diff --git "a/Ins/trainer_state.json" "b/Ins/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/Ins/trainer_state.json"
@@ -0,0 +1,6819 @@
+{
+  "best_metric": 2.729660987854004,
+  "best_model_checkpoint": "./Ins/checkpoint-9678",
+  "epoch": 1.0,
+  "eval_steps": 1000,
+  "global_step": 9678,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0010332713370531101,
+      "grad_norm": 9.612689971923828,
+      "learning_rate": 5.1546391752577315e-05,
+      "loss": 4.2162,
+      "step": 10
+    },
+    {
+      "epoch": 0.0020665426741062203,
+      "grad_norm": 6.56850004196167,
+      "learning_rate": 0.00010309278350515463,
+      "loss": 3.388,
+      "step": 20
+    },
+    {
+      "epoch": 0.0030998140111593306,
+      "grad_norm": 4.049291610717773,
+      "learning_rate": 0.00015463917525773197,
+      "loss": 2.7081,
+      "step": 30
+    },
+    {
+      "epoch": 0.0041330853482124405,
+      "grad_norm": 2.604156970977783,
+      "learning_rate": 0.00020618556701030926,
+      "loss": 2.8688,
+      "step": 40
+    },
+    {
+      "epoch": 0.0051663566852655505,
+      "grad_norm": 1.8748372793197632,
+      "learning_rate": 0.00025773195876288655,
+      "loss": 2.6251,
+      "step": 50
+    },
+    {
+      "epoch": 0.006199628022318661,
+      "grad_norm": 2.638007164001465,
+      "learning_rate": 0.00030927835051546395,
+      "loss": 2.5802,
+      "step": 60
+    },
+    {
+      "epoch": 0.007232899359371771,
+      "grad_norm": 1.8762423992156982,
+      "learning_rate": 0.00036082474226804123,
+      "loss": 2.4978,
+      "step": 70
+    },
+    {
+      "epoch": 0.008266170696424881,
+      "grad_norm": 1.7499455213546753,
+      "learning_rate": 0.0004123711340206185,
+      "loss": 2.4351,
+      "step": 80
+    },
+    {
+      "epoch": 0.009299442033477991,
+      "grad_norm": 1.4334783554077148,
+      "learning_rate": 0.0004639175257731959,
+      "loss": 2.4811,
+      "step": 90
+    },
+    {
+      "epoch": 0.010332713370531101,
+      "grad_norm": 1.4335546493530273,
+      "learning_rate": 0.0004999998790431266,
+      "loss": 2.3653,
+      "step": 100
+    },
+    {
+      "epoch": 0.01136598470758421,
+      "grad_norm": 1.379575490951538,
+      "learning_rate": 0.0004999977287019656,
+      "loss": 2.2769,
+      "step": 110
+    },
+    {
+      "epoch": 0.012399256044637322,
+      "grad_norm": 1.357542872428894,
+      "learning_rate": 0.0004999928904568952,
+      "loss": 2.1347,
+      "step": 120
+    },
+    {
+      "epoch": 0.013432527381690432,
+      "grad_norm": 1.6458231210708618,
+      "learning_rate": 0.0004999853643599349,
+      "loss": 2.1943,
+      "step": 130
+    },
+    {
+      "epoch": 0.014465798718743542,
+      "grad_norm": 1.7004691362380981,
+      "learning_rate": 0.0004999751504920031,
+      "loss": 2.204,
+      "step": 140
+    },
+    {
+      "epoch": 0.015499070055796652,
+      "grad_norm": 1.3042373657226562,
+      "learning_rate": 0.0004999622489629164,
+      "loss": 2.0817,
+      "step": 150
+    },
+    {
+      "epoch": 0.016532341392849762,
+      "grad_norm": 1.5508415699005127,
+      "learning_rate": 0.0004999466599113884,
+      "loss": 2.115,
+      "step": 160
+    },
+    {
+      "epoch": 0.017565612729902872,
+      "grad_norm": 1.3333240747451782,
+      "learning_rate": 0.0004999283835050278,
+      "loss": 2.1762,
+      "step": 170
+    },
+    {
+      "epoch": 0.018598884066955982,
+      "grad_norm": 1.3132269382476807,
+      "learning_rate": 0.0004999074199403377,
+      "loss": 2.0493,
+      "step": 180
+    },
+    {
+      "epoch": 0.019632155404009092,
+      "grad_norm": 1.250373363494873,
+      "learning_rate": 0.000499883769442712,
+      "loss": 2.2234,
+      "step": 190
+    },
+    {
+      "epoch": 0.020665426741062202,
+      "grad_norm": 1.2733889818191528,
+      "learning_rate": 0.0004998574322664343,
+      "loss": 1.9751,
+      "step": 200
+    },
+    {
+      "epoch": 0.02169869807811531,
+      "grad_norm": 0.9156686663627625,
+      "learning_rate": 0.0004998284086946743,
+      "loss": 2.1314,
+      "step": 210
+    },
+    {
+      "epoch": 0.02273196941516842,
+      "grad_norm": 1.0921283960342407,
+      "learning_rate": 0.0004997966990394851,
+      "loss": 1.8402,
+      "step": 220
+    },
+    {
+      "epoch": 0.023765240752221535,
+      "grad_norm": 1.2582165002822876,
+      "learning_rate": 0.0004997623036417998,
+      "loss": 2.1236,
+      "step": 230
+    },
+    {
+      "epoch": 0.024798512089274645,
+      "grad_norm": 1.1085331439971924,
+      "learning_rate": 0.0004997252228714278,
+      "loss": 2.0778,
+      "step": 240
+    },
+    {
+      "epoch": 0.025831783426327755,
+      "grad_norm": 1.096815824508667,
+      "learning_rate": 0.0004996854571270512,
+      "loss": 2.146,
+      "step": 250
+    },
+    {
+      "epoch": 0.026865054763380865,
+      "grad_norm": 1.0710986852645874,
+      "learning_rate": 0.0004996430068362196,
+      "loss": 1.9776,
+      "step": 260
+    },
+    {
+      "epoch": 0.027898326100433975,
+      "grad_norm": 1.4760586023330688,
+      "learning_rate": 0.0004995978724553464,
+      "loss": 1.9536,
+      "step": 270
+    },
+    {
+      "epoch": 0.028931597437487085,
+      "grad_norm": 1.3000599145889282,
+      "learning_rate": 0.0004995500544697036,
+      "loss": 1.7954,
+      "step": 280
+    },
+    {
+      "epoch": 0.029964868774540195,
+      "grad_norm": 1.2158455848693848,
+      "learning_rate": 0.0004994995533934163,
+      "loss": 1.8429,
+      "step": 290
+    },
+    {
+      "epoch": 0.030998140111593304,
+      "grad_norm": 1.1453113555908203,
+      "learning_rate": 0.0004994463697694579,
+      "loss": 2.0769,
+      "step": 300
+    },
+    {
+      "epoch": 0.032031411448646414,
+      "grad_norm": 0.7675238847732544,
+      "learning_rate": 0.0004993905041696433,
+      "loss": 1.903,
+      "step": 310
+    },
+    {
+      "epoch": 0.033064682785699524,
+      "grad_norm": 1.1099224090576172,
+      "learning_rate": 0.0004993319571946235,
+      "loss": 1.9536,
+      "step": 320
+    },
+    {
+      "epoch": 0.034097954122752634,
+      "grad_norm": 1.617551326751709,
+      "learning_rate": 0.0004992707294738786,
+      "loss": 1.9095,
+      "step": 330
+    },
+    {
+      "epoch": 0.035131225459805744,
+      "grad_norm": 1.486296534538269,
+      "learning_rate": 0.0004992068216657115,
+      "loss": 1.8174,
+      "step": 340
+    },
+    {
+      "epoch": 0.036164496796858854,
+      "grad_norm": 1.4763352870941162,
+      "learning_rate": 0.0004991402344572409,
+      "loss": 1.7963,
+      "step": 350
+    },
+    {
+      "epoch": 0.037197768133911964,
+      "grad_norm": 1.2317055463790894,
+      "learning_rate": 0.0004990709685643932,
+      "loss": 1.9319,
+      "step": 360
+    },
+    {
+      "epoch": 0.038231039470965074,
+      "grad_norm": 1.074246883392334,
+      "learning_rate": 0.0004989990247318954,
+      "loss": 1.9447,
+      "step": 370
+    },
+    {
+      "epoch": 0.039264310808018184,
+      "grad_norm": 1.2657012939453125,
+      "learning_rate": 0.0004989244037332671,
+      "loss": 1.7053,
+      "step": 380
+    },
+    {
+      "epoch": 0.040297582145071294,
+      "grad_norm": 1.3143869638442993,
+      "learning_rate": 0.0004988471063708116,
+      "loss": 1.8009,
+      "step": 390
+    },
+    {
+      "epoch": 0.041330853482124404,
+      "grad_norm": 1.381238579750061,
+      "learning_rate": 0.0004987671334756084,
+      "loss": 1.8229,
+      "step": 400
+    },
+    {
+      "epoch": 0.042364124819177514,
+      "grad_norm": 1.2067453861236572,
+      "learning_rate": 0.0004986844859075028,
+      "loss": 1.7401,
+      "step": 410
+    },
+    {
+      "epoch": 0.04339739615623062,
+      "grad_norm": 1.2656382322311401,
+      "learning_rate": 0.0004985991645550979,
+      "loss": 1.6774,
+      "step": 420
+    },
+    {
+      "epoch": 0.04443066749328373,
+      "grad_norm": 0.9306573271751404,
+      "learning_rate": 0.0004985111703357444,
+      "loss": 1.7884,
+      "step": 430
+    },
+    {
+      "epoch": 0.04546393883033684,
+      "grad_norm": 0.9404155611991882,
+      "learning_rate": 0.0004984205041955309,
+      "loss": 1.6522,
+      "step": 440
+    },
+    {
+      "epoch": 0.04649721016738995,
+      "grad_norm": 1.2634170055389404,
+      "learning_rate": 0.0004983271671092738,
+      "loss": 1.7681,
+      "step": 450
+    },
+    {
+      "epoch": 0.04753048150444307,
+      "grad_norm": 1.6461280584335327,
+      "learning_rate": 0.0004982311600805066,
+      "loss": 1.8906,
+      "step": 460
+    },
+    {
+      "epoch": 0.04856375284149618,
+      "grad_norm": 1.242746114730835,
+      "learning_rate": 0.0004981324841414695,
+      "loss": 1.8049,
+      "step": 470
+    },
+    {
+      "epoch": 0.04959702417854929,
+      "grad_norm": 1.0817033052444458,
+      "learning_rate": 0.0004980311403530978,
+      "loss": 1.6272,
+      "step": 480
+    },
+    {
+      "epoch": 0.0506302955156024,
+      "grad_norm": 1.2890691757202148,
+      "learning_rate": 0.0004979271298050107,
+      "loss": 1.6988,
+      "step": 490
+    },
+    {
+      "epoch": 0.05166356685265551,
+      "grad_norm": 1.0929986238479614,
+      "learning_rate": 0.0004978204536155003,
+      "loss": 1.728,
+      "step": 500
+    },
+    {
+      "epoch": 0.05269683818970862,
+      "grad_norm": 1.2483139038085938,
+      "learning_rate": 0.0004977111129315177,
+      "loss": 1.5616,
+      "step": 510
+    },
+    {
+      "epoch": 0.05373010952676173,
+      "grad_norm": 1.0845454931259155,
+      "learning_rate": 0.0004975991089286632,
+      "loss": 1.5409,
+      "step": 520
+    },
+    {
+      "epoch": 0.05476338086381484,
+      "grad_norm": 1.4180911779403687,
+      "learning_rate": 0.0004974844428111715,
+      "loss": 1.7888,
+      "step": 530
+    },
+    {
+      "epoch": 0.05579665220086795,
+      "grad_norm": 1.0860427618026733,
+      "learning_rate": 0.0004973671158118998,
+      "loss": 1.6986,
+      "step": 540
+    },
+    {
+      "epoch": 0.05682992353792106,
+      "grad_norm": 1.1166560649871826,
+      "learning_rate": 0.0004972471291923143,
+      "loss": 1.8097,
+      "step": 550
+    },
+    {
+      "epoch": 0.05786319487497417,
+      "grad_norm": 1.2548375129699707,
+      "learning_rate": 0.0004971244842424768,
+      "loss": 1.6169,
+      "step": 560
+    },
+    {
+      "epoch": 0.05889646621202728,
+      "grad_norm": 0.991226851940155,
+      "learning_rate": 0.0004969991822810307,
+      "loss": 1.6965,
+      "step": 570
+    },
+    {
+      "epoch": 0.05992973754908039,
+      "grad_norm": 1.1663395166397095,
+      "learning_rate": 0.0004968712246551868,
+      "loss": 1.6587,
+      "step": 580
+    },
+    {
+      "epoch": 0.0609630088861335,
+      "grad_norm": 1.3420109748840332,
+      "learning_rate": 0.0004967406127407086,
+      "loss": 1.6837,
+      "step": 590
+    },
+    {
+      "epoch": 0.06199628022318661,
+      "grad_norm": 1.0421655178070068,
+      "learning_rate": 0.0004966073479418982,
+      "loss": 1.637,
+      "step": 600
+    },
+    {
+      "epoch": 0.06302955156023972,
+      "grad_norm": 1.2786924839019775,
+      "learning_rate": 0.0004964714316915803,
+      "loss": 1.5727,
+      "step": 610
+    },
+    {
+      "epoch": 0.06406282289729283,
+      "grad_norm": 1.507780909538269,
+      "learning_rate": 0.0004963328654510877,
+      "loss": 1.6056,
+      "step": 620
+    },
+    {
+      "epoch": 0.06509609423434594,
+      "grad_norm": 1.0936928987503052,
+      "learning_rate": 0.0004961916507102447,
+      "loss": 1.5742,
+      "step": 630
+    },
+    {
+      "epoch": 0.06612936557139905,
+      "grad_norm": 1.2347360849380493,
+      "learning_rate": 0.0004960477889873517,
+      "loss": 1.6692,
+      "step": 640
+    },
+    {
+      "epoch": 0.06716263690845216,
+      "grad_norm": 1.4181699752807617,
+      "learning_rate": 0.0004959012818291688,
+      "loss": 1.6412,
+      "step": 650
+    },
+    {
+      "epoch": 0.06819590824550527,
+      "grad_norm": 1.3150395154953003,
+      "learning_rate": 0.000495752130810899,
+      "loss": 1.4951,
+      "step": 660
+    },
+    {
+      "epoch": 0.06922917958255838,
+      "grad_norm": 1.5858718156814575,
+      "learning_rate": 0.0004956003375361712,
+      "loss": 1.6635,
+      "step": 670
+    },
+    {
+      "epoch": 0.07026245091961149,
+      "grad_norm": 1.1488349437713623,
+      "learning_rate": 0.0004954459036370231,
+      "loss": 1.6101,
+      "step": 680
+    },
+    {
+      "epoch": 0.0712957222566646,
+      "grad_norm": 1.0433565378189087,
+      "learning_rate": 0.0004952888307738839,
+      "loss": 1.7019,
+      "step": 690
+    },
+    {
+      "epoch": 0.07232899359371771,
+      "grad_norm": 1.2093909978866577,
+      "learning_rate": 0.000495129120635556,
+      "loss": 1.6477,
+      "step": 700
+    },
+    {
+      "epoch": 0.07336226493077082,
+      "grad_norm": 1.1268610954284668,
+      "learning_rate": 0.0004949667749391967,
+      "loss": 1.5262,
+      "step": 710
+    },
+    {
+      "epoch": 0.07439553626782393,
+      "grad_norm": 1.0410875082015991,
+      "learning_rate": 0.0004948017954303007,
+      "loss": 1.7201,
+      "step": 720
+    },
+    {
+      "epoch": 0.07542880760487704,
+      "grad_norm": 1.5368582010269165,
+      "learning_rate": 0.0004946341838826803,
+      "loss": 1.7024,
+      "step": 730
+    },
+    {
+      "epoch": 0.07646207894193015,
+      "grad_norm": 0.9913454651832581,
+      "learning_rate": 0.0004944639420984469,
+      "loss": 1.4697,
+      "step": 740
+    },
+    {
+      "epoch": 0.07749535027898326,
+      "grad_norm": 1.1712089776992798,
+      "learning_rate": 0.0004942910719079912,
+      "loss": 1.4866,
+      "step": 750
+    },
+    {
+      "epoch": 0.07852862161603637,
+      "grad_norm": 1.5118815898895264,
+      "learning_rate": 0.0004941155751699639,
+      "loss": 1.6489,
+      "step": 760
+    },
+    {
+      "epoch": 0.07956189295308948,
+      "grad_norm": 0.9319295883178711,
+      "learning_rate": 0.0004939374537712558,
+      "loss": 1.5053,
+      "step": 770
+    },
+    {
+      "epoch": 0.08059516429014259,
+      "grad_norm": 2.0669829845428467,
+      "learning_rate": 0.0004937567096269769,
+      "loss": 1.6412,
+      "step": 780
+    },
+    {
+      "epoch": 0.0816284356271957,
+      "grad_norm": 1.876446008682251,
+      "learning_rate": 0.0004935733446804367,
+      "loss": 1.7186,
+      "step": 790
+    },
+    {
+      "epoch": 0.08266170696424881,
+      "grad_norm": 1.0011149644851685,
+      "learning_rate": 0.0004933873609031224,
+      "loss": 1.4197,
+      "step": 800
+    },
+    {
+      "epoch": 0.08369497830130192,
+      "grad_norm": 84.56254577636719,
+      "learning_rate": 0.000493198760294678,
+      "loss": 2.006,
+      "step": 810
+    },
+    {
+      "epoch": 0.08472824963835503,
+      "grad_norm": 9.466862678527832,
+      "learning_rate": 0.0004930075448828836,
+      "loss": 3.6223,
+      "step": 820
+    },
+    {
+      "epoch": 0.08576152097540814,
+      "grad_norm": 1.8478171825408936,
+      "learning_rate": 0.0004928137167236323,
+      "loss": 1.8625,
+      "step": 830
+    },
+    {
+      "epoch": 0.08679479231246125,
+      "grad_norm": 1.6235121488571167,
+      "learning_rate": 0.0004926172779009089,
+      "loss": 1.6312,
+      "step": 840
+    },
+    {
+      "epoch": 0.08782806364951436,
+      "grad_norm": 1.0582070350646973,
+      "learning_rate": 0.0004924182305267672,
+      "loss": 1.4884,
+      "step": 850
+    },
+    {
+      "epoch": 0.08886133498656747,
+      "grad_norm": 1.6170542240142822,
+      "learning_rate": 0.0004922165767413077,
+      "loss": 1.4772,
+      "step": 860
+    },
+    {
+      "epoch": 0.08989460632362058,
+      "grad_norm": 1.1165497303009033,
+      "learning_rate": 0.0004920123187126539,
+      "loss": 1.396,
+      "step": 870
+    },
+    {
+      "epoch": 0.09092787766067369,
+      "grad_norm": 1.5052683353424072,
+      "learning_rate": 0.0004918054586369299,
+      "loss": 1.7157,
+      "step": 880
+    },
+    {
+      "epoch": 0.0919611489977268,
+      "grad_norm": 1.5653349161148071,
+      "learning_rate": 0.0004915959987382355,
+      "loss": 1.5859,
+      "step": 890
+    },
+    {
+      "epoch": 0.0929944203347799,
+      "grad_norm": 1.4743231534957886,
+      "learning_rate": 0.0004913839412686238,
+      "loss": 1.5155,
+      "step": 900
+    },
+    {
+      "epoch": 0.09402769167183303,
+      "grad_norm": 1.931356430053711,
+      "learning_rate": 0.0004911692885080757,
+      "loss": 1.6259,
+      "step": 910
+    },
+    {
+      "epoch": 0.09506096300888614,
+      "grad_norm": 1.209402322769165,
+      "learning_rate": 0.0004909520427644762,
+      "loss": 1.544,
+      "step": 920
+    },
+    {
+      "epoch": 0.09609423434593925,
+      "grad_norm": 1.3506582975387573,
+      "learning_rate": 0.0004907322063735895,
+      "loss": 1.6404,
+      "step": 930
+    },
+    {
+      "epoch": 0.09712750568299236,
+      "grad_norm": 1.0187022686004639,
+      "learning_rate": 0.0004905097816990331,
+      "loss": 1.4471,
+      "step": 940
+    },
+    {
+      "epoch": 0.09816077702004547,
+      "grad_norm": 1.3293033838272095,
+      "learning_rate": 0.0004902847711322535,
+      "loss": 1.5233,
+      "step": 950
+    },
+    {
+      "epoch": 0.09919404835709858,
+      "grad_norm": 1.0673298835754395,
+      "learning_rate": 0.0004900571770924999,
+      "loss": 1.4875,
+      "step": 960
+    },
+    {
+      "epoch": 0.10022731969415169,
+      "grad_norm": 1.2059558629989624,
+      "learning_rate": 0.000489827002026798,
+      "loss": 1.5143,
+      "step": 970
+    },
+    {
+      "epoch": 0.1012605910312048,
+      "grad_norm": 1.5264915227890015,
+      "learning_rate": 0.0004895942484099241,
+      "loss": 1.3904,
+      "step": 980
+    },
+    {
+      "epoch": 0.10229386236825791,
+      "grad_norm": 1.055259108543396,
+      "learning_rate": 0.0004893589187443786,
+      "loss": 1.5727,
+      "step": 990
+    },
+    {
+      "epoch": 0.10332713370531102,
+      "grad_norm": 1.3388723134994507,
+      "learning_rate": 0.0004891210155603585,
+      "loss": 1.3997,
+      "step": 1000
+    },
+    {
+      "epoch": 0.10436040504236413,
+      "grad_norm": 1.146324872970581,
+      "learning_rate": 0.0004888805414157304,
+      "loss": 1.5489,
+      "step": 1010
+    },
+    {
+      "epoch": 0.10539367637941724,
+      "grad_norm": 1.446535348892212,
+      "learning_rate": 0.0004886374988960036,
+      "loss": 1.5821,
+      "step": 1020
+    },
+    {
+      "epoch": 0.10642694771647035,
+      "grad_norm": 1.3649553060531616,
+      "learning_rate": 0.0004883918906143016,
+      "loss": 1.4677,
+      "step": 1030
+    },
+    {
+      "epoch": 0.10746021905352346,
+      "grad_norm": 1.2442352771759033,
+      "learning_rate": 0.00048814371921133417,
+      "loss": 1.4691,
+      "step": 1040
+    },
+    {
+      "epoch": 0.10849349039057657,
+      "grad_norm": 1.2473433017730713,
+      "learning_rate": 0.00048789298735536904,
+      "loss": 1.6451,
+      "step": 1050
+    },
+    {
+      "epoch": 0.10952676172762968,
+      "grad_norm": 1.7001152038574219,
+      "learning_rate": 0.0004876396977422033,
+      "loss": 1.6799,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11056003306468279,
+      "grad_norm": 1.1802128553390503,
+      "learning_rate": 0.00048738385309513434,
+      "loss": 1.4816,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1115933044017359,
+      "grad_norm": 1.1580235958099365,
+      "learning_rate": 0.0004871254561649303,
+      "loss": 1.681,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11262657573878901,
+      "grad_norm": 1.2179391384124756,
+      "learning_rate": 0.000486864509729801,
+      "loss": 1.507,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11365984707584212,
+      "grad_norm": 1.1689331531524658,
+      "learning_rate": 0.00048660101659536763,
+      "loss": 1.276,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11469311841289523,
+      "grad_norm": 1.1789075136184692,
+      "learning_rate": 0.0004863349795946329,
+      "loss": 1.5065,
+      "step": 1110
+    },
+    {
+      "epoch": 0.11572638974994834,
+      "grad_norm": 0.98433917760849,
+      "learning_rate": 0.00048606640158795034,
+      "loss": 1.286,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11675966108700145,
+      "grad_norm": 0.9510318636894226,
+      "learning_rate": 0.0004857952854629938,
+      "loss": 1.4637,
+      "step": 1130
+    },
+    {
