131958,
+      "learning_rate": 6.562356886959704e-05,
+      "loss": 0.199,
+      "step": 14380
+    },
+    {
+      "epoch": 302.94736842105266,
+      "grad_norm": 0.6117028594017029,
+      "learning_rate": 6.54980621226059e-05,
+      "loss": 0.1878,
+      "step": 14390
+    },
+    {
+      "epoch": 303.1578947368421,
+      "grad_norm": 0.6464748978614807,
+      "learning_rate": 6.537261703617202e-05,
+      "loss": 0.1913,
+      "step": 14400
+    },
+    {
+      "epoch": 303.36842105263156,
+      "grad_norm": 0.4876056909561157,
+      "learning_rate": 6.524723383448607e-05,
+      "loss": 0.1849,
+      "step": 14410
+    },
+    {
+      "epoch": 303.57894736842104,
+      "grad_norm": 0.5367958545684814,
+      "learning_rate": 6.512191274162816e-05,
+      "loss": 0.1915,
+      "step": 14420
+    },
+    {
+      "epoch": 303.7894736842105,
+      "grad_norm": 0.5246376991271973,
+      "learning_rate": 6.499665398156733e-05,
+      "loss": 0.1854,
+      "step": 14430
+    },
+    {
+      "epoch": 304.0,
+      "grad_norm": 0.49785304069519043,
+      "learning_rate": 6.487145777816143e-05,
+      "loss": 0.1839,
+      "step": 14440
+    },
+    {
+      "epoch": 304.2105263157895,
+      "grad_norm": 0.5250376462936401,
+      "learning_rate": 6.474632435515627e-05,
+      "loss": 0.1878,
+      "step": 14450
+    },
+    {
+      "epoch": 304.42105263157896,
+      "grad_norm": 0.48043861985206604,
+      "learning_rate": 6.462125393618561e-05,
+      "loss": 0.1883,
+      "step": 14460
+    },
+    {
+      "epoch": 304.63157894736844,
+      "grad_norm": 0.47180649638175964,
+      "learning_rate": 6.449624674477054e-05,
+      "loss": 0.196,
+      "step": 14470
+    },
+    {
+      "epoch": 304.8421052631579,
+      "grad_norm": 0.5821836590766907,
+      "learning_rate": 6.437130300431924e-05,
+      "loss": 0.1811,
+      "step": 14480
+    },
+    {
+      "epoch": 305.05263157894734,
+      "grad_norm": 0.49136173725128174,
+      "learning_rate": 6.424642293812636e-05,
+      "loss": 0.193,
+      "step": 14490
+    },
+    {
+      "epoch": 305.2631578947368,
+      "grad_norm": 0.45949774980545044,
+      "learning_rate": 6.412160676937288e-05,
+      "loss": 0.1872,
+      "step": 14500
+    },
+    {
+      "epoch": 305.4736842105263,
+      "grad_norm": 0.7283862829208374,
+      "learning_rate": 6.399685472112552e-05,
+      "loss": 0.1949,
+      "step": 14510
+    },
+    {
+      "epoch": 305.6842105263158,
+      "grad_norm": 0.49841785430908203,
+      "learning_rate": 6.387216701633638e-05,
+      "loss": 0.1821,
+      "step": 14520
+    },
+    {
+      "epoch": 305.89473684210526,
+      "grad_norm": 0.47515934705734253,
+      "learning_rate": 6.374754387784262e-05,
+      "loss": 0.196,
+      "step": 14530
+    },
+    {
+      "epoch": 306.10526315789474,
+      "grad_norm": 0.42606112360954285,
+      "learning_rate": 6.362298552836605e-05,
+      "loss": 0.1819,
+      "step": 14540
+    },
+    {
+      "epoch": 306.3157894736842,
+      "grad_norm": 0.5149482488632202,
+      "learning_rate": 6.34984921905126e-05,
+      "loss": 0.1825,
+      "step": 14550
+    },
+    {
+      "epoch": 306.5263157894737,
+      "grad_norm": 0.5274513363838196,
+      "learning_rate": 6.3374064086772e-05,
+      "loss": 0.1868,
+      "step": 14560
+    },
+    {
+      "epoch": 306.7368421052632,
+      "grad_norm": 0.5605398416519165,
+      "learning_rate": 6.324970143951753e-05,
+      "loss": 0.1859,
+      "step": 14570
+    },
+    {
+      "epoch": 306.94736842105266,
+      "grad_norm": 0.6485209465026855,
+      "learning_rate": 6.312540447100534e-05,
+      "loss": 0.1988,
+      "step": 14580
+    },
+    {
+      "epoch": 307.1578947368421,
+      "grad_norm": 0.4830923080444336,
+      "learning_rate": 6.300117340337433e-05,
+      "loss": 0.1932,
+      "step": 14590
+    },
+    {
+      "epoch": 307.36842105263156,
+      "grad_norm": 0.6907581090927124,
+      "learning_rate": 6.287700845864549e-05,
+      "loss": 0.1904,
+      "step": 14600
+    },
+    {
+      "epoch": 307.57894736842104,
+      "grad_norm": 0.5328361988067627,
+      "learning_rate": 6.275290985872177e-05,
+      "loss": 0.1906,
+      "step": 14610
+    },
+    {
+      "epoch": 307.7894736842105,
+      "grad_norm": 0.5460513234138489,
+      "learning_rate": 6.262887782538746e-05,
+      "loss": 0.1901,
+      "step": 14620
+    },
+    {
+      "epoch": 308.0,
+      "grad_norm": 0.5381531119346619,
+      "learning_rate": 6.250491258030791e-05,
+      "loss": 0.1927,
+      "step": 14630
+    },
+    {
+      "epoch": 308.2105263157895,
+      "grad_norm": 0.6306193470954895,
+      "learning_rate": 6.23810143450291e-05,
+      "loss": 0.2011,
+      "step": 14640
+    },
+    {
+      "epoch": 308.42105263157896,
+      "grad_norm": 0.5382252931594849,
+      "learning_rate": 6.225718334097733e-05,
+      "loss": 0.1901,
+      "step": 14650
+    },
+    {
+      "epoch": 308.63157894736844,
+      "grad_norm": 0.5241566300392151,
+      "learning_rate": 6.213341978945859e-05,
+      "loss": 0.171,
+      "step": 14660
+    },
+    {
+      "epoch": 308.8421052631579,
+      "grad_norm": 0.8374577760696411,
+      "learning_rate": 6.200972391165852e-05,
+      "loss": 0.2042,
+      "step": 14670
+    },
+    {
+      "epoch": 309.05263157894734,
+      "grad_norm": 0.516899824142456,
+      "learning_rate": 6.188609592864163e-05,
+      "loss": 0.1779,
+      "step": 14680
+    },
+    {
+      "epoch": 309.2631578947368,
+      "grad_norm": 0.5281073451042175,
+      "learning_rate": 6.176253606135119e-05,
+      "loss": 0.1813,
+      "step": 14690
+    },
+    {
+      "epoch": 309.4736842105263,
+      "grad_norm": 0.5887854099273682,
+      "learning_rate": 6.163904453060869e-05,
+      "loss": 0.1897,
+      "step": 14700
+    },
+    {
+      "epoch": 309.6842105263158,
+      "grad_norm": 0.5451268553733826,
+      "learning_rate": 6.15156215571136e-05,
+      "loss": 0.1968,
+      "step": 14710
+    },
+    {
+      "epoch": 309.89473684210526,
+      "grad_norm": 0.671812891960144,
+      "learning_rate": 6.139226736144273e-05,
+      "loss": 0.2004,
+      "step": 14720
+    },
+    {
+      "epoch": 310.10526315789474,
+      "grad_norm": 0.5144766569137573,
+      "learning_rate": 6.126898216405e-05,
+      "loss": 0.1946,
+      "step": 14730
+    },
+    {
+      "epoch": 310.3157894736842,
+      "grad_norm": 0.5516157150268555,
+      "learning_rate": 6.114576618526611e-05,
+      "loss": 0.1764,
+      "step": 14740
+    },
+    {
+      "epoch": 310.5263157894737,
+      "grad_norm": 0.5606300234794617,
+      "learning_rate": 6.102261964529796e-05,
+      "loss": 0.1935,
+      "step": 14750
+    },
+    {
+      "epoch": 310.7368421052632,
+      "grad_norm": 0.646920382976532,
+      "learning_rate": 6.08995427642284e-05,
+      "loss": 0.1944,
+      "step": 14760
+    },
+    {
+      "epoch": 310.94736842105266,
+      "grad_norm": 0.529812753200531,
+      "learning_rate": 6.077653576201572e-05,
+      "loss": 0.1858,
+      "step": 14770
+    },
+    {
+      "epoch": 311.1578947368421,
+      "grad_norm": 0.42989394068717957,
+      "learning_rate": 6.065359885849345e-05,
+      "loss": 0.1761,
+      "step": 14780
+    },
+    {
+      "epoch": 311.36842105263156,
+      "grad_norm": 0.5556666851043701,
+      "learning_rate": 6.053073227336975e-05,
+      "loss": 0.1884,
+      "step": 14790
+    },
+    {
+      "epoch": 311.57894736842104,
+      "grad_norm": 0.48674774169921875,
+      "learning_rate": 6.040793622622707e-05,
+      "loss": 0.2071,
+      "step": 14800
+    },
+    {
+      "epoch": 311.7894736842105,
+      "grad_norm": 0.48857688903808594,
+      "learning_rate": 6.0285210936521955e-05,
+      "loss": 0.1836,
+      "step": 14810
+    },
+    {
+      "epoch": 312.0,
+      "grad_norm": 0.737316906452179,
+      "learning_rate": 6.016255662358432e-05,
+      "loss": 0.1895,
+      "step": 14820
+    },
+    {
+      "epoch": 312.2105263157895,
+      "grad_norm": 0.49834001064300537,
+      "learning_rate": 6.003997350661732e-05,
+      "loss": 0.1996,
+      "step": 14830
+    },
+    {
+      "epoch": 312.42105263157896,
+      "grad_norm": 0.5362703204154968,
+      "learning_rate": 5.991746180469691e-05,
+      "loss": 0.1813,
+      "step": 14840
+    },
+    {
+      "epoch": 312.63157894736844,
+      "grad_norm": 0.6358118057250977,
+      "learning_rate": 5.979502173677134e-05,
+      "loss": 0.1995,
+      "step": 14850
+    },
+    {
+      "epoch": 312.8421052631579,
+      "grad_norm": 0.5340322256088257,
+      "learning_rate": 5.9672653521660826e-05,
+      "loss": 0.1778,
+      "step": 14860
+    },
+    {
+      "epoch": 313.05263157894734,
+      "grad_norm": 0.44907963275909424,
+      "learning_rate": 5.955035737805725e-05,
+      "loss": 0.1904,
+      "step": 14870
+    },
+    {
+      "epoch": 313.2631578947368,
+      "grad_norm": 0.5180246233940125,
+      "learning_rate": 5.9428133524523646e-05,
+      "loss": 0.1911,
+      "step": 14880
+    },
+    {
+      "epoch": 313.4736842105263,
+      "grad_norm": 0.525974452495575,
+      "learning_rate": 5.930598217949386e-05,
+      "loss": 0.1759,
+      "step": 14890
+    },
+    {
+      "epoch": 313.6842105263158,
+      "grad_norm": 0.5962675213813782,
+      "learning_rate": 5.91839035612721e-05,
+      "loss": 0.1956,
+      "step": 14900
+    },
+    {
+      "epoch": 313.89473684210526,
+      "grad_norm": 0.5130165815353394,
+      "learning_rate": 5.9061897888032747e-05,
+      "loss": 0.1916,
+      "step": 14910
+    },
+    {
+      "epoch": 314.10526315789474,
+      "grad_norm": 0.5901978611946106,
+      "learning_rate": 5.893996537781966e-05,
+      "loss": 0.1812,
+      "step": 14920
+    },
+    {
+      "epoch": 314.3157894736842,
+      "grad_norm": 0.48012682795524597,
+      "learning_rate": 5.8818106248546004e-05,
+      "loss": 0.1936,
+      "step": 14930
+    },
+    {
+      "epoch": 314.5263157894737,
+      "grad_norm": 0.5353314876556396,
+      "learning_rate": 5.8696320717993784e-05,
+      "loss": 0.1792,
+      "step": 14940
+    },
+    {
+      "epoch": 314.7368421052632,
+      "grad_norm": 0.6114587187767029,
+      "learning_rate": 5.857460900381355e-05,
+      "loss": 0.1873,
+      "step": 14950
+    },
+    {
+      "epoch": 314.94736842105266,
+      "grad_norm": 0.654788076877594,
+      "learning_rate": 5.845297132352385e-05,
+      "loss": 0.201,
+      "step": 14960
+    },
+    {
+      "epoch": 315.1578947368421,
+      "grad_norm": 0.6699498295783997,
+      "learning_rate": 5.833140789451086e-05,
+      "loss": 0.1781,
+      "step": 14970
+    },
+    {
+      "epoch": 315.36842105263156,
+      "grad_norm": 0.573408305644989,
+      "learning_rate": 5.8209918934028275e-05,
+      "loss": 0.1975,
+      "step": 14980
+    },
+    {
+      "epoch": 315.57894736842104,
+      "grad_norm": 0.6000211238861084,
+      "learning_rate": 5.808850465919649e-05,
+      "loss": 0.1784,
+      "step": 14990
+    },
+    {
+      "epoch": 315.7894736842105,
+      "grad_norm": 0.6924741864204407,
+      "learning_rate": 5.7967165287002464e-05,
+      "loss": 0.1823,
+      "step": 15000
+    },
+    {
+      "epoch": 316.0,
+      "grad_norm": 0.8868710398674011,
+      "learning_rate": 5.7845901034299424e-05,
+      "loss": 0.2024,
+      "step": 15010
+    },
+    {
+      "epoch": 316.2105263157895,
+      "grad_norm": 0.49164116382598877,
+      "learning_rate": 5.772471211780619e-05,
+      "loss": 0.1755,
+      "step": 15020
+    },
+    {
+      "epoch": 316.42105263157896,
+      "grad_norm": 0.6795914173126221,
+      "learning_rate": 5.760359875410702e-05,
+      "loss": 0.1937,
+      "step": 15030
+    },
+    {
+      "epoch": 316.63157894736844,
+      "grad_norm": 0.6305373311042786,
+      "learning_rate": 5.748256115965109e-05,
+      "loss": 0.1841,
+      "step": 15040
+    },
+    {
+      "epoch": 316.8421052631579,
+      "grad_norm": 0.6585614085197449,
+      "learning_rate": 5.73615995507523e-05,
+      "loss": 0.2104,
+      "step": 15050
+    },
+    {
+      "epoch": 317.05263157894734,
+      "grad_norm": 0.6081206798553467,
+      "learning_rate": 5.724071414358858e-05,
+      "loss": 0.1857,
+      "step": 15060
+    },
+    {
+      "epoch": 317.2631578947368,
+      "grad_norm": 0.4927370846271515,
+      "learning_rate": 5.711990515420176e-05,
+      "loss": 0.1822,
+      "step": 15070
+    },
+    {
+      "epoch": 317.4736842105263,
+      "grad_norm": 0.5095883011817932,
+      "learning_rate": 5.699917279849714e-05,
+      "loss": 0.18,
+      "step": 15080
+    },
+    {
+      "epoch": 317.6842105263158,
+      "grad_norm": 0.7175976037979126,
+      "learning_rate": 5.6878517292242936e-05,
+      "loss": 0.1883,
+      "step": 15090
+    },
+    {
+      "epoch": 317.89473684210526,
+      "grad_norm": 0.5586292147636414,
+      "learning_rate": 5.675793885107019e-05,
+      "loss": 0.1906,
+      "step": 15100
+    },
+    {
+      "epoch": 318.10526315789474,
+      "grad_norm": 0.8061725497245789,
+      "learning_rate": 5.663743769047206e-05,
+      "loss": 0.1899,
+      "step": 15110
+    },
+    {
+      "epoch": 318.3157894736842,
+      "grad_norm": 0.6274729371070862,
+      "learning_rate": 5.651701402580371e-05,
+      "loss": 0.1747,
+      "step": 15120
+    },
+    {
+      "epoch": 318.5263157894737,
+      "grad_norm": 0.4897122383117676,
+      "learning_rate": 5.639666807228175e-05,
+      "loss": 0.1936,
+      "step": 15130
+    },
+    {
+      "epoch": 318.7368421052632,
+      "grad_norm": 0.5890952348709106,
+      "learning_rate": 5.627640004498385e-05,
+      "loss": 0.1928,
+      "step": 15140
+    },
+    {
+      "epoch": 318.94736842105266,
+      "grad_norm": 0.5626781582832336,
+      "learning_rate": 5.6156210158848544e-05,
+      "loss": 0.1967,
+      "step": 15150
+    },
+    {
+      "epoch": 319.1578947368421,
+      "grad_norm": 0.5838659405708313,
+      "learning_rate": 5.603609862867463e-05,
+      "loss": 0.1872,
+      "step": 15160
+    },
+    {
+      "epoch": 319.36842105263156,
+      "grad_norm": 0.4655553698539734,
+      "learning_rate": 5.591606566912082e-05,
+      "loss": 0.1894,
+      "step": 15170
+    },
+    {
+      "epoch": 319.57894736842104,
+      "grad_norm": 0.5683347582817078,
+      "learning_rate": 5.5796111494705584e-05,
+      "loss": 0.1836,
+      "step": 15180
+    },
+    {
+      "epoch": 319.7894736842105,
+      "grad_norm": 0.5712106823921204,
+      "learning_rate": 5.567623631980644e-05,
+      "loss": 0.1925,
+      "step": 15190
+    },
+    {
+      "epoch": 320.0,
+      "grad_norm": 0.5311657190322876,
+      "learning_rate": 5.55564403586597e-05,
+      "loss": 0.1979,
+      "step": 15200
+    },
+    {
+      "epoch": 320.2105263157895,
+      "grad_norm": 0.4528977572917938,
+      "learning_rate": 5.543672382536023e-05,
+      "loss": 0.1969,
+      "step": 15210
+    },
+    {
+      "epoch": 320.42105263157896,
+      "grad_norm": 0.5928493738174438,
+      "learning_rate": 5.5317086933860907e-05,
+      "loss": 0.1889,
+      "step": 15220
+    },
+    {
+      "epoch": 320.63157894736844,
+      "grad_norm": 0.4887818396091461,
+      "learning_rate": 5.519752989797224e-05,
+      "loss": 0.1806,
+      "step": 15230
+    },
+    {
+      "epoch": 320.8421052631579,
+      "grad_norm": 0.5781086683273315,
+      "learning_rate": 5.507805293136198e-05,
+      "loss": 0.1813,
+      "step": 15240
+    },
+    {
+      "epoch": 321.05263157894734,
+      "grad_norm": 0.44095057249069214,
+      "learning_rate": 5.495865624755492e-05,
+      "loss": 0.1859,
+      "step": 15250
+    },
+    {
+      "epoch": 321.2631578947368,
+      "grad_norm": 0.6125297546386719,
+      "learning_rate": 5.4839340059932255e-05,
+      "loss": 0.1851,
+      "step": 15260
+    },
+    {
+      "epoch": 321.4736842105263,
+      "grad_norm": 0.7112295627593994,
+      "learning_rate": 5.472010458173132e-05,
+      "loss": 0.1968,
+      "step": 15270
+    },
+    {
+      "epoch": 321.6842105263158,
+      "grad_norm": 0.5584365129470825,
+      "learning_rate": 5.4600950026045326e-05,
+      "loss": 0.1845,
+      "step": 15280
+    },
+    {
+      "epoch": 321.89473684210526,
+      "grad_norm": 0.5321446061134338,
+      "learning_rate": 5.448187660582276e-05,
+      "loss": 0.1951,
+      "step": 15290
+    },
+    {
+      "epoch": 322.10526315789474,
+      "grad_norm": 0.5082312226295471,
+      "learning_rate": 5.436288453386709e-05,
+      "loss": 0.183,
+      "step": 15300
+    },
+    {
+      "epoch": 322.3157894736842,
+      "grad_norm": 0.47308409214019775,
+      "learning_rate": 5.424397402283644e-05,
+      "loss": 0.1886,
+      "step": 15310
+    },
+    {
+      "epoch": 322.5263157894737,
+      "grad_norm": 0.6271807551383972,
+      "learning_rate": 5.4125145285243194e-05,
+      "loss": 0.1929,
+      "step": 15320
+    },
+    {
+      "epoch": 322.7368421052632,
+      "grad_norm": 0.677312433719635,
+      "learning_rate": 5.400639853345364e-05,
+      "loss": 0.1929,
+      "step": 15330
+    },
+    {
+      "epoch": 322.94736842105266,
+      "grad_norm": 0.3949933648109436,
+      "learning_rate": 5.388773397968736e-05,
+      "loss": 0.1827,
+      "step": 15340
+    },
+    {
+      "epoch": 323.1578947368421,
+      "grad_norm": 0.5227069854736328,
+      "learning_rate": 5.376915183601725e-05,
+      "loss": 0.1821,
+      "step": 15350
+    },
+    {
+      "epoch": 323.36842105263156,
+      "grad_norm": 0.4687119722366333,
+      "learning_rate": 5.36506523143688e-05,
+      "loss": 0.1883,
+      "step": 15360
+    },
+    {
+      "epoch": 323.57894736842104,
+      "grad_norm": 0.6543534398078918,
+      "learning_rate": 5.353223562651986e-05,
+      "loss": 0.1898,
+      "step": 15370
+    },
+    {
+      "epoch": 323.7894736842105,
+      "grad_norm": 0.6527659893035889,
+      "learning_rate": 5.341390198410019e-05,
+      "loss": 0.1915,
+      "step": 15380
+    },
+    {
+      "epoch": 324.0,
+      "grad_norm": 0.5707231760025024,
+      "learning_rate": 5.329565159859131e-05,
+      "loss": 0.1917,
+      "step": 15390
+    },
+    {
+      "epoch": 324.2105263157895,
+      "grad_norm": 0.6318780183792114,
+      "learning_rate": 5.317748468132577e-05,
+      "loss": 0.2033,
+      "step": 15400
+    },
+    {
+      "epoch": 324.42105263157896,
+      "grad_norm": 0.5352064967155457,
+      "learning_rate": 5.305940144348698e-05,
+      "loss": 0.1865,
+      "step": 15410
+    },
+    {
+      "epoch": 324.63157894736844,
+      "grad_norm": 0.4935546815395355,
+      "learning_rate": 5.2941402096108905e-05,
+      "loss": 0.178,
+      "step": 15420
+    },
+    {
+      "epoch": 324.8421052631579,
+      "grad_norm": 0.5082517266273499,
+      "learning_rate": 5.282348685007543e-05,
+      "loss": 0.1892,
+      "step": 15430
+    },
+    {
+      "epoch": 325.05263157894734,
+      "grad_norm": 0.5046442747116089,
+      "learning_rate": 5.2705655916120325e-05,
+      "loss": 0.1926,
+      "step": 15440
+    },
+    {
+      "epoch": 325.2631578947368,
+      "grad_norm": 0.43995267152786255,
+      "learning_rate": 5.258790950482646e-05,
+      "loss": 0.1827,
+      "step": 15450
+    },
+    {
+      "epoch": 325.4736842105263,
+      "grad_norm": 0.5442992448806763,
+      "learning_rate": 5.247024782662586e-05,
+      "loss": 0.1884,
+      "step": 15460
+    },
+    {
+      "epoch": 325.6842105263158,
+      "grad_norm": 0.7053566575050354,
+      "learning_rate": 5.2352671091798997e-05,
+      "loss": 0.1953,
+      "step": 15470
+    },
+    {
+      "epoch": 325.89473684210526,
+      "grad_norm": 0.5656514167785645,
+      "learning_rate": 5.223517951047449e-05,
+      "loss": 0.1848,
+      "step": 15480
+    },
+    {
+      "epoch": 326.10526315789474,
+      "grad_norm": 0.5459962487220764,
+      "learning_rate": 5.2117773292628935e-05,
+      "loss": 0.2066,
+      "step": 15490
+    },
+    {
+      "epoch": 326.3157894736842,
+      "grad_norm": 0.5724794268608093,
+      "learning_rate": 5.200045264808624e-05,
+      "loss": 0.1913,
+      "step": 15500
+    },
+    {
+      "epoch": 326.5263157894737,
+      "grad_norm": 0.5558455586433411,
+      "learning_rate": 5.188321778651739e-05,
+      "loss": 0.185,
+      "step": 15510
+    },
+    {
+      "epoch": 326.7368421052632,
+      "grad_norm": 0.6177617311477661,
+      "learning_rate": 5.176606891744017e-05,