+      "epoch": 0.11779293242405456,
+      "grad_norm": 1.0436633825302124,
+      "learning_rate": 0.0004855216341347259,
+      "loss": 1.4925,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11882620376110767,
+      "grad_norm": 1.4930212497711182,
+      "learning_rate": 0.0004852454505453674,
+      "loss": 1.3586,
+      "step": 1150
+    },
+    {
+      "epoch": 0.11985947509816078,
+      "grad_norm": 0.8224968910217285,
+      "learning_rate": 0.00048496673766436517,
+      "loss": 1.5212,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12089274643521389,
+      "grad_norm": 1.2815650701522827,
+      "learning_rate": 0.00048468549848835996,
+      "loss": 1.4901,
+      "step": 1170
+    },
+    {
+      "epoch": 0.121926017772267,
+      "grad_norm": 1.2067404985427856,
+      "learning_rate": 0.000484401736041155,
+      "loss": 1.5946,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12295928910932011,
+      "grad_norm": 1.530190348625183,
+      "learning_rate": 0.0004841154533736827,
+      "loss": 1.5052,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12399256044637322,
+      "grad_norm": 1.2579395771026611,
+      "learning_rate": 0.0004838266535639722,
+      "loss": 1.4186,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12502583178342633,
+      "grad_norm": 1.5573033094406128,
+      "learning_rate": 0.00048353533971711625,
+      "loss": 1.3459,
+      "step": 1210
+    },
+    {
+      "epoch": 0.12605910312047944,
+      "grad_norm": 1.2861477136611938,
+      "learning_rate": 0.0004832415149652378,
+      "loss": 1.4907,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12709237445753255,
+      "grad_norm": 0.8996832370758057,
+      "learning_rate": 0.00048294518246745643,
+      "loss": 1.3579,
+      "step": 1230
+    },
+    {
+      "epoch": 0.12812564579458566,
+      "grad_norm": 1.4587196111679077,
+      "learning_rate": 0.0004826463454098542,
+      "loss": 1.3771,
+      "step": 1240
+    },
+    {
+      "epoch": 0.12915891713163877,
+      "grad_norm": 1.1128901243209839,
+      "learning_rate": 0.00048234500700544144,
+      "loss": 1.3366,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13019218846869188,
+      "grad_norm": 1.0845853090286255,
+      "learning_rate": 0.0004820411704941223,
+      "loss": 1.3485,
+      "step": 1260
+    },
+    {
+      "epoch": 0.131225459805745,
+      "grad_norm": 1.1594990491867065,
+      "learning_rate": 0.00048173483914265976,
+      "loss": 1.4484,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1322587311427981,
+      "grad_norm": 1.2621511220932007,
+      "learning_rate": 0.0004814260162446408,
+      "loss": 1.4147,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1332920024798512,
+      "grad_norm": 1.6021978855133057,
+      "learning_rate": 0.00048111470512044065,
+      "loss": 1.3724,
+      "step": 1290
+    },
+    {
+      "epoch": 0.13432527381690432,
+      "grad_norm": 1.3557963371276855,
+      "learning_rate": 0.0004808009091171873,
+      "loss": 1.3844,
+      "step": 1300
+    },
+    {
+      "epoch": 0.13535854515395743,
+      "grad_norm": 1.341562271118164,
+      "learning_rate": 0.0004804846316087254,
+      "loss": 1.2895,
+      "step": 1310
+    },
+    {
+      "epoch": 0.13639181649101054,
+      "grad_norm": 1.1911447048187256,
+      "learning_rate": 0.00048016587599557996,
+      "loss": 1.3963,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13742508782806365,
+      "grad_norm": 1.0620297193527222,
+      "learning_rate": 0.0004798446457049201,
+      "loss": 1.3621,
+      "step": 1330
+    },
+    {
+      "epoch": 0.13845835916511676,
+      "grad_norm": 1.1608030796051025,
+      "learning_rate": 0.0004795209441905217,
+      "loss": 1.3658,
+      "step": 1340
+    },
+    {
+      "epoch": 0.13949163050216987,
+      "grad_norm": 1.5855592489242554,
+      "learning_rate": 0.0004791947749327308,
+      "loss": 1.3683,
+      "step": 1350
+    },
+    {
+      "epoch": 0.14052490183922298,
+      "grad_norm": 1.5676512718200684,
+      "learning_rate": 0.0004788661414384255,
+      "loss": 1.3186,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1415581731762761,
+      "grad_norm": 1.149056315422058,
+      "learning_rate": 0.0004785350472409792,
+      "loss": 1.3231,
+      "step": 1370
+    },
+    {
+      "epoch": 0.1425914445133292,
+      "grad_norm": 0.9153057336807251,
+      "learning_rate": 0.00047820149590022153,
+      "loss": 1.2976,
+      "step": 1380
+    },
+    {
+      "epoch": 0.1436247158503823,
+      "grad_norm": 1.2601430416107178,
+      "learning_rate": 0.00047786549100240107,
+      "loss": 1.4205,
+      "step": 1390
+    },
+    {
+      "epoch": 0.14465798718743542,
+      "grad_norm": 1.2112687826156616,
+      "learning_rate": 0.0004775270361601461,
+      "loss": 1.3192,
+      "step": 1400
+    },
+    {
+      "epoch": 0.14569125852448853,
+      "grad_norm": 1.1192986965179443,
+      "learning_rate": 0.00047718613501242604,
+      "loss": 1.355,
+      "step": 1410
+    },
+    {
+      "epoch": 0.14672452986154164,
+      "grad_norm": 1.2099722623825073,
+      "learning_rate": 0.00047684279122451223,
+      "loss": 1.3638,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14775780119859475,
+      "grad_norm": 1.124934434890747,
+      "learning_rate": 0.00047649700848793886,
+      "loss": 1.331,
+      "step": 1430
+    },
+    {
+      "epoch": 0.14879107253564786,
+      "grad_norm": 1.2407630681991577,
+      "learning_rate": 0.00047614879052046254,
+      "loss": 1.5308,
+      "step": 1440
+    },
+    {
+      "epoch": 0.14982434387270097,
+      "grad_norm": 1.2991951704025269,
+      "learning_rate": 0.00047579814106602316,
+      "loss": 1.3365,
+      "step": 1450
+    },
+    {
+      "epoch": 0.15085761520975408,
+      "grad_norm": 1.3195995092391968,
+      "learning_rate": 0.0004754450638947032,
+      "loss": 1.3209,
+      "step": 1460
+    },
+    {
+      "epoch": 0.15189088654680719,
+      "grad_norm": 1.4838430881500244,
+      "learning_rate": 0.000475089562802687,
+      "loss": 1.4705,
+      "step": 1470
+    },
+    {
+      "epoch": 0.1529241578838603,
+      "grad_norm": 1.126265287399292,
+      "learning_rate": 0.0004747316416122205,
+      "loss": 1.3316,
+      "step": 1480
+    },
+    {
+      "epoch": 0.1539574292209134,
+      "grad_norm": 1.0109881162643433,
+      "learning_rate": 0.00047437130417156973,
+      "loss": 1.4045,
+      "step": 1490
+    },
+    {
+      "epoch": 0.15499070055796652,
+      "grad_norm": 0.8355922698974609,
+      "learning_rate": 0.00047400855435497945,
+      "loss": 1.3058,
+      "step": 1500
+    },
+    {
+      "epoch": 0.15602397189501963,
+      "grad_norm": 1.406782865524292,
+      "learning_rate": 0.00047364339606263185,
+      "loss": 1.2732,
+      "step": 1510
+    },
+    {
+      "epoch": 0.15705724323207274,
+      "grad_norm": 0.9180629849433899,
+      "learning_rate": 0.00047327583322060406,
+      "loss": 1.3472,
+      "step": 1520
+    },
+    {
+      "epoch": 0.15809051456912585,
+      "grad_norm": 1.1642565727233887,
+      "learning_rate": 0.00047290586978082654,
+      "loss": 1.4165,
+      "step": 1530
+    },
+    {
+      "epoch": 0.15912378590617896,
+      "grad_norm": 1.6167621612548828,
+      "learning_rate": 0.00047253350972104003,
+      "loss": 1.5038,
+      "step": 1540
+    },
+    {
+      "epoch": 0.16015705724323206,
+      "grad_norm": 1.1963645219802856,
+      "learning_rate": 0.00047215875704475314,
+      "loss": 1.3744,
+      "step": 1550
+    },
+    {
+      "epoch": 0.16119032858028517,
+      "grad_norm": 1.246496319770813,
+      "learning_rate": 0.00047178161578119925,
+      "loss": 1.3314,
+      "step": 1560
+    },
+    {
+      "epoch": 0.16222359991733828,
+      "grad_norm": 1.1050803661346436,
+      "learning_rate": 0.00047140208998529303,
+      "loss": 1.4149,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1632568712543914,
+      "grad_norm": 1.379414439201355,
+      "learning_rate": 0.00047102018373758716,
+      "loss": 1.3817,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1642901425914445,
+      "grad_norm": 1.0111711025238037,
+      "learning_rate": 0.00047063590114422804,
+      "loss": 1.2865,
+      "step": 1590
+    },
+    {
+      "epoch": 0.16532341392849761,
+      "grad_norm": 1.5682810544967651,
+      "learning_rate": 0.00047024924633691197,
+      "loss": 1.317,
+      "step": 1600
+    },
+    {
+      "epoch": 0.16635668526555072,
+      "grad_norm": 0.9005609154701233,
+      "learning_rate": 0.0004698602234728406,
+      "loss": 1.3659,
+      "step": 1610
+    },
+    {
+      "epoch": 0.16738995660260383,
+      "grad_norm": 1.00556218624115,
+      "learning_rate": 0.00046946883673467624,
+      "loss": 1.3373,
+      "step": 1620
+    },
+    {
+      "epoch": 0.16842322793965694,
+      "grad_norm": 1.387304425239563,
+      "learning_rate": 0.00046907509033049683,
+      "loss": 1.4858,
+      "step": 1630
+    },
+    {
+      "epoch": 0.16945649927671005,
+      "grad_norm": 1.2545139789581299,
+      "learning_rate": 0.0004686789884937509,
+      "loss": 1.4221,
+      "step": 1640
+    },
+    {
+      "epoch": 0.17048977061376316,
+      "grad_norm": 1.3092625141143799,
+      "learning_rate": 0.00046828053548321187,
+      "loss": 1.3609,
+      "step": 1650
+    },
+    {
+      "epoch": 0.17152304195081627,
+      "grad_norm": 1.0507373809814453,
+      "learning_rate": 0.0004678797355829322,
+      "loss": 1.3406,
+      "step": 1660
+    },
+    {
+      "epoch": 0.17255631328786938,
+      "grad_norm": 1.0271055698394775,
+      "learning_rate": 0.0004674765931021976,
+      "loss": 1.3098,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1735895846249225,
+      "grad_norm": 1.1777933835983276,
+      "learning_rate": 0.0004670711123754804,
+      "loss": 1.2113,
+      "step": 1680
+    },
+    {
+      "epoch": 0.1746228559619756,
+      "grad_norm": 1.1888736486434937,
+      "learning_rate": 0.00046666329776239306,
+      "loss": 1.3137,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1756561272990287,
+      "grad_norm": 1.4923595190048218,
+      "learning_rate": 0.00046625315364764156,
+      "loss": 1.23,
+      "step": 1700
+    },
+    {
+      "epoch": 0.17668939863608182,
+      "grad_norm": 1.3654290437698364,
+      "learning_rate": 0.00046584068444097766,
+      "loss": 1.3182,
+      "step": 1710
+    },
+    {
+      "epoch": 0.17772266997313493,
+      "grad_norm": 1.5486831665039062,
+      "learning_rate": 0.0004654258945771521,
+      "loss": 1.2854,
+      "step": 1720
+    },
+    {
+      "epoch": 0.17875594131018804,
+      "grad_norm": 1.3204468488693237,
+      "learning_rate": 0.0004650087885158667,
+      "loss": 1.3328,
+      "step": 1730
+    },
+    {
+      "epoch": 0.17978921264724115,
+      "grad_norm": 1.4773411750793457,
+      "learning_rate": 0.000464589370741726,
+      "loss": 1.3741,
+      "step": 1740
+    },
+    {
+      "epoch": 0.18082248398429426,
+      "grad_norm": 1.20607328414917,
+      "learning_rate": 0.00046416764576418993,
+      "loss": 1.26,
+      "step": 1750
+    },
+    {
+      "epoch": 0.18185575532134737,
+      "grad_norm": 1.2402129173278809,
+      "learning_rate": 0.00046374361811752436,
+      "loss": 1.2398,
+      "step": 1760
+    },
+    {
+      "epoch": 0.18288902665840048,
+      "grad_norm": 1.1439170837402344,
+      "learning_rate": 0.00046331729236075305,
+      "loss": 1.4529,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1839222979954536,
+      "grad_norm": 1.268829107284546,
+      "learning_rate": 0.0004628886730776084,
+      "loss": 1.4535,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1849555693325067,
+      "grad_norm": 1.37249755859375,
+      "learning_rate": 0.0004624577648764819,
+      "loss": 1.3249,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1859888406695598,
+      "grad_norm": 1.2565507888793945,
+      "learning_rate": 0.0004620245723903751,
+      "loss": 1.2866,
+      "step": 1800
+    },
+    {
+      "epoch": 0.18702211200661292,
+      "grad_norm": 1.1188476085662842,
+      "learning_rate": 0.0004615891002768494,
+      "loss": 1.1712,
+      "step": 1810
+    },
+    {
+      "epoch": 0.18805538334366606,
+      "grad_norm": 1.3613345623016357,
+      "learning_rate": 0.00046115135321797617,
+      "loss": 1.1788,
+      "step": 1820
+    },
+    {
+      "epoch": 0.18908865468071917,
+      "grad_norm": 1.3465733528137207,
+      "learning_rate": 0.00046071133592028626,
+      "loss": 1.2391,
+      "step": 1830
+    },
+    {
+      "epoch": 0.19012192601777228,
+      "grad_norm": 1.2718288898468018,
+      "learning_rate": 0.00046026905311471946,
+      "loss": 1.2213,
+      "step": 1840
+    },
+    {
+      "epoch": 0.1911551973548254,
+      "grad_norm": 1.372878909111023,
+      "learning_rate": 0.0004598245095565738,
+      "loss": 1.2036,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1921884686918785,
+      "grad_norm": 1.198723316192627,
+      "learning_rate": 0.00045937771002545403,
+      "loss": 1.192,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1932217400289316,
+      "grad_norm": 0.9199055433273315,
+      "learning_rate": 0.00045892865932522077,
+      "loss": 1.2999,
+      "step": 1870
+    },
+    {
+      "epoch": 0.19425501136598472,
+      "grad_norm": 1.3330020904541016,
+      "learning_rate": 0.0004584773622839383,
+      "loss": 1.3587,
+      "step": 1880
+    },
+    {
+      "epoch": 0.19528828270303783,
+      "grad_norm": 1.228247046470642,
+      "learning_rate": 0.0004580238237538232,
+      "loss": 1.4654,
+      "step": 1890
+    },
+    {
+      "epoch": 0.19632155404009094,
+      "grad_norm": 1.6616053581237793,
+      "learning_rate": 0.0004575680486111915,
+      "loss": 1.1583,
+      "step": 1900
+    },
+    {
+      "epoch": 0.19735482537714405,
+      "grad_norm": 1.0149476528167725,
+      "learning_rate": 0.000457110041756407,
+      "loss": 1.2121,
+      "step": 1910
+    },
+    {
+      "epoch": 0.19838809671419716,
+      "grad_norm": 1.6001296043395996,
+      "learning_rate": 0.00045664980811382813,
+      "loss": 1.2938,
+      "step": 1920
+    },
+    {
+      "epoch": 0.19942136805125027,
+      "grad_norm": 1.263951301574707,
+      "learning_rate": 0.00045618735263175495,
+      "loss": 1.1885,
+      "step": 1930
+    },
+    {
+      "epoch": 0.20045463938830338,
+      "grad_norm": 1.4672034978866577,
+      "learning_rate": 0.00045572268028237637,
+      "loss": 1.2981,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2014879107253565,
+      "grad_norm": 1.1751412153244019,
+      "learning_rate": 0.0004552557960617161,
+      "loss": 1.117,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2025211820624096,
+      "grad_norm": 1.131339192390442,
+      "learning_rate": 0.0004547867049895796,
+      "loss": 1.1909,
+      "step": 1960
+    },
+    {
+      "epoch": 0.2035544533994627,
+      "grad_norm": 0.991729736328125,
+      "learning_rate": 0.0004543154121094996,
+      "loss": 1.1328,
+      "step": 1970
+    },
+    {
+      "epoch": 0.20458772473651582,
+      "grad_norm": 0.9073534607887268,
+      "learning_rate": 0.0004538419224886819,
+      "loss": 0.9973,
+      "step": 1980
+    },
+    {
+      "epoch": 0.20562099607356893,
+      "grad_norm": 1.2480231523513794,
+      "learning_rate": 0.00045336624121795124,
+      "loss": 1.2367,
+      "step": 1990
+    },
+    {
+      "epoch": 0.20665426741062204,
+      "grad_norm": 1.28248131275177,
+      "learning_rate": 0.0004528883734116963,
+      "loss": 1.1738,
+      "step": 2000
+    },
+    {
+      "epoch": 0.20768753874767515,
+      "grad_norm": 1.386885166168213,
+      "learning_rate": 0.00045240832420781474,
+      "loss": 1.3788,
+      "step": 2010
+    },
+    {
+      "epoch": 0.20872081008472826,
+      "grad_norm": 1.1158126592636108,
+      "learning_rate": 0.0004519260987676579,
+      "loss": 1.3046,
+      "step": 2020
+    },
+    {
+      "epoch": 0.20975408142178137,
+      "grad_norm": 1.162636399269104,
+      "learning_rate": 0.00045144170227597545,
+      "loss": 1.2355,
+      "step": 2030
+    },
+    {
+      "epoch": 0.21078735275883448,
+      "grad_norm": 1.4077121019363403,
+      "learning_rate": 0.0004509551399408598,
+      "loss": 1.1542,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2118206240958876,
+      "grad_norm": 1.1237014532089233,
+      "learning_rate": 0.00045046641699368953,
+      "loss": 1.3411,
+      "step": 2050
+    },
+    {
+      "epoch": 0.2128538954329407,
+      "grad_norm": 1.1503592729568481,
+      "learning_rate": 0.0004499755386890736,
+      "loss": 1.1783,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2138871667699938,
+      "grad_norm": 1.2852717638015747,
+      "learning_rate": 0.0004494825103047949,
+      "loss": 1.3331,
+      "step": 2070
+    },
+    {
+      "epoch": 0.21492043810704692,
+      "grad_norm": 1.2611557245254517,
+      "learning_rate": 0.0004489873371417534,
+      "loss": 1.3272,
+      "step": 2080
+    },
+    {
+      "epoch": 0.21595370944410003,
+      "grad_norm": 1.4016764163970947,
+      "learning_rate": 0.00044849002452390874,
+      "loss": 1.1658,
+      "step": 2090
+    },
+    {
+      "epoch": 0.21698698078115314,
+      "grad_norm": 1.4920654296875,
+      "learning_rate": 0.0004479905777982238,
+      "loss": 1.2651,
+      "step": 2100
+    },
+    {
+      "epoch": 0.21802025211820625,
+      "grad_norm": 1.1262540817260742,
+      "learning_rate": 0.0004474890023346066,
+      "loss": 1.2347,
+      "step": 2110
+    },
+    {
+      "epoch": 0.21905352345525936,
+      "grad_norm": 1.2998685836791992,
+      "learning_rate": 0.0004469853035258526,
+      "loss": 1.3167,
+      "step": 2120
+    },
+    {
+      "epoch": 0.22008679479231247,
+      "grad_norm": 1.0999696254730225,
+      "learning_rate": 0.0004464794867875871,
+      "loss": 1.1621,
+      "step": 2130
+    },
+    {
+      "epoch": 0.22112006612936558,
+      "grad_norm": 0.9801570177078247,
+      "learning_rate": 0.0004459715575582066,
+      "loss": 1.1631,
+      "step": 2140
+    },
+    {
+      "epoch": 0.2221533374664187,
+      "grad_norm": 1.161098837852478,
+      "learning_rate": 0.00044546152129882054,
+      "loss": 1.1256,
+      "step": 2150
+    },
+    {
+      "epoch": 0.2231866088034718,
+      "grad_norm": 1.2714890241622925,
+      "learning_rate": 0.0004449493834931927,
+      "loss": 1.1182,
+      "step": 2160
+    },
+    {
+      "epoch": 0.2242198801405249,
+      "grad_norm": 0.9561129808425903,
+      "learning_rate": 0.0004444351496476818,
+      "loss": 1.2051,
+      "step": 2170
+    },
+    {
+      "epoch": 0.22525315147757802,
+      "grad_norm": 1.3160920143127441,
+      "learning_rate": 0.0004439188252911828,
+      "loss": 1.2268,
+      "step": 2180
+    },
+    {
+      "epoch": 0.22628642281463113,
+      "grad_norm": 1.1478712558746338,
+      "learning_rate": 0.0004434004159750671,
+      "loss": 1.1928,
+      "step": 2190
+    },
+    {
+      "epoch": 0.22731969415168424,
+      "grad_norm": 1.2426934242248535,
+      "learning_rate": 0.0004428799272731231,
+      "loss": 1.1548,
+      "step": 2200
+    },
+    {
+      "epoch": 0.22835296548873735,
+      "grad_norm": 1.3587703704833984,
+      "learning_rate": 0.0004423573647814961,
+      "loss": 1.1371,
+      "step": 2210
+    },
+    {
+      "epoch": 0.22938623682579046,
+      "grad_norm": 1.0638786554336548,
+      "learning_rate": 0.0004418327341186282,
+      "loss": 1.1668,
+      "step": 2220
+    },
+    {
+      "epoch": 0.23041950816284357,
+      "grad_norm": 1.1600587368011475,
+      "learning_rate": 0.00044130604092519794,
+      "loss": 1.1524,
+      "step": 2230
+    },
+    {
+      "epoch": 0.23145277949989668,
+      "grad_norm": 1.402921199798584,
+      "learning_rate": 0.0004407772908640595,
+      "loss": 1.3963,
+      "step": 2240
+    },
+    {
+      "epoch": 0.2324860508369498,
+      "grad_norm": 1.3043389320373535,
+      "learning_rate": 0.00044024648962018184,
+      "loss": 1.2359,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2335193221740029,
+      "grad_norm": 1.1807844638824463,
+      "learning_rate": 0.0004397136429005879,
+      "loss": 1.366,
+      "step": 2260
+    },
+    {
+      "epoch": 0.234552593511056,
+      "grad_norm": 1.397064208984375,
+      "learning_rate": 0.00043917875643429284,
+      "loss": 1.234,
+      "step": 2270
+    },
+    {
+      "epoch": 0.23558586484810912,
+      "grad_norm": 1.3520923852920532,
+      "learning_rate": 0.0004386418359722426,