+      "loss": 0.1826,
+      "step": 15520
+    },
+    {
+      "epoch": 326.94736842105266,
+      "grad_norm": 0.5286545157432556,
+      "learning_rate": 5.164900625021856e-05,
+      "loss": 0.1786,
+      "step": 15530
+    },
+    {
+      "epoch": 327.1578947368421,
+      "grad_norm": 0.6205462217330933,
+      "learning_rate": 5.153202999406251e-05,
+      "loss": 0.1972,
+      "step": 15540
+    },
+    {
+      "epoch": 327.36842105263156,
+      "grad_norm": 0.5253833532333374,
+      "learning_rate": 5.141514035802755e-05,
+      "loss": 0.1931,
+      "step": 15550
+    },
+    {
+      "epoch": 327.57894736842104,
+      "grad_norm": 0.5076433420181274,
+      "learning_rate": 5.129833755101442e-05,
+      "loss": 0.1911,
+      "step": 15560
+    },
+    {
+      "epoch": 327.7894736842105,
+      "grad_norm": 0.466351717710495,
+      "learning_rate": 5.118162178176873e-05,
+      "loss": 0.1876,
+      "step": 15570
+    },
+    {
+      "epoch": 328.0,
+      "grad_norm": 0.636641800403595,
+      "learning_rate": 5.106499325888041e-05,
+      "loss": 0.1804,
+      "step": 15580
+    },
+    {
+      "epoch": 328.2105263157895,
+      "grad_norm": 0.6042010188102722,
+      "learning_rate": 5.094845219078361e-05,
+      "loss": 0.1795,
+      "step": 15590
+    },
+    {
+      "epoch": 328.42105263157896,
+      "grad_norm": 0.8110350966453552,
+      "learning_rate": 5.083199878575609e-05,
+      "loss": 0.196,
+      "step": 15600
+    },
+    {
+      "epoch": 328.63157894736844,
+      "grad_norm": 0.5079600811004639,
+      "learning_rate": 5.071563325191889e-05,
+      "loss": 0.1775,
+      "step": 15610
+    },
+    {
+      "epoch": 328.8421052631579,
+      "grad_norm": 0.6384349465370178,
+      "learning_rate": 5.0599355797236205e-05,
+      "loss": 0.2021,
+      "step": 15620
+    },
+    {
+      "epoch": 329.05263157894734,
+      "grad_norm": 0.46881303191185,
+      "learning_rate": 5.0483166629514654e-05,
+      "loss": 0.1925,
+      "step": 15630
+    },
+    {
+      "epoch": 329.2631578947368,
+      "grad_norm": 0.7932872176170349,
+      "learning_rate": 5.03670659564031e-05,
+      "loss": 0.1959,
+      "step": 15640
+    },
+    {
+      "epoch": 329.4736842105263,
+      "grad_norm": 0.5118970274925232,
+      "learning_rate": 5.025105398539227e-05,
+      "loss": 0.1719,
+      "step": 15650
+    },
+    {
+      "epoch": 329.6842105263158,
+      "grad_norm": 0.5387006402015686,
+      "learning_rate": 5.0135130923814386e-05,
+      "loss": 0.2077,
+      "step": 15660
+    },
+    {
+      "epoch": 329.89473684210526,
+      "grad_norm": 0.6646759510040283,
+      "learning_rate": 5.001929697884273e-05,
+      "loss": 0.1943,
+      "step": 15670
+    },
+    {
+      "epoch": 330.10526315789474,
+      "grad_norm": 0.7001065611839294,
+      "learning_rate": 4.9903552357491404e-05,
+      "loss": 0.1868,
+      "step": 15680
+    },
+    {
+      "epoch": 330.3157894736842,
+      "grad_norm": 0.5651974081993103,
+      "learning_rate": 4.978789726661472e-05,
+      "loss": 0.1791,
+      "step": 15690
+    },
+    {
+      "epoch": 330.5263157894737,
+      "grad_norm": 0.6005178093910217,
+      "learning_rate": 4.9672331912907174e-05,
+      "loss": 0.1871,
+      "step": 15700
+    },
+    {
+      "epoch": 330.7368421052632,
+      "grad_norm": 0.6612521409988403,
+      "learning_rate": 4.9556856502902745e-05,
+      "loss": 0.1931,
+      "step": 15710
+    },
+    {
+      "epoch": 330.94736842105266,
+      "grad_norm": 0.5340935587882996,
+      "learning_rate": 4.944147124297468e-05,
+      "loss": 0.19,
+      "step": 15720
+    },
+    {
+      "epoch": 331.1578947368421,
+      "grad_norm": 0.6876013278961182,
+      "learning_rate": 4.9326176339335225e-05,
+      "loss": 0.1803,
+      "step": 15730
+    },
+    {
+      "epoch": 331.36842105263156,
+      "grad_norm": 0.7157713770866394,
+      "learning_rate": 4.921097199803503e-05,
+      "loss": 0.1884,
+      "step": 15740
+    },
+    {
+      "epoch": 331.57894736842104,
+      "grad_norm": 0.5469095706939697,
+      "learning_rate": 4.909585842496287e-05,
+      "loss": 0.1763,
+      "step": 15750
+    },
+    {
+      "epoch": 331.7894736842105,
+      "grad_norm": 0.5198671817779541,
+      "learning_rate": 4.8980835825845475e-05,
+      "loss": 0.1859,
+      "step": 15760
+    },
+    {
+      "epoch": 332.0,
+      "grad_norm": 0.6175537705421448,
+      "learning_rate": 4.886590440624682e-05,
+      "loss": 0.2021,
+      "step": 15770
+    },
+    {
+      "epoch": 332.2105263157895,
+      "grad_norm": 0.7941794991493225,
+      "learning_rate": 4.875106437156795e-05,
+      "loss": 0.1942,
+      "step": 15780
+    },
+    {
+      "epoch": 332.42105263157896,
+      "grad_norm": 0.5476126074790955,
+      "learning_rate": 4.863631592704673e-05,
+      "loss": 0.181,
+      "step": 15790
+    },
+    {
+      "epoch": 332.63157894736844,
+      "grad_norm": 0.6549970507621765,
+      "learning_rate": 4.852165927775713e-05,
+      "loss": 0.1829,
+      "step": 15800
+    },
+    {
+      "epoch": 332.8421052631579,
+      "grad_norm": 0.48638707399368286,
+      "learning_rate": 4.840709462860925e-05,
+      "loss": 0.1839,
+      "step": 15810
+    },
+    {
+      "epoch": 333.05263157894734,
+      "grad_norm": 0.42391976714134216,
+      "learning_rate": 4.8292622184348636e-05,
+      "loss": 0.1905,
+      "step": 15820
+    },
+    {
+      "epoch": 333.2631578947368,
+      "grad_norm": 0.5468048453330994,
+      "learning_rate": 4.8178242149556176e-05,
+      "loss": 0.1918,
+      "step": 15830
+    },
+    {
+      "epoch": 333.4736842105263,
+      "grad_norm": 0.8398631811141968,
+      "learning_rate": 4.806395472864749e-05,
+      "loss": 0.179,
+      "step": 15840
+    },
+    {
+      "epoch": 333.6842105263158,
+      "grad_norm": 0.6367696523666382,
+      "learning_rate": 4.79497601258727e-05,
+      "loss": 0.1901,
+      "step": 15850
+    },
+    {
+      "epoch": 333.89473684210526,
+      "grad_norm": 0.5233663320541382,
+      "learning_rate": 4.783565854531615e-05,
+      "loss": 0.1903,
+      "step": 15860
+    },
+    {
+      "epoch": 334.10526315789474,
+      "grad_norm": 0.5648738741874695,
+      "learning_rate": 4.7721650190895826e-05,
+      "loss": 0.1979,
+      "step": 15870
+    },
+    {
+      "epoch": 334.3157894736842,
+      "grad_norm": 0.5198102593421936,
+      "learning_rate": 4.760773526636315e-05,
+      "loss": 0.1795,
+      "step": 15880
+    },
+    {
+      "epoch": 334.5263157894737,
+      "grad_norm": 0.6149677038192749,
+      "learning_rate": 4.7493913975302526e-05,
+      "loss": 0.1966,
+      "step": 15890
+    },
+    {
+      "epoch": 334.7368421052632,
+      "grad_norm": 0.49036896228790283,
+      "learning_rate": 4.738018652113113e-05,
+      "loss": 0.1873,
+      "step": 15900
+    },
+    {
+      "epoch": 334.94736842105266,
+      "grad_norm": 0.5529518127441406,
+      "learning_rate": 4.7266553107098274e-05,
+      "loss": 0.192,
+      "step": 15910
+    },
+    {
+      "epoch": 335.1578947368421,
+      "grad_norm": 0.6180853843688965,
+      "learning_rate": 4.715301393628534e-05,
+      "loss": 0.185,
+      "step": 15920
+    },
+    {
+      "epoch": 335.36842105263156,
+      "grad_norm": 0.5150652527809143,
+      "learning_rate": 4.703956921160528e-05,
+      "loss": 0.1899,
+      "step": 15930
+    },
+    {
+      "epoch": 335.57894736842104,
+      "grad_norm": 0.7054027318954468,
+      "learning_rate": 4.6926219135802173e-05,
+      "loss": 0.1837,
+      "step": 15940
+    },
+    {
+      "epoch": 335.7894736842105,
+      "grad_norm": 0.5216421484947205,
+      "learning_rate": 4.6812963911450934e-05,
+      "loss": 0.1855,
+      "step": 15950
+    },
+    {
+      "epoch": 336.0,
+      "grad_norm": 0.586047887802124,
+      "learning_rate": 4.669980374095709e-05,
+      "loss": 0.1972,
+      "step": 15960
+    },
+    {
+      "epoch": 336.2105263157895,
+      "grad_norm": 0.6370906233787537,
+      "learning_rate": 4.65867388265562e-05,
+      "loss": 0.1911,
+      "step": 15970
+    },
+    {
+      "epoch": 336.42105263157896,
+      "grad_norm": 0.5958567261695862,
+      "learning_rate": 4.647376937031356e-05,
+      "loss": 0.1889,
+      "step": 15980
+    },
+    {
+      "epoch": 336.63157894736844,
+      "grad_norm": 0.45742881298065186,
+      "learning_rate": 4.63608955741239e-05,
+      "loss": 0.1784,
+      "step": 15990
+    },
+    {
+      "epoch": 336.8421052631579,
+      "grad_norm": 0.7190865278244019,
+      "learning_rate": 4.6248117639711044e-05,
+      "loss": 0.1832,
+      "step": 16000
+    },
+    {
+      "epoch": 337.05263157894734,
+      "grad_norm": 0.4909994900226593,
+      "learning_rate": 4.613543576862743e-05,
+      "loss": 0.2005,
+      "step": 16010
+    },
+    {
+      "epoch": 337.2631578947368,
+      "grad_norm": 0.5645287036895752,
+      "learning_rate": 4.6022850162253795e-05,
+      "loss": 0.1745,
+      "step": 16020
+    },
+    {
+      "epoch": 337.4736842105263,
+      "grad_norm": 0.5330631136894226,
+      "learning_rate": 4.591036102179893e-05,
+      "loss": 0.1906,
+      "step": 16030
+    },
+    {
+      "epoch": 337.6842105263158,
+      "grad_norm": 0.6603984832763672,
+      "learning_rate": 4.579796854829911e-05,
+      "loss": 0.2005,
+      "step": 16040
+    },
+    {
+      "epoch": 337.89473684210526,
+      "grad_norm": 0.6532229781150818,
+      "learning_rate": 4.568567294261797e-05,
+      "loss": 0.1868,
+      "step": 16050
+    },
+    {
+      "epoch": 338.10526315789474,
+      "grad_norm": 0.6483712792396545,
+      "learning_rate": 4.55734744054459e-05,
+      "loss": 0.1927,
+      "step": 16060
+    },
+    {
+      "epoch": 338.3157894736842,
+      "grad_norm": 0.592397153377533,
+      "learning_rate": 4.546137313729996e-05,
+      "loss": 0.1991,
+      "step": 16070
+    },
+    {
+      "epoch": 338.5263157894737,
+      "grad_norm": 0.5688551664352417,
+      "learning_rate": 4.534936933852324e-05,
+      "loss": 0.1859,
+      "step": 16080
+    },
+    {
+      "epoch": 338.7368421052632,
+      "grad_norm": 0.710352897644043,
+      "learning_rate": 4.523746320928465e-05,
+      "loss": 0.1932,
+      "step": 16090
+    },
+    {
+      "epoch": 338.94736842105266,
+      "grad_norm": 0.6369743943214417,
+      "learning_rate": 4.5125654949578674e-05,
+      "loss": 0.1927,
+      "step": 16100
+    },
+    {
+      "epoch": 339.1578947368421,
+      "grad_norm": 0.5039718747138977,
+      "learning_rate": 4.5013944759224755e-05,
+      "loss": 0.196,
+      "step": 16110
+    },
+    {
+      "epoch": 339.36842105263156,
+      "grad_norm": 0.5557557940483093,
+      "learning_rate": 4.490233283786709e-05,
+      "loss": 0.171,
+      "step": 16120
+    },
+    {
+      "epoch": 339.57894736842104,
+      "grad_norm": 0.5374524593353271,
+      "learning_rate": 4.479081938497435e-05,
+      "loss": 0.2013,
+      "step": 16130
+    },
+    {
+      "epoch": 339.7894736842105,
+      "grad_norm": 0.48493510484695435,
+      "learning_rate": 4.4679404599839116e-05,
+      "loss": 0.1812,
+      "step": 16140
+    },
+    {
+      "epoch": 340.0,
+      "grad_norm": 0.6007246971130371,
+      "learning_rate": 4.456808868157762e-05,
+      "loss": 0.1925,
+      "step": 16150
+    },
+    {
+      "epoch": 340.2105263157895,
+      "grad_norm": 0.49306368827819824,
+      "learning_rate": 4.445687182912953e-05,
+      "loss": 0.1747,
+      "step": 16160
+    },
+    {
+      "epoch": 340.42105263157896,
+      "grad_norm": 0.5362128615379333,
+      "learning_rate": 4.434575424125741e-05,
+      "loss": 0.1861,
+      "step": 16170
+    },
+    {
+      "epoch": 340.63157894736844,
+      "grad_norm": 0.6607531309127808,
+      "learning_rate": 4.4234736116546364e-05,
+      "loss": 0.1833,
+      "step": 16180
+    },
+    {
+      "epoch": 340.8421052631579,
+      "grad_norm": 0.4933454692363739,
+      "learning_rate": 4.4123817653403756e-05,
+      "loss": 0.1889,
+      "step": 16190
+    },
+    {
+      "epoch": 341.05263157894734,
+      "grad_norm": 0.6923052668571472,
+      "learning_rate": 4.401299905005893e-05,
+      "loss": 0.1935,
+      "step": 16200
+    },
+    {
+      "epoch": 341.2631578947368,
+      "grad_norm": 0.4283411502838135,
+      "learning_rate": 4.390228050456267e-05,
+      "loss": 0.1871,
+      "step": 16210
+    },
+    {
+      "epoch": 341.4736842105263,
+      "grad_norm": 0.5722445249557495,
+      "learning_rate": 4.379166221478697e-05,
+      "loss": 0.1831,
+      "step": 16220
+    },
+    {
+      "epoch": 341.6842105263158,
+      "grad_norm": 0.5370818972587585,
+      "learning_rate": 4.368114437842461e-05,
+      "loss": 0.1974,
+      "step": 16230
+    },
+    {
+      "epoch": 341.89473684210526,
+      "grad_norm": 0.7643104195594788,
+      "learning_rate": 4.357072719298895e-05,
+      "loss": 0.1863,
+      "step": 16240
+    },
+    {
+      "epoch": 342.10526315789474,
+      "grad_norm": 0.6327834129333496,
+      "learning_rate": 4.3460410855813374e-05,
+      "loss": 0.1774,
+      "step": 16250
+    },
+    {
+      "epoch": 342.3157894736842,
+      "grad_norm": 0.5486928820610046,
+      "learning_rate": 4.3350195564051013e-05,
+      "loss": 0.1784,
+      "step": 16260
+    },
+    {
+      "epoch": 342.5263157894737,
+      "grad_norm": 0.4978579580783844,
+      "learning_rate": 4.3240081514674526e-05,
+      "loss": 0.1834,
+      "step": 16270
+    },
+    {
+      "epoch": 342.7368421052632,
+      "grad_norm": 0.6283280253410339,
+      "learning_rate": 4.3130068904475586e-05,
+      "loss": 0.1831,
+      "step": 16280
+    },
+    {
+      "epoch": 342.94736842105266,
+      "grad_norm": 0.7025750875473022,
+      "learning_rate": 4.302015793006451e-05,
+      "loss": 0.2037,
+      "step": 16290
+    },
+    {
+      "epoch": 343.1578947368421,
+      "grad_norm": 0.5586254000663757,
+      "learning_rate": 4.2910348787870094e-05,
+      "loss": 0.2055,
+      "step": 16300
+    },
+    {
+      "epoch": 343.36842105263156,
+      "grad_norm": 0.5136498808860779,
+      "learning_rate": 4.280064167413904e-05,
+      "loss": 0.1858,
+      "step": 16310
+    },
+    {
+      "epoch": 343.57894736842104,
+      "grad_norm": 0.6225966811180115,
+      "learning_rate": 4.2691036784935756e-05,
+      "loss": 0.1814,
+      "step": 16320
+    },
+    {
+      "epoch": 343.7894736842105,
+      "grad_norm": 0.542299747467041,
+      "learning_rate": 4.258153431614193e-05,
+      "loss": 0.1869,
+      "step": 16330
+    },
+    {
+      "epoch": 344.0,
+      "grad_norm": 0.7111690044403076,
+      "learning_rate": 4.247213446345626e-05,
+      "loss": 0.1881,
+      "step": 16340
+    },
+    {
+      "epoch": 344.2105263157895,
+      "grad_norm": 0.4649767577648163,
+      "learning_rate": 4.236283742239401e-05,
+      "loss": 0.1838,
+      "step": 16350
+    },
+    {
+      "epoch": 344.42105263157896,
+      "grad_norm": 0.5829393863677979,
+      "learning_rate": 4.225364338828668e-05,
+      "loss": 0.1988,
+      "step": 16360
+    },
+    {
+      "epoch": 344.63157894736844,
+      "grad_norm": 0.6762527227401733,
+      "learning_rate": 4.214455255628178e-05,
+      "loss": 0.1812,
+      "step": 16370
+    },
+    {
+      "epoch": 344.8421052631579,
+      "grad_norm": 0.5468589663505554,
+      "learning_rate": 4.2035565121342246e-05,
+      "loss": 0.1892,
+      "step": 16380
+    },
+    {
+      "epoch": 345.05263157894734,
+      "grad_norm": 0.4818148612976074,
+      "learning_rate": 4.1926681278246374e-05,
+      "loss": 0.1773,
+      "step": 16390
+    },
+    {
+      "epoch": 345.2631578947368,
+      "grad_norm": 0.46287909150123596,
+      "learning_rate": 4.181790122158716e-05,
+      "loss": 0.1774,
+      "step": 16400
+    },
+    {
+      "epoch": 345.4736842105263,
+      "grad_norm": 0.7505784630775452,
+      "learning_rate": 4.170922514577228e-05,
+      "loss": 0.2055,
+      "step": 16410
+    },
+    {
+      "epoch": 345.6842105263158,
+      "grad_norm": 0.47206437587738037,
+      "learning_rate": 4.160065324502348e-05,
+      "loss": 0.1876,
+      "step": 16420
+    },
+    {
+      "epoch": 345.89473684210526,
+      "grad_norm": 0.4681949317455292,
+      "learning_rate": 4.14921857133763e-05,
+      "loss": 0.1824,
+      "step": 16430
+    },
+    {
+      "epoch": 346.10526315789474,
+      "grad_norm": 0.6797969937324524,
+      "learning_rate": 4.1383822744679866e-05,
+      "loss": 0.1885,
+      "step": 16440
+    },
+    {
+      "epoch": 346.3157894736842,
+      "grad_norm": 0.5295494794845581,
+      "learning_rate": 4.127556453259637e-05,
+      "loss": 0.1743,
+      "step": 16450
+    },
+    {
+      "epoch": 346.5263157894737,
+      "grad_norm": 0.6828837394714355,
+      "learning_rate": 4.116741127060073e-05,
+      "loss": 0.1808,
+      "step": 16460
+    },
+    {
+      "epoch": 346.7368421052632,
+      "grad_norm": 0.6107656955718994,
+      "learning_rate": 4.105936315198043e-05,
+      "loss": 0.1928,
+      "step": 16470
+    },
+    {
+      "epoch": 346.94736842105266,
+      "grad_norm": 0.5245572924613953,
+      "learning_rate": 4.095142036983497e-05,
+      "loss": 0.2069,
+      "step": 16480
+    },
+    {
+      "epoch": 347.1578947368421,
+      "grad_norm": 0.49881359934806824,
+      "learning_rate": 4.0843583117075576e-05,
+      "loss": 0.18,
+      "step": 16490
+    },
+    {
+      "epoch": 347.36842105263156,
+      "grad_norm": 0.8077996969223022,
+      "learning_rate": 4.073585158642488e-05,
+      "loss": 0.19,
+      "step": 16500
+    },
+    {
+      "epoch": 347.57894736842104,
+      "grad_norm": 0.5267335176467896,
+      "learning_rate": 4.062822597041663e-05,
+      "loss": 0.1931,
+      "step": 16510
+    },
+    {
+      "epoch": 347.7894736842105,
+      "grad_norm": 0.672138512134552,
+      "learning_rate": 4.052070646139529e-05,
+      "loss": 0.184,
+      "step": 16520
+    },
+    {
+      "epoch": 348.0,
+      "grad_norm": 0.5576406717300415,
+      "learning_rate": 4.0413293251515574e-05,
+      "loss": 0.1909,
+      "step": 16530
+    },
+    {
+      "epoch": 348.2105263157895,
+      "grad_norm": 0.6533356308937073,
+      "learning_rate": 4.030598653274238e-05,
+      "loss": 0.1859,
+      "step": 16540
+    },
+    {
+      "epoch": 348.42105263157896,
+      "grad_norm": 0.6966098546981812,
+      "learning_rate": 4.019878649685018e-05,
+      "loss": 0.179,
+      "step": 16550
+    },
+    {
+      "epoch": 348.63157894736844,
+      "grad_norm": 0.5067752003669739,
+      "learning_rate": 4.009169333542283e-05,
+      "loss": 0.1943,
+      "step": 16560
+    },
+    {
+      "epoch": 348.8421052631579,
+      "grad_norm": 0.7860790491104126,
+      "learning_rate": 3.998470723985312e-05,
+      "loss": 0.1957,
+      "step": 16570
+    },
+    {
+      "epoch": 349.05263157894734,
+      "grad_norm": 0.6456900835037231,
+      "learning_rate": 3.987782840134263e-05,
+      "loss": 0.1946,
+      "step": 16580
+    },
+    {
+      "epoch": 349.2631578947368,
+      "grad_norm": 0.4159180819988251,
+      "learning_rate": 3.977105701090115e-05,
+      "loss": 0.1864,
+      "step": 16590
+    },
+    {
+      "epoch": 349.4736842105263,
+      "grad_norm": 0.5213875770568848,
+      "learning_rate": 3.96643932593464e-05,
+      "loss": 0.1913,
+      "step": 16600
+    },
+    {
+      "epoch": 349.6842105263158,
+      "grad_norm": 0.6271671652793884,
+      "learning_rate": 3.9557837337303906e-05,
+      "loss": 0.1888,
+      "step": 16610
+    },
+    {
+      "epoch": 349.89473684210526,
+      "grad_norm": 0.7119611501693726,
+      "learning_rate": 3.945138943520628e-05,
+      "loss": 0.1824,
+      "step": 16620
+    },
+    {
+      "epoch": 350.10526315789474,
+      "grad_norm": 0.45116692781448364,
+      "learning_rate": 3.934504974329326e-05,
+      "loss": 0.1866,
+      "step": 16630
+    },
+    {
+      "epoch": 350.3157894736842,
+      "grad_norm": 0.4684094488620758,
+      "learning_rate": 3.9238818451611056e-05,
+      "loss": 0.1807,
+      "step": 16640
+    },
+    {
+      "epoch": 350.5263157894737,
+      "grad_norm": 0.5230817794799805,
+      "learning_rate": 3.913269575001228e-05,
+      "loss": 0.1866,
+      "step": 16650
+    },
+    {
+      "epoch": 350.7368421052632,
+      "grad_norm": 0.7680791616439819,
+      "learning_rate": 3.9026681828155366e-05,
+      "loss": 0.1959,
+      "step": 16660
+    },