+      "loss": 1.1863,
+      "step": 2280
+    },
+    {
+      "epoch": 0.23661913618516223,
+      "grad_norm": 1.3684437274932861,
+      "learning_rate": 0.00043810288728725203,
+      "loss": 1.3008,
+      "step": 2290
+    },
+    {
+      "epoch": 0.23765240752221534,
+      "grad_norm": 1.1893550157546997,
+      "learning_rate": 0.0004375619161739428,
+      "loss": 1.0295,
+      "step": 2300
+    },
+    {
+      "epoch": 0.23868567885926845,
+      "grad_norm": 1.57694673538208,
+      "learning_rate": 0.0004370189284486814,
+      "loss": 1.2778,
+      "step": 2310
+    },
+    {
+      "epoch": 0.23971895019632156,
+      "grad_norm": 1.601629376411438,
+      "learning_rate": 0.0004364739299495162,
+      "loss": 1.0583,
+      "step": 2320
+    },
+    {
+      "epoch": 0.24075222153337467,
+      "grad_norm": 1.3246372938156128,
+      "learning_rate": 0.00043592692653611465,
+      "loss": 1.201,
+      "step": 2330
+    },
+    {
+      "epoch": 0.24178549287042778,
+      "grad_norm": 1.020665168762207,
+      "learning_rate": 0.0004353779240897008,
+      "loss": 1.3127,
+      "step": 2340
+    },
+    {
+      "epoch": 0.24281876420748089,
+      "grad_norm": 1.082587480545044,
+      "learning_rate": 0.00043482692851299146,
+      "loss": 1.225,
+      "step": 2350
+    },
+    {
+      "epoch": 0.243852035544534,
+      "grad_norm": 1.093636155128479,
+      "learning_rate": 0.00043427394573013314,
+      "loss": 1.0587,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2448853068815871,
+      "grad_norm": 1.3400568962097168,
+      "learning_rate": 0.00043371898168663816,
+      "loss": 1.2375,
+      "step": 2370
+    },
+    {
+      "epoch": 0.24591857821864022,
+      "grad_norm": 1.223877191543579,
+      "learning_rate": 0.00043316204234932084,
+      "loss": 1.3076,
+      "step": 2380
+    },
+    {
+      "epoch": 0.24695184955569333,
+      "grad_norm": 1.4696872234344482,
+      "learning_rate": 0.0004326031337062333,
+      "loss": 1.251,
+      "step": 2390
+    },
+    {
+      "epoch": 0.24798512089274644,
+      "grad_norm": 1.3883171081542969,
+      "learning_rate": 0.00043204226176660107,
+      "loss": 1.2215,
+      "step": 2400
+    },
+    {
+      "epoch": 0.24901839222979955,
+      "grad_norm": 1.1045054197311401,
+      "learning_rate": 0.0004314794325607584,
+      "loss": 1.1014,
+      "step": 2410
+    },
+    {
+      "epoch": 0.25005166356685266,
+      "grad_norm": 1.273431658744812,
+      "learning_rate": 0.00043091465214008354,
+      "loss": 1.0845,
+      "step": 2420
+    },
+    {
+      "epoch": 0.25108493490390577,
+      "grad_norm": 1.0618531703948975,
+      "learning_rate": 0.0004303479265769337,
+      "loss": 1.1162,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2521182062409589,
+      "grad_norm": 1.3230130672454834,
+      "learning_rate": 0.0004297792619645797,
+      "loss": 1.0385,
+      "step": 2440
+    },
+    {
+      "epoch": 0.253151477578012,
+      "grad_norm": 1.3020095825195312,
+      "learning_rate": 0.0004292086644171403,
+      "loss": 1.1548,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2541847489150651,
+      "grad_norm": 1.3359605073928833,
+      "learning_rate": 0.0004286361400695169,
+      "loss": 1.3534,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2552180202521182,
+      "grad_norm": 0.9598666429519653,
+      "learning_rate": 0.00042806169507732706,
+      "loss": 1.2432,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2562512915891713,
+      "grad_norm": 0.9668743014335632,
+      "learning_rate": 0.00042748533561683865,
+      "loss": 1.2512,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2572845629262244,
+      "grad_norm": 1.0706520080566406,
+      "learning_rate": 0.0004269070678849034,
+      "loss": 1.136,
+      "step": 2490
+    },
+    {
+      "epoch": 0.25831783426327753,
+      "grad_norm": 1.0470647811889648,
+      "learning_rate": 0.00042632689809889027,
+      "loss": 1.1211,
+      "step": 2500
+    },
+    {
+      "epoch": 0.25935110560033064,
+      "grad_norm": 1.1314566135406494,
+      "learning_rate": 0.0004257448324966183,
+      "loss": 1.0704,
+      "step": 2510
+    },
+    {
+      "epoch": 0.26038437693738375,
+      "grad_norm": 1.085749626159668,
+      "learning_rate": 0.00042516087733629004,
+      "loss": 1.1727,
+      "step": 2520
+    },
+    {
+      "epoch": 0.26141764827443686,
+      "grad_norm": 1.2759054899215698,
+      "learning_rate": 0.00042457503889642396,
+      "loss": 1.2864,
+      "step": 2530
+    },
+    {
+      "epoch": 0.26245091961149,
+      "grad_norm": 1.2468546628952026,
+      "learning_rate": 0.0004239873234757871,
+      "loss": 1.0734,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2634841909485431,
+      "grad_norm": 1.2701605558395386,
+      "learning_rate": 0.00042339773739332706,
+      "loss": 1.1816,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2645174622855962,
+      "grad_norm": 1.0027897357940674,
+      "learning_rate": 0.00042280628698810447,
+      "loss": 1.0726,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2655507336226493,
+      "grad_norm": 1.4462809562683105,
+      "learning_rate": 0.0004222129786192245,
+      "loss": 1.278,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2665840049597024,
+      "grad_norm": 1.241769552230835,
+      "learning_rate": 0.0004216178186657686,
+      "loss": 1.1829,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2676172762967555,
+      "grad_norm": 1.6216635704040527,
+      "learning_rate": 0.00042102081352672594,
+      "loss": 1.1479,
+      "step": 2590
+    },
+    {
+      "epoch": 0.26865054763380863,
+      "grad_norm": 4.33831787109375,
+      "learning_rate": 0.0004204219696209248,
+      "loss": 1.1709,
+      "step": 2600
+    },
+    {
+      "epoch": 0.26968381897086174,
+      "grad_norm": 0.952393114566803,
+      "learning_rate": 0.00041982129338696296,
+      "loss": 1.1018,
+      "step": 2610
+    },
+    {
+      "epoch": 0.27071709030791485,
+      "grad_norm": 1.4484752416610718,
+      "learning_rate": 0.0004192187912831393,
+      "loss": 1.081,
+      "step": 2620
+    },
+    {
+      "epoch": 0.27175036164496796,
+      "grad_norm": 1.0843290090560913,
+      "learning_rate": 0.0004186144697873835,
+      "loss": 1.251,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2727836329820211,
+      "grad_norm": 1.5165774822235107,
+      "learning_rate": 0.0004180083353971871,
+      "loss": 1.134,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2738169043190742,
+      "grad_norm": 1.2171909809112549,
+      "learning_rate": 0.00041740039462953343,
+      "loss": 1.1763,
+      "step": 2650
+    },
+    {
+      "epoch": 0.2748501756561273,
+      "grad_norm": 1.0561660528182983,
+      "learning_rate": 0.00041679065402082724,
+      "loss": 1.1178,
+      "step": 2660
+    },
+    {
+      "epoch": 0.2758834469931804,
+      "grad_norm": 1.2738356590270996,
+      "learning_rate": 0.0004161791201268247,
+      "loss": 1.2379,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2769167183302335,
+      "grad_norm": 1.127266764640808,
+      "learning_rate": 0.000415565799522563,
+      "loss": 1.2347,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2779499896672866,
+      "grad_norm": 1.4078705310821533,
+      "learning_rate": 0.0004149506988022894,
+      "loss": 1.2155,
+      "step": 2690
+    },
+    {
+      "epoch": 0.27898326100433973,
+      "grad_norm": 0.9871230125427246,
+      "learning_rate": 0.0004143338245793906,
+      "loss": 1.1583,
+      "step": 2700
+    },
+    {
+      "epoch": 0.28001653234139284,
+      "grad_norm": 0.8731614947319031,
+      "learning_rate": 0.0004137151834863213,
+      "loss": 1.171,
+      "step": 2710
+    },
+    {
+      "epoch": 0.28104980367844595,
+      "grad_norm": 1.0814101696014404,
+      "learning_rate": 0.0004130947821745332,
+      "loss": 1.2112,
+      "step": 2720
+    },
+    {
+      "epoch": 0.28208307501549906,
+      "grad_norm": 1.695854902267456,
+      "learning_rate": 0.0004124726273144034,
+      "loss": 1.0269,
+      "step": 2730
+    },
+    {
+      "epoch": 0.2831163463525522,
+      "grad_norm": 0.9004189372062683,
+      "learning_rate": 0.00041184872559516244,
+      "loss": 1.1974,
+      "step": 2740
+    },
+    {
+      "epoch": 0.2841496176896053,
+      "grad_norm": 1.034590482711792,
+      "learning_rate": 0.0004112230837248229,
+      "loss": 1.1068,
+      "step": 2750
+    },
+    {
+      "epoch": 0.2851828890266584,
+      "grad_norm": 1.4124444723129272,
+      "learning_rate": 0.0004105957084301069,
+      "loss": 1.0922,
+      "step": 2760
+    },
+    {
+      "epoch": 0.2862161603637115,
+      "grad_norm": 1.4991061687469482,
+      "learning_rate": 0.00040996660645637345,
+      "loss": 1.2392,
+      "step": 2770
+    },
+    {
+      "epoch": 0.2872494317007646,
+      "grad_norm": 1.316277027130127,
+      "learning_rate": 0.00040933578456754687,
+      "loss": 1.0433,
+      "step": 2780
+    },
+    {
+      "epoch": 0.2882827030378177,
+      "grad_norm": 1.6624524593353271,
+      "learning_rate": 0.00040870324954604323,
+      "loss": 0.9623,
+      "step": 2790
+    },
+    {
+      "epoch": 0.28931597437487083,
+      "grad_norm": 1.2658953666687012,
+      "learning_rate": 0.00040806900819269766,
+      "loss": 1.108,
+      "step": 2800
+    },
+    {
+      "epoch": 0.29034924571192394,
+      "grad_norm": 1.3295769691467285,
+      "learning_rate": 0.00040743306732669133,
+      "loss": 1.137,
+      "step": 2810
+    },
+    {
+      "epoch": 0.29138251704897705,
+      "grad_norm": 1.6526551246643066,
+      "learning_rate": 0.0004067954337854783,
+      "loss": 1.1281,
+      "step": 2820
+    },
+    {
+      "epoch": 0.29241578838603016,
+      "grad_norm": 1.1587738990783691,
+      "learning_rate": 0.0004061561144247115,
+      "loss": 1.0086,
+      "step": 2830
+    },
+    {
+      "epoch": 0.29344905972308327,
+      "grad_norm": 1.454256534576416,
+      "learning_rate": 0.0004055151161181694,
+      "loss": 1.0723,
+      "step": 2840
+    },
+    {
+      "epoch": 0.2944823310601364,
+      "grad_norm": 1.476521611213684,
+      "learning_rate": 0.00040487244575768193,
+      "loss": 1.2486,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2955156023971895,
+      "grad_norm": 0.9193058609962463,
+      "learning_rate": 0.00040422811025305676,
+      "loss": 1.1071,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2965488737342426,
+      "grad_norm": 1.8984912633895874,
+      "learning_rate": 0.0004035821165320045,
+      "loss": 1.1724,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2975821450712957,
+      "grad_norm": 1.7232333421707153,
+      "learning_rate": 0.00040293447154006435,
+      "loss": 1.174,
+      "step": 2880
+    },
+    {
+      "epoch": 0.2986154164083488,
+      "grad_norm": 0.9074342846870422,
+      "learning_rate": 0.0004022851822405297,
+      "loss": 1.2448,
+      "step": 2890
+    },
+    {
+      "epoch": 0.29964868774540193,
+      "grad_norm": 1.2896414995193481,
+      "learning_rate": 0.0004016342556143727,
+      "loss": 1.1207,
+      "step": 2900
+    },
+    {
+      "epoch": 0.30068195908245504,
+      "grad_norm": 1.264952301979065,
+      "learning_rate": 0.0004009816986601702,
+      "loss": 1.0868,
+      "step": 2910
+    },
+    {
+      "epoch": 0.30171523041950815,
+      "grad_norm": 1.1345736980438232,
+      "learning_rate": 0.00040032751839402727,
+      "loss": 1.1881,
+      "step": 2920
+    },
+    {
+      "epoch": 0.30274850175656126,
+      "grad_norm": 1.0581121444702148,
+      "learning_rate": 0.00039967172184950263,
+      "loss": 1.1796,
+      "step": 2930
+    },
+    {
+      "epoch": 0.30378177309361437,
+      "grad_norm": 1.1877319812774658,
+      "learning_rate": 0.0003990143160775327,
+      "loss": 1.108,
+      "step": 2940
+    },
+    {
+      "epoch": 0.3048150444306675,
+      "grad_norm": 1.0478935241699219,
+      "learning_rate": 0.0003983553081463559,
+      "loss": 0.9986,
+      "step": 2950
+    },
+    {
+      "epoch": 0.3058483157677206,
+      "grad_norm": 1.1299808025360107,
+      "learning_rate": 0.0003976947051414366,
+      "loss": 1.0221,
+      "step": 2960
+    },
+    {
+      "epoch": 0.3068815871047737,
+      "grad_norm": 1.0618709325790405,
+      "learning_rate": 0.00039703251416538883,
+      "loss": 0.9024,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3079148584418268,
+      "grad_norm": 1.099435806274414,
+      "learning_rate": 0.0003963687423379003,
+      "loss": 1.2754,
+      "step": 2980
+    },
+    {
+      "epoch": 0.3089481297788799,
+      "grad_norm": 1.1851657629013062,
+      "learning_rate": 0.00039570339679565527,
+      "loss": 1.1154,
+      "step": 2990
+    },
+    {
+      "epoch": 0.30998140111593303,
+      "grad_norm": 1.1812468767166138,
+      "learning_rate": 0.00039503648469225826,
+      "loss": 0.9979,
+      "step": 3000
+    },
+    {
+      "epoch": 0.31101467245298614,
+      "grad_norm": 1.1941215991973877,
+      "learning_rate": 0.00039436801319815717,
+      "loss": 1.0625,
+      "step": 3010
+    },
+    {
+      "epoch": 0.31204794379003925,
+      "grad_norm": 0.9969115853309631,
+      "learning_rate": 0.00039369798950056567,
+      "loss": 1.189,
+      "step": 3020
+    },
+    {
+      "epoch": 0.31308121512709236,
+      "grad_norm": 1.1364282369613647,
+      "learning_rate": 0.00039302642080338664,
+      "loss": 1.107,
+      "step": 3030
+    },
+    {
+      "epoch": 0.31411448646414547,
+      "grad_norm": 1.234078288078308,
+      "learning_rate": 0.0003923533143271341,
+      "loss": 1.238,
+      "step": 3040
+    },
+    {
+      "epoch": 0.3151477578011986,
+      "grad_norm": 1.0862330198287964,
+      "learning_rate": 0.0003916786773088559,
+      "loss": 1.1151,
+      "step": 3050
+    },
+    {
+      "epoch": 0.3161810291382517,
+      "grad_norm": 0.7923426032066345,
+      "learning_rate": 0.0003910025170020558,
+      "loss": 1.1031,
+      "step": 3060
+    },
+    {
+      "epoch": 0.3172143004753048,
+      "grad_norm": 1.6818230152130127,
+      "learning_rate": 0.0003903248406766158,
+      "loss": 1.0546,
+      "step": 3070
+    },
+    {
+      "epoch": 0.3182475718123579,
+      "grad_norm": 1.0749332904815674,
+      "learning_rate": 0.00038964565561871723,
+      "loss": 1.1412,
+      "step": 3080
+    },
+    {
+      "epoch": 0.319280843149411,
+      "grad_norm": 1.6896083354949951,
+      "learning_rate": 0.00038896496913076327,
+      "loss": 1.0668,
+      "step": 3090
+    },
+    {
+      "epoch": 0.32031411448646413,
+      "grad_norm": 1.24112069606781,
+      "learning_rate": 0.0003882827885312998,
+      "loss": 1.2632,
+      "step": 3100
+    },
+    {
+      "epoch": 0.32134738582351724,
+      "grad_norm": 1.2017937898635864,
+      "learning_rate": 0.00038759912115493724,
+      "loss": 1.0969,
+      "step": 3110
+    },
+    {
+      "epoch": 0.32238065716057035,
+      "grad_norm": 1.20210599899292,
+      "learning_rate": 0.00038691397435227097,
+      "loss": 0.9202,
+      "step": 3120
+    },
+    {
+      "epoch": 0.32341392849762346,
+      "grad_norm": 1.408257246017456,
+      "learning_rate": 0.0003862273554898032,
+      "loss": 1.1583,
+      "step": 3130
+    },
+    {
+      "epoch": 0.32444719983467657,
+      "grad_norm": 1.5274341106414795,
+      "learning_rate": 0.000385539271949863,
+      "loss": 1.0382,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3254804711717297,
+      "grad_norm": 1.1086093187332153,
+      "learning_rate": 0.00038484973113052736,
+      "loss": 1.1232,
+      "step": 3150
+    },
+    {
+      "epoch": 0.3265137425087828,
+      "grad_norm": 1.35590660572052,
+      "learning_rate": 0.0003841587404455413,
+      "loss": 1.1019,
+      "step": 3160
+    },
+    {
+      "epoch": 0.3275470138458359,
+      "grad_norm": 1.0598667860031128,
+      "learning_rate": 0.000383466307324239,
+      "loss": 1.1557,
+      "step": 3170
+    },
+    {
+      "epoch": 0.328580285182889,
+      "grad_norm": 1.1116646528244019,
+      "learning_rate": 0.00038277243921146253,
+      "loss": 1.0634,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3296135565199421,
+      "grad_norm": 1.4738123416900635,
+      "learning_rate": 0.0003820771435674829,
+      "loss": 1.0067,
+      "step": 3190
+    },
+    {
+      "epoch": 0.33064682785699523,
+      "grad_norm": 1.3569648265838623,
+      "learning_rate": 0.00038138042786791987,
+      "loss": 1.0798,
+      "step": 3200
+    },
+    {
+      "epoch": 0.33168009919404834,
+      "grad_norm": 1.073232650756836,
+      "learning_rate": 0.00038068229960366054,
+      "loss": 1.0349,
+      "step": 3210
+    },
+    {
+      "epoch": 0.33271337053110145,
+      "grad_norm": 1.4980833530426025,
+      "learning_rate": 0.0003799827662807801,
+      "loss": 1.1281,
+      "step": 3220
+    },
+    {
+      "epoch": 0.33374664186815456,
+      "grad_norm": 1.4308544397354126,
+      "learning_rate": 0.0003792818354204601,
+      "loss": 1.1056,
+      "step": 3230
+    },
+    {
+      "epoch": 0.33477991320520767,
+      "grad_norm": 1.5736769437789917,
+      "learning_rate": 0.0003785795145589085,
+      "loss": 1.0928,
+      "step": 3240
+    },
+    {
+      "epoch": 0.3358131845422608,
+      "grad_norm": 1.5231635570526123,
+      "learning_rate": 0.0003778758112472776,
+      "loss": 1.2857,
+      "step": 3250
+    },
+    {
+      "epoch": 0.3368464558793139,
+      "grad_norm": 1.0374634265899658,
+      "learning_rate": 0.00037717073305158376,
+      "loss": 1.1738,
+      "step": 3260
+    },
+    {
+      "epoch": 0.337879727216367,
+      "grad_norm": 1.1238751411437988,
+      "learning_rate": 0.0003764642875526256,
+      "loss": 0.9849,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3389129985534201,
+      "grad_norm": 1.039632797241211,
+      "learning_rate": 0.0003757564823459025,
+      "loss": 1.059,
+      "step": 3280
+    },
+    {
+      "epoch": 0.3399462698904732,
+      "grad_norm": 1.1179864406585693,
+      "learning_rate": 0.0003750473250415334,
+      "loss": 0.9745,
+      "step": 3290
+    },
+    {
+      "epoch": 0.34097954122752633,
+      "grad_norm": 1.1467711925506592,
+      "learning_rate": 0.0003743368232641741,
+      "loss": 1.1077,
+      "step": 3300
+    },
+    {
+      "epoch": 0.34201281256457944,
+      "grad_norm": 1.0984638929367065,
+      "learning_rate": 0.00037362498465293604,
+      "loss": 1.0454,
+      "step": 3310
+    },
+    {
+      "epoch": 0.34304608390163255,
+      "grad_norm": 1.0687668323516846,
+      "learning_rate": 0.000372911816861304,
+      "loss": 1.073,
+      "step": 3320
+    },
+    {
+      "epoch": 0.34407935523868566,
+      "grad_norm": 1.6367201805114746,
+      "learning_rate": 0.00037219732755705353,
+      "loss": 1.1094,
+      "step": 3330
+    },
+    {
+      "epoch": 0.34511262657573877,
+      "grad_norm": 1.2905350923538208,
+      "learning_rate": 0.0003714815244221689,
+      "loss": 1.1536,
+      "step": 3340
+    },
+    {
+      "epoch": 0.3461458979127919,
+      "grad_norm": 0.9039587378501892,
+      "learning_rate": 0.00037076441515276003,
+      "loss": 1.1906,
+      "step": 3350
+    },
+    {
+      "epoch": 0.347179169249845,
+      "grad_norm": 1.3063287734985352,
+      "learning_rate": 0.0003700460074589804,
+      "loss": 1.1448,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3482124405868981,
+      "grad_norm": 1.5665302276611328,
+      "learning_rate": 0.00036932630906494346,
+      "loss": 1.2537,
+      "step": 3370
+    },
+    {
+      "epoch": 0.3492457119239512,
+      "grad_norm": 1.0483524799346924,
+      "learning_rate": 0.0003686053277086401,
+      "loss": 1.1135,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3502789832610043,
+      "grad_norm": 1.2478828430175781,
+      "learning_rate": 0.0003678830711418551,
+      "loss": 1.0816,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3513122545980574,
+      "grad_norm": 1.14175283908844,
+      "learning_rate": 0.00036715954713008406,
+      "loss": 1.1217,
+      "step": 3400
+    },
+    {
+      "epoch": 0.35234552593511054,
+      "grad_norm": 1.0000718832015991,
+      "learning_rate": 0.0003664347634524497,
+      "loss": 1.1127,
+      "step": 3410
+    },
+    {
+      "epoch": 0.35337879727216365,
+      "grad_norm": 1.5337084531784058,
+      "learning_rate": 0.00036570872790161834,
+      "loss": 1.0619,
+      "step": 3420
+    },
+    {
+      "epoch": 0.35441206860921676,
+      "grad_norm": 0.878587543964386,