+    {
+      "epoch": 350.94736842105266,
+      "grad_norm": 0.47278884053230286,
+      "learning_rate": 3.892077687550435e-05,
+      "loss": 0.1774,
+      "step": 16670
+    },
+    {
+      "epoch": 351.1578947368421,
+      "grad_norm": 0.6973139643669128,
+      "learning_rate": 3.8814981081328615e-05,
+      "loss": 0.1845,
+      "step": 16680
+    },
+    {
+      "epoch": 351.36842105263156,
+      "grad_norm": 0.5785807967185974,
+      "learning_rate": 3.8709294634702376e-05,
+      "loss": 0.2071,
+      "step": 16690
+    },
+    {
+      "epoch": 351.57894736842104,
+      "grad_norm": 0.5238916277885437,
+      "learning_rate": 3.8603717724504404e-05,
+      "loss": 0.1744,
+      "step": 16700
+    },
+    {
+      "epoch": 351.7894736842105,
+      "grad_norm": 0.7316076159477234,
+      "learning_rate": 3.8498250539417835e-05,
+      "loss": 0.1834,
+      "step": 16710
+    },
+    {
+      "epoch": 352.0,
+      "grad_norm": 0.5799602270126343,
+      "learning_rate": 3.8392893267929597e-05,
+      "loss": 0.1788,
+      "step": 16720
+    },
+    {
+      "epoch": 352.2105263157895,
+      "grad_norm": 0.6016362905502319,
+      "learning_rate": 3.8287646098330166e-05,
+      "loss": 0.1821,
+      "step": 16730
+    },
+    {
+      "epoch": 352.42105263157896,
+      "grad_norm": 0.5646636486053467,
+      "learning_rate": 3.818250921871338e-05,
+      "loss": 0.1875,
+      "step": 16740
+    },
+    {
+      "epoch": 352.63157894736844,
+      "grad_norm": 0.5801871418952942,
+      "learning_rate": 3.807748281697583e-05,
+      "loss": 0.1926,
+      "step": 16750
+    },
+    {
+      "epoch": 352.8421052631579,
+      "grad_norm": 0.4862273037433624,
+      "learning_rate": 3.797256708081678e-05,
+      "loss": 0.1938,
+      "step": 16760
+    },
+    {
+      "epoch": 353.05263157894734,
+      "grad_norm": 0.5285743474960327,
+      "learning_rate": 3.786776219773759e-05,
+      "loss": 0.1796,
+      "step": 16770
+    },
+    {
+      "epoch": 353.2631578947368,
+      "grad_norm": 0.5707828998565674,
+      "learning_rate": 3.776306835504166e-05,
+      "loss": 0.187,
+      "step": 16780
+    },
+    {
+      "epoch": 353.4736842105263,
+      "grad_norm": 0.5975477695465088,
+      "learning_rate": 3.7658485739833824e-05,
+      "loss": 0.1862,
+      "step": 16790
+    },
+    {
+      "epoch": 353.6842105263158,
+      "grad_norm": 0.6158720850944519,
+      "learning_rate": 3.7554014539020134e-05,
+      "loss": 0.1942,
+      "step": 16800
+    },
+    {
+      "epoch": 353.89473684210526,
+      "grad_norm": 0.5053333044052124,
+      "learning_rate": 3.7449654939307635e-05,
+      "loss": 0.1784,
+      "step": 16810
+    },
+    {
+      "epoch": 354.10526315789474,
+      "grad_norm": 0.5054594874382019,
+      "learning_rate": 3.7345407127203826e-05,
+      "loss": 0.1938,
+      "step": 16820
+    },
+    {
+      "epoch": 354.3157894736842,
+      "grad_norm": 0.5698423385620117,
+      "learning_rate": 3.724127128901644e-05,
+      "loss": 0.183,
+      "step": 16830
+    },
+    {
+      "epoch": 354.5263157894737,
+      "grad_norm": 0.5378226637840271,
+      "learning_rate": 3.713724761085308e-05,
+      "loss": 0.182,
+      "step": 16840
+    },
+    {
+      "epoch": 354.7368421052632,
+      "grad_norm": 0.6348969340324402,
+      "learning_rate": 3.703333627862099e-05,
+      "loss": 0.1928,
+      "step": 16850
+    },
+    {
+      "epoch": 354.94736842105266,
+      "grad_norm": 0.6424091458320618,
+      "learning_rate": 3.692953747802649e-05,
+      "loss": 0.1792,
+      "step": 16860
+    },
+    {
+      "epoch": 355.1578947368421,
+      "grad_norm": 0.5656790733337402,
+      "learning_rate": 3.683621492536592e-05,
+      "loss": 0.1836,
+      "step": 16870
+    },
+    {
+      "epoch": 355.36842105263156,
+      "grad_norm": 0.5092337727546692,
+      "learning_rate": 3.6732630445783543e-05,
+      "loss": 0.1819,
+      "step": 16880
+    },
+    {
+      "epoch": 355.57894736842104,
+      "grad_norm": 0.5355526804924011,
+      "learning_rate": 3.662915903524888e-05,
+      "loss": 0.2005,
+      "step": 16890
+    },
+    {
+      "epoch": 355.7894736842105,
+      "grad_norm": 0.5522583723068237,
+      "learning_rate": 3.6525800878682084e-05,
+      "loss": 0.176,
+      "step": 16900
+    },
+    {
+      "epoch": 356.0,
+      "grad_norm": 0.5521705150604248,
+      "learning_rate": 3.642255616080101e-05,
+      "loss": 0.2027,
+      "step": 16910
+    },
+    {
+      "epoch": 356.2105263157895,
+      "grad_norm": 0.6535980701446533,
+      "learning_rate": 3.631942506612064e-05,
+      "loss": 0.176,
+      "step": 16920
+    },
+    {
+      "epoch": 356.42105263157896,
+      "grad_norm": 0.7658635973930359,
+      "learning_rate": 3.6216407778953033e-05,
+      "loss": 0.1811,
+      "step": 16930
+    },
+    {
+      "epoch": 356.63157894736844,
+      "grad_norm": 0.6077464818954468,
+      "learning_rate": 3.61135044834067e-05,
+      "loss": 0.1914,
+      "step": 16940
+    },
+    {
+      "epoch": 356.8421052631579,
+      "grad_norm": 0.5396264791488647,
+      "learning_rate": 3.601071536338661e-05,
+      "loss": 0.1869,
+      "step": 16950
+    },
+    {
+      "epoch": 357.05263157894734,
+      "grad_norm": 0.5491655468940735,
+      "learning_rate": 3.590804060259354e-05,
+      "loss": 0.189,
+      "step": 16960
+    },
+    {
+      "epoch": 357.2631578947368,
+      "grad_norm": 0.562882125377655,
+      "learning_rate": 3.5805480384523895e-05,
+      "loss": 0.1792,
+      "step": 16970
+    },
+    {
+      "epoch": 357.4736842105263,
+      "grad_norm": 0.6570281982421875,
+      "learning_rate": 3.570303489246949e-05,
+      "loss": 0.1885,
+      "step": 16980
+    },
+    {
+      "epoch": 357.6842105263158,
+      "grad_norm": 0.5322051048278809,
+      "learning_rate": 3.5600704309516997e-05,
+      "loss": 0.1805,
+      "step": 16990
+    },
+    {
+      "epoch": 357.89473684210526,
+      "grad_norm": 0.6352676153182983,
+      "learning_rate": 3.549848881854772e-05,
+      "loss": 0.1897,
+      "step": 17000
+    },
+    {
+      "epoch": 358.10526315789474,
+      "grad_norm": 0.6159738302230835,
+      "learning_rate": 3.539638860223738e-05,
+      "loss": 0.1955,
+      "step": 17010
+    },
+    {
+      "epoch": 358.3157894736842,
+      "grad_norm": 0.5578885078430176,
+      "learning_rate": 3.52944038430556e-05,
+      "loss": 0.1817,
+      "step": 17020
+    },
+    {
+      "epoch": 358.5263157894737,
+      "grad_norm": 0.5949670672416687,
+      "learning_rate": 3.519253472326562e-05,
+      "loss": 0.195,
+      "step": 17030
+    },
+    {
+      "epoch": 358.7368421052632,
+      "grad_norm": 0.4901580512523651,
+      "learning_rate": 3.509078142492418e-05,
+      "loss": 0.187,
+      "step": 17040
+    },
+    {
+      "epoch": 358.94736842105266,
+      "grad_norm": 0.5938405990600586,
+      "learning_rate": 3.498914412988083e-05,
+      "loss": 0.1811,
+      "step": 17050
+    },
+    {
+      "epoch": 359.1578947368421,
+      "grad_norm": 0.5262094736099243,
+      "learning_rate": 3.488762301977796e-05,
+      "loss": 0.1899,
+      "step": 17060
+    },
+    {
+      "epoch": 359.36842105263156,
+      "grad_norm": 0.41206714510917664,
+      "learning_rate": 3.47862182760502e-05,
+      "loss": 0.1825,
+      "step": 17070
+    },
+    {
+      "epoch": 359.57894736842104,
+      "grad_norm": 0.4352954626083374,
+      "learning_rate": 3.468493007992433e-05,
+      "loss": 0.1758,
+      "step": 17080
+    },
+    {
+      "epoch": 359.7894736842105,
+      "grad_norm": 0.6213468909263611,
+      "learning_rate": 3.458375861241874e-05,
+      "loss": 0.1799,
+      "step": 17090
+    },
+    {
+      "epoch": 360.0,
+      "grad_norm": 0.7452707290649414,
+      "learning_rate": 3.448270405434323e-05,
+      "loss": 0.1959,
+      "step": 17100
+    },
+    {
+      "epoch": 360.2105263157895,
+      "grad_norm": 0.7887313365936279,
+      "learning_rate": 3.438176658629873e-05,
+      "loss": 0.1806,
+      "step": 17110
+    },
+    {
+      "epoch": 360.42105263157896,
+      "grad_norm": 0.6283901333808899,
+      "learning_rate": 3.428094638867684e-05,
+      "loss": 0.1865,
+      "step": 17120
+    },
+    {
+      "epoch": 360.63157894736844,
+      "grad_norm": 0.6054136157035828,
+      "learning_rate": 3.418024364165959e-05,
+      "loss": 0.1943,
+      "step": 17130
+    },
+    {
+      "epoch": 360.8421052631579,
+      "grad_norm": 0.5418136119842529,
+      "learning_rate": 3.4079658525219106e-05,
+      "loss": 0.1799,
+      "step": 17140
+    },
+    {
+      "epoch": 361.05263157894734,
+      "grad_norm": 0.4626767039299011,
+      "learning_rate": 3.397919121911734e-05,
+      "loss": 0.1906,
+      "step": 17150
+    },
+    {
+      "epoch": 361.2631578947368,
+      "grad_norm": 0.4851849377155304,
+      "learning_rate": 3.38788419029056e-05,
+      "loss": 0.1909,
+      "step": 17160
+    },
+    {
+      "epoch": 361.4736842105263,
+      "grad_norm": 0.4757574796676636,
+      "learning_rate": 3.377861075592442e-05,
+      "loss": 0.1731,
+      "step": 17170
+    },
+    {
+      "epoch": 361.6842105263158,
+      "grad_norm": 0.6023973822593689,
+      "learning_rate": 3.367849795730314e-05,
+      "loss": 0.1866,
+      "step": 17180
+    },
+    {
+      "epoch": 361.89473684210526,
+      "grad_norm": 0.49852490425109863,
+      "learning_rate": 3.357850368595955e-05,
+      "loss": 0.2079,
+      "step": 17190
+    },
+    {
+      "epoch": 362.10526315789474,
+      "grad_norm": 0.569017231464386,
+      "learning_rate": 3.3478628120599573e-05,
+      "loss": 0.1862,
+      "step": 17200
+    },
+    {
+      "epoch": 362.3157894736842,
+      "grad_norm": 0.5501569509506226,
+      "learning_rate": 3.337887143971711e-05,
+      "loss": 0.1872,
+      "step": 17210
+    },
+    {
+      "epoch": 362.5263157894737,
+      "grad_norm": 0.7906637787818909,
+      "learning_rate": 3.3279233821593494e-05,
+      "loss": 0.1925,
+      "step": 17220
+    },
+    {
+      "epoch": 362.7368421052632,
+      "grad_norm": 0.6459226608276367,
+      "learning_rate": 3.3179715444297286e-05,
+      "loss": 0.1832,
+      "step": 17230
+    },
+    {
+      "epoch": 362.94736842105266,
+      "grad_norm": 0.6349130272865295,
+      "learning_rate": 3.308031648568396e-05,
+      "loss": 0.202,
+      "step": 17240
+    },
+    {
+      "epoch": 363.1578947368421,
+      "grad_norm": 0.4548916816711426,
+      "learning_rate": 3.298103712339562e-05,
+      "loss": 0.1833,
+      "step": 17250
+    },
+    {
+      "epoch": 363.36842105263156,
+      "grad_norm": 0.53446364402771,
+      "learning_rate": 3.288187753486056e-05,
+      "loss": 0.1857,
+      "step": 17260
+    },
+    {
+      "epoch": 363.57894736842104,
+      "grad_norm": 0.8272064328193665,
+      "learning_rate": 3.2782837897293e-05,
+      "loss": 0.1796,
+      "step": 17270
+    },
+    {
+      "epoch": 363.7894736842105,
+      "grad_norm": 0.538015604019165,
+      "learning_rate": 3.268391838769286e-05,
+      "loss": 0.1971,
+      "step": 17280
+    },
+    {
+      "epoch": 364.0,
+      "grad_norm": 0.6953075528144836,
+      "learning_rate": 3.258511918284538e-05,
+      "loss": 0.1908,
+      "step": 17290
+    },
+    {
+      "epoch": 364.2105263157895,
+      "grad_norm": 0.64228755235672,
+      "learning_rate": 3.248644045932074e-05,
+      "loss": 0.1865,
+      "step": 17300
+    },
+    {
+      "epoch": 364.42105263157896,
+      "grad_norm": 0.4800601899623871,
+      "learning_rate": 3.2387882393473766e-05,
+      "loss": 0.1773,
+      "step": 17310
+    },
+    {
+      "epoch": 364.63157894736844,
+      "grad_norm": 0.8892537951469421,
+      "learning_rate": 3.228944516144379e-05,
+      "loss": 0.1917,
+      "step": 17320
+    },
+    {
+      "epoch": 364.8421052631579,
+      "grad_norm": 0.6665611863136292,
+      "learning_rate": 3.219112893915405e-05,
+      "loss": 0.1877,
+      "step": 17330
+    },
+    {
+      "epoch": 365.05263157894734,
+      "grad_norm": 0.5820897221565247,
+      "learning_rate": 3.209293390231155e-05,
+      "loss": 0.1975,
+      "step": 17340
+    },
+    {
+      "epoch": 365.2631578947368,
+      "grad_norm": 0.6017282009124756,
+      "learning_rate": 3.199486022640681e-05,
+      "loss": 0.1879,
+      "step": 17350
+    },
+    {
+      "epoch": 365.4736842105263,
+      "grad_norm": 0.5204026103019714,
+      "learning_rate": 3.189690808671336e-05,
+      "loss": 0.1886,
+      "step": 17360
+    },
+    {
+      "epoch": 365.6842105263158,
+      "grad_norm": 0.5471468567848206,
+      "learning_rate": 3.1799077658287534e-05,
+      "loss": 0.1852,
+      "step": 17370
+    },
+    {
+      "epoch": 365.89473684210526,
+      "grad_norm": 0.5290389060974121,
+      "learning_rate": 3.170136911596822e-05,
+      "loss": 0.1863,
+      "step": 17380
+    },
+    {
+      "epoch": 366.10526315789474,
+      "grad_norm": 0.5705314874649048,
+      "learning_rate": 3.160378263437639e-05,
+      "loss": 0.1946,
+      "step": 17390
+    },
+    {
+      "epoch": 366.3157894736842,
+      "grad_norm": 0.5175087451934814,
+      "learning_rate": 3.150631838791489e-05,
+      "loss": 0.1771,
+      "step": 17400
+    },
+    {
+      "epoch": 366.5263157894737,
+      "grad_norm": 0.5701705813407898,
+      "learning_rate": 3.1408976550768156e-05,
+      "loss": 0.1878,
+      "step": 17410
+    },
+    {
+      "epoch": 366.7368421052632,
+      "grad_norm": 0.7232944965362549,
+      "learning_rate": 3.131175729690187e-05,
+      "loss": 0.1968,
+      "step": 17420
+    },
+    {
+      "epoch": 366.94736842105266,
+      "grad_norm": 0.5719231963157654,
+      "learning_rate": 3.1214660800062567e-05,
+      "loss": 0.1725,
+      "step": 17430
+    },
+    {
+      "epoch": 367.1578947368421,
+      "grad_norm": 0.542506217956543,
+      "learning_rate": 3.111768723377741e-05,
+      "loss": 0.1909,
+      "step": 17440
+    },
+    {
+      "epoch": 367.36842105263156,
+      "grad_norm": 0.5347987413406372,
+      "learning_rate": 3.1020836771353926e-05,
+      "loss": 0.1861,
+      "step": 17450
+    },
+    {
+      "epoch": 367.57894736842104,
+      "grad_norm": 0.5752246975898743,
+      "learning_rate": 3.092410958587958e-05,
+      "loss": 0.1868,
+      "step": 17460
+    },
+    {
+      "epoch": 367.7894736842105,
+      "grad_norm": 0.5393999814987183,
+      "learning_rate": 3.082750585022153e-05,
+      "loss": 0.1806,
+      "step": 17470
+    },
+    {
+      "epoch": 368.0,
+      "grad_norm": 0.5871136784553528,
+      "learning_rate": 3.073102573702629e-05,
+      "loss": 0.193,
+      "step": 17480
+    },
+    {
+      "epoch": 368.2105263157895,
+      "grad_norm": 0.6469846963882446,
+      "learning_rate": 3.063466941871952e-05,
+      "loss": 0.1846,
+      "step": 17490
+    },
+    {
+      "epoch": 368.42105263157896,
+      "grad_norm": 0.5563372373580933,
+      "learning_rate": 3.0538437067505565e-05,
+      "loss": 0.1901,
+      "step": 17500
+    },
+    {
+      "epoch": 368.63157894736844,
+      "grad_norm": 0.5169476270675659,
+      "learning_rate": 3.0442328855367197e-05,
+      "loss": 0.1779,
+      "step": 17510
+    },
+    {
+      "epoch": 368.8421052631579,
+      "grad_norm": 0.5597511529922485,
+      "learning_rate": 3.0346344954065408e-05,
+      "loss": 0.1934,
+      "step": 17520
+    },
+    {
+      "epoch": 369.05263157894734,
+      "grad_norm": 0.5994619727134705,
+      "learning_rate": 3.0250485535139028e-05,
+      "loss": 0.182,
+      "step": 17530
+    },
+    {
+      "epoch": 369.2631578947368,
+      "grad_norm": 0.5678781270980835,
+      "learning_rate": 3.0154750769904317e-05,
+      "loss": 0.1866,
+      "step": 17540
+    },
+    {
+      "epoch": 369.4736842105263,
+      "grad_norm": 0.5461990237236023,
+      "learning_rate": 3.005914082945488e-05,
+      "loss": 0.1935,
+      "step": 17550
+    },
+    {
+      "epoch": 369.6842105263158,
+      "grad_norm": 0.5570061802864075,
+      "learning_rate": 2.996365588466117e-05,
+      "loss": 0.1912,
+      "step": 17560
+    },
+    {
+      "epoch": 369.89473684210526,
+      "grad_norm": 0.5612841844558716,
+      "learning_rate": 2.9868296106170236e-05,
+      "loss": 0.1772,
+      "step": 17570
+    },
+    {
+      "epoch": 370.10526315789474,
+      "grad_norm": 0.5573035478591919,
+      "learning_rate": 2.9773061664405455e-05,
+      "loss": 0.1834,
+      "step": 17580
+    },
+    {
+      "epoch": 370.3157894736842,
+      "grad_norm": 0.5268656611442566,
+      "learning_rate": 2.9677952729566284e-05,
+      "loss": 0.1801,
+      "step": 17590
+    },
+    {
+      "epoch": 370.5263157894737,
+      "grad_norm": 0.4846976697444916,
+      "learning_rate": 2.958296947162775e-05,
+      "loss": 0.1878,
+      "step": 17600
+    },
+    {
+      "epoch": 370.7368421052632,
+      "grad_norm": 0.625234067440033,
+      "learning_rate": 2.9488112060340333e-05,
+      "loss": 0.1863,
+      "step": 17610
+    },
+    {
+      "epoch": 370.94736842105266,
+      "grad_norm": 0.7202187180519104,
+      "learning_rate": 2.9393380665229666e-05,
+      "loss": 0.1998,
+      "step": 17620
+    },
+    {
+      "epoch": 371.1578947368421,
+      "grad_norm": 0.8498916029930115,
+      "learning_rate": 2.9298775455596027e-05,
+      "loss": 0.1814,
+      "step": 17630
+    },
+    {
+      "epoch": 371.36842105263156,
+      "grad_norm": 0.5455896854400635,
+      "learning_rate": 2.920429660051436e-05,
+      "loss": 0.1823,
+      "step": 17640
+    },
+    {
+      "epoch": 371.57894736842104,
+      "grad_norm": 0.6018717288970947,
+      "learning_rate": 2.910994426883361e-05,
+      "loss": 0.1923,
+      "step": 17650
+    },
+    {
+      "epoch": 371.7894736842105,
+      "grad_norm": 0.8865271806716919,
+      "learning_rate": 2.9015718629176758e-05,
+      "loss": 0.1908,
+      "step": 17660
+    },
+    {
+      "epoch": 372.0,
+      "grad_norm": 0.7926931381225586,
+      "learning_rate": 2.8921619849940286e-05,
+      "loss": 0.1847,
+      "step": 17670
+    },
+    {
+      "epoch": 372.2105263157895,
+      "grad_norm": 0.531242847442627,
+      "learning_rate": 2.8827648099293925e-05,
+      "loss": 0.18,
+      "step": 17680
+    },
+    {
+      "epoch": 372.42105263157896,
+      "grad_norm": 0.7653034925460815,
+      "learning_rate": 2.8733803545180492e-05,
+      "loss": 0.1898,
+      "step": 17690
+    },
+    {
+      "epoch": 372.63157894736844,
+      "grad_norm": 0.5573826432228088,
+      "learning_rate": 2.86400863553154e-05,
+      "loss": 0.1743,
+      "step": 17700
+    },
+    {
+      "epoch": 372.8421052631579,
+      "grad_norm": 0.5950312614440918,
+      "learning_rate": 2.854649669718642e-05,
+      "loss": 0.2007,
+      "step": 17710
+    },
+    {
+      "epoch": 373.05263157894734,
+      "grad_norm": 0.48360878229141235,
+      "learning_rate": 2.845303473805352e-05,
+      "loss": 0.1765,
+      "step": 17720
+    },
+    {
+      "epoch": 373.2631578947368,
+      "grad_norm": 0.6277357935905457,
+      "learning_rate": 2.8359700644948327e-05,
+      "loss": 0.1944,
+      "step": 17730
+    },
+    {
+      "epoch": 373.4736842105263,
+      "grad_norm": 0.6063629984855652,
+      "learning_rate": 2.8266494584673987e-05,
+      "loss": 0.1862,
+      "step": 17740
+    },
+    {
+      "epoch": 373.6842105263158,
+      "grad_norm": 0.5055272579193115,
+      "learning_rate": 2.817341672380489e-05,
+      "loss": 0.1818,
+      "step": 17750
+    },
+    {
+      "epoch": 373.89473684210526,
+      "grad_norm": 0.5888772010803223,
+      "learning_rate": 2.8080467228686203e-05,
+      "loss": 0.1894,
+      "step": 17760
+    },
+    {
+      "epoch": 374.10526315789474,
+      "grad_norm": 0.650392472743988,
+      "learning_rate": 2.798764626543382e-05,
+      "loss": 0.1881,
+      "step": 17770
+    },
+    {
+      "epoch": 374.3157894736842,
+      "grad_norm": 0.8182873725891113,
+      "learning_rate": 2.7894953999933783e-05,
+      "loss": 0.1782,
+      "step": 17780
+    },
+    {
+      "epoch": 374.5263157894737,
+      "grad_norm": 0.5956078171730042,
+      "learning_rate": 2.7802390597842264e-05,
+      "loss": 0.2024,
+      "step": 17790
+    },
+    {
+      "epoch": 374.7368421052632,
+      "grad_norm": 0.5271019339561462,
+      "learning_rate": 2.7709956224585033e-05,
+      "loss": 0.1798,
+      "step": 17800
+    },
+    {
+      "epoch": 374.94736842105266,