+      "learning_rate": 0.00036498144828371604,
+      "loss": 1.1094,
+      "step": 3430
+    },
+    {
+      "epoch": 0.35544533994626987,
+      "grad_norm": 1.3913145065307617,
+      "learning_rate": 0.0003642529324182449,
+      "loss": 0.9874,
+      "step": 3440
+    },
+    {
+      "epoch": 0.356478611283323,
+      "grad_norm": 1.388934850692749,
+      "learning_rate": 0.0003635231881379985,
+      "loss": 1.0177,
+      "step": 3450
+    },
+    {
+      "epoch": 0.3575118826203761,
+      "grad_norm": 1.2259199619293213,
+      "learning_rate": 0.00036279222328897826,
+      "loss": 1.0533,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3585451539574292,
+      "grad_norm": 1.561023235321045,
+      "learning_rate": 0.0003620600457303085,
+      "loss": 1.1231,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3595784252944823,
+      "grad_norm": 1.6612300872802734,
+      "learning_rate": 0.0003613266633341528,
+      "loss": 1.0971,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3606116966315354,
+      "grad_norm": 1.3272353410720825,
+      "learning_rate": 0.00036059208398562793,
+      "loss": 1.024,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3616449679685885,
+      "grad_norm": 1.2441381216049194,
+      "learning_rate": 0.00035985631558272074,
+      "loss": 1.0369,
+      "step": 3500
+    },
+    {
+      "epoch": 0.36267823930564164,
+      "grad_norm": 1.1805377006530762,
+      "learning_rate": 0.00035911936603620225,
+      "loss": 0.9626,
+      "step": 3510
+    },
+    {
+      "epoch": 0.36371151064269475,
+      "grad_norm": 1.3905142545700073,
+      "learning_rate": 0.00035838124326954254,
+      "loss": 1.1415,
+      "step": 3520
+    },
+    {
+      "epoch": 0.36474478197974786,
+      "grad_norm": 1.2646347284317017,
+      "learning_rate": 0.0003576419552188261,
+      "loss": 1.0101,
+      "step": 3530
+    },
+    {
+      "epoch": 0.36577805331680097,
+      "grad_norm": 1.1667832136154175,
+      "learning_rate": 0.000356901509832666,
+      "loss": 0.9862,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3668113246538541,
+      "grad_norm": 0.8571506142616272,
+      "learning_rate": 0.00035615991507211887,
+      "loss": 0.9644,
+      "step": 3550
+    },
+    {
+      "epoch": 0.3678445959909072,
+      "grad_norm": 0.9799628853797913,
+      "learning_rate": 0.00035541717891059886,
+      "loss": 0.8918,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3688778673279603,
+      "grad_norm": 1.0068583488464355,
+      "learning_rate": 0.0003546733093337924,
+      "loss": 0.935,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3699111386650134,
+      "grad_norm": 0.8947041034698486,
+      "learning_rate": 0.0003539283143395719,
+      "loss": 1.0752,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3709444100020665,
+      "grad_norm": 1.0797079801559448,
+      "learning_rate": 0.00035318220193790976,
+      "loss": 0.9534,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3719776813391196,
+      "grad_norm": 1.0746815204620361,
+      "learning_rate": 0.0003524349801507929,
+      "loss": 0.9862,
+      "step": 3600
+    },
+    {
+      "epoch": 0.37301095267617274,
+      "grad_norm": 1.0419459342956543,
+      "learning_rate": 0.0003516866570121356,
+      "loss": 1.0606,
+      "step": 3610
+    },
+    {
+      "epoch": 0.37404422401322585,
+      "grad_norm": 1.1848483085632324,
+      "learning_rate": 0.0003509372405676937,
+      "loss": 0.97,
+      "step": 3620
+    },
+    {
+      "epoch": 0.375077495350279,
+      "grad_norm": 1.2240206003189087,
+      "learning_rate": 0.00035018673887497807,
+      "loss": 1.0237,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3761107666873321,
+      "grad_norm": 1.6243517398834229,
+      "learning_rate": 0.0003494351600031677,
+      "loss": 1.0748,
+      "step": 3640
+    },
+    {
+      "epoch": 0.37714403802438523,
+      "grad_norm": 1.4265719652175903,
+      "learning_rate": 0.00034868251203302314,
+      "loss": 1.0639,
+      "step": 3650
+    },
+    {
+      "epoch": 0.37817730936143834,
+      "grad_norm": 1.0866336822509766,
+      "learning_rate": 0.00034792880305679967,
+      "loss": 0.9635,
+      "step": 3660
+    },
+    {
+      "epoch": 0.37921058069849145,
+      "grad_norm": 1.0087511539459229,
+      "learning_rate": 0.00034717404117816,
+      "loss": 1.1082,
+      "step": 3670
+    },
+    {
+      "epoch": 0.38024385203554456,
+      "grad_norm": 0.9332922697067261,
+      "learning_rate": 0.0003464182345120877,
+      "loss": 1.0258,
+      "step": 3680
+    },
+    {
+      "epoch": 0.38127712337259767,
+      "grad_norm": 0.9739323854446411,
+      "learning_rate": 0.000345661391184799,
+      "loss": 1.0168,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3823103947096508,
+      "grad_norm": 0.912225067615509,
+      "learning_rate": 0.00034490351933365673,
+      "loss": 1.0996,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3833436660467039,
+      "grad_norm": 0.853449285030365,
+      "learning_rate": 0.0003441446271070816,
+      "loss": 0.9759,
+      "step": 3710
+    },
+    {
+      "epoch": 0.384376937383757,
+      "grad_norm": 1.4455146789550781,
+      "learning_rate": 0.0003433847226644653,
+      "loss": 0.9917,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3854102087208101,
+      "grad_norm": 0.9426379799842834,
+      "learning_rate": 0.0003426238141760826,
+      "loss": 1.1549,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3864434800578632,
+      "grad_norm": 1.3207985162734985,
+      "learning_rate": 0.00034186190982300323,
+      "loss": 1.0742,
+      "step": 3740
+    },
+    {
+      "epoch": 0.38747675139491633,
+      "grad_norm": 0.9599164128303528,
+      "learning_rate": 0.00034109901779700474,
+      "loss": 1.0011,
+      "step": 3750
+    },
+    {
+      "epoch": 0.38851002273196944,
+      "grad_norm": 1.0271940231323242,
+      "learning_rate": 0.0003403351463004831,
+      "loss": 0.9508,
+      "step": 3760
+    },
+    {
+      "epoch": 0.38954329406902255,
+      "grad_norm": 1.680550456047058,
+      "learning_rate": 0.0003395703035463659,
+      "loss": 1.134,
+      "step": 3770
+    },
+    {
+      "epoch": 0.39057656540607566,
+      "grad_norm": 1.4388446807861328,
+      "learning_rate": 0.00033880449775802306,
+      "loss": 1.1811,
+      "step": 3780
+    },
+    {
+      "epoch": 0.39160983674312877,
+      "grad_norm": 1.1560338735580444,
+      "learning_rate": 0.00033803773716917896,
+      "loss": 0.8677,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3926431080801819,
+      "grad_norm": 1.414915680885315,
+      "learning_rate": 0.00033727003002382346,
+      "loss": 1.0965,
+      "step": 3800
+    },
+    {
+      "epoch": 0.393676379417235,
+      "grad_norm": 1.1397249698638916,
+      "learning_rate": 0.0003365013845761239,
+      "loss": 1.0057,
+      "step": 3810
+    },
+    {
+      "epoch": 0.3947096507542881,
+      "grad_norm": 1.470551609992981,
+      "learning_rate": 0.00033573180909033567,
+      "loss": 1.1104,
+      "step": 3820
+    },
+    {
+      "epoch": 0.3957429220913412,
+      "grad_norm": 0.9758129119873047,
+      "learning_rate": 0.0003349613118407138,
+      "loss": 1.0244,
+      "step": 3830
+    },
+    {
+      "epoch": 0.3967761934283943,
+      "grad_norm": 0.9483941793441772,
+      "learning_rate": 0.0003341899011114238,
+      "loss": 1.0993,
+      "step": 3840
+    },
+    {
+      "epoch": 0.39780946476544743,
+      "grad_norm": 1.0213124752044678,
+      "learning_rate": 0.00033341758519645273,
+      "loss": 0.9506,
+      "step": 3850
+    },
+    {
+      "epoch": 0.39884273610250054,
+      "grad_norm": 1.0135273933410645,
+      "learning_rate": 0.0003326443723995199,
+      "loss": 0.993,
+      "step": 3860
+    },
+    {
+      "epoch": 0.39987600743955365,
+      "grad_norm": 0.9621232748031616,
+      "learning_rate": 0.0003318702710339875,
+      "loss": 0.9969,
+      "step": 3870
+    },
+    {
+      "epoch": 0.40090927877660676,
+      "grad_norm": 1.1387019157409668,
+      "learning_rate": 0.0003310952894227718,
+      "loss": 1.0203,
+      "step": 3880
+    },
+    {
+      "epoch": 0.40194255011365987,
+      "grad_norm": 1.406870722770691,
+      "learning_rate": 0.0003303194358982526,
+      "loss": 1.0166,
+      "step": 3890
+    },
+    {
+      "epoch": 0.402975821450713,
+      "grad_norm": 1.339579701423645,
+      "learning_rate": 0.00032954271880218465,
+      "loss": 0.9749,
+      "step": 3900
+    },
+    {
+      "epoch": 0.4040090927877661,
+      "grad_norm": 1.2406803369522095,
+      "learning_rate": 0.00032876514648560757,
+      "loss": 0.943,
+      "step": 3910
+    },
+    {
+      "epoch": 0.4050423641248192,
+      "grad_norm": 1.172455072402954,
+      "learning_rate": 0.00032798672730875603,
+      "loss": 0.9874,
+      "step": 3920
+    },
+    {
+      "epoch": 0.4060756354618723,
+      "grad_norm": 1.2804698944091797,
+      "learning_rate": 0.0003272074696409697,
+      "loss": 0.8747,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4071089067989254,
+      "grad_norm": 1.3023015260696411,
+      "learning_rate": 0.00032642738186060375,
+      "loss": 1.0686,
+      "step": 3940
+    },
+    {
+      "epoch": 0.40814217813597853,
+      "grad_norm": 1.0053542852401733,
+      "learning_rate": 0.00032564647235493833,
+      "loss": 1.074,
+      "step": 3950
+    },
+    {
+      "epoch": 0.40917544947303164,
+      "grad_norm": 1.0846545696258545,
+      "learning_rate": 0.0003248647495200886,
+      "loss": 1.0043,
+      "step": 3960
+    },
+    {
+      "epoch": 0.41020872081008475,
+      "grad_norm": 1.0012749433517456,
+      "learning_rate": 0.00032408222176091427,
+      "loss": 1.0302,
+      "step": 3970
+    },
+    {
+      "epoch": 0.41124199214713786,
+      "grad_norm": 1.074317455291748,
+      "learning_rate": 0.0003232988974909296,
+      "loss": 0.9389,
+      "step": 3980
+    },
+    {
+      "epoch": 0.41227526348419097,
+      "grad_norm": 1.5110340118408203,
+      "learning_rate": 0.00032251478513221254,
+      "loss": 0.9772,
+      "step": 3990
+    },
+    {
+      "epoch": 0.4133085348212441,
+      "grad_norm": 0.9564982652664185,
+      "learning_rate": 0.00032172989311531427,
+      "loss": 1.0754,
+      "step": 4000
+    },
+    {
+      "epoch": 0.4143418061582972,
+      "grad_norm": 1.091377854347229,
+      "learning_rate": 0.0003209442298791689,
+      "loss": 0.9343,
+      "step": 4010
+    },
+    {
+      "epoch": 0.4153750774953503,
+      "grad_norm": 1.3888299465179443,
+      "learning_rate": 0.00032015780387100227,
+      "loss": 0.9448,
+      "step": 4020
+    },
+    {
+      "epoch": 0.4164083488324034,
+      "grad_norm": 0.9059675335884094,
+      "learning_rate": 0.0003193706235462412,
+      "loss": 0.9877,
+      "step": 4030
+    },
+    {
+      "epoch": 0.4174416201694565,
+      "grad_norm": 1.5517827272415161,
+      "learning_rate": 0.0003185826973684231,
+      "loss": 1.0814,
+      "step": 4040
+    },
+    {
+      "epoch": 0.4184748915065096,
+      "grad_norm": 1.1768487691879272,
+      "learning_rate": 0.00031779403380910427,
+      "loss": 1.0654,
+      "step": 4050
+    },
+    {
+      "epoch": 0.41950816284356274,
+      "grad_norm": 1.0996865034103394,
+      "learning_rate": 0.0003170046413477692,
+      "loss": 1.0507,
+      "step": 4060
+    },
+    {
+      "epoch": 0.42054143418061585,
+      "grad_norm": 1.7401593923568726,
+      "learning_rate": 0.00031621452847173923,
+      "loss": 1.1611,
+      "step": 4070
+    },
+    {
+      "epoch": 0.42157470551766896,
+      "grad_norm": 1.4233949184417725,
+      "learning_rate": 0.00031542370367608176,
+      "loss": 1.0043,
+      "step": 4080
+    },
+    {
+      "epoch": 0.42260797685472207,
+      "grad_norm": 1.1471216678619385,
+      "learning_rate": 0.00031463217546351803,
+      "loss": 1.0389,
+      "step": 4090
+    },
+    {
+      "epoch": 0.4236412481917752,
+      "grad_norm": 0.9187898635864258,
+      "learning_rate": 0.0003138399523443326,
+      "loss": 0.9048,
+      "step": 4100
+    },
+    {
+      "epoch": 0.4246745195288283,
+      "grad_norm": 1.0712145566940308,
+      "learning_rate": 0.0003130470428362811,
+      "loss": 0.9538,
+      "step": 4110
+    },
+    {
+      "epoch": 0.4257077908658814,
+      "grad_norm": 1.3978677988052368,
+      "learning_rate": 0.0003122534554644995,
+      "loss": 0.9093,
+      "step": 4120
+    },
+    {
+      "epoch": 0.4267410622029345,
+      "grad_norm": 0.9700504541397095,
+      "learning_rate": 0.0003114591987614117,
+      "loss": 0.984,
+      "step": 4130
+    },
+    {
+      "epoch": 0.4277743335399876,
+      "grad_norm": 1.1521259546279907,
+      "learning_rate": 0.0003106642812666379,
+      "loss": 0.9965,
+      "step": 4140
+    },
+    {
+      "epoch": 0.4288076048770407,
+      "grad_norm": 1.2085431814193726,
+      "learning_rate": 0.0003098687115269034,
+      "loss": 0.9018,
+      "step": 4150
+    },
+    {
+      "epoch": 0.42984087621409384,
+      "grad_norm": 1.1607948541641235,
+      "learning_rate": 0.0003090724980959457,
+      "loss": 1.0776,
+      "step": 4160
+    },
+    {
+      "epoch": 0.43087414755114695,
+      "grad_norm": 1.4290366172790527,
+      "learning_rate": 0.0003082756495344234,
+      "loss": 0.9008,
+      "step": 4170
+    },
+    {
+      "epoch": 0.43190741888820006,
+      "grad_norm": 1.3393192291259766,
+      "learning_rate": 0.00030747817440982385,
+      "loss": 0.9175,
+      "step": 4180
+    },
+    {
+      "epoch": 0.43294069022525317,
+      "grad_norm": 1.4794620275497437,
+      "learning_rate": 0.00030668008129637104,
+      "loss": 1.0164,
+      "step": 4190
+    },
+    {
+      "epoch": 0.4339739615623063,
+      "grad_norm": 0.5630869269371033,
+      "learning_rate": 0.00030588137877493317,
+      "loss": 1.0511,
+      "step": 4200
+    },
+    {
+      "epoch": 0.4350072328993594,
+      "grad_norm": 1.0285252332687378,
+      "learning_rate": 0.0003050820754329309,
+      "loss": 0.8834,
+      "step": 4210
+    },
+    {
+      "epoch": 0.4360405042364125,
+      "grad_norm": 1.4534152746200562,
+      "learning_rate": 0.00030428217986424444,
+      "loss": 1.0248,
+      "step": 4220
+    },
+    {
+      "epoch": 0.4370737755734656,
+      "grad_norm": 1.7549463510513306,
+      "learning_rate": 0.0003034817006691219,
+      "loss": 0.9278,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4381070469105187,
+      "grad_norm": 1.1008073091506958,
+      "learning_rate": 0.0003026806464540859,
+      "loss": 1.0172,
+      "step": 4240
+    },
+    {
+      "epoch": 0.4391403182475718,
+      "grad_norm": 1.042400598526001,
+      "learning_rate": 0.0003018790258318417,
+      "loss": 0.9703,
+      "step": 4250
+    },
+    {
+      "epoch": 0.44017358958462494,
+      "grad_norm": 1.428467035293579,
+      "learning_rate": 0.00030107684742118466,
+      "loss": 0.9187,
+      "step": 4260
+    },
+    {
+      "epoch": 0.44120686092167805,
+      "grad_norm": 0.8269082903862,
+      "learning_rate": 0.0003002741198469068,
+      "loss": 0.8501,
+      "step": 4270
+    },
+    {
+      "epoch": 0.44224013225873116,
+      "grad_norm": 1.0836302042007446,
+      "learning_rate": 0.00029947085173970506,
+      "loss": 0.9252,
+      "step": 4280
+    },
+    {
+      "epoch": 0.44327340359578427,
+      "grad_norm": 0.8747875690460205,
+      "learning_rate": 0.00029866705173608776,
+      "loss": 0.8576,
+      "step": 4290
+    },
+    {
+      "epoch": 0.4443066749328374,
+      "grad_norm": 1.261354923248291,
+      "learning_rate": 0.0002978627284782821,
+      "loss": 0.8515,
+      "step": 4300
+    },
+    {
+      "epoch": 0.4453399462698905,
+      "grad_norm": 1.2424492835998535,
+      "learning_rate": 0.0002970578906141411,
+      "loss": 0.9274,
+      "step": 4310
+    },
+    {
+      "epoch": 0.4463732176069436,
+      "grad_norm": 1.6803146600723267,
+      "learning_rate": 0.00029625254679705094,
+      "loss": 1.1236,
+      "step": 4320
+    },
+    {
+      "epoch": 0.4474064889439967,
+      "grad_norm": 1.284347653388977,
+      "learning_rate": 0.00029544670568583734,
+      "loss": 0.9698,
+      "step": 4330
+    },
+    {
+      "epoch": 0.4484397602810498,
+      "grad_norm": 0.7693723440170288,
+      "learning_rate": 0.00029464037594467284,
+      "loss": 0.9518,
+      "step": 4340
+    },
+    {
+      "epoch": 0.4494730316181029,
+      "grad_norm": 1.0315730571746826,
+      "learning_rate": 0.00029383356624298383,
+      "loss": 1.0873,
+      "step": 4350
+    },
+    {
+      "epoch": 0.45050630295515603,
+      "grad_norm": 0.8411365747451782,
+      "learning_rate": 0.0002930262852553568,
+      "loss": 1.0025,
+      "step": 4360
+    },
+    {
+      "epoch": 0.45153957429220914,
+      "grad_norm": 0.8453785181045532,
+      "learning_rate": 0.0002922185416614456,
+      "loss": 0.8811,
+      "step": 4370
+    },
+    {
+      "epoch": 0.45257284562926225,
+      "grad_norm": 1.2322771549224854,
+      "learning_rate": 0.0002914103441458775,
+      "loss": 1.0246,
+      "step": 4380
+    },
+    {
+      "epoch": 0.45360611696631536,
+      "grad_norm": 1.2167178392410278,
+      "learning_rate": 0.0002906017013981609,
+      "loss": 0.9645,
+      "step": 4390
+    },
+    {
+      "epoch": 0.4546393883033685,
+      "grad_norm": 1.0350911617279053,
+      "learning_rate": 0.0002897926221125906,
+      "loss": 0.9242,
+      "step": 4400
+    },
+    {
+      "epoch": 0.4556726596404216,
+      "grad_norm": 1.1375268697738647,
+      "learning_rate": 0.0002889831149881551,
+      "loss": 0.945,
+      "step": 4410
+    },
+    {
+      "epoch": 0.4567059309774747,
+      "grad_norm": 1.209620475769043,
+      "learning_rate": 0.0002881731887284429,
+      "loss": 1.0773,
+      "step": 4420
+    },
+    {
+      "epoch": 0.4577392023145278,
+      "grad_norm": 1.2932454347610474,
+      "learning_rate": 0.0002873628520415491,
+      "loss": 0.8517,
+      "step": 4430
+    },
+    {
+      "epoch": 0.4587724736515809,
+      "grad_norm": 1.0030730962753296,
+      "learning_rate": 0.0002865521136399814,
+      "loss": 0.8911,
+      "step": 4440
+    },
+    {
+      "epoch": 0.459805744988634,
+      "grad_norm": 0.9559937119483948,
+      "learning_rate": 0.0002857409822405666,
+      "loss": 0.9505,
+      "step": 4450
+    },
+    {
+      "epoch": 0.46083901632568713,
+      "grad_norm": 0.934845507144928,
+      "learning_rate": 0.00028492946656435704,
+      "loss": 1.0858,
+      "step": 4460
+    },
+    {
+      "epoch": 0.46187228766274024,
+      "grad_norm": 1.2203209400177002,
+      "learning_rate": 0.00028411757533653657,
+      "loss": 0.9919,
+      "step": 4470
+    },
+    {
+      "epoch": 0.46290555899979335,
+      "grad_norm": 1.3204402923583984,
+      "learning_rate": 0.000283305317286327,
+      "loss": 0.9198,
+      "step": 4480
+    },
+    {
+      "epoch": 0.46393883033684646,
+      "grad_norm": 0.7426913380622864,
+      "learning_rate": 0.0002824927011468939,
+      "loss": 1.0825,
+      "step": 4490
+    },
+    {
+      "epoch": 0.4649721016738996,
+      "grad_norm": 0.8991280198097229,
+      "learning_rate": 0.00028167973565525325,
+      "loss": 0.9528,
+      "step": 4500
+    },
+    {
+      "epoch": 0.4660053730109527,
+      "grad_norm": 0.697215735912323,
+      "learning_rate": 0.00028086642955217666,
+      "loss": 0.8118,
+      "step": 4510
+    },
+    {
+      "epoch": 0.4670386443480058,
+      "grad_norm": 1.4364171028137207,
+      "learning_rate": 0.0002800527915820984,
+      "loss": 0.8903,
+      "step": 4520
+    },
+    {
+      "epoch": 0.4680719156850589,
+      "grad_norm": 0.9677620530128479,
+      "learning_rate": 0.00027923883049302067,
+      "loss": 0.9083,
+      "step": 4530
+    },
+    {
+      "epoch": 0.469105187022112,
+      "grad_norm": 1.240580439567566,
+      "learning_rate": 0.00027842455503641965,
+      "loss": 0.8643,
+      "step": 4540
+    },
+    {
+      "epoch": 0.4701384583591651,
+      "grad_norm": 1.102919101715088,
+      "learning_rate": 0.00027760997396715167,
+      "loss": 1.0802,
+      "step": 4550
+    },
+    {
+      "epoch": 0.47117172969621823,
+      "grad_norm": 0.921911895275116,
+      "learning_rate": 0.0002767950960433588,
+      "loss": 0.8717,
+      "step": 4560
+    },
+    {
+      "epoch": 0.47220500103327134,
+      "grad_norm": 1.3429372310638428,
+      "learning_rate": 0.000275979930026375,
+      "loss": 0.9584,
+      "step": 4570
+    },
+    {
+      "epoch": 0.47323827237032445,