+      "grad_norm": 0.6090193390846252,
+      "learning_rate": 2.7617651045357307e-05,
+      "loss": 0.1864,
+      "step": 17810
+    },
+    {
+      "epoch": 375.1578947368421,
+      "grad_norm": 0.5264062285423279,
+      "learning_rate": 2.7525475225123377e-05,
+      "loss": 0.1925,
+      "step": 17820
+    },
+    {
+      "epoch": 375.36842105263156,
+      "grad_norm": 0.8413964509963989,
+      "learning_rate": 2.7433428928616444e-05,
+      "loss": 0.1828,
+      "step": 17830
+    },
+    {
+      "epoch": 375.57894736842104,
+      "grad_norm": 0.5953909158706665,
+      "learning_rate": 2.7341512320338125e-05,
+      "loss": 0.1952,
+      "step": 17840
+    },
+    {
+      "epoch": 375.7894736842105,
+      "grad_norm": 0.6618128418922424,
+      "learning_rate": 2.7249725564558294e-05,
+      "loss": 0.1924,
+      "step": 17850
+    },
+    {
+      "epoch": 376.0,
+      "grad_norm": 0.5150593519210815,
+      "learning_rate": 2.7158068825314798e-05,
+      "loss": 0.1819,
+      "step": 17860
+    },
+    {
+      "epoch": 376.2105263157895,
+      "grad_norm": 0.6715971827507019,
+      "learning_rate": 2.7066542266413042e-05,
+      "loss": 0.1881,
+      "step": 17870
+    },
+    {
+      "epoch": 376.42105263157896,
+      "grad_norm": 0.6898446679115295,
+      "learning_rate": 2.6975146051425892e-05,
+      "loss": 0.1784,
+      "step": 17880
+    },
+    {
+      "epoch": 376.63157894736844,
+      "grad_norm": 0.549727737903595,
+      "learning_rate": 2.6883880343693146e-05,
+      "loss": 0.1982,
+      "step": 17890
+    },
+    {
+      "epoch": 376.8421052631579,
+      "grad_norm": 0.5464901328086853,
+      "learning_rate": 2.6792745306321464e-05,
+      "loss": 0.2045,
+      "step": 17900
+    },
+    {
+      "epoch": 377.05263157894734,
+      "grad_norm": 0.6344653367996216,
+      "learning_rate": 2.670174110218393e-05,
+      "loss": 0.1812,
+      "step": 17910
+    },
+    {
+      "epoch": 377.2631578947368,
+      "grad_norm": 0.543652355670929,
+      "learning_rate": 2.6610867893919768e-05,
+      "loss": 0.1899,
+      "step": 17920
+    },
+    {
+      "epoch": 377.4736842105263,
+      "grad_norm": 0.6504817605018616,
+      "learning_rate": 2.6520125843934184e-05,
+      "loss": 0.1917,
+      "step": 17930
+    },
+    {
+      "epoch": 377.6842105263158,
+      "grad_norm": 0.511242687702179,
+      "learning_rate": 2.6429515114397928e-05,
+      "loss": 0.1858,
+      "step": 17940
+    },
+    {
+      "epoch": 377.89473684210526,
+      "grad_norm": 0.4847296178340912,
+      "learning_rate": 2.633903586724703e-05,
+      "loss": 0.1846,
+      "step": 17950
+    },
+    {
+      "epoch": 378.10526315789474,
+      "grad_norm": 0.5976632237434387,
+      "learning_rate": 2.624868826418262e-05,
+      "loss": 0.1861,
+      "step": 17960
+    },
+    {
+      "epoch": 378.3157894736842,
+      "grad_norm": 0.5360187292098999,
+      "learning_rate": 2.6158472466670502e-05,
+      "loss": 0.1818,
+      "step": 17970
+    },
+    {
+      "epoch": 378.5263157894737,
+      "grad_norm": 0.5664785504341125,
+      "learning_rate": 2.6068388635940888e-05,
+      "loss": 0.1807,
+      "step": 17980
+    },
+    {
+      "epoch": 378.7368421052632,
+      "grad_norm": 0.6407875418663025,
+      "learning_rate": 2.597843693298826e-05,
+      "loss": 0.1823,
+      "step": 17990
+    },
+    {
+      "epoch": 378.94736842105266,
+      "grad_norm": 0.6425795555114746,
+      "learning_rate": 2.5888617518570834e-05,
+      "loss": 0.1914,
+      "step": 18000
+    },
+    {
+      "epoch": 379.1578947368421,
+      "grad_norm": 0.7084444165229797,
+      "learning_rate": 2.5798930553210533e-05,
+      "loss": 0.1944,
+      "step": 18010
+    },
+    {
+      "epoch": 379.36842105263156,
+      "grad_norm": 0.6060276627540588,
+      "learning_rate": 2.5709376197192437e-05,
+      "loss": 0.1833,
+      "step": 18020
+    },
+    {
+      "epoch": 379.57894736842104,
+      "grad_norm": 0.55902498960495,
+      "learning_rate": 2.5619954610564767e-05,
+      "loss": 0.1859,
+      "step": 18030
+    },
+    {
+      "epoch": 379.7894736842105,
+      "grad_norm": 0.4850768744945526,
+      "learning_rate": 2.5530665953138356e-05,
+      "loss": 0.1866,
+      "step": 18040
+    },
+    {
+      "epoch": 380.0,
+      "grad_norm": 0.6732696890830994,
+      "learning_rate": 2.54415103844865e-05,
+      "loss": 0.1842,
+      "step": 18050
+    },
+    {
+      "epoch": 380.2105263157895,
+      "grad_norm": 0.48501572012901306,
+      "learning_rate": 2.53524880639447e-05,
+      "loss": 0.173,
+      "step": 18060
+    },
+    {
+      "epoch": 380.42105263157896,
+      "grad_norm": 0.4836499094963074,
+      "learning_rate": 2.526359915061025e-05,
+      "loss": 0.1913,
+      "step": 18070
+    },
+    {
+      "epoch": 380.63157894736844,
+      "grad_norm": 0.8080970048904419,
+      "learning_rate": 2.5174843803342062e-05,
+      "loss": 0.1905,
+      "step": 18080
+    },
+    {
+      "epoch": 380.8421052631579,
+      "grad_norm": 0.530227780342102,
+      "learning_rate": 2.5086222180760298e-05,
+      "loss": 0.187,
+      "step": 18090
+    },
+    {
+      "epoch": 381.05263157894734,
+      "grad_norm": 0.45946136116981506,
+      "learning_rate": 2.499773444124621e-05,
+      "loss": 0.1861,
+      "step": 18100
+    },
+    {
+      "epoch": 381.2631578947368,
+      "grad_norm": 0.5182830095291138,
+      "learning_rate": 2.4909380742941703e-05,
+      "loss": 0.1856,
+      "step": 18110
+    },
+    {
+      "epoch": 381.4736842105263,
+      "grad_norm": 0.8414605259895325,
+      "learning_rate": 2.482116124374918e-05,
+      "loss": 0.1873,
+      "step": 18120
+    },
+    {
+      "epoch": 381.6842105263158,
+      "grad_norm": 0.6385777592658997,
+      "learning_rate": 2.473307610133121e-05,
+      "loss": 0.194,
+      "step": 18130
+    },
+    {
+      "epoch": 381.89473684210526,
+      "grad_norm": 0.6704742312431335,
+      "learning_rate": 2.464512547311021e-05,
+      "loss": 0.1955,
+      "step": 18140
+    },
+    {
+      "epoch": 382.10526315789474,
+      "grad_norm": 0.5813697576522827,
+      "learning_rate": 2.45573095162682e-05,
+      "loss": 0.1752,
+      "step": 18150
+    },
+    {
+      "epoch": 382.3157894736842,
+      "grad_norm": 0.6703978180885315,
+      "learning_rate": 2.4469628387746523e-05,
+      "loss": 0.1811,
+      "step": 18160
+    },
+    {
+      "epoch": 382.5263157894737,
+      "grad_norm": 0.5987458825111389,
+      "learning_rate": 2.438208224424561e-05,
+      "loss": 0.1813,
+      "step": 18170
+    },
+    {
+      "epoch": 382.7368421052632,
+      "grad_norm": 0.678972601890564,
+      "learning_rate": 2.42946712422246e-05,
+      "loss": 0.1964,
+      "step": 18180
+    },
+    {
+      "epoch": 382.94736842105266,
+      "grad_norm": 0.5534781813621521,
+      "learning_rate": 2.420739553790109e-05,
+      "loss": 0.1738,
+      "step": 18190
+    },
+    {
+      "epoch": 383.1578947368421,
+      "grad_norm": 0.4981260299682617,
+      "learning_rate": 2.412025528725097e-05,
+      "loss": 0.1803,
+      "step": 18200
+    },
+    {
+      "epoch": 383.36842105263156,
+      "grad_norm": 0.5293849110603333,
+      "learning_rate": 2.4033250646007976e-05,
+      "loss": 0.193,
+      "step": 18210
+    },
+    {
+      "epoch": 383.57894736842104,
+      "grad_norm": 0.4817885160446167,
+      "learning_rate": 2.3946381769663484e-05,
+      "loss": 0.1799,
+      "step": 18220
+    },
+    {
+      "epoch": 383.7894736842105,
+      "grad_norm": 0.4696321189403534,
+      "learning_rate": 2.3859648813466274e-05,
+      "loss": 0.1982,
+      "step": 18230
+    },
+    {
+      "epoch": 384.0,
+      "grad_norm": 0.761279821395874,
+      "learning_rate": 2.377305193242224e-05,
+      "loss": 0.1971,
+      "step": 18240
+    },
+    {
+      "epoch": 384.2105263157895,
+      "grad_norm": 0.5053116679191589,
+      "learning_rate": 2.3686591281294034e-05,
+      "loss": 0.1784,
+      "step": 18250
+    },
+    {
+      "epoch": 384.42105263157896,
+      "grad_norm": 0.6713317036628723,
+      "learning_rate": 2.3600267014600796e-05,
+      "loss": 0.1904,
+      "step": 18260
+    },
+    {
+      "epoch": 384.63157894736844,
+      "grad_norm": 0.49430567026138306,
+      "learning_rate": 2.3514079286618085e-05,
+      "loss": 0.178,
+      "step": 18270
+    },
+    {
+      "epoch": 384.8421052631579,
+      "grad_norm": 0.5539741516113281,
+      "learning_rate": 2.3428028251377278e-05,
+      "loss": 0.1921,
+      "step": 18280
+    },
+    {
+      "epoch": 385.05263157894734,
+      "grad_norm": 0.5688350200653076,
+      "learning_rate": 2.3342114062665533e-05,
+      "loss": 0.1851,
+      "step": 18290
+    },
+    {
+      "epoch": 385.2631578947368,
+      "grad_norm": 0.46466580033302307,
+      "learning_rate": 2.3256336874025463e-05,
+      "loss": 0.1808,
+      "step": 18300
+    },
+    {
+      "epoch": 385.4736842105263,
+      "grad_norm": 0.5379757285118103,
+      "learning_rate": 2.3170696838754814e-05,
+      "loss": 0.1842,
+      "step": 18310
+    },
+    {
+      "epoch": 385.6842105263158,
+      "grad_norm": 0.5040892362594604,
+      "learning_rate": 2.308519410990618e-05,
+      "loss": 0.1789,
+      "step": 18320
+    },
+    {
+      "epoch": 385.89473684210526,
+      "grad_norm": 0.48588046431541443,
+      "learning_rate": 2.2999828840286807e-05,
+      "loss": 0.191,
+      "step": 18330
+    },
+    {
+      "epoch": 386.10526315789474,
+      "grad_norm": 0.7209267020225525,
+      "learning_rate": 2.291460118245832e-05,
+      "loss": 0.1982,
+      "step": 18340
+    },
+    {
+      "epoch": 386.3157894736842,
+      "grad_norm": 0.5708928108215332,
+      "learning_rate": 2.282951128873628e-05,
+      "loss": 0.1913,
+      "step": 18350
+    },
+    {
+      "epoch": 386.5263157894737,
+      "grad_norm": 0.6246204376220703,
+      "learning_rate": 2.2744559311190185e-05,
+      "loss": 0.1827,
+      "step": 18360
+    },
+    {
+      "epoch": 386.7368421052632,
+      "grad_norm": 0.6227989792823792,
+      "learning_rate": 2.2659745401643005e-05,
+      "loss": 0.175,
+      "step": 18370
+    },
+    {
+      "epoch": 386.94736842105266,
+      "grad_norm": 0.5730044841766357,
+      "learning_rate": 2.2575069711670928e-05,
+      "loss": 0.179,
+      "step": 18380
+    },
+    {
+      "epoch": 387.1578947368421,
+      "grad_norm": 0.6064084768295288,
+      "learning_rate": 2.2490532392603103e-05,
+      "loss": 0.1846,
+      "step": 18390
+    },
+    {
+      "epoch": 387.36842105263156,
+      "grad_norm": 0.5600056052207947,
+      "learning_rate": 2.2406133595521495e-05,
+      "loss": 0.1718,
+      "step": 18400
+    },
+    {
+      "epoch": 387.57894736842104,
+      "grad_norm": 0.6896880269050598,
+      "learning_rate": 2.23218734712604e-05,
+      "loss": 0.1965,
+      "step": 18410
+    },
+    {
+      "epoch": 387.7894736842105,
+      "grad_norm": 0.4734235107898712,
+      "learning_rate": 2.2237752170406333e-05,
+      "loss": 0.1891,
+      "step": 18420
+    },
+    {
+      "epoch": 388.0,
+      "grad_norm": 1.0163463354110718,
+      "learning_rate": 2.2153769843297667e-05,
+      "loss": 0.1877,
+      "step": 18430
+    },
+    {
+      "epoch": 388.2105263157895,
+      "grad_norm": 0.6150470972061157,
+      "learning_rate": 2.2069926640024486e-05,
+      "loss": 0.1841,
+      "step": 18440
+    },
+    {
+      "epoch": 388.42105263157896,
+      "grad_norm": 0.4762638211250305,
+      "learning_rate": 2.1986222710428163e-05,
+      "loss": 0.1835,
+      "step": 18450
+    },
+    {
+      "epoch": 388.63157894736844,
+      "grad_norm": 0.5731767416000366,
+      "learning_rate": 2.190265820410117e-05,
+      "loss": 0.2089,
+      "step": 18460
+    },
+    {
+      "epoch": 388.8421052631579,
+      "grad_norm": 0.600663959980011,
+      "learning_rate": 2.1819233270386852e-05,
+      "loss": 0.1831,
+      "step": 18470
+    },
+    {
+      "epoch": 389.05263157894734,
+      "grad_norm": 0.6098585724830627,
+      "learning_rate": 2.1735948058379118e-05,
+      "loss": 0.1861,
+      "step": 18480
+    },
+    {
+      "epoch": 389.2631578947368,
+      "grad_norm": 0.6013163328170776,
+      "learning_rate": 2.1652802716922126e-05,
+      "loss": 0.1851,
+      "step": 18490
+    },
+    {
+      "epoch": 389.4736842105263,
+      "grad_norm": 0.5845993757247925,
+      "learning_rate": 2.1569797394610048e-05,
+      "loss": 0.1787,
+      "step": 18500
+    },
+    {
+      "epoch": 389.6842105263158,
+      "grad_norm": 0.635697066783905,
+      "learning_rate": 2.1486932239786916e-05,
+      "loss": 0.1819,
+      "step": 18510
+    },
+    {
+      "epoch": 389.89473684210526,
+      "grad_norm": 0.49835336208343506,
+      "learning_rate": 2.1404207400546173e-05,
+      "loss": 0.1815,
+      "step": 18520
+    },
+    {
+      "epoch": 390.10526315789474,
+      "grad_norm": 0.4586002230644226,
+      "learning_rate": 2.1321623024730474e-05,
+      "loss": 0.1867,
+      "step": 18530
+    },
+    {
+      "epoch": 390.3157894736842,
+      "grad_norm": 0.6160049438476562,
+      "learning_rate": 2.1239179259931563e-05,
+      "loss": 0.1922,
+      "step": 18540
+    },
+    {
+      "epoch": 390.5263157894737,
+      "grad_norm": 0.5551902651786804,
+      "learning_rate": 2.1156876253489766e-05,
+      "loss": 0.1907,
+      "step": 18550
+    },
+    {
+      "epoch": 390.7368421052632,
+      "grad_norm": 0.7020893692970276,
+      "learning_rate": 2.10747141524939e-05,
+      "loss": 0.1816,
+      "step": 18560
+    },
+    {
+      "epoch": 390.94736842105266,
+      "grad_norm": 0.5538049936294556,
+      "learning_rate": 2.0992693103781e-05,
+      "loss": 0.183,
+      "step": 18570
+    },
+    {
+      "epoch": 391.1578947368421,
+      "grad_norm": 0.5073581337928772,
+      "learning_rate": 2.0910813253935936e-05,
+      "loss": 0.1808,
+      "step": 18580
+    },
+    {
+      "epoch": 391.36842105263156,
+      "grad_norm": 0.6061038374900818,
+      "learning_rate": 2.082907474929131e-05,
+      "loss": 0.1868,
+      "step": 18590
+    },
+    {
+      "epoch": 391.57894736842104,
+      "grad_norm": 0.541469156742096,
+      "learning_rate": 2.0747477735927044e-05,
+      "loss": 0.1778,
+      "step": 18600
+    },
+    {
+      "epoch": 391.7894736842105,
+      "grad_norm": 0.6137720346450806,
+      "learning_rate": 2.066602235967029e-05,
+      "loss": 0.1857,
+      "step": 18610
+    },
+    {
+      "epoch": 392.0,
+      "grad_norm": 0.7164705991744995,
+      "learning_rate": 2.0584708766094963e-05,
+      "loss": 0.1914,
+      "step": 18620
+    },
+    {
+      "epoch": 392.2105263157895,
+      "grad_norm": 0.6752933859825134,
+      "learning_rate": 2.050353710052164e-05,
+      "loss": 0.1878,
+      "step": 18630
+    },
+    {
+      "epoch": 392.42105263157896,
+      "grad_norm": 0.5490244030952454,
+      "learning_rate": 2.0422507508017286e-05,
+      "loss": 0.1797,
+      "step": 18640
+    },
+    {
+      "epoch": 392.63157894736844,
+      "grad_norm": 0.639519989490509,
+      "learning_rate": 2.0341620133394902e-05,
+      "loss": 0.1836,
+      "step": 18650
+    },
+    {
+      "epoch": 392.8421052631579,
+      "grad_norm": 0.6504760980606079,
+      "learning_rate": 2.0260875121213297e-05,
+      "loss": 0.1853,
+      "step": 18660
+    },
+    {
+      "epoch": 393.05263157894734,
+      "grad_norm": 0.7008463740348816,
+      "learning_rate": 2.0180272615776985e-05,
+      "loss": 0.1989,
+      "step": 18670
+    },
+    {
+      "epoch": 393.2631578947368,
+      "grad_norm": 0.5902712941169739,
+      "learning_rate": 2.009981276113565e-05,
+      "loss": 0.1822,
+      "step": 18680
+    },
+    {
+      "epoch": 393.4736842105263,
+      "grad_norm": 0.5219281911849976,
+      "learning_rate": 2.0019495701084102e-05,
+      "loss": 0.1871,
+      "step": 18690
+    },
+    {
+      "epoch": 393.6842105263158,
+      "grad_norm": 0.6408507823944092,
+      "learning_rate": 1.9939321579161994e-05,
+      "loss": 0.1794,
+      "step": 18700
+    },
+    {
+      "epoch": 393.89473684210526,
+      "grad_norm": 0.5509892106056213,
+      "learning_rate": 1.985929053865342e-05,
+      "loss": 0.1868,
+      "step": 18710
+    },
+    {
+      "epoch": 394.10526315789474,
+      "grad_norm": 0.5968159437179565,
+      "learning_rate": 1.977940272258688e-05,
+      "loss": 0.1922,
+      "step": 18720
+    },
+    {
+      "epoch": 394.3157894736842,
+      "grad_norm": 0.6544321179389954,
+      "learning_rate": 1.969965827373481e-05,
+      "loss": 0.1899,
+      "step": 18730
+    },
+    {
+      "epoch": 394.5263157894737,
+      "grad_norm": 0.9849665760993958,
+      "learning_rate": 1.9620057334613516e-05,
+      "loss": 0.1883,
+      "step": 18740
+    },
+    {
+      "epoch": 394.7368421052632,
+      "grad_norm": 0.5996710658073425,
+      "learning_rate": 1.954060004748276e-05,
+      "loss": 0.1782,
+      "step": 18750
+    },
+    {
+      "epoch": 394.94736842105266,
+      "grad_norm": 0.7064588069915771,
+      "learning_rate": 1.9461286554345592e-05,
+      "loss": 0.1827,
+      "step": 18760
+    },
+    {
+      "epoch": 395.1578947368421,
+      "grad_norm": 0.531457781791687,
+      "learning_rate": 1.9382116996948074e-05,
+      "loss": 0.1759,
+      "step": 18770
+    },
+    {
+      "epoch": 395.36842105263156,
+      "grad_norm": 0.5805523991584778,
+      "learning_rate": 1.930309151677907e-05,
+      "loss": 0.1811,
+      "step": 18780
+    },
+    {
+      "epoch": 395.57894736842104,
+      "grad_norm": 0.6076486706733704,
+      "learning_rate": 1.922421025506992e-05,
+      "loss": 0.1919,
+      "step": 18790
+    },
+    {
+      "epoch": 395.7894736842105,
+      "grad_norm": 0.46915534138679504,
+      "learning_rate": 1.9145473352794197e-05,
+      "loss": 0.1876,
+      "step": 18800
+    },
+    {
+      "epoch": 396.0,
+      "grad_norm": 0.6902257800102234,
+      "learning_rate": 1.9066880950667565e-05,
+      "loss": 0.1893,
+      "step": 18810
+    },
+    {
+      "epoch": 396.2105263157895,
+      "grad_norm": 0.5422173738479614,
+      "learning_rate": 1.8988433189147325e-05,
+      "loss": 0.1844,
+      "step": 18820
+    },
+    {
+      "epoch": 396.42105263157896,
+      "grad_norm": 0.42955440282821655,
+      "learning_rate": 1.891013020843242e-05,
+      "loss": 0.1751,
+      "step": 18830
+    },
+    {
+      "epoch": 396.63157894736844,
+      "grad_norm": 0.5527623891830444,
+      "learning_rate": 1.8831972148462906e-05,
+      "loss": 0.1857,
+      "step": 18840
+    },
+    {
+      "epoch": 396.8421052631579,
+      "grad_norm": 0.5967190861701965,
+      "learning_rate": 1.8753959148919964e-05,
+      "loss": 0.1952,
+      "step": 18850
+    },
+    {
+      "epoch": 397.05263157894734,
+      "grad_norm": 0.4433288276195526,
+      "learning_rate": 1.8676091349225444e-05,
+      "loss": 0.1828,
+      "step": 18860
+    },
+    {
+      "epoch": 397.2631578947368,
+      "grad_norm": 0.6135048270225525,
+      "learning_rate": 1.8598368888541706e-05,
+      "loss": 0.1861,
+      "step": 18870
+    },
+    {
+      "epoch": 397.4736842105263,
+      "grad_norm": 0.55300372838974,
+      "learning_rate": 1.852079190577145e-05,
+      "loss": 0.1795,
+      "step": 18880
+    },
+    {
+      "epoch": 397.6842105263158,
+      "grad_norm": 0.5510004162788391,
+      "learning_rate": 1.8443360539557285e-05,
+      "loss": 0.1923,
+      "step": 18890
+    },
+    {
+      "epoch": 397.89473684210526,
+      "grad_norm": 0.5126235485076904,
+      "learning_rate": 1.8366074928281607e-05,
+      "loss": 0.1815,
+      "step": 18900
+    },
+    {
+      "epoch": 398.10526315789474,
+      "grad_norm": 0.5166956186294556,
+      "learning_rate": 1.8288935210066373e-05,
+      "loss": 0.1909,
+      "step": 18910
+    },
+    {
+      "epoch": 398.3157894736842,
+      "grad_norm": 0.5784156322479248,
+      "learning_rate": 1.8211941522772736e-05,
+      "loss": 0.1917,
+      "step": 18920
+    },
+    {
+      "epoch": 398.5263157894737,
+      "grad_norm": 0.4633670449256897,
+      "learning_rate": 1.8135094004000884e-05,
+      "loss": 0.1796,
+      "step": 18930
+    },
+    {
+      "epoch": 398.7368421052632,
+      "grad_norm": 0.5912382006645203,
+      "learning_rate": 1.805839279108984e-05,
+      "loss": 0.1799,
+      "step": 18940
+    },
+    {
+      "epoch": 398.94736842105266,