+      "grad_norm": 1.0819973945617676,
+      "learning_rate": 0.0002751644846806315,
+      "loss": 0.7652,
+      "step": 4580
+    },
+    {
+      "epoch": 0.47427154370737756,
+      "grad_norm": 0.9730139970779419,
+      "learning_rate": 0.00027434876877356287,
+      "loss": 0.8873,
+      "step": 4590
+    },
+    {
+      "epoch": 0.4753048150444307,
+      "grad_norm": 0.9837844371795654,
+      "learning_rate": 0.00027353279107551276,
+      "loss": 0.9581,
+      "step": 4600
+    },
+    {
+      "epoch": 0.4763380863814838,
+      "grad_norm": 1.1134988069534302,
+      "learning_rate": 0.0002727165603596394,
+      "loss": 0.9409,
+      "step": 4610
+    },
+    {
+      "epoch": 0.4773713577185369,
+      "grad_norm": 0.7778913378715515,
+      "learning_rate": 0.00027190008540182135,
+      "loss": 0.9099,
+      "step": 4620
+    },
+    {
+      "epoch": 0.47840462905559,
+      "grad_norm": 1.166568636894226,
+      "learning_rate": 0.00027108337498056356,
+      "loss": 0.8565,
+      "step": 4630
+    },
+    {
+      "epoch": 0.4794379003926431,
+      "grad_norm": 0.7446126341819763,
+      "learning_rate": 0.0002702664378769021,
+      "loss": 0.8251,
+      "step": 4640
+    },
+    {
+      "epoch": 0.4804711717296962,
+      "grad_norm": 1.5371274948120117,
+      "learning_rate": 0.0002694492828743106,
+      "loss": 0.9475,
+      "step": 4650
+    },
+    {
+      "epoch": 0.48150444306674933,
+      "grad_norm": 0.7517269253730774,
+      "learning_rate": 0.0002686319187586054,
+      "loss": 0.8798,
+      "step": 4660
+    },
+    {
+      "epoch": 0.48253771440380244,
+      "grad_norm": 1.4352338314056396,
+      "learning_rate": 0.00026781435431785116,
+      "loss": 1.0462,
+      "step": 4670
+    },
+    {
+      "epoch": 0.48357098574085555,
+      "grad_norm": 1.1265496015548706,
+      "learning_rate": 0.0002669965983422663,
+      "loss": 0.9107,
+      "step": 4680
+    },
+    {
+      "epoch": 0.48460425707790866,
+      "grad_norm": 0.8634279370307922,
+      "learning_rate": 0.00026617865962412865,
+      "loss": 0.8822,
+      "step": 4690
+    },
+    {
+      "epoch": 0.48563752841496177,
+      "grad_norm": 1.0112096071243286,
+      "learning_rate": 0.0002653605469576809,
+      "loss": 0.8842,
+      "step": 4700
+    },
+    {
+      "epoch": 0.4866707997520149,
+      "grad_norm": 0.8251860737800598,
+      "learning_rate": 0.0002645422691390358,
+      "loss": 0.908,
+      "step": 4710
+    },
+    {
+      "epoch": 0.487704071089068,
+      "grad_norm": 0.7483717203140259,
+      "learning_rate": 0.0002637238349660819,
+      "loss": 0.8912,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4887373424261211,
+      "grad_norm": 0.8353511691093445,
+      "learning_rate": 0.0002629052532383888,
+      "loss": 0.7702,
+      "step": 4730
+    },
+    {
+      "epoch": 0.4897706137631742,
+      "grad_norm": 0.9910038113594055,
+      "learning_rate": 0.00026208653275711265,
+      "loss": 0.8797,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4908038851002273,
+      "grad_norm": 1.1162399053573608,
+      "learning_rate": 0.00026126768232490115,
+      "loss": 0.9161,
+      "step": 4750
+    },
+    {
+      "epoch": 0.49183715643728043,
+      "grad_norm": 0.8416900038719177,
+      "learning_rate": 0.00026044871074579955,
+      "loss": 1.0406,
+      "step": 4760
+    },
+    {
+      "epoch": 0.49287042777433354,
+      "grad_norm": 1.2481132745742798,
+      "learning_rate": 0.0002596296268251556,
+      "loss": 1.0545,
+      "step": 4770
+    },
+    {
+      "epoch": 0.49390369911138665,
+      "grad_norm": 1.115768551826477,
+      "learning_rate": 0.0002588104393695245,
+      "loss": 0.9128,
+      "step": 4780
+    },
+    {
+      "epoch": 0.49493697044843976,
+      "grad_norm": 0.7396090030670166,
+      "learning_rate": 0.0002579911571865752,
+      "loss": 1.0822,
+      "step": 4790
+    },
+    {
+      "epoch": 0.49597024178549287,
+      "grad_norm": 0.9519025087356567,
+      "learning_rate": 0.0002571717890849946,
+      "loss": 0.8659,
+      "step": 4800
+    },
+    {
+      "epoch": 0.497003513122546,
+      "grad_norm": 0.7343184947967529,
+      "learning_rate": 0.0002563523438743939,
+      "loss": 0.8773,
+      "step": 4810
+    },
+    {
+      "epoch": 0.4980367844595991,
+      "grad_norm": 1.4175959825515747,
+      "learning_rate": 0.0002555328303652129,
+      "loss": 0.7956,
+      "step": 4820
+    },
+    {
+      "epoch": 0.4990700557966522,
+      "grad_norm": 1.0520236492156982,
+      "learning_rate": 0.000254713257368626,
+      "loss": 0.9604,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5001033271337053,
+      "grad_norm": 0.7733720541000366,
+      "learning_rate": 0.0002538936336964471,
+      "loss": 0.7715,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5011365984707584,
+      "grad_norm": 1.4437015056610107,
+      "learning_rate": 0.0002530739681610349,
+      "loss": 0.9329,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5021698698078115,
+      "grad_norm": 0.892091691493988,
+      "learning_rate": 0.00025225426957519825,
+      "loss": 0.8434,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5032031411448646,
+      "grad_norm": 0.9304814338684082,
+      "learning_rate": 0.0002514345467521014,
+      "loss": 0.8912,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5042364124819178,
+      "grad_norm": 1.063408374786377,
+      "learning_rate": 0.00025061480850516914,
+      "loss": 0.9901,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5052696838189709,
+      "grad_norm": 1.0481808185577393,
+      "learning_rate": 0.00024979506364799207,
+      "loss": 0.89,
+      "step": 4890
+    },
+    {
+      "epoch": 0.506302955156024,
+      "grad_norm": 1.2767094373703003,
+      "learning_rate": 0.000248975320994232,
+      "loss": 1.1239,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5073362264930771,
+      "grad_norm": 1.0354382991790771,
+      "learning_rate": 0.00024815558935752677,
+      "loss": 0.8866,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5083694978301302,
+      "grad_norm": 0.7483660578727722,
+      "learning_rate": 0.0002473358775513959,
+      "loss": 0.8817,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5094027691671833,
+      "grad_norm": 0.6301681399345398,
+      "learning_rate": 0.000246516194389146,
+      "loss": 0.8991,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5104360405042364,
+      "grad_norm": 1.0671608448028564,
+      "learning_rate": 0.0002456965486837752,
+      "loss": 0.7891,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5114693118412895,
+      "grad_norm": 1.3131459951400757,
+      "learning_rate": 0.00024487694924787935,
+      "loss": 0.9388,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5125025831783426,
+      "grad_norm": 1.2082393169403076,
+      "learning_rate": 0.00024405740489355634,
+      "loss": 0.8558,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5135358545153957,
+      "grad_norm": 0.915107786655426,
+      "learning_rate": 0.0002432379244323124,
+      "loss": 0.8703,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5145691258524488,
+      "grad_norm": 1.2518354654312134,
+      "learning_rate": 0.00024241851667496645,
+      "loss": 0.9959,
+      "step": 4980
+    },
+    {
+      "epoch": 0.515602397189502,
+      "grad_norm": 1.0495281219482422,
+      "learning_rate": 0.00024159919043155556,
+      "loss": 1.0497,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5166356685265551,
+      "grad_norm": 1.469854712486267,
+      "learning_rate": 0.0002407799545112409,
+      "loss": 0.7946,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5176689398636082,
+      "grad_norm": 0.7165640592575073,
+      "learning_rate": 0.0002399608177222122,
+      "loss": 0.7768,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5187022112006613,
+      "grad_norm": 0.7148188352584839,
+      "learning_rate": 0.00023914178887159327,
+      "loss": 0.8901,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5197354825377144,
+      "grad_norm": 1.2880741357803345,
+      "learning_rate": 0.0002383228767653474,
+      "loss": 1.0241,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5207687538747675,
+      "grad_norm": 1.3181527853012085,
+      "learning_rate": 0.0002375040902081831,
+      "loss": 0.9248,
+      "step": 5040
+    },
+    {
+      "epoch": 0.5218020252118206,
+      "grad_norm": 1.5515258312225342,
+      "learning_rate": 0.00023668543800345852,
+      "loss": 0.8286,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5228352965488737,
+      "grad_norm": 1.0438510179519653,
+      "learning_rate": 0.0002358669289530875,
+      "loss": 0.7546,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5238685678859268,
+      "grad_norm": 1.179828405380249,
+      "learning_rate": 0.00023504857185744453,
+      "loss": 0.8099,
+      "step": 5070
+    },
+    {
+      "epoch": 0.52490183922298,
+      "grad_norm": 1.0748212337493896,
+      "learning_rate": 0.00023423037551527085,
+      "loss": 0.8692,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5259351105600331,
+      "grad_norm": 0.8671577572822571,
+      "learning_rate": 0.00023341234872357872,
+      "loss": 0.8423,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5269683818970862,
+      "grad_norm": 1.147182583808899,
+      "learning_rate": 0.00023259450027755764,
+      "loss": 0.8823,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5280016532341393,
+      "grad_norm": 1.0983150005340576,
+      "learning_rate": 0.0002317768389704799,
+      "loss": 0.879,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5290349245711924,
+      "grad_norm": 1.5440839529037476,
+      "learning_rate": 0.00023095937359360525,
+      "loss": 0.9303,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5300681959082455,
+      "grad_norm": 1.0175645351409912,
+      "learning_rate": 0.00023014211293608718,
+      "loss": 0.8823,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5311014672452986,
+      "grad_norm": 1.1853641271591187,
+      "learning_rate": 0.00022932506578487782,
+      "loss": 0.9467,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5321347385823517,
+      "grad_norm": 0.9132509827613831,
+      "learning_rate": 0.00022850824092463413,
+      "loss": 0.7992,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5331680099194048,
+      "grad_norm": 0.8958359360694885,
+      "learning_rate": 0.0002276916471376227,
+      "loss": 0.9003,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5342012812564579,
+      "grad_norm": 0.9055427312850952,
+      "learning_rate": 0.00022687529320362587,
+      "loss": 0.8922,
+      "step": 5170
+    },
+    {
+      "epoch": 0.535234552593511,
+      "grad_norm": 1.2413357496261597,
+      "learning_rate": 0.00022605918789984707,
+      "loss": 0.8021,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5362678239305642,
+      "grad_norm": 1.096956729888916,
+      "learning_rate": 0.00022524334000081664,
+      "loss": 0.7335,
+      "step": 5190
+    },
+    {
+      "epoch": 0.5373010952676173,
+      "grad_norm": 1.0181100368499756,
+      "learning_rate": 0.00022442775827829722,
+      "loss": 0.9217,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5383343666046704,
+      "grad_norm": 0.9444778561592102,
+      "learning_rate": 0.0002236124515011897,
+      "loss": 0.8444,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5393676379417235,
+      "grad_norm": 0.9442963004112244,
+      "learning_rate": 0.00022279742843543887,
+      "loss": 0.7901,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5404009092787766,
+      "grad_norm": 1.3280583620071411,
+      "learning_rate": 0.00022198269784393904,
+      "loss": 0.9,
+      "step": 5230
+    },
+    {
+      "epoch": 0.5414341806158297,
+      "grad_norm": 0.7821478843688965,
+      "learning_rate": 0.00022116826848643994,
+      "loss": 0.8782,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5424674519528828,
+      "grad_norm": 0.942866861820221,
+      "learning_rate": 0.00022035414911945252,
+      "loss": 0.8268,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5435007232899359,
+      "grad_norm": 1.0075554847717285,
+      "learning_rate": 0.0002195403484961549,
+      "loss": 0.7923,
+      "step": 5260
+    },
+    {
+      "epoch": 0.544533994626989,
+      "grad_norm": 1.0974137783050537,
+      "learning_rate": 0.00021872687536629806,
+      "loss": 0.8108,
+      "step": 5270
+    },
+    {
+      "epoch": 0.5455672659640421,
+      "grad_norm": 0.8785361647605896,
+      "learning_rate": 0.00021791373847611186,
+      "loss": 0.8236,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5466005373010953,
+      "grad_norm": 1.300614356994629,
+      "learning_rate": 0.00021710094656821098,
+      "loss": 0.9875,
+      "step": 5290
+    },
+    {
+      "epoch": 0.5476338086381484,
+      "grad_norm": 0.9605757594108582,
+      "learning_rate": 0.0002162885083815011,
+      "loss": 0.8824,
+      "step": 5300
+    },
+    {
+      "epoch": 0.5486670799752015,
+      "grad_norm": 0.8986537456512451,
+      "learning_rate": 0.0002154764326510847,
+      "loss": 0.9041,
+      "step": 5310
+    },
+    {
+      "epoch": 0.5497003513122546,
+      "grad_norm": 0.8150350451469421,
+      "learning_rate": 0.0002146647281081671,
+      "loss": 0.8553,
+      "step": 5320
+    },
+    {
+      "epoch": 0.5507336226493077,
+      "grad_norm": 1.396665334701538,
+      "learning_rate": 0.00021385340347996297,
+      "loss": 0.9266,
+      "step": 5330
+    },
+    {
+      "epoch": 0.5517668939863608,
+      "grad_norm": 1.1270840167999268,
+      "learning_rate": 0.00021304246748960208,
+      "loss": 0.8426,
+      "step": 5340
+    },
+    {
+      "epoch": 0.5528001653234139,
+      "grad_norm": 1.4210362434387207,
+      "learning_rate": 0.00021223192885603571,
+      "loss": 0.9085,
+      "step": 5350
+    },
+    {
+      "epoch": 0.553833436660467,
+      "grad_norm": 0.9129871726036072,
+      "learning_rate": 0.00021142179629394283,
+      "loss": 0.8978,
+      "step": 5360
+    },
+    {
+      "epoch": 0.5548667079975201,
+      "grad_norm": 1.0395097732543945,
+      "learning_rate": 0.00021061207851363661,
+      "loss": 0.9224,
+      "step": 5370
+    },
+    {
+      "epoch": 0.5558999793345732,
+      "grad_norm": 0.5978155136108398,
+      "learning_rate": 0.00020980278422097045,
+      "loss": 0.789,
+      "step": 5380
+    },
+    {
+      "epoch": 0.5569332506716264,
+      "grad_norm": 0.8347190022468567,
+      "learning_rate": 0.0002089939221172446,
+      "loss": 0.8651,
+      "step": 5390
+    },
+    {
+      "epoch": 0.5579665220086795,
+      "grad_norm": 1.0471965074539185,
+      "learning_rate": 0.00020818550089911234,
+      "loss": 0.9259,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5589997933457326,
+      "grad_norm": 0.8594735264778137,
+      "learning_rate": 0.00020737752925848707,
+      "loss": 0.7926,
+      "step": 5410
+    },
+    {
+      "epoch": 0.5600330646827857,
+      "grad_norm": 0.9445158243179321,
+      "learning_rate": 0.00020657001588244816,
+      "loss": 0.9561,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5610663360198388,
+      "grad_norm": 0.7682790160179138,
+      "learning_rate": 0.00020576296945314787,
+      "loss": 0.8636,
+      "step": 5430
+    },
+    {
+      "epoch": 0.5620996073568919,
+      "grad_norm": 0.8511012196540833,
+      "learning_rate": 0.00020495639864771814,
+      "loss": 0.877,
+      "step": 5440
+    },
+    {
+      "epoch": 0.563132878693945,
+      "grad_norm": 1.3375914096832275,
+      "learning_rate": 0.00020415031213817704,
+      "loss": 0.8865,
+      "step": 5450
+    },
+    {
+      "epoch": 0.5641661500309981,
+      "grad_norm": 1.4135067462921143,
+      "learning_rate": 0.0002033447185913356,
+      "loss": 0.8816,
+      "step": 5460
+    },
+    {
+      "epoch": 0.5651994213680512,
+      "grad_norm": 1.0292867422103882,
+      "learning_rate": 0.0002025396266687045,
+      "loss": 0.7509,
+      "step": 5470
+    },
+    {
+      "epoch": 0.5662326927051043,
+      "grad_norm": 0.8321288228034973,
+      "learning_rate": 0.00020173504502640164,
+      "loss": 0.8376,
+      "step": 5480
+    },
+    {
+      "epoch": 0.5672659640421575,
+      "grad_norm": 0.7995291948318481,
+      "learning_rate": 0.000200930982315058,
+      "loss": 0.822,
+      "step": 5490
+    },
+    {
+      "epoch": 0.5682992353792106,
+      "grad_norm": 0.7237703800201416,
+      "learning_rate": 0.0002001274471797252,
+      "loss": 0.9819,
+      "step": 5500
+    },
+    {
+      "epoch": 0.5693325067162637,
+      "grad_norm": 1.2666767835617065,
+      "learning_rate": 0.0001993244482597829,
+      "loss": 0.7477,
+      "step": 5510
+    },
+    {
+      "epoch": 0.5703657780533168,
+      "grad_norm": 1.1628350019454956,
+      "learning_rate": 0.00019852199418884527,
+      "loss": 0.7239,
+      "step": 5520
+    },
+    {
+      "epoch": 0.5713990493903699,
+      "grad_norm": 1.1754180192947388,
+      "learning_rate": 0.00019772009359466828,
+      "loss": 0.8326,
+      "step": 5530
+    },
+    {
+      "epoch": 0.572432320727423,
+      "grad_norm": 0.9641863703727722,
+      "learning_rate": 0.00019691875509905718,
+      "loss": 0.779,
+      "step": 5540
+    },
+    {
+      "epoch": 0.5734655920644761,
+      "grad_norm": 1.4042216539382935,
+      "learning_rate": 0.00019611798731777397,
+      "loss": 0.8895,
+      "step": 5550
+    },
+    {
+      "epoch": 0.5744988634015292,
+      "grad_norm": 0.8444200158119202,
+      "learning_rate": 0.00019531779886044424,
+      "loss": 0.8636,
+      "step": 5560
+    },
+    {
+      "epoch": 0.5755321347385823,
+      "grad_norm": 0.8476575016975403,
+      "learning_rate": 0.0001945181983304647,
+      "loss": 0.802,
+      "step": 5570
+    },
+    {
+      "epoch": 0.5765654060756354,
+      "grad_norm": 1.086524486541748,
+      "learning_rate": 0.00019371919432491108,
+      "loss": 0.8356,
+      "step": 5580
+    },
+    {
+      "epoch": 0.5775986774126886,
+      "grad_norm": 0.8224913477897644,
+      "learning_rate": 0.00019292079543444557,
+      "loss": 0.8772,
+      "step": 5590
+    },
+    {
+      "epoch": 0.5786319487497417,
+      "grad_norm": 0.8469100594520569,
+      "learning_rate": 0.00019212301024322396,
+      "loss": 0.883,
+      "step": 5600
+    },
+    {
+      "epoch": 0.5796652200867948,
+      "grad_norm": 0.6162629127502441,
+      "learning_rate": 0.00019132584732880383,
+      "loss": 0.7449,
+      "step": 5610
+    },
+    {
+      "epoch": 0.5806984914238479,
+      "grad_norm": 1.1471900939941406,
+      "learning_rate": 0.00019052931526205267,
+      "loss": 0.8848,
+      "step": 5620
+    },
+    {
+      "epoch": 0.581731762760901,
+      "grad_norm": 1.4094953536987305,
+      "learning_rate": 0.00018973342260705458,
+      "loss": 0.7942,
+      "step": 5630
+    },
+    {
+      "epoch": 0.5827650340979541,
+      "grad_norm": 0.8406280279159546,
+      "learning_rate": 0.00018893817792101935,
+      "loss": 0.8205,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5837983054350072,
+      "grad_norm": 1.5820422172546387,
+      "learning_rate": 0.00018814358975418984,
+      "loss": 0.733,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5848315767720603,
+      "grad_norm": 1.0317678451538086,
+      "learning_rate": 0.00018734966664975023,
+      "loss": 0.8889,
+      "step": 5660
+    },
+    {
+      "epoch": 0.5858648481091134,
+      "grad_norm": 1.1359983682632446,
+      "learning_rate": 0.00018655641714373413,
+      "loss": 0.897,
+      "step": 5670
+    },
+    {
+      "epoch": 0.5868981194461665,
+      "grad_norm": 1.0122281312942505,
+      "learning_rate": 0.00018576384976493282,
+      "loss": 0.7574,
+      "step": 5680
+    },
+    {
+      "epoch": 0.5879313907832197,
+      "grad_norm": 0.9288797378540039,
+      "learning_rate": 0.00018497197303480348,
+      "loss": 0.7553,
+      "step": 5690
+    },
+    {
+      "epoch": 0.5889646621202728,
+      "grad_norm": 0.8775319457054138,
+      "learning_rate": 0.00018418079546737783,
+      "loss": 0.74,
+      "step": 5700
+    },
+    {
+      "epoch": 0.5899979334573259,
+      "grad_norm": 1.075302243232727,
+      "learning_rate": 0.0001833903255691702,
+      "loss": 0.8761,
+      "step": 5710
+    },
+    {
+      "epoch": 0.591031204794379,
+      "grad_norm": 1.0737923383712769,
+      "learning_rate": 0.00018260057183908625,
+      "loss": 0.8817,
+      "step": 5720
+    },
+    {
+      "epoch": 0.5920644761314321,
+      "grad_norm": 1.1469508409500122,
+      "learning_rate": 0.00018181154276833178,
+      "loss": 0.7806,
+      "step": 5730
+    },
+    {
+      "epoch": 0.5930977474684852,
+      "grad_norm": 0.8804445266723633,
+      "learning_rate": 0.00018102324684032115,
+      "loss": 0.7683,
+      "step": 5740
+    },
+    {
+      "epoch": 0.5941310188055383,
+      "grad_norm": 0.9593983888626099,
+      "learning_rate": 0.00018023569253058613,
+      "loss": 0.8427,
+      "step": 5750
+    },
+    {
+      "epoch": 0.5951642901425914,
+      "grad_norm": 0.7528694272041321,
+      "learning_rate": 0.00017944888830668488,
+      "loss": 0.8697,
+      "step": 5760
+    },
+    {
+      "epoch": 0.5961975614796445,
+      "grad_norm": 1.0688844919204712,
+      "learning_rate": 0.00017866284262811093,
+      "loss": 0.779,
+      "step": 5770
+    },
+    {
+      "epoch": 0.5972308328166976,
+      "grad_norm": 1.181410312652588,
+      "learning_rate": 0.00017787756394620202,
+      "loss": 0.8387,
+      "step": 5780
+    },
+    {
+      "epoch": 0.5982641041537508,
+      "grad_norm": 0.7557913064956665,
+      "learning_rate": 0.0001770930607040494,
+      "loss": 0.7897,
+      "step": 5790
+    },
+    {
+      "epoch": 0.5992973754908039,
+      "grad_norm": 0.9230046272277832,
+      "learning_rate": 0.00017630934133640704,
+      "loss": 0.7056,
+      "step": 5800
+    },
+    {
+      "epoch": 0.600330646827857,
+      "grad_norm": 0.839525043964386,
+      "learning_rate": 0.00017552641426960092,
+      "loss": 0.8133,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6013639181649101,
+      "grad_norm": 1.102609395980835,
+      "learning_rate": 0.0001747442879214384,
+      "loss": 0.884,
+      "step": 5820
+    },
+    {
+      "epoch": 0.6023971895019632,
+      "grad_norm": 0.8469570875167847,
+      "learning_rate": 0.0001739629707011177,
+      "loss": 0.7398,
+      "step": 5830
+    },
+    {
+      "epoch": 0.6034304608390163,
+      "grad_norm": 1.2042063474655151,
+      "learning_rate": 0.0001731824710091377,
+      "loss": 0.8074,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6044637321760694,
+      "grad_norm": 2.6056320667266846,
+      "learning_rate": 0.00017240279723720732,
+      "loss": 0.8092,
+      "step": 5850
+    },
+    {
+      "epoch": 0.6054970035131225,
+      "grad_norm": 0.788226306438446,
+      "learning_rate": 0.00017162395776815532,
+      "loss": 0.8346,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6065302748501756,
+      "grad_norm": 0.8313378095626831,
+      "learning_rate": 0.00017084596097584044,
+      "loss": 0.7632,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6075635461872287,
+      "grad_norm": 0.9668551087379456,
+      "learning_rate": 0.00017006881522506123,
+      "loss": 0.7813,
+      "step": 5880
+    },
+    {
+      "epoch": 0.6085968175242819,
+      "grad_norm": 1.2980097532272339,
+      "learning_rate": 0.00016929252887146597,
+      "loss": 0.633,
+      "step": 5890
+    },
+    {
+      "epoch": 0.609630088861335,
+      "grad_norm": 1.296339511871338,
+      "learning_rate": 0.00016851711026146294,
+      "loss": 0.8067,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6106633601983881,
+      "grad_norm": 0.8966313600540161,
+      "learning_rate": 0.00016774256773213075,
+      "loss": 0.8509,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6116966315354412,
+      "grad_norm": 1.408347249031067,
+      "learning_rate": 0.00016696890961112867,
+      "loss": 0.6981,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6127299028724943,
+      "grad_norm": 1.421396017074585,
+      "learning_rate": 0.00016619614421660702,
+      "loss": 0.7626,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6137631742095474,
+      "grad_norm": 1.4539517164230347,
+      "learning_rate": 0.0001654242798571177,
+      "loss": 0.7558,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6147964455466005,
+      "grad_norm": 1.0731654167175293,
+      "learning_rate": 0.0001646533248315252,
+      "loss": 0.7626,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6158297168836536,
+      "grad_norm": 1.100475788116455,
+      "learning_rate": 0.00016388328742891677,
+      "loss": 0.7243,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6168629882207067,
+      "grad_norm": 1.0175180435180664,
+      "learning_rate": 0.00016311417592851396,
+      "loss": 0.8071,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6178962595577598,
+      "grad_norm": 0.9055405259132385,
+      "learning_rate": 0.00016234599859958287,
+      "loss": 0.7203,
+      "step": 5980
+    },
+    {
+      "epoch": 0.618929530894813,
+      "grad_norm": 1.0334464311599731,
+      "learning_rate": 0.00016157876370134617,
+      "loss": 0.706,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6199628022318661,
+      "grad_norm": 1.1770230531692505,
+      "learning_rate": 0.00016081247948289353,
+      "loss": 0.7802,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6209960735689192,
+      "grad_norm": 0.8190944790840149,
+      "learning_rate": 0.00016004715418309304,
+      "loss": 0.7762,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6220293449059723,
+      "grad_norm": 1.0050560235977173,
+      "learning_rate": 0.0001592827960305029,
+      "loss": 0.7886,
+      "step": 6020
+    },
+    {
+      "epoch": 0.6230626162430254,
+      "grad_norm": 1.1464440822601318,
+      "learning_rate": 0.000158519413243283,
+      "loss": 0.8016,
+      "step": 6030
+    },
+    {
+      "epoch": 0.6240958875800785,
+      "grad_norm": 0.8896029591560364,
+      "learning_rate": 0.00015775701402910626,
+      "loss": 0.802,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6251291589171316,
+      "grad_norm": 0.7307144403457642,
+      "learning_rate": 0.00015699560658507012,
+      "loss": 0.7755,
+      "step": 6050
+    },
+    {
+      "epoch": 0.6261624302541847,
+      "grad_norm": 0.767782986164093,
+      "learning_rate": 0.00015623519909760953,
+      "loss": 0.697,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6271957015912378,
+      "grad_norm": 0.8861032724380493,
+      "learning_rate": 0.00015547579974240766,
+      "loss": 0.8005,
+      "step": 6070
+    },
+    {
+      "epoch": 0.6282289729282909,
+      "grad_norm": 0.9863024950027466,
+      "learning_rate": 0.00015471741668430863,
+      "loss": 0.8324,
+      "step": 6080
+    },
+    {
+      "epoch": 0.629262244265344,
+      "grad_norm": 0.8046024441719055,
+      "learning_rate": 0.0001539600580772296,
+      "loss": 0.8615,
+      "step": 6090
+    },
+    {
+      "epoch": 0.6302955156023972,
+      "grad_norm": 1.283008098602295,
+      "learning_rate": 0.0001532037320640734,
+      "loss": 0.8317,
+      "step": 6100
+    },
+    {
+      "epoch": 0.6313287869394503,
+      "grad_norm": 0.9851896166801453,
+      "learning_rate": 0.00015244844677664036,
+      "loss": 0.7516,
+      "step": 6110
+    },
+    {
+      "epoch": 0.6323620582765034,
+      "grad_norm": 1.4205191135406494,
+      "learning_rate": 0.00015169421033554137,
+      "loss": 0.7692,
+      "step": 6120
+    },
+    {
+      "epoch": 0.6333953296135565,
+      "grad_norm": 0.9991669058799744,
+      "learning_rate": 0.0001509410308501104,
+      "loss": 0.7453,
+      "step": 6130
+    },
+    {
+      "epoch": 0.6344286009506096,
+      "grad_norm": 0.723731279373169,
+      "learning_rate": 0.00015018891641831744,
+      "loss": 0.8026,
+      "step": 6140
+    },
+    {
+      "epoch": 0.6354618722876627,
+      "grad_norm": 0.5217877626419067,
+      "learning_rate": 0.00014943787512668118,
+      "loss": 0.7759,
+      "step": 6150
+    },
+    {
+      "epoch": 0.6364951436247158,
+      "grad_norm": 0.6816266179084778,
+      "learning_rate": 0.0001486879150501822,
+      "loss": 0.7659,
+      "step": 6160
+    },
+    {
+      "epoch": 0.6375284149617689,
+      "grad_norm": 0.5098093152046204,
+      "learning_rate": 0.00014793904425217648,
+      "loss": 0.8352,
+      "step": 6170
+    },
+    {
+      "epoch": 0.638561686298822,
+      "grad_norm": 1.0227922201156616,
+      "learning_rate": 0.00014719127078430795,
+      "loss": 0.8162,
+      "step": 6180
+    },
+    {
+      "epoch": 0.6395949576358751,
+      "grad_norm": 1.5966726541519165,
+      "learning_rate": 0.00014644460268642263,
+      "loss": 0.7707,
+      "step": 6190
+    },
+    {
+      "epoch": 0.6406282289729283,
+      "grad_norm": 0.6314913630485535,
+      "learning_rate": 0.00014569904798648176,
+      "loss": 0.7286,
+      "step": 6200
+    },
+    {
+      "epoch": 0.6416615003099814,
+      "grad_norm": 1.072977900505066,
+      "learning_rate": 0.00014495461470047594,
+      "loss": 0.7554,
+      "step": 6210
+    },
+    {
+      "epoch": 0.6426947716470345,
+      "grad_norm": 1.458894968032837,
+      "learning_rate": 0.0001442113108323383,
+      "loss": 0.8975,
+      "step": 6220
+    },
+    {
+      "epoch": 0.6437280429840876,
+      "grad_norm": 0.7281268239021301,
+      "learning_rate": 0.00014346914437385905,
+      "loss": 0.6558,
+      "step": 6230
+    },
+    {
+      "epoch": 0.6447613143211407,
+      "grad_norm": 0.9581575393676758,
+      "learning_rate": 0.00014272812330459893,
+      "loss": 0.7082,
+      "step": 6240
+    },
+    {
+      "epoch": 0.6457945856581938,
+      "grad_norm": 1.271925687789917,
+      "learning_rate": 0.00014198825559180432,
+      "loss": 0.8919,
+      "step": 6250
+    },
+    {
+      "epoch": 0.6468278569952469,
+      "grad_norm": 0.9335398077964783,
+      "learning_rate": 0.00014124954919032064,
+      "loss": 0.7847,
+      "step": 6260
+    },
+    {
+      "epoch": 0.6478611283323,
+      "grad_norm": 0.7705923318862915,
+      "learning_rate": 0.00014051201204250726,
+      "loss": 0.8197,
+      "step": 6270
+    },
+    {
+      "epoch": 0.6488943996693531,
+      "grad_norm": 1.433953881263733,
+      "learning_rate": 0.00013977565207815238,
+      "loss": 0.8028,
+      "step": 6280
+    },
+    {
+      "epoch": 0.6499276710064062,
+      "grad_norm": 1.4676005840301514,
+      "learning_rate": 0.0001390404772143872,
+      "loss": 0.6803,
+      "step": 6290
+    },
+    {
+      "epoch": 0.6509609423434594,
+      "grad_norm": 0.8155161738395691,
+      "learning_rate": 0.00013830649535560093,
+      "loss": 0.8868,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6519942136805125,
+      "grad_norm": 0.799152672290802,
+      "learning_rate": 0.00013757371439335604,
+      "loss": 0.716,
+      "step": 6310
+    },
+    {
+      "epoch": 0.6530274850175656,
+      "grad_norm": 1.120835304260254,
+      "learning_rate": 0.0001368421422063036,
+      "loss": 0.7953,
+      "step": 6320
+    },
+    {
+      "epoch": 0.6540607563546187,
+      "grad_norm": 0.7950730919837952,
+      "learning_rate": 0.00013611178666009794,
+      "loss": 0.7389,
+      "step": 6330
+    },
+    {
+      "epoch": 0.6550940276916718,
+      "grad_norm": 0.9817484617233276,
+      "learning_rate": 0.0001353826556073125,
+      "loss": 0.7451,
+      "step": 6340
+    },
+    {
+      "epoch": 0.6561272990287249,
+      "grad_norm": 0.7330102324485779,
+      "learning_rate": 0.00013465475688735526,
+      "loss": 0.7386,
+      "step": 6350
+    },
+    {
+      "epoch": 0.657160570365778,
+      "grad_norm": 0.7877906560897827,
+      "learning_rate": 0.00013392809832638486,
+      "loss": 0.7596,
+      "step": 6360
+    },
+    {
+      "epoch": 0.6581938417028311,
+      "grad_norm": 0.6060128211975098,
+      "learning_rate": 0.00013320268773722553,
+      "loss": 0.6482,
+      "step": 6370
+    },
+    {
+      "epoch": 0.6592271130398842,
+      "grad_norm": 1.8573722839355469,
+      "learning_rate": 0.00013247853291928395,
+      "loss": 0.7416,
+      "step": 6380
+    },
+    {
+      "epoch": 0.6602603843769373,
+      "grad_norm": 0.8654054403305054,
+      "learning_rate": 0.00013175564165846532,
+      "loss": 0.7027,
+      "step": 6390
+    },
+    {
+      "epoch": 0.6612936557139905,
+      "grad_norm": 0.7953044772148132,
+      "learning_rate": 0.00013103402172708918,
+      "loss": 0.7331,
+      "step": 6400
+    },
+    {
+      "epoch": 0.6623269270510436,
+      "grad_norm": 0.8240858316421509,
+      "learning_rate": 0.0001303136808838061,
+      "loss": 0.7007,
+      "step": 6410
+    },
+    {
+      "epoch": 0.6633601983880967,
+      "grad_norm": 0.8028026819229126,
+      "learning_rate": 0.00012959462687351435,
+      "loss": 0.7626,
+      "step": 6420
+    },
+    {
+      "epoch": 0.6643934697251498,
+      "grad_norm": 1.4915801286697388,
+      "learning_rate": 0.0001288768674272765,
+      "loss": 0.8101,
+      "step": 6430
+    },
+    {
+      "epoch": 0.6654267410622029,
+      "grad_norm": 0.8312230110168457,
+      "learning_rate": 0.0001281604102622364,
+      "loss": 0.7186,
+      "step": 6440
+    },
+    {
+      "epoch": 0.666460012399256,
+      "grad_norm": 0.8355895280838013,
+      "learning_rate": 0.000127445263081536,
+      "loss": 0.756,
+      "step": 6450
+    },
+    {
+      "epoch": 0.6674932837363091,
+      "grad_norm": 0.7141729593276978,
+      "learning_rate": 0.0001267314335742327,
+      "loss": 0.707,
+      "step": 6460
+    },
+    {
+      "epoch": 0.6685265550733622,
+      "grad_norm": 1.2061463594436646,
+      "learning_rate": 0.000126018929415217,
+      "loss": 0.6433,
+      "step": 6470
+    },
+    {
+      "epoch": 0.6695598264104153,
+      "grad_norm": 0.9426934123039246,
+      "learning_rate": 0.00012530775826512916,
+      "loss": 0.7131,
+      "step": 6480
+    },
+    {
+      "epoch": 0.6705930977474684,
+      "grad_norm": 0.5171589255332947,
+      "learning_rate": 0.00012459792777027754,
+      "loss": 0.6755,
+      "step": 6490
+    },
+    {
+      "epoch": 0.6716263690845216,
+      "grad_norm": 0.813228189945221,
+      "learning_rate": 0.00012388944556255614,
+      "loss": 0.7517,
+      "step": 6500
+    },
+    {
+      "epoch": 0.6726596404215747,
+      "grad_norm": 1.0219289064407349,
+      "learning_rate": 0.0001231823192593625,
+      "loss": 0.647,
+      "step": 6510
+    },
+    {
+      "epoch": 0.6736929117586278,
+      "grad_norm": 1.7192949056625366,
+      "learning_rate": 0.00012247655646351597,
+      "loss": 0.7193,
+      "step": 6520
+    },
+    {
+      "epoch": 0.6747261830956809,
+      "grad_norm": 0.8978095054626465,
+      "learning_rate": 0.00012177216476317565,
+      "loss": 0.7004,
+      "step": 6530
+    },
+    {
+      "epoch": 0.675759454432734,
+      "grad_norm": 1.2175655364990234,
+      "learning_rate": 0.00012106915173175942,
+      "loss": 0.7798,
+      "step": 6540
+    },
+    {
+      "epoch": 0.6767927257697871,
+      "grad_norm": 0.7553932070732117,
+      "learning_rate": 0.00012036752492786171,
+      "loss": 0.7077,
+      "step": 6550
+    },
+    {
+      "epoch": 0.6778259971068402,
+      "grad_norm": 1.2669484615325928,
+      "learning_rate": 0.0001196672918951728,
+      "loss": 0.6207,
+      "step": 6560
+    },
+    {
+      "epoch": 0.6788592684438933,
+      "grad_norm": 1.0848954916000366,
+      "learning_rate": 0.00011896846016239752,
+      "loss": 0.7216,
+      "step": 6570
+    },
+    {
+      "epoch": 0.6798925397809464,
+      "grad_norm": 1.306950569152832,
+      "learning_rate": 0.00011827103724317426,
+      "loss": 0.8071,
+      "step": 6580
+    },
+    {
+      "epoch": 0.6809258111179995,
+      "grad_norm": 1.175700306892395,
+      "learning_rate": 0.00011757503063599436,
+      "loss": 0.7327,
+      "step": 6590
+    },
+    {
+      "epoch": 0.6819590824550527,
+      "grad_norm": 1.052412509918213,
+      "learning_rate": 0.00011688044782412124,
+      "loss": 0.7375,
+      "step": 6600
+    },
+    {
+      "epoch": 0.6829923537921058,
+      "grad_norm": 0.922924280166626,
+      "learning_rate": 0.00011618729627551037,
+      "loss": 0.7765,
+      "step": 6610
+    },
+    {
+      "epoch": 0.6840256251291589,
+      "grad_norm": 0.6624295711517334,
+      "learning_rate": 0.00011549558344272834,
+      "loss": 0.7365,
+      "step": 6620
+    },
+    {
+      "epoch": 0.685058896466212,
+      "grad_norm": 0.9047833681106567,
+      "learning_rate": 0.00011480531676287322,
+      "loss": 0.7549,
+      "step": 6630
+    },
+    {
+      "epoch": 0.6860921678032651,
+      "grad_norm": 0.8293854594230652,
+      "learning_rate": 0.00011411650365749454,
+      "loss": 0.7531,
+      "step": 6640
+    },
+    {
+      "epoch": 0.6871254391403182,
+      "grad_norm": 0.865106463432312,
+      "learning_rate": 0.00011342915153251332,
+      "loss": 0.6403,
+      "step": 6650
+    },
+    {
+      "epoch": 0.6881587104773713,
+      "grad_norm": 1.0484962463378906,
+      "learning_rate": 0.0001127432677781425,
+      "loss": 0.7709,
+      "step": 6660
+    },
+    {
+      "epoch": 0.6891919818144244,
+      "grad_norm": 0.9296818375587463,
+      "learning_rate": 0.00011205885976880753,
+      "loss": 0.6943,
+      "step": 6670
+    },
+    {
+      "epoch": 0.6902252531514775,
+      "grad_norm": 0.9692133069038391,
+      "learning_rate": 0.00011137593486306727,
+      "loss": 0.6684,
+      "step": 6680
+    },
+    {
+      "epoch": 0.6912585244885306,
+      "grad_norm": 1.2010059356689453,
+      "learning_rate": 0.00011069450040353448,
+      "loss": 0.8146,
+      "step": 6690
+    },
+    {
+      "epoch": 0.6922917958255838,
+      "grad_norm": 0.8989507555961609,
+      "learning_rate": 0.00011001456371679707,
+      "loss": 0.6553,
+      "step": 6700
+    },
+    {
+      "epoch": 0.6933250671626369,
+      "grad_norm": 1.166821837425232,
+      "learning_rate": 0.00010933613211333943,
+      "loss": 0.76,
+      "step": 6710
+    },
+    {
+      "epoch": 0.69435833849969,
+      "grad_norm": 0.8655950427055359,
+      "learning_rate": 0.00010865921288746362,
+      "loss": 0.7237,
+      "step": 6720
+    },
+    {
+      "epoch": 0.6953916098367431,
+      "grad_norm": 1.0120173692703247,
+      "learning_rate": 0.00010798381331721108,
+      "loss": 0.8003,
+      "step": 6730
+    },
+    {
+      "epoch": 0.6964248811737962,
+      "grad_norm": 1.4091392755508423,
+      "learning_rate": 0.0001073099406642844,
+      "loss": 0.8101,
+      "step": 6740
+    },
+    {
+      "epoch": 0.6974581525108493,
+      "grad_norm": 0.8245342969894409,
+      "learning_rate": 0.00010663760217396906,
+      "loss": 0.6408,
+      "step": 6750
+    },
+    {
+      "epoch": 0.6984914238479024,
+      "grad_norm": 0.7788679003715515,
+      "learning_rate": 0.00010596680507505596,
+      "loss": 0.7559,
+      "step": 6760
+    },
+    {
+      "epoch": 0.6995246951849555,
+      "grad_norm": 0.5751708745956421,
+      "learning_rate": 0.00010529755657976315,
+      "loss": 0.6268,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7005579665220086,
+      "grad_norm": 0.9208303689956665,
+      "learning_rate": 0.00010462986388365837,
+      "loss": 0.7369,
+      "step": 6780
+    },
+    {
+      "epoch": 0.7015912378590617,
+      "grad_norm": 0.8211421966552734,
+      "learning_rate": 0.00010396373416558233,
+      "loss": 0.7131,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7026245091961149,
+      "grad_norm": 0.6540753245353699,
+      "learning_rate": 0.00010329917458757068,
+      "loss": 0.7293,
+      "step": 6800
+    },
+    {
+      "epoch": 0.703657780533168,
+      "grad_norm": 1.3494755029678345,
+      "learning_rate": 0.0001026361922947775,
+      "loss": 0.6745,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7046910518702211,
+      "grad_norm": 0.8039237856864929,
+      "learning_rate": 0.00010197479441539827,
+      "loss": 0.6993,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7057243232072742,
+      "grad_norm": 0.8956775069236755,
+      "learning_rate": 0.00010131498806059352,
+      "loss": 0.7667,
+      "step": 6830
+    },
+    {
+      "epoch": 0.7067575945443273,
+      "grad_norm": 0.7461487650871277,
+      "learning_rate": 0.00010065678032441208,
+      "loss": 0.6487,
+      "step": 6840
+    },
+    {
+      "epoch": 0.7077908658813804,
+      "grad_norm": 1.2220842838287354,
+      "learning_rate": 0.00010000017828371458,
+      "loss": 0.6455,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7088241372184335,
+      "grad_norm": 0.9173330068588257,
+      "learning_rate": 9.934518899809799e-05,
+      "loss": 0.6879,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7098574085554866,
+      "grad_norm": 0.6598500609397888,
+      "learning_rate": 9.869181950981948e-05,
+      "loss": 0.7175,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7108906798925397,