+      "grad_norm": 0.8275567889213562,
+      "learning_rate": 1.798183802111706e-05,
+      "loss": 0.1967,
+      "step": 18950
+    },
+    {
+      "epoch": 399.1578947368421,
+      "grad_norm": 0.5267394781112671,
+      "learning_rate": 1.7905429830898378e-05,
+      "loss": 0.1839,
+      "step": 18960
+    },
+    {
+      "epoch": 399.36842105263156,
+      "grad_norm": 0.5726824998855591,
+      "learning_rate": 1.782916835698758e-05,
+      "loss": 0.1909,
+      "step": 18970
+    },
+    {
+      "epoch": 399.57894736842104,
+      "grad_norm": 0.7753190398216248,
+      "learning_rate": 1.7753053735676317e-05,
+      "loss": 0.2027,
+      "step": 18980
+    },
+    {
+      "epoch": 399.7894736842105,
+      "grad_norm": 0.637991726398468,
+      "learning_rate": 1.7677086102993744e-05,
+      "loss": 0.1696,
+      "step": 18990
+    },
+    {
+      "epoch": 400.0,
+      "grad_norm": 0.7141334414482117,
+      "learning_rate": 1.7601265594706316e-05,
+      "loss": 0.1854,
+      "step": 19000
+    },
+    {
+      "epoch": 400.2105263157895,
+      "grad_norm": 0.8016759753227234,
+      "learning_rate": 1.752559234631762e-05,
+      "loss": 0.187,
+      "step": 19010
+    },
+    {
+      "epoch": 400.42105263157896,
+      "grad_norm": 0.5627360939979553,
+      "learning_rate": 1.7450066493067997e-05,
+      "loss": 0.1826,
+      "step": 19020
+    },
+    {
+      "epoch": 400.63157894736844,
+      "grad_norm": 0.6535181403160095,
+      "learning_rate": 1.7374688169934385e-05,
+      "loss": 0.1826,
+      "step": 19030
+    },
+    {
+      "epoch": 400.8421052631579,
+      "grad_norm": 0.6111413836479187,
+      "learning_rate": 1.7299457511630057e-05,
+      "loss": 0.1889,
+      "step": 19040
+    },
+    {
+      "epoch": 401.05263157894734,
+      "grad_norm": 0.5383073687553406,
+      "learning_rate": 1.722437465260445e-05,
+      "loss": 0.1904,
+      "step": 19050
+    },
+    {
+      "epoch": 401.2631578947368,
+      "grad_norm": 0.8181249499320984,
+      "learning_rate": 1.7149439727042736e-05,
+      "loss": 0.1775,
+      "step": 19060
+    },
+    {
+      "epoch": 401.4736842105263,
+      "grad_norm": 0.6523582339286804,
+      "learning_rate": 1.7074652868865814e-05,
+      "loss": 0.1846,
+      "step": 19070
+    },
+    {
+      "epoch": 401.6842105263158,
+      "grad_norm": 0.8838443160057068,
+      "learning_rate": 1.7000014211729964e-05,
+      "loss": 0.1955,
+      "step": 19080
+    },
+    {
+      "epoch": 401.89473684210526,
+      "grad_norm": 0.614313006401062,
+      "learning_rate": 1.692552388902653e-05,
+      "loss": 0.1855,
+      "step": 19090
+    },
+    {
+      "epoch": 402.10526315789474,
+      "grad_norm": 0.5090711116790771,
+      "learning_rate": 1.6851182033881795e-05,
+      "loss": 0.1891,
+      "step": 19100
+    },
+    {
+      "epoch": 402.3157894736842,
+      "grad_norm": 0.5899013876914978,
+      "learning_rate": 1.677698877915669e-05,
+      "loss": 0.1798,
+      "step": 19110
+    },
+    {
+      "epoch": 402.5263157894737,
+      "grad_norm": 0.550887405872345,
+      "learning_rate": 1.6702944257446627e-05,
+      "loss": 0.1776,
+      "step": 19120
+    },
+    {
+      "epoch": 402.7368421052632,
+      "grad_norm": 0.5447081923484802,
+      "learning_rate": 1.6629048601081167e-05,
+      "loss": 0.1982,
+      "step": 19130
+    },
+    {
+      "epoch": 402.94736842105266,
+      "grad_norm": 0.5017291903495789,
+      "learning_rate": 1.655530194212379e-05,
+      "loss": 0.181,
+      "step": 19140
+    },
+    {
+      "epoch": 403.1578947368421,
+      "grad_norm": 0.6135120391845703,
+      "learning_rate": 1.648170441237179e-05,
+      "loss": 0.1986,
+      "step": 19150
+    },
+    {
+      "epoch": 403.36842105263156,
+      "grad_norm": 0.5479750633239746,
+      "learning_rate": 1.640825614335586e-05,
+      "loss": 0.182,
+      "step": 19160
+    },
+    {
+      "epoch": 403.57894736842104,
+      "grad_norm": 0.486905962228775,
+      "learning_rate": 1.6334957266339933e-05,
+      "loss": 0.1747,
+      "step": 19170
+    },
+    {
+      "epoch": 403.7894736842105,
+      "grad_norm": 0.5777673125267029,
+      "learning_rate": 1.6261807912321037e-05,
+      "loss": 0.1829,
+      "step": 19180
+    },
+    {
+      "epoch": 404.0,
+      "grad_norm": 0.6578062772750854,
+      "learning_rate": 1.6188808212028916e-05,
+      "loss": 0.1882,
+      "step": 19190
+    },
+    {
+      "epoch": 404.2105263157895,
+      "grad_norm": 0.7375777959823608,
+      "learning_rate": 1.611595829592587e-05,
+      "loss": 0.1956,
+      "step": 19200
+    },
+    {
+      "epoch": 404.42105263157896,
+      "grad_norm": 0.5458284020423889,
+      "learning_rate": 1.6043258294206487e-05,
+      "loss": 0.1829,
+      "step": 19210
+    },
+    {
+      "epoch": 404.63157894736844,
+      "grad_norm": 0.4725562334060669,
+      "learning_rate": 1.5970708336797503e-05,
+      "loss": 0.1706,
+      "step": 19220
+    },
+    {
+      "epoch": 404.8421052631579,
+      "grad_norm": 0.6534218192100525,
+      "learning_rate": 1.589830855335742e-05,
+      "loss": 0.1896,
+      "step": 19230
+    },
+    {
+      "epoch": 405.05263157894734,
+      "grad_norm": 0.44919002056121826,
+      "learning_rate": 1.582605907327638e-05,
+      "loss": 0.1904,
+      "step": 19240
+    },
+    {
+      "epoch": 405.2631578947368,
+      "grad_norm": 0.5909680128097534,
+      "learning_rate": 1.5753960025675963e-05,
+      "loss": 0.1879,
+      "step": 19250
+    },
+    {
+      "epoch": 405.4736842105263,
+      "grad_norm": 0.6542550921440125,
+      "learning_rate": 1.5682011539408826e-05,
+      "loss": 0.1772,
+      "step": 19260
+    },
+    {
+      "epoch": 405.6842105263158,
+      "grad_norm": 0.5599654316902161,
+      "learning_rate": 1.561021374305859e-05,
+      "loss": 0.1792,
+      "step": 19270
+    },
+    {
+      "epoch": 405.89473684210526,
+      "grad_norm": 0.6703615188598633,
+      "learning_rate": 1.553856676493953e-05,
+      "loss": 0.2035,
+      "step": 19280
+    },
+    {
+      "epoch": 406.10526315789474,
+      "grad_norm": 0.6065393090248108,
+      "learning_rate": 1.5467070733096466e-05,
+      "loss": 0.1823,
+      "step": 19290
+    },
+    {
+      "epoch": 406.3157894736842,
+      "grad_norm": 0.6260120272636414,
+      "learning_rate": 1.5395725775304348e-05,
+      "loss": 0.1947,
+      "step": 19300
+    },
+    {
+      "epoch": 406.5263157894737,
+      "grad_norm": 0.5651423335075378,
+      "learning_rate": 1.5324532019068195e-05,
+      "loss": 0.1923,
+      "step": 19310
+    },
+    {
+      "epoch": 406.7368421052632,
+      "grad_norm": 0.5632277131080627,
+      "learning_rate": 1.5253489591622837e-05,
+      "loss": 0.1795,
+      "step": 19320
+    },
+    {
+      "epoch": 406.94736842105266,
+      "grad_norm": 0.5379475951194763,
+      "learning_rate": 1.5182598619932576e-05,
+      "loss": 0.1904,
+      "step": 19330
+    },
+    {
+      "epoch": 407.1578947368421,
+      "grad_norm": 0.5060853362083435,
+      "learning_rate": 1.511185923069105e-05,
+      "loss": 0.1758,
+      "step": 19340
+    },
+    {
+      "epoch": 407.36842105263156,
+      "grad_norm": 0.5877968668937683,
+      "learning_rate": 1.5041271550321079e-05,
+      "loss": 0.1813,
+      "step": 19350
+    },
+    {
+      "epoch": 407.57894736842104,
+      "grad_norm": 0.7655476927757263,
+      "learning_rate": 1.497083570497424e-05,
+      "loss": 0.1915,
+      "step": 19360
+    },
+    {
+      "epoch": 407.7894736842105,
+      "grad_norm": 0.5808361172676086,
+      "learning_rate": 1.4900551820530828e-05,
+      "loss": 0.1809,
+      "step": 19370
+    },
+    {
+      "epoch": 408.0,
+      "grad_norm": 0.5417425036430359,
+      "learning_rate": 1.4830420022599523e-05,
+      "loss": 0.191,
+      "step": 19380
+    },
+    {
+      "epoch": 408.2105263157895,
+      "grad_norm": 0.564896821975708,
+      "learning_rate": 1.4760440436517253e-05,
+      "loss": 0.1742,
+      "step": 19390
+    },
+    {
+      "epoch": 408.42105263157896,
+      "grad_norm": 0.42806559801101685,
+      "learning_rate": 1.4690613187348867e-05,
+      "loss": 0.1943,
+      "step": 19400
+    },
+    {
+      "epoch": 408.63157894736844,
+      "grad_norm": 0.5757007002830505,
+      "learning_rate": 1.4620938399886963e-05,
+      "loss": 0.1872,
+      "step": 19410
+    },
+    {
+      "epoch": 408.8421052631579,
+      "grad_norm": 0.5994120240211487,
+      "learning_rate": 1.4551416198651701e-05,
+      "loss": 0.1847,
+      "step": 19420
+    },
+    {
+      "epoch": 409.05263157894734,
+      "grad_norm": 0.6407523155212402,
+      "learning_rate": 1.448204670789054e-05,
+      "loss": 0.1928,
+      "step": 19430
+    },
+    {
+      "epoch": 409.2631578947368,
+      "grad_norm": 0.5221366882324219,
+      "learning_rate": 1.4412830051578009e-05,
+      "loss": 0.1834,
+      "step": 19440
+    },
+    {
+      "epoch": 409.4736842105263,
+      "grad_norm": 0.5237503051757812,
+      "learning_rate": 1.4343766353415444e-05,
+      "loss": 0.1758,
+      "step": 19450
+    },
+    {
+      "epoch": 409.6842105263158,
+      "grad_norm": 0.6575828790664673,
+      "learning_rate": 1.4274855736830938e-05,
+      "loss": 0.1866,
+      "step": 19460
+    },
+    {
+      "epoch": 409.89473684210526,
+      "grad_norm": 0.47361230850219727,
+      "learning_rate": 1.4206098324978912e-05,
+      "loss": 0.1845,
+      "step": 19470
+    },
+    {
+      "epoch": 410.10526315789474,
+      "grad_norm": 0.5174840688705444,
+      "learning_rate": 1.4137494240739979e-05,
+      "loss": 0.1986,
+      "step": 19480
+    },
+    {
+      "epoch": 410.3157894736842,
+      "grad_norm": 0.48306819796562195,
+      "learning_rate": 1.4069043606720811e-05,
+      "loss": 0.1861,
+      "step": 19490
+    },
+    {
+      "epoch": 410.5263157894737,
+      "grad_norm": 0.4556773602962494,
+      "learning_rate": 1.4000746545253774e-05,
+      "loss": 0.1815,
+      "step": 19500
+    },
+    {
+      "epoch": 410.7368421052632,
+      "grad_norm": 0.5436505079269409,
+      "learning_rate": 1.3932603178396752e-05,
+      "loss": 0.1896,
+      "step": 19510
+    },
+    {
+      "epoch": 410.94736842105266,
+      "grad_norm": 0.6052523851394653,
+      "learning_rate": 1.3864613627933042e-05,
+      "loss": 0.181,
+      "step": 19520
+    },
+    {
+      "epoch": 411.1578947368421,
+      "grad_norm": 0.5917760729789734,
+      "learning_rate": 1.3796778015370959e-05,
+      "loss": 0.1892,
+      "step": 19530
+    },
+    {
+      "epoch": 411.36842105263156,
+      "grad_norm": 0.5070799589157104,
+      "learning_rate": 1.372909646194377e-05,
+      "loss": 0.1841,
+      "step": 19540
+    },
+    {
+      "epoch": 411.57894736842104,
+      "grad_norm": 0.5299696922302246,
+      "learning_rate": 1.366156908860936e-05,
+      "loss": 0.1858,
+      "step": 19550
+    },
+    {
+      "epoch": 411.7894736842105,
+      "grad_norm": 0.5153812766075134,
+      "learning_rate": 1.359419601605012e-05,
+      "loss": 0.1798,
+      "step": 19560
+    },
+    {
+      "epoch": 412.0,
+      "grad_norm": 0.7871899604797363,
+      "learning_rate": 1.3526977364672644e-05,
+      "loss": 0.1864,
+      "step": 19570
+    },
+    {
+      "epoch": 412.2105263157895,
+      "grad_norm": 0.6095196008682251,
+      "learning_rate": 1.3459913254607537e-05,
+      "loss": 0.1817,
+      "step": 19580
+    },
+    {
+      "epoch": 412.42105263157896,
+      "grad_norm": 0.5391228795051575,
+      "learning_rate": 1.3393003805709281e-05,
+      "loss": 0.182,
+      "step": 19590
+    },
+    {
+      "epoch": 412.63157894736844,
+      "grad_norm": 0.9291274547576904,
+      "learning_rate": 1.332624913755588e-05,
+      "loss": 0.1953,
+      "step": 19600
+    },
+    {
+      "epoch": 412.8421052631579,
+      "grad_norm": 0.5795119404792786,
+      "learning_rate": 1.3259649369448768e-05,
+      "loss": 0.1871,
+      "step": 19610
+    },
+    {
+      "epoch": 413.05263157894734,
+      "grad_norm": 0.7376877069473267,
+      "learning_rate": 1.3193204620412481e-05,
+      "loss": 0.199,
+      "step": 19620
+    },
+    {
+      "epoch": 413.2631578947368,
+      "grad_norm": 0.6380026936531067,
+      "learning_rate": 1.312691500919463e-05,
+      "loss": 0.1857,
+      "step": 19630
+    },
+    {
+      "epoch": 413.4736842105263,
+      "grad_norm": 0.5857008099555969,
+      "learning_rate": 1.3060780654265447e-05,
+      "loss": 0.2078,
+      "step": 19640
+    },
+    {
+      "epoch": 413.6842105263158,
+      "grad_norm": 0.6055216789245605,
+      "learning_rate": 1.299480167381778e-05,
+      "loss": 0.1767,
+      "step": 19650
+    },
+    {
+      "epoch": 413.89473684210526,
+      "grad_norm": 0.714047372341156,
+      "learning_rate": 1.2928978185766727e-05,
+      "loss": 0.1731,
+      "step": 19660
+    },
+    {
+      "epoch": 414.10526315789474,
+      "grad_norm": 0.767358124256134,
+      "learning_rate": 1.2863310307749577e-05,
+      "loss": 0.2019,
+      "step": 19670
+    },
+    {
+      "epoch": 414.3157894736842,
+      "grad_norm": 0.5826098322868347,
+      "learning_rate": 1.2797798157125441e-05,
+      "loss": 0.1889,
+      "step": 19680
+    },
+    {
+      "epoch": 414.5263157894737,
+      "grad_norm": 0.46154114603996277,
+      "learning_rate": 1.2732441850975185e-05,
+      "loss": 0.1743,
+      "step": 19690
+    },
+    {
+      "epoch": 414.7368421052632,
+      "grad_norm": 0.5386494398117065,
+      "learning_rate": 1.2667241506101124e-05,
+      "loss": 0.1811,
+      "step": 19700
+    },
+    {
+      "epoch": 414.94736842105266,
+      "grad_norm": 0.590414822101593,
+      "learning_rate": 1.2602197239026814e-05,
+      "loss": 0.1926,
+      "step": 19710
+    },
+    {
+      "epoch": 415.1578947368421,
+      "grad_norm": 0.722994327545166,
+      "learning_rate": 1.2537309165996913e-05,
+      "loss": 0.1967,
+      "step": 19720
+    },
+    {
+      "epoch": 415.36842105263156,
+      "grad_norm": 0.5225125551223755,
+      "learning_rate": 1.247257740297696e-05,
+      "loss": 0.1861,
+      "step": 19730
+    },
+    {
+      "epoch": 415.57894736842104,
+      "grad_norm": 0.4299336075782776,
+      "learning_rate": 1.2408002065653091e-05,
+      "loss": 0.188,
+      "step": 19740
+    },
+    {
+      "epoch": 415.7894736842105,
+      "grad_norm": 0.5853255987167358,
+      "learning_rate": 1.234358326943188e-05,
+      "loss": 0.1818,
+      "step": 19750
+    },
+    {
+      "epoch": 416.0,
+      "grad_norm": 0.5214622020721436,
+      "learning_rate": 1.2279321129440202e-05,
+      "loss": 0.1748,
+      "step": 19760
+    },
+    {
+      "epoch": 416.2105263157895,
+      "grad_norm": 0.6937418580055237,
+      "learning_rate": 1.221521576052489e-05,
+      "loss": 0.1787,
+      "step": 19770
+    },
+    {
+      "epoch": 416.42105263157896,
+      "grad_norm": 0.6764062643051147,
+      "learning_rate": 1.2151267277252665e-05,
+      "loss": 0.1894,
+      "step": 19780
+    },
+    {
+      "epoch": 416.63157894736844,
+      "grad_norm": 0.5873314738273621,
+      "learning_rate": 1.2087475793909798e-05,
+      "loss": 0.1764,
+      "step": 19790
+    },
+    {
+      "epoch": 416.8421052631579,
+      "grad_norm": 0.9404758810997009,
+      "learning_rate": 1.2023841424502048e-05,
+      "loss": 0.209,
+      "step": 19800
+    },
+    {
+      "epoch": 417.05263157894734,
+      "grad_norm": 0.6922588348388672,
+      "learning_rate": 1.1960364282754344e-05,
+      "loss": 0.183,
+      "step": 19810
+    },
+    {
+      "epoch": 417.2631578947368,
+      "grad_norm": 0.8207189440727234,
+      "learning_rate": 1.1897044482110586e-05,
+      "loss": 0.1839,
+      "step": 19820
+    },
+    {
+      "epoch": 417.4736842105263,
+      "grad_norm": 0.5307655930519104,
+      "learning_rate": 1.1833882135733599e-05,
+      "loss": 0.1842,
+      "step": 19830
+    },
+    {
+      "epoch": 417.6842105263158,
+      "grad_norm": 0.6499212980270386,
+      "learning_rate": 1.1770877356504683e-05,
+      "loss": 0.184,
+      "step": 19840
+    },
+    {
+      "epoch": 417.89473684210526,
+      "grad_norm": 0.5272055268287659,
+      "learning_rate": 1.1708030257023595e-05,
+      "loss": 0.1798,
+      "step": 19850
+    },
+    {
+      "epoch": 418.10526315789474,
+      "grad_norm": 0.5772253274917603,
+      "learning_rate": 1.164534094960833e-05,
+      "loss": 0.169,
+      "step": 19860
+    },
+    {
+      "epoch": 418.3157894736842,
+      "grad_norm": 0.6082757711410522,
+      "learning_rate": 1.1582809546294816e-05,
+      "loss": 0.1939,
+      "step": 19870
+    },
+    {
+      "epoch": 418.5263157894737,
+      "grad_norm": 0.8696257472038269,
+      "learning_rate": 1.15204361588368e-05,
+      "loss": 0.1822,
+      "step": 19880
+    },
+    {
+      "epoch": 418.7368421052632,
+      "grad_norm": 0.7336968183517456,
+      "learning_rate": 1.1458220898705663e-05,
+      "loss": 0.1893,
+      "step": 19890
+    },
+    {
+      "epoch": 418.94736842105266,
+      "grad_norm": 0.7387612462043762,
+      "learning_rate": 1.1396163877090148e-05,
+      "loss": 0.1973,
+      "step": 19900
+    },
+    {
+      "epoch": 419.1578947368421,
+      "grad_norm": 0.660318911075592,
+      "learning_rate": 1.1334265204896233e-05,
+      "loss": 0.1809,
+      "step": 19910
+    },
+    {
+      "epoch": 419.36842105263156,
+      "grad_norm": 0.6014663577079773,
+      "learning_rate": 1.1272524992746846e-05,
+      "loss": 0.192,
+      "step": 19920
+    },
+    {
+      "epoch": 419.57894736842104,
+      "grad_norm": 0.6727678775787354,
+      "learning_rate": 1.1210943350981806e-05,
+      "loss": 0.1865,
+      "step": 19930
+    },
+    {
+      "epoch": 419.7894736842105,
+      "grad_norm": 0.5072178244590759,
+      "learning_rate": 1.1149520389657463e-05,
+      "loss": 0.176,
+      "step": 19940
+    },
+    {
+      "epoch": 420.0,
+      "grad_norm": 0.5579133033752441,
+      "learning_rate": 1.1088256218546611e-05,
+      "loss": 0.1804,
+      "step": 19950
+    },
+    {
+      "epoch": 420.2105263157895,
+      "grad_norm": 0.46790534257888794,
+      "learning_rate": 1.1027150947138232e-05,
+      "loss": 0.1986,
+      "step": 19960
+    },
+    {
+      "epoch": 420.42105263157896,
+      "grad_norm": 0.6098372936248779,
+      "learning_rate": 1.0966204684637405e-05,
+      "loss": 0.1805,
+      "step": 19970
+    },
+    {
+      "epoch": 420.63157894736844,
+      "grad_norm": 0.734483003616333,
+      "learning_rate": 1.0905417539964946e-05,
+      "loss": 0.1899,
+      "step": 19980
+    },
+    {
+      "epoch": 420.8421052631579,
+      "grad_norm": 0.5049176216125488,
+      "learning_rate": 1.0844789621757335e-05,
+      "loss": 0.173,
+      "step": 19990
+    },
+    {
+      "epoch": 421.05263157894734,
+      "grad_norm": 0.5509609580039978,
+      "learning_rate": 1.0784321038366529e-05,
+      "loss": 0.186,
+      "step": 20000
+    },
+    {
+      "epoch": 421.2631578947368,
+      "grad_norm": 0.6344820261001587,
+      "learning_rate": 1.0724011897859653e-05,
+      "loss": 0.1726,
+      "step": 20010
+    },
+    {
+      "epoch": 421.4736842105263,
+      "grad_norm": 0.5593404769897461,
+      "learning_rate": 1.0663862308018924e-05,
+      "loss": 0.1833,
+      "step": 20020
+    },
+    {
+      "epoch": 421.6842105263158,
+      "grad_norm": 0.6860533356666565,
+      "learning_rate": 1.060387237634145e-05,
+      "loss": 0.1867,
+      "step": 20030
+    },
+    {
+      "epoch": 421.89473684210526,
+      "grad_norm": 0.5627481937408447,
+      "learning_rate": 1.054404221003894e-05,
+      "loss": 0.1808,
+      "step": 20040
+    },
+    {
+      "epoch": 422.10526315789474,
+      "grad_norm": 0.5001289248466492,
+      "learning_rate": 1.0484371916037606e-05,
+      "loss": 0.1808,
+      "step": 20050
+    },
+    {
+      "epoch": 422.3157894736842,
+      "grad_norm": 0.5672506093978882,
+      "learning_rate": 1.0424861600977898e-05,
+      "loss": 0.174,
+      "step": 20060
+    },
+    {
+      "epoch": 422.5263157894737,
+      "grad_norm": 0.6734273433685303,
+      "learning_rate": 1.0365511371214465e-05,
+      "loss": 0.192,
+      "step": 20070
+    },
+    {
+      "epoch": 422.7368421052632,
+      "grad_norm": 0.4520343840122223,
+      "learning_rate": 1.0306321332815761e-05,
+      "loss": 0.1846,
+      "step": 20080
+    },
+    {
+      "epoch": 422.94736842105266,