+      "grad_norm": 1.0425814390182495,
+      "learning_rate": 9.804007684372038e-05,
+      "loss": 0.7662,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7119239512295928,
+      "grad_norm": 1.0084232091903687,
+      "learning_rate": 9.738996800715094e-05,
+      "loss": 0.7261,
+      "step": 6890
+    },
+    {
+      "epoch": 0.712957222566646,
+      "grad_norm": 0.9493532180786133,
+      "learning_rate": 9.674149998989523e-05,
+      "loss": 0.7575,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7139904939036991,
+      "grad_norm": 0.6258692741394043,
+      "learning_rate": 9.609467976409525e-05,
+      "loss": 0.6136,
+      "step": 6910
+    },
+    {
+      "epoch": 0.7150237652407522,
+      "grad_norm": 0.6811854243278503,
+      "learning_rate": 9.544951428417667e-05,
+      "loss": 0.8021,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7160570365778053,
+      "grad_norm": 0.8507707118988037,
+      "learning_rate": 9.480601048677371e-05,
+      "loss": 0.644,
+      "step": 6930
+    },
+    {
+      "epoch": 0.7170903079148584,
+      "grad_norm": 1.1403719186782837,
+      "learning_rate": 9.416417529065488e-05,
+      "loss": 0.7257,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7181235792519115,
+      "grad_norm": 0.9672065377235413,
+      "learning_rate": 9.352401559664817e-05,
+      "loss": 0.6545,
+      "step": 6950
+    },
+    {
+      "epoch": 0.7191568505889646,
+      "grad_norm": 1.221876859664917,
+      "learning_rate": 9.288553828756702e-05,
+      "loss": 0.6231,
+      "step": 6960
+    },
+    {
+      "epoch": 0.7201901219260177,
+      "grad_norm": 0.8762854337692261,
+      "learning_rate": 9.224875022813645e-05,
+      "loss": 0.7172,
+      "step": 6970
+    },
+    {
+      "epoch": 0.7212233932630708,
+      "grad_norm": 0.811069667339325,
+      "learning_rate": 9.161365826491913e-05,
+      "loss": 0.6319,
+      "step": 6980
+    },
+    {
+      "epoch": 0.7222566646001239,
+      "grad_norm": 1.3847732543945312,
+      "learning_rate": 9.098026922624175e-05,
+      "loss": 0.8109,
+      "step": 6990
+    },
+    {
+      "epoch": 0.723289935937177,
+      "grad_norm": 1.142921805381775,
+      "learning_rate": 9.03485899221216e-05,
+      "loss": 0.7571,
+      "step": 7000
+    },
+    {
+      "epoch": 0.7243232072742302,
+      "grad_norm": 0.5236563086509705,
+      "learning_rate": 8.971862714419368e-05,
+      "loss": 0.6592,
+      "step": 7010
+    },
+    {
+      "epoch": 0.7253564786112833,
+      "grad_norm": 1.2576050758361816,
+      "learning_rate": 8.909038766563703e-05,
+      "loss": 0.7714,
+      "step": 7020
+    },
+    {
+      "epoch": 0.7263897499483364,
+      "grad_norm": 0.9121869802474976,
+      "learning_rate": 8.846387824110247e-05,
+      "loss": 0.8312,
+      "step": 7030
+    },
+    {
+      "epoch": 0.7274230212853895,
+      "grad_norm": 1.3254677057266235,
+      "learning_rate": 8.783910560663968e-05,
+      "loss": 0.7242,
+      "step": 7040
+    },
+    {
+      "epoch": 0.7284562926224426,
+      "grad_norm": 1.342081904411316,
+      "learning_rate": 8.721607647962496e-05,
+      "loss": 0.7367,
+      "step": 7050
+    },
+    {
+      "epoch": 0.7294895639594957,
+      "grad_norm": 1.3918049335479736,
+      "learning_rate": 8.659479755868882e-05,
+      "loss": 0.8162,
+      "step": 7060
+    },
+    {
+      "epoch": 0.7305228352965488,
+      "grad_norm": 0.9307500720024109,
+      "learning_rate": 8.597527552364415e-05,
+      "loss": 0.6592,
+      "step": 7070
+    },
+    {
+      "epoch": 0.7315561066336019,
+      "grad_norm": 0.958733081817627,
+      "learning_rate": 8.535751703541417e-05,
+      "loss": 0.7096,
+      "step": 7080
+    },
+    {
+      "epoch": 0.732589377970655,
+      "grad_norm": 0.6930942535400391,
+      "learning_rate": 8.474152873596122e-05,
+      "loss": 0.6513,
+      "step": 7090
+    },
+    {
+      "epoch": 0.7336226493077082,
+      "grad_norm": 1.3499219417572021,
+      "learning_rate": 8.412731724821482e-05,
+      "loss": 0.7495,
+      "step": 7100
+    },
+    {
+      "epoch": 0.7346559206447613,
+      "grad_norm": 0.9743627905845642,
+      "learning_rate": 8.351488917600077e-05,
+      "loss": 0.6989,
+      "step": 7110
+    },
+    {
+      "epoch": 0.7356891919818144,
+      "grad_norm": 0.8228883147239685,
+      "learning_rate": 8.290425110397018e-05,
+      "loss": 0.7556,
+      "step": 7120
+    },
+    {
+      "epoch": 0.7367224633188675,
+      "grad_norm": 0.9745365381240845,
+      "learning_rate": 8.229540959752851e-05,
+      "loss": 0.6756,
+      "step": 7130
+    },
+    {
+      "epoch": 0.7377557346559206,
+      "grad_norm": 0.9700762629508972,
+      "learning_rate": 8.168837120276512e-05,
+      "loss": 0.7298,
+      "step": 7140
+    },
+    {
+      "epoch": 0.7387890059929737,
+      "grad_norm": 1.2264478206634521,
+      "learning_rate": 8.108314244638267e-05,
+      "loss": 0.7307,
+      "step": 7150
+    },
+    {
+      "epoch": 0.7398222773300268,
+      "grad_norm": 1.1239992380142212,
+      "learning_rate": 8.047972983562748e-05,
+      "loss": 0.7043,
+      "step": 7160
+    },
+    {
+      "epoch": 0.7408555486670799,
+      "grad_norm": 0.8346306681632996,
+      "learning_rate": 7.987813985821882e-05,
+      "loss": 0.7194,
+      "step": 7170
+    },
+    {
+      "epoch": 0.741888820004133,
+      "grad_norm": 0.8321127891540527,
+      "learning_rate": 7.927837898227966e-05,
+      "loss": 0.7095,
+      "step": 7180
+    },
+    {
+      "epoch": 0.7429220913411861,
+      "grad_norm": 0.9774494767189026,
+      "learning_rate": 7.868045365626702e-05,
+      "loss": 0.7328,
+      "step": 7190
+    },
+    {
+      "epoch": 0.7439553626782393,
+      "grad_norm": 0.8818146586418152,
+      "learning_rate": 7.808437030890258e-05,
+      "loss": 0.6084,
+      "step": 7200
+    },
+    {
+      "epoch": 0.7449886340152924,
+      "grad_norm": 0.6830877661705017,
+      "learning_rate": 7.749013534910362e-05,
+      "loss": 0.6242,
+      "step": 7210
+    },
+    {
+      "epoch": 0.7460219053523455,
+      "grad_norm": 0.7373770475387573,
+      "learning_rate": 7.689775516591397e-05,
+      "loss": 0.6592,
+      "step": 7220
+    },
+    {
+      "epoch": 0.7470551766893986,
+      "grad_norm": 1.5116766691207886,
+      "learning_rate": 7.63072361284357e-05,
+      "loss": 0.7781,
+      "step": 7230
+    },
+    {
+      "epoch": 0.7480884480264517,
+      "grad_norm": 1.11928129196167,
+      "learning_rate": 7.571858458576017e-05,
+      "loss": 0.8028,
+      "step": 7240
+    },
+    {
+      "epoch": 0.7491217193635048,
+      "grad_norm": 0.5922040939331055,
+      "learning_rate": 7.513180686690002e-05,
+      "loss": 0.6601,
+      "step": 7250
+    },
+    {
+      "epoch": 0.750154990700558,
+      "grad_norm": 0.9148753881454468,
+      "learning_rate": 7.454690928072088e-05,
+      "loss": 0.7152,
+      "step": 7260
+    },
+    {
+      "epoch": 0.7511882620376111,
+      "grad_norm": 0.7024368047714233,
+      "learning_rate": 7.396389811587415e-05,
+      "loss": 0.6906,
+      "step": 7270
+    },
+    {
+      "epoch": 0.7522215333746642,
+      "grad_norm": 1.3108892440795898,
+      "learning_rate": 7.338277964072864e-05,
+      "loss": 0.7471,
+      "step": 7280
+    },
+    {
+      "epoch": 0.7532548047117174,
+      "grad_norm": 0.6525394320487976,
+      "learning_rate": 7.280356010330361e-05,
+      "loss": 0.7364,
+      "step": 7290
+    },
+    {
+      "epoch": 0.7542880760487705,
+      "grad_norm": 1.344499111175537,
+      "learning_rate": 7.222624573120137e-05,
+      "loss": 0.7023,
+      "step": 7300
+    },
+    {
+      "epoch": 0.7553213473858236,
+      "grad_norm": 0.7209261059761047,
+      "learning_rate": 7.165084273154074e-05,
+      "loss": 0.6748,
+      "step": 7310
+    },
+    {
+      "epoch": 0.7563546187228767,
+      "grad_norm": 1.4905532598495483,
+      "learning_rate": 7.10773572908898e-05,
+      "loss": 0.6376,
+      "step": 7320
+    },
+    {
+      "epoch": 0.7573878900599298,
+      "grad_norm": 0.4700772762298584,
+      "learning_rate": 7.050579557519938e-05,
+      "loss": 0.6694,
+      "step": 7330
+    },
+    {
+      "epoch": 0.7584211613969829,
+      "grad_norm": 0.6527320146560669,
+      "learning_rate": 6.993616372973739e-05,
+      "loss": 0.6764,
+      "step": 7340
+    },
+    {
+      "epoch": 0.759454432734036,
+      "grad_norm": 0.7010611295700073,
+      "learning_rate": 6.936846787902207e-05,
+      "loss": 0.658,
+      "step": 7350
+    },
+    {
+      "epoch": 0.7604877040710891,
+      "grad_norm": 1.1044212579727173,
+      "learning_rate": 6.880271412675645e-05,
+      "loss": 0.6569,
+      "step": 7360
+    },
+    {
+      "epoch": 0.7615209754081422,
+      "grad_norm": 0.8779295682907104,
+      "learning_rate": 6.823890855576261e-05,
+      "loss": 0.6382,
+      "step": 7370
+    },
+    {
+      "epoch": 0.7625542467451953,
+      "grad_norm": 0.9512449502944946,
+      "learning_rate": 6.76770572279167e-05,
+      "loss": 0.6453,
+      "step": 7380
+    },
+    {
+      "epoch": 0.7635875180822485,
+      "grad_norm": 0.782164990901947,
+      "learning_rate": 6.711716618408281e-05,
+      "loss": 0.6539,
+      "step": 7390
+    },
+    {
+      "epoch": 0.7646207894193016,
+      "grad_norm": 1.7972410917282104,
+      "learning_rate": 6.655924144404906e-05,
+      "loss": 0.6142,
+      "step": 7400
+    },
+    {
+      "epoch": 0.7656540607563547,
+      "grad_norm": 0.9399009346961975,
+      "learning_rate": 6.600328900646222e-05,
+      "loss": 0.7511,
+      "step": 7410
+    },
+    {
+      "epoch": 0.7666873320934078,
+      "grad_norm": 2.6008989810943604,
+      "learning_rate": 6.544931484876368e-05,
+      "loss": 0.5901,
+      "step": 7420
+    },
+    {
+      "epoch": 0.7677206034304609,
+      "grad_norm": 1.3178836107254028,
+      "learning_rate": 6.489732492712466e-05,
+      "loss": 0.6217,
+      "step": 7430
+    },
+    {
+      "epoch": 0.768753874767514,
+      "grad_norm": 1.1651490926742554,
+      "learning_rate": 6.434732517638254e-05,
+      "loss": 0.6472,
+      "step": 7440
+    },
+    {
+      "epoch": 0.7697871461045671,
+      "grad_norm": 1.072582721710205,
+      "learning_rate": 6.37993215099771e-05,
+      "loss": 0.6595,
+      "step": 7450
+    },
+    {
+      "epoch": 0.7708204174416202,
+      "grad_norm": 1.5162521600723267,
+      "learning_rate": 6.325331981988653e-05,
+      "loss": 0.7025,
+      "step": 7460
+    },
+    {
+      "epoch": 0.7718536887786733,
+      "grad_norm": 0.804498016834259,
+      "learning_rate": 6.27093259765645e-05,
+      "loss": 0.7168,
+      "step": 7470
+    },
+    {
+      "epoch": 0.7728869601157264,
+      "grad_norm": 1.0188984870910645,
+      "learning_rate": 6.216734582887678e-05,
+      "loss": 0.6158,
+      "step": 7480
+    },
+    {
+      "epoch": 0.7739202314527796,
+      "grad_norm": 0.9018194675445557,
+      "learning_rate": 6.162738520403871e-05,
+      "loss": 0.5752,
+      "step": 7490
+    },
+    {
+      "epoch": 0.7749535027898327,
+      "grad_norm": 0.6925551891326904,
+      "learning_rate": 6.108944990755202e-05,
+      "loss": 0.6934,
+      "step": 7500
+    },
+    {
+      "epoch": 0.7759867741268858,
+      "grad_norm": 1.3633153438568115,
+      "learning_rate": 6.0553545723142864e-05,
+      "loss": 0.6985,
+      "step": 7510
+    },
+    {
+      "epoch": 0.7770200454639389,
+      "grad_norm": 1.114071249961853,
+      "learning_rate": 6.001967841269937e-05,
+      "loss": 0.6031,
+      "step": 7520
+    },
+    {
+      "epoch": 0.778053316800992,
+      "grad_norm": 0.7959581613540649,
+      "learning_rate": 5.9487853716209794e-05,
+      "loss": 0.6645,
+      "step": 7530
+    },
+    {
+      "epoch": 0.7790865881380451,
+      "grad_norm": 0.8598949313163757,
+      "learning_rate": 5.89580773517008e-05,
+      "loss": 0.7458,
+      "step": 7540
+    },
+    {
+      "epoch": 0.7801198594750982,
+      "grad_norm": 1.4160740375518799,
+      "learning_rate": 5.843035501517596e-05,
+      "loss": 0.6635,
+      "step": 7550
+    },
+    {
+      "epoch": 0.7811531308121513,
+      "grad_norm": 1.0915048122406006,
+      "learning_rate": 5.790469238055465e-05,
+      "loss": 0.6052,
+      "step": 7560
+    },
+    {
+      "epoch": 0.7821864021492044,
+      "grad_norm": 0.8317732214927673,
+      "learning_rate": 5.738109509961076e-05,
+      "loss": 0.7007,
+      "step": 7570
+    },
+    {
+      "epoch": 0.7832196734862575,
+      "grad_norm": 0.9893842339515686,
+      "learning_rate": 5.685956880191218e-05,
+      "loss": 0.5915,
+      "step": 7580
+    },
+    {
+      "epoch": 0.7842529448233106,
+      "grad_norm": 1.2170053720474243,
+      "learning_rate": 5.634011909476008e-05,
+      "loss": 0.6921,
+      "step": 7590
+    },
+    {
+      "epoch": 0.7852862161603638,
+      "grad_norm": 1.261479377746582,
+      "learning_rate": 5.582275156312885e-05,
+      "loss": 0.6338,
+      "step": 7600
+    },
+    {
+      "epoch": 0.7863194874974169,
+      "grad_norm": 1.0150582790374756,
+      "learning_rate": 5.5307471769605875e-05,
+      "loss": 0.637,
+      "step": 7610
+    },
+    {
+      "epoch": 0.78735275883447,
+      "grad_norm": 1.148558497428894,
+      "learning_rate": 5.479428525433167e-05,
+      "loss": 0.6243,
+      "step": 7620
+    },
+    {
+      "epoch": 0.7883860301715231,
+      "grad_norm": 0.6112420558929443,
+      "learning_rate": 5.4283197534940684e-05,
+      "loss": 0.579,
+      "step": 7630
+    },
+    {
+      "epoch": 0.7894193015085762,
+      "grad_norm": 1.1648898124694824,
+      "learning_rate": 5.377421410650149e-05,
+      "loss": 0.6978,
+      "step": 7640
+    },
+    {
+      "epoch": 0.7904525728456293,
+      "grad_norm": 1.3245562314987183,
+      "learning_rate": 5.326734044145801e-05,
+      "loss": 0.5724,
+      "step": 7650
+    },
+    {
+      "epoch": 0.7914858441826824,
+      "grad_norm": 1.0554862022399902,
+      "learning_rate": 5.276258198957051e-05,
+      "loss": 0.7884,
+      "step": 7660
+    },
+    {
+      "epoch": 0.7925191155197355,
+      "grad_norm": 1.1452159881591797,
+      "learning_rate": 5.225994417785726e-05,
+      "loss": 0.7309,
+      "step": 7670
+    },
+    {
+      "epoch": 0.7935523868567886,
+      "grad_norm": 0.7298494577407837,
+      "learning_rate": 5.175943241053582e-05,
+      "loss": 0.6804,
+      "step": 7680
+    },
+    {
+      "epoch": 0.7945856581938417,
+      "grad_norm": 1.0335783958435059,
+      "learning_rate": 5.1261052068965306e-05,
+      "loss": 0.6306,
+      "step": 7690
+    },
+    {
+      "epoch": 0.7956189295308949,
+      "grad_norm": 0.6662817001342773,
+      "learning_rate": 5.0764808511588155e-05,
+      "loss": 0.6118,
+      "step": 7700
+    },
+    {
+      "epoch": 0.796652200867948,
+      "grad_norm": 0.8202494978904724,
+      "learning_rate": 5.027070707387296e-05,
+      "loss": 0.6764,
+      "step": 7710
+    },
+    {
+      "epoch": 0.7976854722050011,
+      "grad_norm": 1.5033270120620728,
+      "learning_rate": 4.977875306825672e-05,
+      "loss": 0.7725,
+      "step": 7720
+    },
+    {
+      "epoch": 0.7987187435420542,
+      "grad_norm": 0.7202991843223572,
+      "learning_rate": 4.928895178408782e-05,
+      "loss": 0.685,
+      "step": 7730
+    },
+    {
+      "epoch": 0.7997520148791073,
+      "grad_norm": 0.8215998411178589,
+      "learning_rate": 4.880130848756925e-05,
+      "loss": 0.6536,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8007852862161604,
+      "grad_norm": 1.0507868528366089,
+      "learning_rate": 4.8315828421701986e-05,
+      "loss": 0.6238,
+      "step": 7750
+    },
+    {
+      "epoch": 0.8018185575532135,
+      "grad_norm": 1.0970897674560547,
+      "learning_rate": 4.783251680622847e-05,
+      "loss": 0.6302,
+      "step": 7760
+    },
+    {
+      "epoch": 0.8028518288902666,
+      "grad_norm": 1.2945756912231445,
+      "learning_rate": 4.7351378837576565e-05,
+      "loss": 0.5465,
+      "step": 7770
+    },
+    {
+      "epoch": 0.8038851002273197,
+      "grad_norm": 1.6131089925765991,
+      "learning_rate": 4.687241968880393e-05,
+      "loss": 0.6698,
+      "step": 7780
+    },
+    {
+      "epoch": 0.8049183715643728,
+      "grad_norm": 1.0525102615356445,
+      "learning_rate": 4.639564450954201e-05,
+      "loss": 0.6425,
+      "step": 7790
+    },
+    {
+      "epoch": 0.805951642901426,
+      "grad_norm": 1.0598100423812866,
+      "learning_rate": 4.5921058425940745e-05,
+      "loss": 0.5235,
+      "step": 7800
+    },
+    {
+      "epoch": 0.8069849142384791,
+      "grad_norm": 1.4403691291809082,
+      "learning_rate": 4.544866654061364e-05,
+      "loss": 0.7243,
+      "step": 7810
+    },
+    {
+      "epoch": 0.8080181855755322,
+      "grad_norm": 0.995022714138031,
+      "learning_rate": 4.497847393258292e-05,
+      "loss": 0.6382,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8090514569125853,
+      "grad_norm": 1.2074267864227295,
+      "learning_rate": 4.451048565722468e-05,
+      "loss": 0.644,
+      "step": 7830
+    },
+    {
+      "epoch": 0.8100847282496384,
+      "grad_norm": 1.038830041885376,
+      "learning_rate": 4.404470674621463e-05,
+      "loss": 0.6955,
+      "step": 7840
+    },
+    {
+      "epoch": 0.8111179995866915,
+      "grad_norm": 1.028713345527649,
+      "learning_rate": 4.3581142207474225e-05,
+      "loss": 0.5929,
+      "step": 7850
+    },
+    {
+      "epoch": 0.8121512709237446,
+      "grad_norm": 1.214220404624939,
+      "learning_rate": 4.311979702511645e-05,
+      "loss": 0.6807,
+      "step": 7860
+    },
+    {
+      "epoch": 0.8131845422607977,
+      "grad_norm": 0.91230708360672,
+      "learning_rate": 4.266067615939234e-05,
+      "loss": 0.5622,
+      "step": 7870
+    },
+    {
+      "epoch": 0.8142178135978508,
+      "grad_norm": 0.7656465172767639,
+      "learning_rate": 4.220378454663784e-05,
+      "loss": 0.7191,
+      "step": 7880
+    },
+    {
+      "epoch": 0.815251084934904,
+      "grad_norm": 0.9677980542182922,
+      "learning_rate": 4.1749127099220684e-05,
+      "loss": 0.6268,
+      "step": 7890
+    },
+    {
+      "epoch": 0.8162843562719571,
+      "grad_norm": 0.8608049750328064,
+      "learning_rate": 4.129670870548738e-05,
+      "loss": 0.608,
+      "step": 7900
+    },
+    {
+      "epoch": 0.8173176276090102,
+      "grad_norm": 1.3923230171203613,
+      "learning_rate": 4.084653422971077e-05,
+      "loss": 0.6626,
+      "step": 7910
+    },
+    {
+      "epoch": 0.8183508989460633,
+      "grad_norm": 1.0703678131103516,
+      "learning_rate": 4.039860851203775e-05,
+      "loss": 0.5827,
+      "step": 7920
+    },
+    {
+      "epoch": 0.8193841702831164,
+      "grad_norm": 0.9771378040313721,
+      "learning_rate": 3.9952936368437275e-05,
+      "loss": 0.7703,
+      "step": 7930
+    },
+    {
+      "epoch": 0.8204174416201695,
+      "grad_norm": 1.0725332498550415,
+      "learning_rate": 3.950952259064841e-05,
+      "loss": 0.6425,
+      "step": 7940
+    },
+    {
+      "epoch": 0.8214507129572226,
+      "grad_norm": 1.1629067659378052,
+      "learning_rate": 3.9068371946128945e-05,
+      "loss": 0.6658,
+      "step": 7950
+    },
+    {
+      "epoch": 0.8224839842942757,
+      "grad_norm": 0.9839202165603638,
+      "learning_rate": 3.862948917800424e-05,
+      "loss": 0.5883,
+      "step": 7960
+    },
+    {
+      "epoch": 0.8235172556313288,
+      "grad_norm": 0.8677191138267517,
+      "learning_rate": 3.8192879005015974e-05,
+      "loss": 0.6664,
+      "step": 7970
+    },
+    {
+      "epoch": 0.8245505269683819,
+      "grad_norm": 0.8415852189064026,
+      "learning_rate": 3.775854612147153e-05,
+      "loss": 0.5629,
+      "step": 7980
+    },
+    {
+      "epoch": 0.825583798305435,
+      "grad_norm": 0.7151034474372864,
+      "learning_rate": 3.7326495197193555e-05,
+      "loss": 0.5575,
+      "step": 7990
+    },
+    {
+      "epoch": 0.8266170696424882,
+      "grad_norm": 1.2307301759719849,
+      "learning_rate": 3.689673087746975e-05,
+      "loss": 0.6356,
+      "step": 8000
+    },
+    {
+      "epoch": 0.8276503409795413,
+      "grad_norm": 2.110441207885742,
+      "learning_rate": 3.646925778300286e-05,
+      "loss": 0.7123,
+      "step": 8010
+    },
+    {
+      "epoch": 0.8286836123165944,
+      "grad_norm": 0.47642096877098083,
+      "learning_rate": 3.604408050986105e-05,
+      "loss": 0.6165,
+      "step": 8020
+    },
+    {
+      "epoch": 0.8297168836536475,