+      "grad_norm": 0.6919784545898438,
+      "learning_rate": 1.0247291591563956e-05,
+      "loss": 0.2028,
+      "step": 20090
+    },
+    {
+      "epoch": 423.1578947368421,
+      "grad_norm": 0.7424619197845459,
+      "learning_rate": 1.018842225295481e-05,
+      "loss": 0.187,
+      "step": 20100
+    },
+    {
+      "epoch": 423.36842105263156,
+      "grad_norm": 0.6489419937133789,
+      "learning_rate": 1.0129713422197362e-05,
+      "loss": 0.1847,
+      "step": 20110
+    },
+    {
+      "epoch": 423.57894736842104,
+      "grad_norm": 0.6978447437286377,
+      "learning_rate": 1.0071165204213794e-05,
+      "loss": 0.1783,
+      "step": 20120
+    },
+    {
+      "epoch": 423.7894736842105,
+      "grad_norm": 0.6544562578201294,
+      "learning_rate": 1.0012777703639275e-05,
+      "loss": 0.1866,
+      "step": 20130
+    },
+    {
+      "epoch": 424.0,
+      "grad_norm": 1.0255416631698608,
+      "learning_rate": 9.954551024821767e-06,
+      "loss": 0.1943,
+      "step": 20140
+    },
+    {
+      "epoch": 424.2105263157895,
+      "grad_norm": 0.44083118438720703,
+      "learning_rate": 9.896485271821755e-06,
+      "loss": 0.1805,
+      "step": 20150
+    },
+    {
+      "epoch": 424.42105263157896,
+      "grad_norm": 0.4890379011631012,
+      "learning_rate": 9.838580548412135e-06,
+      "loss": 0.1836,
+      "step": 20160
+    },
+    {
+      "epoch": 424.63157894736844,
+      "grad_norm": 0.5661543607711792,
+      "learning_rate": 9.780836958078087e-06,
+      "loss": 0.189,
+      "step": 20170
+    },
+    {
+      "epoch": 424.8421052631579,
+      "grad_norm": 0.555959939956665,
+      "learning_rate": 9.72325460401674e-06,
+      "loss": 0.1852,
+      "step": 20180
+    },
+    {
+      "epoch": 425.05263157894734,
+      "grad_norm": 0.6740918755531311,
+      "learning_rate": 9.665833589137085e-06,
+      "loss": 0.1896,
+      "step": 20190
+    },
+    {
+      "epoch": 425.2631578947368,
+      "grad_norm": 0.5326783061027527,
+      "learning_rate": 9.608574016059823e-06,
+      "loss": 0.1747,
+      "step": 20200
+    },
+    {
+      "epoch": 425.4736842105263,
+      "grad_norm": 0.6140498518943787,
+      "learning_rate": 9.551475987117065e-06,
+      "loss": 0.1852,
+      "step": 20210
+    },
+    {
+      "epoch": 425.6842105263158,
+      "grad_norm": 0.7099624276161194,
+      "learning_rate": 9.49453960435226e-06,
+      "loss": 0.1775,
+      "step": 20220
+    },
+    {
+      "epoch": 425.89473684210526,
+      "grad_norm": 0.6613169312477112,
+      "learning_rate": 9.437764969519935e-06,
+      "loss": 0.1855,
+      "step": 20230
+    },
+    {
+      "epoch": 426.10526315789474,
+      "grad_norm": 0.5651838779449463,
+      "learning_rate": 9.381152184085595e-06,
+      "loss": 0.1926,
+      "step": 20240
+    },
+    {
+      "epoch": 426.3157894736842,
+      "grad_norm": 0.5982670783996582,
+      "learning_rate": 9.32470134922544e-06,
+      "loss": 0.1869,
+      "step": 20250
+    },
+    {
+      "epoch": 426.5263157894737,
+      "grad_norm": 0.5540538430213928,
+      "learning_rate": 9.26841256582629e-06,
+      "loss": 0.1862,
+      "step": 20260
+    },
+    {
+      "epoch": 426.7368421052632,
+      "grad_norm": 0.7269121408462524,
+      "learning_rate": 9.212285934485332e-06,
+      "loss": 0.1826,
+      "step": 20270
+    },
+    {
+      "epoch": 426.94736842105266,
+      "grad_norm": 0.9267326593399048,
+      "learning_rate": 9.15632155550994e-06,
+      "loss": 0.1919,
+      "step": 20280
+    },
+    {
+      "epoch": 427.1578947368421,
+      "grad_norm": 0.600455641746521,
+      "learning_rate": 9.10051952891754e-06,
+      "loss": 0.1857,
+      "step": 20290
+    },
+    {
+      "epoch": 427.36842105263156,
+      "grad_norm": 0.5266342759132385,
+      "learning_rate": 9.044879954435381e-06,
+      "loss": 0.1725,
+      "step": 20300
+    },
+    {
+      "epoch": 427.57894736842104,
+      "grad_norm": 0.8109986186027527,
+      "learning_rate": 8.989402931500434e-06,
+      "loss": 0.1943,
+      "step": 20310
+    },
+    {
+      "epoch": 427.7894736842105,
+      "grad_norm": 0.5752395391464233,
+      "learning_rate": 8.934088559259135e-06,
+      "loss": 0.1944,
+      "step": 20320
+    },
+    {
+      "epoch": 428.0,
+      "grad_norm": 0.6188866496086121,
+      "learning_rate": 8.878936936567195e-06,
+      "loss": 0.1817,
+      "step": 20330
+    },
+    {
+      "epoch": 428.2105263157895,
+      "grad_norm": 0.6044108271598816,
+      "learning_rate": 8.823948161989549e-06,
+      "loss": 0.1799,
+      "step": 20340
+    },
+    {
+      "epoch": 428.42105263157896,
+      "grad_norm": 0.5862536430358887,
+      "learning_rate": 8.76912233380005e-06,
+      "loss": 0.1798,
+      "step": 20350
+    },
+    {
+      "epoch": 428.63157894736844,
+      "grad_norm": 0.6204822063446045,
+      "learning_rate": 8.714459549981302e-06,
+      "loss": 0.1862,
+      "step": 20360
+    },
+    {
+      "epoch": 428.8421052631579,
+      "grad_norm": 0.5744066834449768,
+      "learning_rate": 8.65995990822459e-06,
+      "loss": 0.1869,
+      "step": 20370
+    },
+    {
+      "epoch": 429.05263157894734,
+      "grad_norm": 0.6363000273704529,
+      "learning_rate": 8.60562350592964e-06,
+      "loss": 0.1959,
+      "step": 20380
+    },
+    {
+      "epoch": 429.2631578947368,
+      "grad_norm": 0.5891885757446289,
+      "learning_rate": 8.551450440204379e-06,
+      "loss": 0.1804,
+      "step": 20390
+    },
+    {
+      "epoch": 429.4736842105263,
+      "grad_norm": 0.5133945345878601,
+      "learning_rate": 8.497440807864853e-06,
+      "loss": 0.1825,
+      "step": 20400
+    },
+    {
+      "epoch": 429.6842105263158,
+      "grad_norm": 0.578368067741394,
+      "learning_rate": 8.443594705435054e-06,
+      "loss": 0.1823,
+      "step": 20410
+    },
+    {
+      "epoch": 429.89473684210526,
+      "grad_norm": 0.6351949572563171,
+      "learning_rate": 8.389912229146702e-06,
+      "loss": 0.1895,
+      "step": 20420
+    },
+    {
+      "epoch": 430.10526315789474,
+      "grad_norm": 0.731504499912262,
+      "learning_rate": 8.336393474939042e-06,
+      "loss": 0.1819,
+      "step": 20430
+    },
+    {
+      "epoch": 430.3157894736842,
+      "grad_norm": 0.5696659088134766,
+      "learning_rate": 8.28303853845882e-06,
+      "loss": 0.1879,
+      "step": 20440
+    },
+    {
+      "epoch": 430.5263157894737,
+      "grad_norm": 0.5984363555908203,
+      "learning_rate": 8.22984751505993e-06,
+      "loss": 0.1897,
+      "step": 20450
+    },
+    {
+      "epoch": 430.7368421052632,
+      "grad_norm": 0.7230535745620728,
+      "learning_rate": 8.17682049980334e-06,
+      "loss": 0.1815,
+      "step": 20460
+    },
+    {
+      "epoch": 430.94736842105266,
+      "grad_norm": 0.4991123378276825,
+      "learning_rate": 8.123957587456966e-06,
+      "loss": 0.1735,
+      "step": 20470
+    },
+    {
+      "epoch": 431.1578947368421,
+      "grad_norm": 0.5288248062133789,
+      "learning_rate": 8.07125887249537e-06,
+      "loss": 0.1841,
+      "step": 20480
+    },
+    {
+      "epoch": 431.36842105263156,
+      "grad_norm": 0.6192090511322021,
+      "learning_rate": 8.018724449099724e-06,
+      "loss": 0.1746,
+      "step": 20490
+    },
+    {
+      "epoch": 431.57894736842104,
+      "grad_norm": 0.7012197971343994,
+      "learning_rate": 7.966354411157529e-06,
+      "loss": 0.1866,
+      "step": 20500
+    },
+    {
+      "epoch": 431.7894736842105,
+      "grad_norm": 0.7939274311065674,
+      "learning_rate": 7.914148852262582e-06,
+      "loss": 0.1971,
+      "step": 20510
+    },
+    {
+      "epoch": 432.0,
+      "grad_norm": 0.5685390830039978,
+      "learning_rate": 7.862107865714641e-06,
+      "loss": 0.1882,
+      "step": 20520
+    },
+    {
+      "epoch": 432.2105263157895,
+      "grad_norm": 0.6198951601982117,
+      "learning_rate": 7.810231544519386e-06,
+      "loss": 0.1767,
+      "step": 20530
+    },
+    {
+      "epoch": 432.42105263157896,
+      "grad_norm": 0.4967138469219208,
+      "learning_rate": 7.758519981388257e-06,
+      "loss": 0.1871,
+      "step": 20540
+    },
+    {
+      "epoch": 432.63157894736844,
+      "grad_norm": 0.6240216493606567,
+      "learning_rate": 7.70697326873816e-06,
+      "loss": 0.1821,
+      "step": 20550
+    },
+    {
+      "epoch": 432.8421052631579,
+      "grad_norm": 0.43565139174461365,
+      "learning_rate": 7.65559149869144e-06,
+      "loss": 0.1816,
+      "step": 20560
+    },
+    {
+      "epoch": 433.05263157894734,
+      "grad_norm": 0.5389068722724915,
+      "learning_rate": 7.604374763075639e-06,
+      "loss": 0.1962,
+      "step": 20570
+    },
+    {
+      "epoch": 433.2631578947368,
+      "grad_norm": 0.4656810462474823,
+      "learning_rate": 7.553323153423409e-06,
+      "loss": 0.1835,
+      "step": 20580
+    },
+    {
+      "epoch": 433.4736842105263,
+      "grad_norm": 0.5822871327400208,
+      "learning_rate": 7.502436760972198e-06,
+      "loss": 0.1879,
+      "step": 20590
+    },
+    {
+      "epoch": 433.6842105263158,
+      "grad_norm": 0.6489611864089966,
+      "learning_rate": 7.451715676664284e-06,
+      "loss": 0.1833,
+      "step": 20600
+    },
+    {
+      "epoch": 433.89473684210526,
+      "grad_norm": 0.60507732629776,
+      "learning_rate": 7.401159991146445e-06,
+      "loss": 0.1779,
+      "step": 20610
+    },
+    {
+      "epoch": 434.10526315789474,
+      "grad_norm": 0.5765190720558167,
+      "learning_rate": 7.3507697947699075e-06,
+      "loss": 0.1823,
+      "step": 20620
+    },
+    {
+      "epoch": 434.3157894736842,
+      "grad_norm": 0.5958349108695984,
+      "learning_rate": 7.30054517759009e-06,
+      "loss": 0.1959,
+      "step": 20630
+    },
+    {
+      "epoch": 434.5263157894737,
+      "grad_norm": 0.5677564144134521,
+      "learning_rate": 7.250486229366582e-06,
+      "loss": 0.1747,
+      "step": 20640
+    },
+    {
+      "epoch": 434.7368421052632,
+      "grad_norm": 0.4779718816280365,
+      "learning_rate": 7.2005930395627975e-06,
+      "loss": 0.1864,
+      "step": 20650
+    },
+    {
+      "epoch": 434.94736842105266,
+      "grad_norm": 0.6373128294944763,
+      "learning_rate": 7.1508656973459655e-06,
+      "loss": 0.1881,
+      "step": 20660
+    },
+    {
+      "epoch": 435.1578947368421,
+      "grad_norm": 0.5015693306922913,
+      "learning_rate": 7.101304291586897e-06,
+      "loss": 0.1923,
+      "step": 20670
+    },
+    {
+      "epoch": 435.36842105263156,
+      "grad_norm": 0.9822379946708679,
+      "learning_rate": 7.051908910859884e-06,
+      "loss": 0.19,
+      "step": 20680
+    },
+    {
+      "epoch": 435.57894736842104,
+      "grad_norm": 0.5332909822463989,
+      "learning_rate": 7.002679643442478e-06,
+      "loss": 0.1895,
+      "step": 20690
+    },
+    {
+      "epoch": 435.7894736842105,
+      "grad_norm": 0.6863967776298523,
+      "learning_rate": 6.953616577315336e-06,
+      "loss": 0.1824,
+      "step": 20700
+    },
+    {
+      "epoch": 436.0,
+      "grad_norm": 0.8219735026359558,
+      "learning_rate": 6.904719800162141e-06,
+      "loss": 0.1844,
+      "step": 20710
+    },
+    {
+      "epoch": 436.2105263157895,
+      "grad_norm": 0.43607479333877563,
+      "learning_rate": 6.855989399369345e-06,
+      "loss": 0.1911,
+      "step": 20720
+    },
+    {
+      "epoch": 436.42105263157896,
+      "grad_norm": 0.5909510254859924,
+      "learning_rate": 6.807425462026096e-06,
+      "loss": 0.1902,
+      "step": 20730
+    },
+    {
+      "epoch": 436.63157894736844,
+      "grad_norm": 0.8456542491912842,
+      "learning_rate": 6.75902807492399e-06,
+      "loss": 0.1841,
+      "step": 20740
+    },
+    {
+      "epoch": 436.8421052631579,
+      "grad_norm": 0.558668315410614,
+      "learning_rate": 6.71079732455705e-06,
+      "loss": 0.1807,
+      "step": 20750
+    },
+    {
+      "epoch": 437.05263157894734,
+      "grad_norm": 0.581362783908844,
+      "learning_rate": 6.662733297121415e-06,
+      "loss": 0.1739,
+      "step": 20760
+    },
+    {
+      "epoch": 437.2631578947368,
+      "grad_norm": 0.5437993407249451,
+      "learning_rate": 6.614836078515285e-06,
+      "loss": 0.1792,
+      "step": 20770
+    },
+    {
+      "epoch": 437.4736842105263,
+      "grad_norm": 0.824394166469574,
+      "learning_rate": 6.5671057543387985e-06,
+      "loss": 0.197,
+      "step": 20780
+    },
+    {
+      "epoch": 437.6842105263158,
+      "grad_norm": 0.4408915042877197,
+      "learning_rate": 6.519542409893753e-06,
+      "loss": 0.1877,
+      "step": 20790
+    },
+    {
+      "epoch": 437.89473684210526,
+      "grad_norm": 0.5344646573066711,
+      "learning_rate": 6.472146130183554e-06,
+      "loss": 0.1764,
+      "step": 20800
+    },
+    {
+      "epoch": 438.10526315789474,
+      "grad_norm": 0.5722460150718689,
+      "learning_rate": 6.424916999913055e-06,
+      "loss": 0.1904,
+      "step": 20810
+    },
+    {
+      "epoch": 438.3157894736842,
+      "grad_norm": 0.6273982524871826,
+      "learning_rate": 6.377855103488373e-06,
+      "loss": 0.187,
+      "step": 20820
+    },
+    {
+      "epoch": 438.5263157894737,
+      "grad_norm": 0.49412593245506287,
+      "learning_rate": 6.330960525016716e-06,
+      "loss": 0.1766,
+      "step": 20830
+    },
+    {
+      "epoch": 438.7368421052632,
+      "grad_norm": 0.5377347469329834,
+      "learning_rate": 6.284233348306334e-06,
+      "loss": 0.1842,
+      "step": 20840
+    },
+    {
+      "epoch": 438.94736842105266,
+      "grad_norm": 0.6035183668136597,
+      "learning_rate": 6.237673656866238e-06,
+      "loss": 0.1907,
+      "step": 20850
+    },
+    {
+      "epoch": 439.1578947368421,
+      "grad_norm": 0.719294548034668,
+      "learning_rate": 6.19128153390619e-06,
+      "loss": 0.1936,
+      "step": 20860
+    },
+    {
+      "epoch": 439.36842105263156,
+      "grad_norm": 0.5597818493843079,
+      "learning_rate": 6.145057062336379e-06,
+      "loss": 0.1819,
+      "step": 20870
+    },
+    {
+      "epoch": 439.57894736842104,
+      "grad_norm": 0.5979058742523193,
+      "learning_rate": 6.099000324767479e-06,
+      "loss": 0.1813,
+      "step": 20880
+    },
+    {
+      "epoch": 439.7894736842105,
+      "grad_norm": 0.5603199601173401,
+      "learning_rate": 6.053111403510336e-06,
+      "loss": 0.1794,
+      "step": 20890
+    },
+    {
+      "epoch": 440.0,
+      "grad_norm": 1.0742629766464233,
+      "learning_rate": 6.011954925113683e-06,
+      "loss": 0.1906,
+      "step": 20900
+    },
+    {
+      "epoch": 440.2105263157895,
+      "grad_norm": 0.6296944618225098,
+      "learning_rate": 5.966385080540993e-06,
+      "loss": 0.1886,
+      "step": 20910
+    },
+    {
+      "epoch": 440.42105263157896,
+      "grad_norm": 0.5945459008216858,
+      "learning_rate": 5.920983289285009e-06,
+      "loss": 0.1768,
+      "step": 20920
+    },
+    {
+      "epoch": 440.63157894736844,
+      "grad_norm": 0.504224419593811,
+      "learning_rate": 5.8757496324860715e-06,
+      "loss": 0.1877,
+      "step": 20930
+    },
+    {
+      "epoch": 440.8421052631579,
+      "grad_norm": 0.6713933944702148,
+      "learning_rate": 5.8306841909840816e-06,
+      "loss": 0.1903,
+      "step": 20940
+    },
+    {
+      "epoch": 441.05263157894734,
+      "grad_norm": 0.5135478973388672,
+      "learning_rate": 5.78578704531828e-06,
+      "loss": 0.1901,
+      "step": 20950
+    },
+    {
+      "epoch": 441.2631578947368,
+      "grad_norm": 0.6023509502410889,
+      "learning_rate": 5.74105827572714e-06,
+      "loss": 0.1932,
+      "step": 20960
+    },
+    {
+      "epoch": 441.4736842105263,
+      "grad_norm": 0.5241458415985107,
+      "learning_rate": 5.696497962148218e-06,
+      "loss": 0.1797,
+      "step": 20970
+    },
+    {
+      "epoch": 441.6842105263158,
+      "grad_norm": 0.5880340933799744,
+      "learning_rate": 5.652106184218042e-06,
+      "loss": 0.1931,
+      "step": 20980
+    },
+    {
+      "epoch": 441.89473684210526,
+      "grad_norm": 0.7315283417701721,
+      "learning_rate": 5.607883021271898e-06,
+      "loss": 0.1866,
+      "step": 20990
+    },
+    {
+      "epoch": 442.10526315789474,
+      "grad_norm": 0.6057814359664917,
+      "learning_rate": 5.5638285523437525e-06,
+      "loss": 0.1749,
+      "step": 21000
+    },
+    {
+      "epoch": 442.3157894736842,
+      "grad_norm": 0.7352088689804077,
+      "learning_rate": 5.519942856166105e-06,
+      "loss": 0.1888,
+      "step": 21010
+    },
+    {
+      "epoch": 442.5263157894737,
+      "grad_norm": 0.7723200917243958,
+      "learning_rate": 5.4762260111697714e-06,
+      "loss": 0.1831,
+      "step": 21020
+    },
+    {
+      "epoch": 442.7368421052632,
+      "grad_norm": 0.6795921921730042,
+      "learning_rate": 5.432678095483878e-06,
+      "loss": 0.1801,
+      "step": 21030
+    },
+    {
+      "epoch": 442.94736842105266,
+      "grad_norm": 0.607151448726654,
+      "learning_rate": 5.389299186935592e-06,
+      "loss": 0.1814,
+      "step": 21040
+    },
+    {
+      "epoch": 443.1578947368421,
+      "grad_norm": 0.6719841361045837,
+      "learning_rate": 5.3460893630500664e-06,
+      "loss": 0.1824,
+      "step": 21050
+    },
+    {
+      "epoch": 443.36842105263156,
+      "grad_norm": 0.6265808343887329,
+      "learning_rate": 5.3030487010502615e-06,
+      "loss": 0.1916,
+      "step": 21060
+    },
+    {
+      "epoch": 443.57894736842104,
+      "grad_norm": 0.6264698505401611,
+      "learning_rate": 5.260177277856804e-06,
+      "loss": 0.1817,
+      "step": 21070
+    },
+    {
+      "epoch": 443.7894736842105,
+      "grad_norm": 0.5315179824829102,
+      "learning_rate": 5.217475170087893e-06,
+      "loss": 0.1863,
+      "step": 21080
+    },
+    {
+      "epoch": 444.0,
+      "grad_norm": 0.7230284214019775,
+      "learning_rate": 5.174942454059128e-06,
+      "loss": 0.1912,
+      "step": 21090
+    },
+    {
+      "epoch": 444.2105263157895,
+      "grad_norm": 0.6552962064743042,
+      "learning_rate": 5.132579205783339e-06,
+      "loss": 0.182,
+      "step": 21100
+    },
+    {
+      "epoch": 444.42105263157896,
+      "grad_norm": 0.6705915331840515,
+      "learning_rate": 5.0903855009705514e-06,
+      "loss": 0.1784,
+      "step": 21110
+    },
+    {
+      "epoch": 444.63157894736844,
+      "grad_norm": 0.6432520747184753,
+      "learning_rate": 5.048361415027736e-06,
+      "loss": 0.1842,
+      "step": 21120
+    },
+    {
+      "epoch": 444.8421052631579,
+      "grad_norm": 0.5475180745124817,
+      "learning_rate": 5.0065070230587485e-06,
+      "loss": 0.2005,
+      "step": 21130
+    },
+    {
+      "epoch": 445.05263157894734,
+      "grad_norm": 0.7086638808250427,
+      "learning_rate": 4.964822399864189e-06,
+      "loss": 0.1755,
+      "step": 21140
+    },
+    {
+      "epoch": 445.2631578947368,
+      "grad_norm": 0.4936695098876953,
+      "learning_rate": 4.92330761994122e-06,
+      "loss": 0.1741,
+      "step": 21150
+    },
+    {
+      "epoch": 445.4736842105263,
+      "grad_norm": 0.5482287406921387,
+      "learning_rate": 4.8819627574835045e-06,
+      "loss": 0.185,
+      "step": 21160
+    },
+    {
+      "epoch": 445.6842105263158,
+      "grad_norm": 0.560600996017456,
+      "learning_rate": 4.840787886380993e-06,
+      "loss": 0.1802,
+      "step": 21170
+    },
+    {
+      "epoch": 445.89473684210526,
+      "grad_norm": 0.8080334067344666,
+      "learning_rate": 4.799783080219889e-06,
+      "loss": 0.1843,
+      "step": 21180
+    },
+    {
+      "epoch": 446.10526315789474,
+      "grad_norm": 0.4889693260192871,
+      "learning_rate": 4.758948412282404e-06,
+      "loss": 0.1859,
+      "step": 21190
+    },
+    {
+      "epoch": 446.3157894736842,
+      "grad_norm": 0.4943747818470001,
+      "learning_rate": 4.7182839555467095e-06,
+      "loss": 0.1825,
+      "step": 21200
+    },
+    {
+      "epoch": 446.5263157894737,
+      "grad_norm": 0.6152059435844421,
+      "learning_rate": 4.677789782686781e-06,
+      "loss": 0.1936,
+      "step": 21210
+    },
+    {
+      "epoch": 446.7368421052632,
+      "grad_norm": 0.6645772457122803,
+      "learning_rate": 4.6374659660722854e-06,
+      "loss": 0.1878,
+      "step": 21220
+    },
+    {
+      "epoch": 446.94736842105266,
+      "grad_norm": 0.6091228127479553,
+      "learning_rate": 4.597312577768431e-06,
+      "loss": 0.182,
+      "step": 21230
+    },
+    {
+      "epoch": 447.1578947368421,
+      "grad_norm": 0.5711859464645386,
+      "learning_rate": 4.557329689535794e-06,
+      "loss": 0.1766,
+      "step": 21240
+    },
+    {
+      "epoch": 447.36842105263156,
+      "grad_norm": 0.5759438872337341,
+      "learning_rate": 4.517517372830315e-06,
+      "loss": 0.1906,
+      "step": 21250
+    },
+    {
+      "epoch": 447.57894736842104,
+      "grad_norm": 0.6798800230026245,
+      "learning_rate": 4.477875698803025e-06,
+      "loss": 0.179,
+      "step": 21260
+    },
+    {
+      "epoch": 447.7894736842105,
+      "grad_norm": 0.6980090141296387,
+      "learning_rate": 4.438404738300061e-06,
+      "loss": 0.1753,
+      "step": 21270
+    },
+    {
+      "epoch": 448.0,
+      "grad_norm": 0.8547975420951843,
+      "learning_rate": 4.399104561862411e-06,
+      "loss": 0.1939,
+      "step": 21280
+    },
+    {
+      "epoch": 448.2105263157895,
+      "grad_norm": 0.7332622408866882,
+      "learning_rate": 4.359975239725878e-06,
+      "loss": 0.1973,
+      "step": 21290
+    },
+    {
+      "epoch": 448.42105263157896,
+      "grad_norm": 0.6926571130752563,
+      "learning_rate": 4.321016841820879e-06,
+      "loss": 0.1766,
+      "step": 21300
+    },
+    {
+      "epoch": 448.63157894736844,
+      "grad_norm": 0.6712207198143005,
+      "learning_rate": 4.2822294377724e-06,
+      "loss": 0.176,
+      "step": 21310
+    },
+    {
+      "epoch": 448.8421052631579,
+      "grad_norm": 0.5506131649017334,
+      "learning_rate": 4.243613096899823e-06,
+      "loss": 0.1944,
+      "step": 21320
+    },
+    {
+      "epoch": 449.05263157894734,
+      "grad_norm": 0.4963815212249756,
+      "learning_rate": 4.2090047061760115e-06,
+      "loss": 0.1906,
+      "step": 21330
+    },
+    {
+      "epoch": 449.2631578947368,
+      "grad_norm": 0.5332702398300171,
+      "learning_rate": 4.1707135752175e-06,
+      "loss": 0.1775,
+      "step": 21340
+    },
+    {
+      "epoch": 449.4736842105263,
+      "grad_norm": 0.5114326477050781,
+      "learning_rate": 4.1325937067318245e-06,
+      "loss": 0.1875,
+      "step": 21350
+    },
+    {
+      "epoch": 449.6842105263158,
+      "grad_norm": 0.599938690662384,
+      "learning_rate": 4.094645168845379e-06,
+      "loss": 0.1852,
+      "step": 21360
+    },
+    {
+      "epoch": 449.89473684210526,
+      "grad_norm": 0.6404860019683838,
+      "learning_rate": 4.056868029378314e-06,
+      "loss": 0.1804,
+      "step": 21370
+    },
+    {
+      "epoch": 450.10526315789474,
+      "grad_norm": 0.6003533601760864,
+      "learning_rate": 4.019262355844533e-06,
+      "loss": 0.1909,
+      "step": 21380
+    },
+    {
+      "epoch": 450.3157894736842,
+      "grad_norm": 0.5360564589500427,
+      "learning_rate": 3.981828215451477e-06,
+      "loss": 0.1886,
+      "step": 21390
+    },
+    {
+      "epoch": 450.5263157894737,
+      "grad_norm": 0.5857632756233215,
+      "learning_rate": 3.944565675099999e-06,
+      "loss": 0.1885,
+      "step": 21400
+    },
+    {
+      "epoch": 450.7368421052632,
+      "grad_norm": 0.4886971712112427,
+      "learning_rate": 3.907474801384326e-06,
+      "loss": 0.171,
+      "step": 21410
+    },
+    {
+      "epoch": 450.94736842105266,
+      "grad_norm": 0.6512584090232849,
+      "learning_rate": 3.870555660591846e-06,
+      "loss": 0.1885,
+      "step": 21420
+    },
+    {
+      "epoch": 451.1578947368421,
+      "grad_norm": 0.5677419304847717,
+      "learning_rate": 3.833808318703058e-06,
+      "loss": 0.1778,
+      "step": 21430
+    },
+    {
+      "epoch": 451.36842105263156,
+      "grad_norm": 0.5320295095443726,
+      "learning_rate": 3.797232841391407e-06,
+      "loss": 0.1895,
+      "step": 21440
+    },
+    {
+      "epoch": 451.57894736842104,
+      "grad_norm": 0.47199392318725586,
+      "learning_rate": 3.760829294023227e-06,
+      "loss": 0.1821,
+      "step": 21450
+    },
+    {
+      "epoch": 451.7894736842105,
+      "grad_norm": 0.626926064491272,
+      "learning_rate": 3.724597741657543e-06,
+      "loss": 0.1892,
+      "step": 21460
+    },
+    {
+      "epoch": 452.0,
+      "grad_norm": 0.5806630849838257,
+      "learning_rate": 3.688538249046003e-06,
+      "loss": 0.1848,
+      "step": 21470
+    },
+    {
+      "epoch": 452.2105263157895,
+      "grad_norm": 0.46587905287742615,
+      "learning_rate": 3.652650880632802e-06,
+      "loss": 0.1711,
+      "step": 21480
+    },
+    {
+      "epoch": 452.42105263157896,
+      "grad_norm": 0.8197913765907288,
+      "learning_rate": 3.616935700554458e-06,
+      "loss": 0.1931,
+      "step": 21490
+    },
+    {
+      "epoch": 452.63157894736844,
+      "grad_norm": 0.5315291285514832,
+      "learning_rate": 3.5813927726397913e-06,
+      "loss": 0.1861,
+      "step": 21500
+    },
+    {
+      "epoch": 452.8421052631579,
+      "grad_norm": 0.6884347796440125,
+      "learning_rate": 3.546022160409779e-06,
+      "loss": 0.1856,
+      "step": 21510
+    },
+    {
+      "epoch": 453.05263157894734,
+      "grad_norm": 0.8013739585876465,
+      "learning_rate": 3.5108239270774446e-06,
+      "loss": 0.1943,
+      "step": 21520
+    },
+    {
+      "epoch": 453.2631578947368,
+      "grad_norm": 0.720817506313324,
+      "learning_rate": 3.4757981355477363e-06,
+      "loss": 0.1862,
+      "step": 21530
+    },
+    {
+      "epoch": 453.4736842105263,
+      "grad_norm": 0.7499425411224365,
+      "learning_rate": 3.4409448484174157e-06,
+      "loss": 0.1908,
+      "step": 21540
+    },
+    {
+      "epoch": 453.6842105263158,
+      "grad_norm": 0.7999539375305176,
+      "learning_rate": 3.4062641279749674e-06,
+      "loss": 0.1806,
+      "step": 21550
+    },
+    {
+      "epoch": 453.89473684210526,
+      "grad_norm": 0.490826278924942,
+      "learning_rate": 3.3717560362004574e-06,
+      "loss": 0.1734,
+      "step": 21560
+    },
+    {
+      "epoch": 454.10526315789474,
+      "grad_norm": 0.5968190431594849,
+      "learning_rate": 3.3374206347654426e-06,
+      "loss": 0.2056,
+      "step": 21570
+    },
+    {
+      "epoch": 454.3157894736842,
+      "grad_norm": 0.5196442604064941,
+      "learning_rate": 3.3032579850328595e-06,
+      "loss": 0.1799,
+      "step": 21580
+    },
+    {
+      "epoch": 454.5263157894737,
+      "grad_norm": 0.6417738795280457,
+      "learning_rate": 3.269268148056892e-06,
+      "loss": 0.1891,
+      "step": 21590
+    },
+    {
+      "epoch": 454.7368421052632,
+      "grad_norm": 0.7515450119972229,
+      "learning_rate": 3.235451184582894e-06,
+      "loss": 0.1913,
+      "step": 21600
+    },
+    {
+      "epoch": 454.94736842105266,
+      "grad_norm": 0.5569813847541809,
+      "learning_rate": 3.201807155047254e-06,
+      "loss": 0.1714,
+      "step": 21610
+    },
+    {
+      "epoch": 455.1578947368421,
+      "grad_norm": 0.7771729230880737,
+      "learning_rate": 3.168336119577331e-06,
+      "loss": 0.1859,
+      "step": 21620
+    },
+    {
+      "epoch": 455.36842105263156,
+      "grad_norm": 0.5104111433029175,
+      "learning_rate": 3.1350381379912753e-06,
+      "loss": 0.1872,
+      "step": 21630
+    },
+    {
+      "epoch": 455.57894736842104,
+      "grad_norm": 0.46235164999961853,
+      "learning_rate": 3.1019132697979623e-06,
+      "loss": 0.1865,
+      "step": 21640
+    },
+    {
+      "epoch": 455.7894736842105,
+      "grad_norm": 0.5363319516181946,
+      "learning_rate": 3.068961574196938e-06,
+      "loss": 0.1858,
+      "step": 21650
+    },
+    {
+      "epoch": 456.0,
+      "grad_norm": 0.567591667175293,
+      "learning_rate": 3.036183110078217e-06,
+      "loss": 0.1714,
+      "step": 21660
+    },
+    {
+      "epoch": 456.2105263157895,
+      "grad_norm": 0.7822842001914978,
+      "learning_rate": 3.003577936022195e-06,
+      "loss": 0.1813,
+      "step": 21670
+    },
+    {
+      "epoch": 456.42105263157896,
+      "grad_norm": 0.5344955325126648,
+      "learning_rate": 2.9711461102996383e-06,
+      "loss": 0.1761,
+      "step": 21680
+    },
+    {
+      "epoch": 456.63157894736844,
+      "grad_norm": 0.721997082233429,
+      "learning_rate": 2.9388876908714834e-06,
+      "loss": 0.1899,
+      "step": 21690
+    },
+    {
+      "epoch": 456.8421052631579,
+      "grad_norm": 0.7189823389053345,
+      "learning_rate": 2.906802735388736e-06,
+      "loss": 0.1792,
+      "step": 21700
+    },
+    {
+      "epoch": 457.05263157894734,
+      "grad_norm": 0.545745849609375,
+      "learning_rate": 2.8748913011924174e-06,
+      "loss": 0.188,
+      "step": 21710
+    },
+    {
+      "epoch": 457.2631578947368,
+      "grad_norm": 0.5995526313781738,
+      "learning_rate": 2.84315344531344e-06,
+      "loss": 0.1845,
+      "step": 21720
+    },
+    {
+      "epoch": 457.4736842105263,
+      "grad_norm": 0.5464062094688416,
+      "learning_rate": 2.8115892244724993e-06,
+      "loss": 0.1966,
+      "step": 21730
+    },
+    {
+      "epoch": 457.6842105263158,
+      "grad_norm": 0.7825399041175842,
+      "learning_rate": 2.780198695079972e-06,
+      "loss": 0.1815,
+      "step": 21740
+    },
+    {
+      "epoch": 457.89473684210526,
+      "grad_norm": 0.5444468259811401,
+      "learning_rate": 2.7489819132358265e-06,
+      "loss": 0.1789,
+      "step": 21750
+    },
+    {
+      "epoch": 458.10526315789474,
+      "grad_norm": 0.5473366379737854,
+      "learning_rate": 2.7179389347295137e-06,
+      "loss": 0.1948,
+      "step": 21760
+    },
+    {
+      "epoch": 458.3157894736842,
+      "grad_norm": 0.5430154800415039,
+      "learning_rate": 2.6870698150398664e-06,
+      "loss": 0.1827,
+      "step": 21770
+    },
+    {
+      "epoch": 458.5263157894737,
+      "grad_norm": 0.6059640645980835,
+      "learning_rate": 2.6563746093349996e-06,
+      "loss": 0.1981,
+      "step": 21780
+    },
+    {
+      "epoch": 458.7368421052632,
+      "grad_norm": 0.594071626663208,
+      "learning_rate": 2.625853372472231e-06,
+      "loss": 0.181,
+      "step": 21790
+    },
+    {
+      "epoch": 458.94736842105266,
+      "grad_norm": 0.6014748811721802,
+      "learning_rate": 2.5955061589979734e-06,
+      "loss": 0.1857,
+      "step": 21800
+    },
+    {
+      "epoch": 459.1578947368421,
+      "grad_norm": 0.6244844794273376,
+      "learning_rate": 2.565333023147587e-06,
+      "loss": 0.1709,
+      "step": 21810
+    },
+    {
+      "epoch": 459.36842105263156,
+      "grad_norm": 0.5881552696228027,
+      "learning_rate": 2.5353340188453923e-06,
+      "loss": 0.181,
+      "step": 21820
+    },
+    {
+      "epoch": 459.57894736842104,
+      "grad_norm": 0.6716004610061646,
+      "learning_rate": 2.5055091997044587e-06,
+      "loss": 0.1802,
+      "step": 21830
+    },
+    {
+      "epoch": 459.7894736842105,
+      "grad_norm": 0.7322579622268677,
+      "learning_rate": 2.475858619026572e-06,
+      "loss": 0.1886,
+      "step": 21840
+    },
+    {
+      "epoch": 460.0,
+      "grad_norm": 0.5972902178764343,
+      "learning_rate": 2.4463823298021103e-06,
+      "loss": 0.1982,
+      "step": 21850
+    },
+    {
+      "epoch": 460.2105263157895,
+      "grad_norm": 0.6081510782241821,
+      "learning_rate": 2.417080384710013e-06,
+      "loss": 0.1767,
+      "step": 21860
+    },
+    {
+      "epoch": 460.42105263157896,
+      "grad_norm": 0.6108368039131165,
+      "learning_rate": 2.387952836117602e-06,
+      "loss": 0.1859,
+      "step": 21870
+    },
+    {
+      "epoch": 460.63157894736844,
+      "grad_norm": 0.5147649049758911,
+      "learning_rate": 2.3589997360805025e-06,
+      "loss": 0.182,
+      "step": 21880
+    },
+    {
+      "epoch": 460.8421052631579,
+      "grad_norm": 0.5353897213935852,
+      "learning_rate": 2.330221136342625e-06,
+      "loss": 0.1764,
+      "step": 21890
+    },
+    {
+      "epoch": 461.05263157894734,
+      "grad_norm": 0.6796343922615051,
+      "learning_rate": 2.3016170883359835e-06,
+      "loss": 0.2011,
+      "step": 21900
+    },
+    {
+      "epoch": 461.2631578947368,
+      "grad_norm": 0.6354873776435852,
+      "learning_rate": 2.273187643180652e-06,
+      "loss": 0.1889,
+      "step": 21910
+    },
+    {
+      "epoch": 461.4736842105263,
+      "grad_norm": 0.543091356754303,
+      "learning_rate": 2.2449328516846556e-06,
+      "loss": 0.1804,
+      "step": 21920
+    },
+    {
+      "epoch": 461.6842105263158,
+      "grad_norm": 0.5700305104255676,
+      "learning_rate": 2.216852764343902e-06,
+      "loss": 0.1827,
+      "step": 21930
+    },
+    {
+      "epoch": 461.89473684210526,
+      "grad_norm": 0.569500207901001,
+      "learning_rate": 2.1889474313420477e-06,
+      "loss": 0.1872,
+      "step": 21940
+    },
+    {
+      "epoch": 462.10526315789474,
+      "grad_norm": 0.6591985821723938,
+      "learning_rate": 2.1612169025504446e-06,
+      "loss": 0.1858,
+      "step": 21950
+    },
+    {
+      "epoch": 462.3157894736842,
+      "grad_norm": 0.5844663381576538,
+      "learning_rate": 2.1336612275280497e-06,
+      "loss": 0.1866,
+      "step": 21960
+    },
+    {
+      "epoch": 462.5263157894737,
+      "grad_norm": 0.5001121759414673,
+      "learning_rate": 2.1062804555213255e-06,
+      "loss": 0.1761,
+      "step": 21970
+    },
+    {
+      "epoch": 462.7368421052632,
+      "grad_norm": 0.5647687911987305,
+      "learning_rate": 2.079074635464129e-06,
+      "loss": 0.1863,
+      "step": 21980
+    },
+    {
+      "epoch": 462.94736842105266,
+      "grad_norm": 0.6637557744979858,
+      "learning_rate": 2.0520438159777e-06,
+      "loss": 0.1864,
+      "step": 21990
+    },
+    {
+      "epoch": 463.1578947368421,
+      "grad_norm": 0.9768401384353638,
+      "learning_rate": 2.0251880453704963e-06,
+      "loss": 0.1822,
+      "step": 22000
+    },
+    {
+      "epoch": 463.36842105263156,
+      "grad_norm": 0.6367493271827698,
+      "learning_rate": 1.998507371638114e-06,
+      "loss": 0.1967,
+      "step": 22010
+    },
+    {
+      "epoch": 463.57894736842104,
+      "grad_norm": 0.6616984605789185,
+      "learning_rate": 1.972001842463245e-06,
+      "loss": 0.1843,
+      "step": 22020
+    },
+    {
+      "epoch": 463.7894736842105,
+      "grad_norm": 1.017850399017334,
+      "learning_rate": 1.945671505215574e-06,
+      "loss": 0.1819,
+      "step": 22030
+    },
+    {
+      "epoch": 464.0,
+      "grad_norm": 0.6348969340324402,
+      "learning_rate": 1.9195164069516936e-06,
+      "loss": 0.1739,
+      "step": 22040
+    },
+    {
+      "epoch": 464.2105263157895,
+      "grad_norm": 0.8872740864753723,
+      "learning_rate": 1.8935365944149908e-06,
+      "loss": 0.1858,
+      "step": 22050
+    },
+    {
+      "epoch": 464.42105263157896,
+      "grad_norm": 0.7518464922904968,
+      "learning_rate": 1.867732114035614e-06,
+      "loss": 0.1923,
+      "step": 22060
+    },
+    {
+      "epoch": 464.63157894736844,
+      "grad_norm": 0.5332704782485962,
+      "learning_rate": 1.8421030119303407e-06,
+      "loss": 0.1698,
+      "step": 22070
+    },
+    {
+      "epoch": 464.8421052631579,
+      "grad_norm": 0.6198592185974121,
+      "learning_rate": 1.8166493339025426e-06,
+      "loss": 0.1872,
+      "step": 22080
+    },
+    {
+      "epoch": 465.05263157894734,
+      "grad_norm": 0.5435618162155151,
+      "learning_rate": 1.791371125442065e-06,
+      "loss": 0.1871,
+      "step": 22090
+    },
+    {
+      "epoch": 465.2631578947368,
+      "grad_norm": 0.5456479787826538,
+      "learning_rate": 1.7662684317251598e-06,
+      "loss": 0.1755,
+      "step": 22100
+    },
+    {
+      "epoch": 465.4736842105263,
+      "grad_norm": 0.5278346538543701,
+      "learning_rate": 1.7413412976144294e-06,
+      "loss": 0.1989,
+      "step": 22110
+    },
+    {
+      "epoch": 465.6842105263158,
+      "grad_norm": 0.5534194707870483,
+      "learning_rate": 1.7165897676586717e-06,
+      "loss": 0.175,
+      "step": 22120
+    },
+    {
+      "epoch": 465.89473684210526,
+      "grad_norm": 0.6580941081047058,
+      "learning_rate": 1.6920138860929246e-06,
+      "loss": 0.1784,
+      "step": 22130
+    },
+    {
+      "epoch": 466.10526315789474,
+      "grad_norm": 0.6645970344543457,
+      "learning_rate": 1.6676136968382328e-06,
+      "loss": 0.1935,
+      "step": 22140
+    },
+    {
+      "epoch": 466.3157894736842,
+      "grad_norm": 0.5245230197906494,
+      "learning_rate": 1.643389243501725e-06,
+      "loss": 0.1813,
+      "step": 22150
+    },
+    {
+      "epoch": 466.5263157894737,
+      "grad_norm": 0.5805260539054871,
+      "learning_rate": 1.619340569376404e-06,
+      "loss": 0.1731,
+      "step": 22160
+    },
+    {
+      "epoch": 466.7368421052632,
+      "grad_norm": 0.6286890506744385,
+      "learning_rate": 1.5954677174411681e-06,
+      "loss": 0.1829,
+      "step": 22170
+    },
+    {
+      "epoch": 466.94736842105266,
+      "grad_norm": 0.5264356732368469,
+      "learning_rate": 1.5717707303606555e-06,
+      "loss": 0.1884,
+      "step": 22180
+    },
+    {
+      "epoch": 467.1578947368421,
+      "grad_norm": 0.5247817039489746,
+      "learning_rate": 1.548249650485234e-06,
+      "loss": 0.1781,
+      "step": 22190
+    },
+    {
+      "epoch": 467.36842105263156,
+      "grad_norm": 0.8821797966957092,
+      "learning_rate": 1.5249045198508893e-06,
+      "loss": 0.1869,
+      "step": 22200
+    },
+    {
+      "epoch": 467.57894736842104,
+      "grad_norm": 0.5761975049972534,
+      "learning_rate": 1.5017353801791589e-06,
+      "loss": 0.1823,
+      "step": 22210
+    },
+    {
+      "epoch": 467.7894736842105,
+      "grad_norm": 0.5740618109703064,
+      "learning_rate": 1.4787422728770316e-06,
+      "loss": 0.1841,
+      "step": 22220
+    },
+    {
+      "epoch": 468.0,
+      "grad_norm": 0.5789286494255066,
+      "learning_rate": 1.4559252390369483e-06,
+      "loss": 0.1863,
+      "step": 22230
+    },
+    {
+      "epoch": 468.2105263157895,
+      "grad_norm": 0.4390312731266022,
+      "learning_rate": 1.433284319436623e-06,
+      "loss": 0.1852,
+      "step": 22240
+    },
+    {
+      "epoch": 468.42105263157896,
+      "grad_norm": 0.7425456047058105,
+      "learning_rate": 1.4108195545390557e-06,
+      "loss": 0.1847,
+      "step": 22250
+    },
+    {
+      "epoch": 468.63157894736844,
+      "grad_norm": 0.5220855474472046,
+      "learning_rate": 1.388530984492431e-06,
+      "loss": 0.1767,
+      "step": 22260
+    },
+    {
+      "epoch": 468.8421052631579,
+      "grad_norm": 0.8341360092163086,
+      "learning_rate": 1.36641864913003e-06,
+      "loss": 0.1947,
+      "step": 22270
+    },
+    {
+      "epoch": 469.05263157894734,
+      "grad_norm": 0.5250037312507629,
+      "learning_rate": 1.3444825879701973e-06,
+      "loss": 0.1934,
+      "step": 22280
+    },
+    {
+      "epoch": 469.2631578947368,
+      "grad_norm": 0.5445212721824646,
+      "learning_rate": 1.3227228402162061e-06,
+      "loss": 0.1716,
+      "step": 22290
+    },
+    {
+      "epoch": 469.4736842105263,
+      "grad_norm": 0.6371374130249023,
+      "learning_rate": 1.301139444756272e-06,
+      "loss": 0.1891,
+      "step": 22300
+    },
+    {
+      "epoch": 469.6842105263158,
+      "grad_norm": 0.5438109636306763,
+      "learning_rate": 1.279732440163417e-06,
+      "loss": 0.1868,
+      "step": 22310
+    },
+    {
+      "epoch": 469.89473684210526,
+      "grad_norm": 0.9809474349021912,
+      "learning_rate": 1.2585018646954273e-06,
+      "loss": 0.1901,
+      "step": 22320
+    },
+    {
+      "epoch": 470.10526315789474,
+      "grad_norm": 0.5445967316627502,
+      "learning_rate": 1.237447756294785e-06,
+      "loss": 0.1909,
+      "step": 22330
+    },
+    {
+      "epoch": 470.3157894736842,
+      "grad_norm": 0.5317168235778809,
+      "learning_rate": 1.216570152588603e-06,
+      "loss": 0.1852,
+      "step": 22340
+    },
+    {
+      "epoch": 470.5263157894737,
+      "grad_norm": 0.9151925444602966,
+      "learning_rate": 1.195869090888524e-06,
+      "loss": 0.1809,
+      "step": 22350
+    },
+    {
+      "epoch": 470.7368421052632,
+      "grad_norm": 0.6518405675888062,
+      "learning_rate": 1.1753446081907205e-06,
+      "loss": 0.1738,
+      "step": 22360
+    },
+    {
+      "epoch": 470.94736842105266,
+      "grad_norm": 0.7550083994865417,
+      "learning_rate": 1.1549967411757734e-06,
+      "loss": 0.1979,
+      "step": 22370
+    },
+    {
+      "epoch": 471.1578947368421,
+      "grad_norm": 0.5095014572143555,
+      "learning_rate": 1.134825526208605e-06,
+      "loss": 0.1843,
+      "step": 22380
+    },
+    {
+      "epoch": 471.36842105263156,
+      "grad_norm": 0.7665607333183289,
+      "learning_rate": 1.1148309993384454e-06,
+      "loss": 0.1755,
+      "step": 22390
+    },
+    {
+      "epoch": 471.57894736842104,
+      "grad_norm": 0.5469737648963928,
+      "learning_rate": 1.0950131962987774e-06,
+      "loss": 0.1837,
+      "step": 22400
+    },
+    {
+      "epoch": 471.7894736842105,
+      "grad_norm": 0.6690005660057068,
+      "learning_rate": 1.0753721525072147e-06,
+      "loss": 0.1928,
+      "step": 22410
+    },
+    {
+      "epoch": 472.0,
+      "grad_norm": 0.8379321098327637,
+      "learning_rate": 1.0559079030654895e-06,
+      "loss": 0.1947,
+      "step": 22420
+    },
+    {
+      "epoch": 472.2105263157895,
+      "grad_norm": 0.5236616134643555,
+      "learning_rate": 1.0366204827593652e-06,
+      "loss": 0.1857,
+      "step": 22430
+    },
+    {
+      "epoch": 472.42105263157896,
+      "grad_norm": 0.5310667753219604,
+      "learning_rate": 1.0175099260586018e-06,
+      "loss": 0.1799,
+      "step": 22440
+    },
+    {
+      "epoch": 472.63157894736844,
+      "grad_norm": 0.6408274173736572,
+      "learning_rate": 9.985762671168576e-07,
+      "loss": 0.1812,
+      "step": 22450
+    },
+    {
+      "epoch": 472.8421052631579,
+      "grad_norm": 0.533226728439331,
+      "learning_rate": 9.798195397716315e-07,
+      "loss": 0.1825,
+      "step": 22460
+    },
+    {
+      "epoch": 473.05263157894734,
+      "grad_norm": 0.5859424471855164,
+      "learning_rate": 9.61239777544276e-07,
+      "loss": 0.1839,
+      "step": 22470
+    },
+    {
+      "epoch": 473.2631578947368,
+      "grad_norm": 0.5590562224388123,
+      "learning_rate": 9.428370136398079e-07,
+      "loss": 0.182,
+      "step": 22480
+    },
+    {
+      "epoch": 473.4736842105263,
+      "grad_norm": 0.48322606086730957,
+      "learning_rate": 9.246112809469521e-07,
+      "loss": 0.1899,
+      "step": 22490
+    },
+    {
+      "epoch": 473.6842105263158,
+      "grad_norm": 0.7842751741409302,
+      "learning_rate": 9.065626120380643e-07,
+      "loss": 0.1917,
+      "step": 22500
+    },
+    {
+      "epoch": 473.89473684210526,
+      "grad_norm": 0.6499484181404114,
+      "learning_rate": 8.886910391690206e-07,
+      "loss": 0.1824,
+      "step": 22510
+    },
+    {
+      "epoch": 474.10526315789474,
+      "grad_norm": 0.5859502553939819,
+      "learning_rate": 8.709965942792386e-07,
+      "loss": 0.1774,
+      "step": 22520
+    },
+    {
+      "epoch": 474.3157894736842,
+      "grad_norm": 0.552681028842926,
+      "learning_rate": 8.53479308991556e-07,
+      "loss": 0.1796,
+      "step": 22530
+    },
+    {
+      "epoch": 474.5263157894737,
+      "grad_norm": 0.6872820854187012,
+      "learning_rate": 8.361392146121972e-07,
+      "loss": 0.1819,
+      "step": 22540
+    },
+    {
+      "epoch": 474.7368421052632,
+      "grad_norm": 0.5292307734489441,
+      "learning_rate": 8.189763421307284e-07,
+      "loss": 0.1897,
+      "step": 22550
+    },
+    {
+      "epoch": 474.94736842105266,
+      "grad_norm": 0.4911171495914459,
+      "learning_rate": 8.019907222199807e-07,
+      "loss": 0.1706,
+      "step": 22560
+    },
+    {
+      "epoch": 475.1578947368421,
+      "grad_norm": 0.6938400268554688,
+      "learning_rate": 7.851823852360163e-07,
+      "loss": 0.1977,
+      "step": 22570
+    },
+    {
+      "epoch": 475.36842105263156,
+      "grad_norm": 0.5019783973693848,
+      "learning_rate": 7.685513612180506e-07,
+      "loss": 0.1746,
+      "step": 22580
+    },
+    {
+      "epoch": 475.57894736842104,
+      "grad_norm": 0.5230517983436584,
+      "learning_rate": 7.520976798884194e-07,
+      "loss": 0.1827,
+      "step": 22590
+    },
+    {
+      "epoch": 475.7894736842105,
+      "grad_norm": 0.5872176885604858,
+      "learning_rate": 7.35821370652523e-07,
+      "loss": 0.1827,
+      "step": 22600
+    },
+    {
+      "epoch": 476.0,
+      "grad_norm": 0.861818253993988,
+      "learning_rate": 7.197224625987819e-07,
+      "loss": 0.1873,
+      "step": 22610
+    },
+    {
+      "epoch": 476.2105263157895,
+      "grad_norm": 0.5507122278213501,
+      "learning_rate": 7.038009844985149e-07,
+      "loss": 0.1974,
+      "step": 22620
+    },
+    {
+      "epoch": 476.42105263157896,
+      "grad_norm": 0.5509169101715088,
+      "learning_rate": 6.880569648060275e-07,
+      "loss": 0.1734,
+      "step": 22630
+    },
+    {
+      "epoch": 476.63157894736844,
+      "grad_norm": 0.4984310269355774,
+      "learning_rate": 6.724904316584124e-07,
+      "loss": 0.1843,
+      "step": 22640
+    },
+    {
+      "epoch": 476.8421052631579,
+      "grad_norm": 0.6479700207710266,
+      "learning_rate": 6.571014128755937e-07,
+      "loss": 0.1784,
+      "step": 22650
+    },
+    {
+      "epoch": 477.05263157894734,
+      "grad_norm": 0.49195757508277893,
+      "learning_rate": 6.418899359602381e-07,
+      "loss": 0.1862,
+      "step": 22660
+    },
+    {
+      "epoch": 477.2631578947368,
+      "grad_norm": 0.6385235786437988,
+      "learning_rate": 6.26856028097722e-07,
+      "loss": 0.1802,
+      "step": 22670
+    },
+    {
+      "epoch": 477.4736842105263,
+      "grad_norm": 0.599315881729126,
+      "learning_rate": 6.119997161560975e-07,
+      "loss": 0.1821,
+      "step": 22680
+    },
+    {
+      "epoch": 477.6842105263158,
+      "grad_norm": 0.5851359963417053,
+      "learning_rate": 5.973210266859708e-07,
+      "loss": 0.1878,
+      "step": 22690
+    },
+    {
+      "epoch": 477.89473684210526,
+      "grad_norm": 0.5468791127204895,
+      "learning_rate": 5.828199859205574e-07,
+      "loss": 0.1882,
+      "step": 22700
+    },
+    {
+      "epoch": 478.10526315789474,
+      "grad_norm": 0.7265982031822205,
+      "learning_rate": 5.684966197755715e-07,
+      "loss": 0.1906,
+      "step": 22710
+    },
+    {
+      "epoch": 478.3157894736842,
+      "grad_norm": 0.5022777915000916,
+      "learning_rate": 5.543509538491809e-07,
+      "loss": 0.1756,
+      "step": 22720
+    },
+    {
+      "epoch": 478.5263157894737,
+      "grad_norm": 0.501018762588501,
+      "learning_rate": 5.403830134219856e-07,
+      "loss": 0.1757,
+      "step": 22730
+    },
+    {
+      "epoch": 478.7368421052632,
+      "grad_norm": 0.5156716108322144,
+      "learning_rate": 5.265928234569617e-07,
+      "loss": 0.1961,
+      "step": 22740
+    },
+    {
+      "epoch": 478.94736842105266,
+      "grad_norm": 0.6293845772743225,
+      "learning_rate": 5.129804085994284e-07,
+      "loss": 0.1811,
+      "step": 22750
+    },
+    {
+      "epoch": 479.1578947368421,
+      "grad_norm": 0.6754159927368164,
+      "learning_rate": 4.995457931769477e-07,
+      "loss": 0.1861,
+      "step": 22760
+    },
+    {
+      "epoch": 479.36842105263156,
+      "grad_norm": 0.5567587018013,
+      "learning_rate": 4.862890011993915e-07,
+      "loss": 0.1847,
+      "step": 22770
+    },
+    {
+      "epoch": 479.57894736842104,
+      "grad_norm": 0.713141143321991,
+      "learning_rate": 4.732100563587638e-07,
+      "loss": 0.1737,
+      "step": 22780
+    },
+    {
+      "epoch": 479.7894736842105,
+      "grad_norm": 0.5813489556312561,
+      "learning_rate": 4.6030898202928943e-07,
+      "loss": 0.1858,
+      "step": 22790
+    },
+    {
+      "epoch": 480.0,
+      "grad_norm": 0.5960988998413086,
+      "learning_rate": 4.475858012672474e-07,
+      "loss": 0.1832,
+      "step": 22800
+    },
+    {
+      "epoch": 480.2105263157895,
+      "grad_norm": 0.7617952227592468,
+      "learning_rate": 4.350405368110488e-07,
+      "loss": 0.19,
+      "step": 22810
+    },
+    {
+      "epoch": 480.42105263157896,
+      "grad_norm": 0.47482919692993164,
+      "learning_rate": 4.226732110811149e-07,
+      "loss": 0.1753,
+      "step": 22820
+    },
+    {
+      "epoch": 480.63157894736844,
+      "grad_norm": 0.8491657972335815,
+      "learning_rate": 4.1048384617985435e-07,
+      "loss": 0.1899,
+      "step": 22830
+    },
+    {
+      "epoch": 480.8421052631579,
+      "grad_norm": 0.5300630927085876,
+      "learning_rate": 3.984724638916415e-07,
+      "loss": 0.1823,
+      "step": 22840
+    },
+    {
+      "epoch": 481.05263157894734,
+      "grad_norm": 0.4608703553676605,
+      "learning_rate": 3.866390856827495e-07,
+      "loss": 0.1805,
+      "step": 22850
+    },
+    {
+      "epoch": 481.2631578947368,
+      "grad_norm": 0.6642935276031494,
+      "learning_rate": 3.749837327013728e-07,
+      "loss": 0.1783,
+      "step": 22860
+    },
+    {
+      "epoch": 481.4736842105263,
+      "grad_norm": 0.6894769072532654,
+      "learning_rate": 3.635064257774934e-07,
+      "loss": 0.1898,
+      "step": 22870
+    },
+    {
+      "epoch": 481.6842105263158,
+      "grad_norm": 0.5321648120880127,
+      "learning_rate": 3.5220718542292583e-07,
+      "loss": 0.1824,
+      "step": 22880
+    },
+    {
+      "epoch": 481.89473684210526,
+      "grad_norm": 0.6732971668243408,
+      "learning_rate": 3.410860318312614e-07,
+      "loss": 0.186,
+      "step": 22890
+    },
+    {
+      "epoch": 482.10526315789474,
+      "grad_norm": 0.7677638530731201,
+      "learning_rate": 3.301429848777793e-07,
+      "loss": 0.1756,
+      "step": 22900
+    },
+    {
+      "epoch": 482.3157894736842,
+      "grad_norm": 0.7483295798301697,
+      "learning_rate": 3.19378064119491e-07,
+      "loss": 0.1943,
+      "step": 22910
+    },
+    {
+      "epoch": 482.5263157894737,
+      "grad_norm": 0.49219393730163574,
+      "learning_rate": 3.087912887950517e-07,
+      "loss": 0.1763,
+      "step": 22920
+    },
+    {
+      "epoch": 482.7368421052632,
+      "grad_norm": 0.5193999409675598,
+      "learning_rate": 2.983826778247489e-07,
+      "loss": 0.1858,
+      "step": 22930
+    },
+    {
+      "epoch": 482.94736842105266,
+      "grad_norm": 0.6139624118804932,
+      "learning_rate": 2.881522498104472e-07,
+      "loss": 0.1908,
+      "step": 22940
+    },
+    {
+      "epoch": 483.1578947368421,
+      "grad_norm": 0.6963362097740173,
+      "learning_rate": 2.781000230356101e-07,
+      "loss": 0.2133,
+      "step": 22950
+    },
+    {
+      "epoch": 483.36842105263156,
+      "grad_norm": 0.5296799540519714,
+      "learning_rate": 2.682260154651672e-07,
+      "loss": 0.1714,
+      "step": 22960
+    },
+    {
+      "epoch": 483.57894736842104,
+      "grad_norm": 0.5533150434494019,
+      "learning_rate": 2.5853024474556953e-07,
+      "loss": 0.194,
+      "step": 22970
+    },
+    {
+      "epoch": 483.7894736842105,
+      "grad_norm": 0.55152827501297,
+      "learning_rate": 2.4901272820475605e-07,
+      "loss": 0.1859,
+      "step": 22980
+    },
+    {
+      "epoch": 484.0,
+      "grad_norm": 1.465172290802002,
+      "learning_rate": 2.3967348285205416e-07,
+      "loss": 0.1798,
+      "step": 22990
+    },
+    {
+      "epoch": 484.2105263157895,
+      "grad_norm": 0.6252908706665039,
+      "learning_rate": 2.3051252537820145e-07,
+      "loss": 0.1796,
+      "step": 23000
+    },
+    {
+      "epoch": 484.42105263157896,
+      "grad_norm": 0.4841025173664093,
+      "learning_rate": 2.2152987215534604e-07,
+      "loss": 0.1842,
+      "step": 23010
+    },
+    {
+      "epoch": 484.63157894736844,
+      "grad_norm": 0.5873754024505615,
+      "learning_rate": 2.1272553923691317e-07,
+      "loss": 0.1686,
+      "step": 23020
+    },
+    {
+      "epoch": 484.8421052631579,
+      "grad_norm": 0.7624306678771973,
+      "learning_rate": 2.0409954235769414e-07,
+      "loss": 0.1874,
+      "step": 23030
+    },
+    {
+      "epoch": 485.05263157894734,
+      "grad_norm": 0.6849631667137146,
+      "learning_rate": 1.9565189693373508e-07,
+      "loss": 0.187,
+      "step": 23040
+    },
+    {
+      "epoch": 485.2631578947368,
+      "grad_norm": 0.5415328741073608,
+      "learning_rate": 1.8738261806234837e-07,
+      "loss": 0.1719,
+      "step": 23050
+    },
+    {
+      "epoch": 485.4736842105263,
+      "grad_norm": 0.6806001663208008,
+      "learning_rate": 1.7929172052207898e-07,
+      "loss": 0.1908,
+      "step": 23060
+    },
+    {
+      "epoch": 485.6842105263158,
+      "grad_norm": 0.8779856562614441,
+      "learning_rate": 1.713792187726604e-07,
+      "loss": 0.1895,
+      "step": 23070
+    },
+    {
+      "epoch": 485.89473684210526,
+      "grad_norm": 0.6320582032203674,
+      "learning_rate": 1.6364512695503654e-07,
+      "loss": 0.1895,
+      "step": 23080
+    },
+    {
+      "epoch": 486.10526315789474,
+      "grad_norm": 0.5849341154098511,
+      "learning_rate": 1.5608945889127314e-07,
+      "loss": 0.1835,
+      "step": 23090
+    },
+    {
+      "epoch": 486.3157894736842,
+      "grad_norm": 0.49242666363716125,
+      "learning_rate": 1.4871222808456874e-07,
+      "loss": 0.1755,
+      "step": 23100
+    },
+    {
+      "epoch": 486.5263157894737,
+      "grad_norm": 0.6123622059822083,
+      "learning_rate": 1.415134477192437e-07,
+      "loss": 0.1888,
+      "step": 23110
+    },
+    {
+      "epoch": 486.7368421052632,
+      "grad_norm": 0.6847497224807739,
+      "learning_rate": 1.3449313066067337e-07,
+      "loss": 0.1761,
+      "step": 23120
+    },
+    {
+      "epoch": 486.94736842105266,
+      "grad_norm": 0.6061682105064392,
+      "learning_rate": 1.2765128945531057e-07,
+      "loss": 0.1884,
+      "step": 23130
+    },
+    {
+      "epoch": 487.1578947368421,
+      "grad_norm": 0.5884981155395508,
+      "learning_rate": 1.209879363306299e-07,
+      "loss": 0.188,
+      "step": 23140
+    },
+    {
+      "epoch": 487.36842105263156,
+      "grad_norm": 0.5741125345230103,
+      "learning_rate": 1.145030831951277e-07,
+      "loss": 0.1862,
+      "step": 23150
+    },
+    {
+      "epoch": 487.57894736842104,
+      "grad_norm": 0.6427663564682007,
+      "learning_rate": 1.0819674163828897e-07,
+      "loss": 0.1893,
+      "step": 23160
+    },
+    {
+      "epoch": 487.7894736842105,
+      "grad_norm": 0.5162391066551208,
+      "learning_rate": 1.0206892293055382e-07,
+      "loss": 0.1799,
+      "step": 23170
+    },
+    {
+      "epoch": 488.0,
+      "grad_norm": 0.8966327905654907,
+      "learning_rate": 9.611963802335089e-08,
+      "loss": 0.1892,
+      "step": 23180
+    },
+    {
+      "epoch": 488.2105263157895,
+      "grad_norm": 0.7478238940238953,
+      "learning_rate": 9.034889754900855e-08,
+      "loss": 0.1786,
+      "step": 23190
+    },
+    {
+      "epoch": 488.42105263157896,
+      "grad_norm": 0.6039992570877075,
+      "learning_rate": 8.475671182076595e-08,
+      "loss": 0.1848,
+      "step": 23200
+    },
+    {
+      "epoch": 488.63157894736844,
+      "grad_norm": 0.6707776784896851,
+      "learning_rate": 7.934309083278413e-08,
+      "loss": 0.1813,
+      "step": 23210
+    },
+    {
+      "epoch": 488.8421052631579,
+      "grad_norm": 0.5292564034461975,
+      "learning_rate": 7.410804426005724e-08,
+      "loss": 0.196,
+      "step": 23220
+    },
+    {
+      "epoch": 489.05263157894734,
+      "grad_norm": 0.6198012232780457,
+      "learning_rate": 6.905158145847913e-08,
+      "loss": 0.1719,
+      "step": 23230
+    },
+    {
+      "epoch": 489.2631578947368,
+      "grad_norm": 0.5900288224220276,
+      "learning_rate": 6.417371146476559e-08,
+      "loss": 0.172,
+      "step": 23240
+    },
+    {
+      "epoch": 489.4736842105263,
+      "grad_norm": 0.5824329257011414,
+      "learning_rate": 5.947444299646554e-08,
+      "loss": 0.1932,
+      "step": 23250
+    },
+    {
+      "epoch": 489.6842105263158,
+      "grad_norm": 0.4687194526195526,
+      "learning_rate": 5.495378445192767e-08,
+      "loss": 0.1689,
+      "step": 23260
+    },
+    {
+      "epoch": 489.89473684210526,
+      "grad_norm": 0.7158029675483704,
+      "learning_rate": 5.0611743910300436e-08,
+      "loss": 0.1915,
+      "step": 23270
+    },
+    {
+      "epoch": 490.10526315789474,
+      "grad_norm": 0.6645452380180359,
+      "learning_rate": 4.644832913152097e-08,
+      "loss": 0.1968,
+      "step": 23280
+    },
+    {
+      "epoch": 490.3157894736842,
+      "grad_norm": 0.6376003623008728,
+      "learning_rate": 4.246354755628179e-08,
+      "loss": 0.1806,
+      "step": 23290
+    },
+    {
+      "epoch": 490.5263157894737,
+      "grad_norm": 0.8974951505661011,
+      "learning_rate": 3.8657406306030764e-08,
+      "loss": 0.1905,
+      "step": 23300
+    },
+    {
+      "epoch": 490.7368421052632,
+      "grad_norm": 0.5921673774719238,
+      "learning_rate": 3.502991218296003e-08,
+      "loss": 0.1861,
+      "step": 23310
+    },
+    {
+      "epoch": 490.94736842105266,
+      "grad_norm": 0.892542839050293,
+      "learning_rate": 3.1581071670006015e-08,
+      "loss": 0.1832,
+      "step": 23320
+    },
+    {
+      "epoch": 491.1578947368421,
+      "grad_norm": 0.5395411849021912,
+      "learning_rate": 2.8310890930782763e-08,
+      "loss": 0.1806,
+      "step": 23330
+    },
+    {
+      "epoch": 491.36842105263156,
+      "grad_norm": 0.6547475457191467,
+      "learning_rate": 2.5219375809637514e-08,
+      "loss": 0.192,
+      "step": 23340
+    },
+    {
+      "epoch": 491.57894736842104,
+      "grad_norm": 0.6222972273826599,
+      "learning_rate": 2.230653183162845e-08,
+      "loss": 0.1725,
+      "step": 23350
+    },
+    {
+      "epoch": 491.7894736842105,
+      "grad_norm": 0.6540040969848633,
+      "learning_rate": 1.9572364202458115e-08,
+      "loss": 0.1761,
+      "step": 23360
+    },
+    {
+      "epoch": 492.0,
+      "grad_norm": 0.9256670475006104,
+      "learning_rate": 1.7016877808539998e-08,
+      "loss": 0.1914,
+      "step": 23370
+    },
+    {
+      "epoch": 492.2105263157895,
+      "grad_norm": 0.6202569007873535,
+      "learning_rate": 1.4640077216931946e-08,
+      "loss": 0.1839,
+      "step": 23380
+    },
+    {
+      "epoch": 492.42105263157896,
+      "grad_norm": 0.6193927526473999,
+      "learning_rate": 1.2441966675380556e-08,
+      "loss": 0.1789,
+      "step": 23390
+    },
+    {
+      "epoch": 492.63157894736844,
+      "grad_norm": 0.4895412027835846,
+      "learning_rate": 1.0422550112243468e-08,
+      "loss": 0.1863,
+      "step": 23400
+    },
+    {
+      "epoch": 492.8421052631579,
+      "grad_norm": 0.5214998722076416,
+      "learning_rate": 8.581831136555973e-09,
+      "loss": 0.1762,
+      "step": 23410
+    },
+    {
+      "epoch": 493.05263157894734,
+      "grad_norm": 0.6743729114532471,
+      "learning_rate": 6.919813037986611e-09,
+      "loss": 0.1918,
+      "step": 23420
+    },
+    {
+      "epoch": 493.2631578947368,
+      "grad_norm": 0.5977574586868286,
+      "learning_rate": 5.436498786826061e-09,
+      "loss": 0.1914,
+      "step": 23430
+    },
+    {
+      "epoch": 493.4736842105263,
+      "grad_norm": 0.6423200368881226,
+      "learning_rate": 4.1318910339982475e-09,
+      "loss": 0.1975,
+      "step": 23440
+    },
+    {
+      "epoch": 493.6842105263158,
+      "grad_norm": 0.4789294898509979,
+      "learning_rate": 3.005992111038136e-09,
+      "loss": 0.1658,
+      "step": 23450
+    },
+    {
+      "epoch": 493.89473684210526,
+      "grad_norm": 0.5794366002082825,
+      "learning_rate": 2.058804030125039e-09,
+      "loss": 0.1798,
+      "step": 23460
+    },
+    {
+      "epoch": 494.10526315789474,
+      "grad_norm": 0.621591329574585,
+      "learning_rate": 1.3591339329321884e-09,
+      "loss": 0.1856,
+      "step": 23470
+    },
+    {
+      "epoch": 494.3157894736842,
+      "grad_norm": 0.5210126638412476,
+      "learning_rate": 7.515008515257549e-10,
+      "loss": 0.1857,
+      "step": 23480
+    },
+    {
+      "epoch": 494.5263157894737,
+      "grad_norm": 0.5082560181617737,
+      "learning_rate": 3.225826413100208e-10,
+      "loss": 0.1706,
+      "step": 23490
+    },
+    {
+      "epoch": 494.7368421052632,
+      "grad_norm": 0.5357773303985596,
+      "learning_rate": 7.238006881626901e-11,
+      "loss": 0.1852,
+      "step": 23500
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 23500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 500,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.305535571859456e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}