+      "grad_norm": 0.69282066822052,
+      "learning_rate": 3.5621203629428386e-05,
+      "loss": 0.6196,
+      "step": 8030
+    },
+    {
+      "epoch": 0.8307501549907006,
+      "grad_norm": 1.5054296255111694,
+      "learning_rate": 3.520063168835605e-05,
+      "loss": 0.749,
+      "step": 8040
+    },
+    {
+      "epoch": 0.8317834263277537,
+      "grad_norm": 1.0589144229888916,
+      "learning_rate": 3.478236920851283e-05,
+      "loss": 0.6056,
+      "step": 8050
+    },
+    {
+      "epoch": 0.8328166976648068,
+      "grad_norm": 0.8430522680282593,
+      "learning_rate": 3.4366420686937044e-05,
+      "loss": 0.6502,
+      "step": 8060
+    },
+    {
+      "epoch": 0.8338499690018599,
+      "grad_norm": 0.8304775357246399,
+      "learning_rate": 3.395279059578798e-05,
+      "loss": 0.6363,
+      "step": 8070
+    },
+    {
+      "epoch": 0.834883240338913,
+      "grad_norm": 0.5285193920135498,
+      "learning_rate": 3.3541483382297734e-05,
+      "loss": 0.751,
+      "step": 8080
+    },
+    {
+      "epoch": 0.8359165116759661,
+      "grad_norm": 0.9843568205833435,
+      "learning_rate": 3.313250346872362e-05,
+      "loss": 0.5978,
+      "step": 8090
+    },
+    {
+      "epoch": 0.8369497830130193,
+      "grad_norm": 1.0727510452270508,
+      "learning_rate": 3.272585525230032e-05,
+      "loss": 0.528,
+      "step": 8100
+    },
+    {
+      "epoch": 0.8379830543500724,
+      "grad_norm": 0.9402909874916077,
+      "learning_rate": 3.232154310519298e-05,
+      "loss": 0.5649,
+      "step": 8110
+    },
+    {
+      "epoch": 0.8390163256871255,
+      "grad_norm": 0.6825330257415771,
+      "learning_rate": 3.1919571374449894e-05,
+      "loss": 0.6619,
+      "step": 8120
+    },
+    {
+      "epoch": 0.8400495970241786,
+      "grad_norm": 1.605404019355774,
+      "learning_rate": 3.151994438195582e-05,
+      "loss": 0.7536,
+      "step": 8130
+    },
+    {
+      "epoch": 0.8410828683612317,
+      "grad_norm": 0.9325007796287537,
+      "learning_rate": 3.112266642438566e-05,
+      "loss": 0.7271,
+      "step": 8140
+    },
+    {
+      "epoch": 0.8421161396982848,
+      "grad_norm": 1.083452820777893,
+      "learning_rate": 3.072774177315807e-05,
+      "loss": 0.6664,
+      "step": 8150
+    },
+    {
+      "epoch": 0.8431494110353379,
+      "grad_norm": 0.6602939963340759,
+      "learning_rate": 3.0335174674389732e-05,
+      "loss": 0.5289,
+      "step": 8160
+    },
+    {
+      "epoch": 0.844182682372391,
+      "grad_norm": 1.497051477432251,
+      "learning_rate": 2.9944969348849493e-05,
+      "loss": 0.5557,
+      "step": 8170
+    },
+    {
+      "epoch": 0.8452159537094441,
+      "grad_norm": 0.7010759115219116,
+      "learning_rate": 2.9557129991913235e-05,
+      "loss": 0.5932,
+      "step": 8180
+    },
+    {
+      "epoch": 0.8462492250464972,
+      "grad_norm": 0.93887859582901,
+      "learning_rate": 2.917166077351846e-05,
+      "loss": 0.5421,
+      "step": 8190
+    },
+    {
+      "epoch": 0.8472824963835504,
+      "grad_norm": 1.242521047592163,
+      "learning_rate": 2.8788565838119757e-05,
+      "loss": 0.5551,
+      "step": 8200
+    },
+    {
+      "epoch": 0.8483157677206035,
+      "grad_norm": 1.1588674783706665,
+      "learning_rate": 2.840784930464399e-05,
+      "loss": 0.5647,
+      "step": 8210
+    },
+    {
+      "epoch": 0.8493490390576566,
+      "grad_norm": 1.381052851676941,
+      "learning_rate": 2.8029515266446232e-05,
+      "loss": 0.6187,
+      "step": 8220
+    },
+    {
+      "epoch": 0.8503823103947097,
+      "grad_norm": 1.1169166564941406,
+      "learning_rate": 2.765356779126557e-05,
+      "loss": 0.6038,
+      "step": 8230
+    },
+    {
+      "epoch": 0.8514155817317628,
+      "grad_norm": 1.2974213361740112,
+      "learning_rate": 2.7280010921181487e-05,
+      "loss": 0.7017,
+      "step": 8240
+    },
+    {
+      "epoch": 0.8524488530688159,
+      "grad_norm": 0.9372439384460449,
+      "learning_rate": 2.690884867257032e-05,
+      "loss": 0.5887,
+      "step": 8250
+    },
+    {
+      "epoch": 0.853482124405869,
+      "grad_norm": 0.8471134305000305,
+      "learning_rate": 2.6540085036062227e-05,
+      "loss": 0.5159,
+      "step": 8260
+    },
+    {
+      "epoch": 0.8545153957429221,
+      "grad_norm": 1.077082633972168,
+      "learning_rate": 2.617372397649814e-05,
+      "loss": 0.6389,
+      "step": 8270
+    },
+    {
+      "epoch": 0.8555486670799752,
+      "grad_norm": 1.1442515850067139,
+      "learning_rate": 2.5809769432886986e-05,
+      "loss": 0.6233,
+      "step": 8280
+    },
+    {
+      "epoch": 0.8565819384170283,
+      "grad_norm": 1.59578537940979,
+      "learning_rate": 2.5448225318363828e-05,
+      "loss": 0.6336,
+      "step": 8290
+    },
+    {
+      "epoch": 0.8576152097540815,
+      "grad_norm": 0.7789118885993958,
+      "learning_rate": 2.5089095520147266e-05,
+      "loss": 0.6106,
+      "step": 8300
+    },
+    {
+      "epoch": 0.8586484810911346,
+      "grad_norm": 0.8877026438713074,
+      "learning_rate": 2.4732383899497972e-05,
+      "loss": 0.6226,
+      "step": 8310
+    },
+    {
+      "epoch": 0.8596817524281877,
+      "grad_norm": 0.9339987635612488,
+      "learning_rate": 2.4378094291676978e-05,
+      "loss": 0.5998,
+      "step": 8320
+    },
+    {
+      "epoch": 0.8607150237652408,
+      "grad_norm": 0.45565560460090637,
+      "learning_rate": 2.4026230505904633e-05,
+      "loss": 0.5463,
+      "step": 8330
+    },
+    {
+      "epoch": 0.8617482951022939,
+      "grad_norm": 0.7657277584075928,
+      "learning_rate": 2.367679632531955e-05,
+      "loss": 0.6507,
+      "step": 8340
+    },
+    {
+      "epoch": 0.862781566439347,
+      "grad_norm": 1.4094293117523193,
+      "learning_rate": 2.3329795506937728e-05,
+      "loss": 0.5541,
+      "step": 8350
+    },
+    {
+      "epoch": 0.8638148377764001,
+      "grad_norm": 0.5916699171066284,
+      "learning_rate": 2.2985231781612466e-05,
+      "loss": 0.6416,
+      "step": 8360
+    },
+    {
+      "epoch": 0.8648481091134532,
+      "grad_norm": 0.7877978086471558,
+      "learning_rate": 2.2643108853994317e-05,
+      "loss": 0.6413,
+      "step": 8370
+    },
+    {
+      "epoch": 0.8658813804505063,
+      "grad_norm": 1.3543034791946411,
+      "learning_rate": 2.2303430402490805e-05,
+      "loss": 0.6255,
+      "step": 8380
+    },
+    {
+      "epoch": 0.8669146517875594,
+      "grad_norm": 1.2339602708816528,
+      "learning_rate": 2.1966200079227257e-05,
+      "loss": 0.6086,
+      "step": 8390
+    },
+    {
+      "epoch": 0.8679479231246126,
+      "grad_norm": 1.563103199005127,
+      "learning_rate": 2.1631421510007575e-05,
+      "loss": 0.6084,
+      "step": 8400
+    },
+    {
+      "epoch": 0.8689811944616657,
+      "grad_norm": 1.1265037059783936,
+      "learning_rate": 2.129909829427479e-05,
+      "loss": 0.5604,
+      "step": 8410
+    },
+    {
+      "epoch": 0.8700144657987188,
+      "grad_norm": 1.4551678895950317,
+      "learning_rate": 2.0969234005072916e-05,
+      "loss": 0.6835,
+      "step": 8420
+    },
+    {
+      "epoch": 0.8710477371357719,
+      "grad_norm": 0.9663675427436829,
+      "learning_rate": 2.064183218900817e-05,
+      "loss": 0.6,
+      "step": 8430
+    },
+    {
+      "epoch": 0.872081008472825,
+      "grad_norm": 1.2805395126342773,
+      "learning_rate": 2.0316896366211118e-05,
+      "loss": 0.6471,
+      "step": 8440
+    },
+    {
+      "epoch": 0.8731142798098781,
+      "grad_norm": 0.9442639946937561,
+      "learning_rate": 1.9994430030298496e-05,
+      "loss": 0.6104,
+      "step": 8450
+    },
+    {
+      "epoch": 0.8741475511469312,
+      "grad_norm": 1.5761253833770752,
+      "learning_rate": 1.967443664833596e-05,
+      "loss": 0.6629,
+      "step": 8460
+    },
+    {
+      "epoch": 0.8751808224839843,
+      "grad_norm": 1.1689879894256592,
+      "learning_rate": 1.9356919660800553e-05,
+      "loss": 0.6161,
+      "step": 8470
+    },
+    {
+      "epoch": 0.8762140938210374,
+      "grad_norm": 1.1315240859985352,
+      "learning_rate": 1.9041882481543955e-05,
+      "loss": 0.574,
+      "step": 8480
+    },
+    {
+      "epoch": 0.8772473651580905,
+      "grad_norm": 1.0627079010009766,
+      "learning_rate": 1.8729328497755578e-05,
+      "loss": 0.6021,
+      "step": 8490
+    },
+    {
+      "epoch": 0.8782806364951437,
+      "grad_norm": 0.9222522377967834,
+      "learning_rate": 1.8419261069926197e-05,
+      "loss": 0.5026,
+      "step": 8500
+    },
+    {
+      "epoch": 0.8793139078321968,
+      "grad_norm": 0.92143315076828,
+      "learning_rate": 1.8111683531812002e-05,
+      "loss": 0.5853,
+      "step": 8510
+    },
+    {
+      "epoch": 0.8803471791692499,
+      "grad_norm": 1.042581558227539,
+      "learning_rate": 1.7806599190398455e-05,
+      "loss": 0.5239,
+      "step": 8520
+    },
+    {
+      "epoch": 0.881380450506303,
+      "grad_norm": 1.405364751815796,
+      "learning_rate": 1.750401132586496e-05,
+      "loss": 0.571,
+      "step": 8530
+    },
+    {
+      "epoch": 0.8824137218433561,
+      "grad_norm": 1.2399895191192627,
+      "learning_rate": 1.720392319154948e-05,
+      "loss": 0.5704,
+      "step": 8540
+    },
+    {
+      "epoch": 0.8834469931804092,
+      "grad_norm": 0.9870766401290894,
+      "learning_rate": 1.6906338013913585e-05,
+      "loss": 0.5944,
+      "step": 8550
+    },
+    {
+      "epoch": 0.8844802645174623,
+      "grad_norm": 0.8951932787895203,
+      "learning_rate": 1.6611258992507817e-05,
+      "loss": 0.5861,
+      "step": 8560
+    },
+    {
+      "epoch": 0.8855135358545154,
+      "grad_norm": 1.8214749097824097,
+      "learning_rate": 1.6318689299937213e-05,
+      "loss": 0.6643,
+      "step": 8570
+    },
+    {
+      "epoch": 0.8865468071915685,
+      "grad_norm": 0.8029395341873169,
+      "learning_rate": 1.6028632081827173e-05,
+      "loss": 0.5645,
+      "step": 8580
+    },
+    {
+      "epoch": 0.8875800785286216,
+      "grad_norm": 0.43516650795936584,
+      "learning_rate": 1.574109045678987e-05,
+      "loss": 0.5996,
+      "step": 8590
+    },
+    {
+      "epoch": 0.8886133498656748,
+      "grad_norm": 1.673608422279358,
+      "learning_rate": 1.545606751639034e-05,
+      "loss": 0.6781,
+      "step": 8600
+    },
+    {
+      "epoch": 0.8896466212027279,
+      "grad_norm": 1.0479223728179932,
+      "learning_rate": 1.5173566325113575e-05,
+      "loss": 0.5522,
+      "step": 8610
+    },
+    {
+      "epoch": 0.890679892539781,
+      "grad_norm": 1.2102528810501099,
+      "learning_rate": 1.4893589920331352e-05,
+      "loss": 0.5926,
+      "step": 8620
+    },
+    {
+      "epoch": 0.8917131638768341,
+      "grad_norm": 1.0137587785720825,
+      "learning_rate": 1.4616141312269737e-05,
+      "loss": 0.5393,
+      "step": 8630
+    },
+    {
+      "epoch": 0.8927464352138872,
+      "grad_norm": 2.0007264614105225,
+      "learning_rate": 1.4341223483976624e-05,
+      "loss": 0.6629,
+      "step": 8640
+    },
+    {
+      "epoch": 0.8937797065509403,
+      "grad_norm": 0.966791570186615,
+      "learning_rate": 1.406883939128964e-05,
+      "loss": 0.6255,
+      "step": 8650
+    },
+    {
+      "epoch": 0.8948129778879934,
+      "grad_norm": 1.436010479927063,
+      "learning_rate": 1.3798991962804541e-05,
+      "loss": 0.6226,
+      "step": 8660
+    },
+    {
+      "epoch": 0.8958462492250465,
+      "grad_norm": 1.060235857963562,
+      "learning_rate": 1.3531684099843483e-05,
+      "loss": 0.6685,
+      "step": 8670
+    },
+    {
+      "epoch": 0.8968795205620996,
+      "grad_norm": 0.578377902507782,
+      "learning_rate": 1.3266918676423962e-05,
+      "loss": 0.5681,
+      "step": 8680
+    },
+    {
+      "epoch": 0.8979127918991527,
+      "grad_norm": 0.5564404726028442,
+      "learning_rate": 1.300469853922795e-05,
+      "loss": 0.587,
+      "step": 8690
+    },
+    {
+      "epoch": 0.8989460632362059,
+      "grad_norm": 0.9987608194351196,
+      "learning_rate": 1.2745026507571145e-05,
+      "loss": 0.65,
+      "step": 8700
+    },
+    {
+      "epoch": 0.899979334573259,
+      "grad_norm": 1.4747377634048462,
+      "learning_rate": 1.2487905373372799e-05,
+      "loss": 0.6024,
+      "step": 8710
+    },
+    {
+      "epoch": 0.9010126059103121,
+      "grad_norm": 0.7746207118034363,
+      "learning_rate": 1.223333790112563e-05,
+      "loss": 0.6037,
+      "step": 8720
+    },
+    {
+      "epoch": 0.9020458772473652,
+      "grad_norm": 1.0953818559646606,
+      "learning_rate": 1.1981326827866124e-05,
+      "loss": 0.6469,
+      "step": 8730
+    },
+    {
+      "epoch": 0.9030791485844183,
+      "grad_norm": 1.1232373714447021,
+      "learning_rate": 1.1731874863145143e-05,
+      "loss": 0.591,
+      "step": 8740
+    },
+    {
+      "epoch": 0.9041124199214714,
+      "grad_norm": 2.062878131866455,
+      "learning_rate": 1.148498468899864e-05,
+      "loss": 0.6622,
+      "step": 8750
+    },
+    {
+      "epoch": 0.9051456912585245,
+      "grad_norm": 1.1355712413787842,
+      "learning_rate": 1.1240658959918965e-05,
+      "loss": 0.762,
+      "step": 8760
+    },
+    {
+      "epoch": 0.9061789625955776,
+      "grad_norm": 1.6199957132339478,
+      "learning_rate": 1.0998900302826382e-05,
+      "loss": 0.6976,
+      "step": 8770
+    },
+    {
+      "epoch": 0.9072122339326307,
+      "grad_norm": 1.111836552619934,
+      "learning_rate": 1.0759711317040649e-05,
+      "loss": 0.5116,
+      "step": 8780
+    },
+    {
+      "epoch": 0.9082455052696838,
+      "grad_norm": 1.650025725364685,
+      "learning_rate": 1.0523094574253101e-05,
+      "loss": 0.5384,
+      "step": 8790
+    },
+    {
+      "epoch": 0.909278776606737,
+      "grad_norm": 0.9917317628860474,
+      "learning_rate": 1.0289052618499218e-05,
+      "loss": 0.5526,
+      "step": 8800
+    },
+    {
+      "epoch": 0.9103120479437901,
+      "grad_norm": 0.618635356426239,
+      "learning_rate": 1.005758796613096e-05,
+      "loss": 0.6995,
+      "step": 8810
+    },
+    {
+      "epoch": 0.9113453192808432,
+      "grad_norm": 1.289839506149292,
+      "learning_rate": 9.828703105789983e-06,
+      "loss": 0.6453,
+      "step": 8820
+    },
+    {
+      "epoch": 0.9123785906178963,
+      "grad_norm": 1.1484878063201904,
+      "learning_rate": 9.602400498380542e-06,
+      "loss": 0.5949,
+      "step": 8830
+    },
+    {
+      "epoch": 0.9134118619549494,
+      "grad_norm": 1.1618013381958008,
+      "learning_rate": 9.378682577043524e-06,
+      "loss": 0.633,
+      "step": 8840
+    },
+    {
+      "epoch": 0.9144451332920025,
+      "grad_norm": 0.8447564840316772,
+      "learning_rate": 9.157551747129844e-06,
+      "loss": 0.6921,
+      "step": 8850
+    },
+    {
+      "epoch": 0.9154784046290556,
+      "grad_norm": 2.2064759731292725,
+      "learning_rate": 8.939010386174783e-06,
+      "loss": 0.5906,
+      "step": 8860
+    },
+    {
+      "epoch": 0.9165116759661087,
+      "grad_norm": 1.3931576013565063,
+      "learning_rate": 8.723060843872393e-06,
+      "loss": 0.6567,
+      "step": 8870
+    },
+    {
+      "epoch": 0.9175449473031618,
+      "grad_norm": 0.8289423584938049,
+      "learning_rate": 8.50970544205032e-06,
+      "loss": 0.6287,
+      "step": 8880
+    },
+    {
+      "epoch": 0.9185782186402149,
+      "grad_norm": 0.911371111869812,
+      "learning_rate": 8.298946474644575e-06,
+      "loss": 0.541,
+      "step": 8890
+    },
+    {
+      "epoch": 0.919611489977268,
+      "grad_norm": 1.081002950668335,
+      "learning_rate": 8.090786207675171e-06,
+      "loss": 0.6585,
+      "step": 8900
+    },
+    {
+      "epoch": 0.9206447613143212,
+      "grad_norm": 1.1589564085006714,
+      "learning_rate": 7.885226879221691e-06,
+      "loss": 0.644,
+      "step": 8910
+    },
+    {
+      "epoch": 0.9216780326513743,
+      "grad_norm": 0.6146034002304077,
+      "learning_rate": 7.682270699399057e-06,
+      "loss": 0.525,
+      "step": 8920
+    },
+    {
+      "epoch": 0.9227113039884274,
+      "grad_norm": 0.8697605729103088,
+      "learning_rate": 7.481919850333946e-06,
+      "loss": 0.5737,
+      "step": 8930
+    },
+    {
+      "epoch": 0.9237445753254805,
+      "grad_norm": 1.1086236238479614,
+      "learning_rate": 7.284176486141214e-06,
+      "loss": 0.5475,
+      "step": 8940
+    },
+    {
+      "epoch": 0.9247778466625336,
+      "grad_norm": 1.8016564846038818,
+      "learning_rate": 7.0890427329008964e-06,
+      "loss": 0.6598,
+      "step": 8950
+    },
+    {
+      "epoch": 0.9258111179995867,
+      "grad_norm": 1.1818935871124268,
+      "learning_rate": 6.896520688635111e-06,
+      "loss": 0.634,
+      "step": 8960
+    },
+    {
+      "epoch": 0.9268443893366398,
+      "grad_norm": 1.2800990343093872,
+      "learning_rate": 6.706612423285685e-06,
+      "loss": 0.6006,
+      "step": 8970
+    },
+    {
+      "epoch": 0.9278776606736929,
+      "grad_norm": 1.0055065155029297,
+      "learning_rate": 6.519319978691845e-06,
+      "loss": 0.5678,
+      "step": 8980
+    },
+    {
+      "epoch": 0.928910932010746,
+      "grad_norm": 2.050903081893921,
+      "learning_rate": 6.334645368568315e-06,
+      "loss": 0.6205,
+      "step": 8990
+    },
+    {
+      "epoch": 0.9299442033477991,
+      "grad_norm": 1.3951302766799927,
+      "learning_rate": 6.152590578483497e-06,
+      "loss": 0.5148,
+      "step": 9000
+    },
+    {
+      "epoch": 0.9309774746848523,
+      "grad_norm": 2.197643995285034,
+      "learning_rate": 5.973157565838327e-06,
+      "loss": 0.6931,
+      "step": 9010
+    },
+    {
+      "epoch": 0.9320107460219054,
+      "grad_norm": 1.161935567855835,
+      "learning_rate": 5.796348259845096e-06,
+      "loss": 0.5684,
+      "step": 9020
+    },
+    {
+      "epoch": 0.9330440173589585,
+      "grad_norm": 1.7656304836273193,
+      "learning_rate": 5.622164561506826e-06,
+      "loss": 0.6114,
+      "step": 9030
+    },
+    {
+      "epoch": 0.9340772886960116,
+      "grad_norm": 0.9781146049499512,
+      "learning_rate": 5.450608343596647e-06,
+      "loss": 0.5743,
+      "step": 9040
+    },
+    {
+      "epoch": 0.9351105600330647,
+      "grad_norm": 0.7141128182411194,
+      "learning_rate": 5.281681450637843e-06,
+      "loss": 0.5756,
+      "step": 9050
+    },
+    {
+      "epoch": 0.9361438313701178,
+      "grad_norm": 1.8938722610473633,
+      "learning_rate": 5.115385698883979e-06,
+      "loss": 0.5679,
+      "step": 9060
+    },
+    {
+      "epoch": 0.9371771027071709,
+      "grad_norm": 1.8188812732696533,
+      "learning_rate": 4.951722876299275e-06,
+      "loss": 0.744,
+      "step": 9070
+    },
+    {
+      "epoch": 0.938210374044224,
+      "grad_norm": 0.8902860879898071,
+      "learning_rate": 4.790694742539487e-06,
+      "loss": 0.6274,
+      "step": 9080
+    },
+    {
+      "epoch": 0.9392436453812771,
+      "grad_norm": 2.0708553791046143,
+      "learning_rate": 4.632303028932943e-06,
+      "loss": 0.6173,
+      "step": 9090
+    },
+    {
+      "epoch": 0.9402769167183302,
+      "grad_norm": 1.710466742515564,
+      "learning_rate": 4.476549438461958e-06,
+      "loss": 0.6267,
+      "step": 9100
+    },
+    {
+      "epoch": 0.9413101880553834,
+      "grad_norm": 0.8895155787467957,
+      "learning_rate": 4.323435645744445e-06,
+      "loss": 0.7292,
+      "step": 9110
+    },
+    {
+      "epoch": 0.9423434593924365,
+      "grad_norm": 1.9920309782028198,
+      "learning_rate": 4.172963297015997e-06,
+      "loss": 0.6778,
+      "step": 9120
+    },
+    {
+      "epoch": 0.9433767307294896,
+      "grad_norm": 1.4172946214675903,
+      "learning_rate": 4.025134010112258e-06,
+      "loss": 0.5764,
+      "step": 9130
+    },
+    {
+      "epoch": 0.9444100020665427,
+      "grad_norm": 0.9763116836547852,
+      "learning_rate": 3.879949374451269e-06,
+      "loss": 0.5936,
+      "step": 9140
+    },
+    {
+      "epoch": 0.9454432734035958,
+      "grad_norm": 0.7206150889396667,
+      "learning_rate": 3.737410951016623e-06,
+      "loss": 0.6432,
+      "step": 9150
+    },
+    {
+      "epoch": 0.9464765447406489,
+      "grad_norm": 1.7434405088424683,
+      "learning_rate": 3.5975202723405874e-06,
+      "loss": 0.6962,
+      "step": 9160
+    },
+    {
+      "epoch": 0.947509816077702,
+      "grad_norm": 1.5172473192214966,
+      "learning_rate": 3.4602788424876464e-06,
+      "loss": 0.6048,
+      "step": 9170
+    },
+    {
+      "epoch": 0.9485430874147551,
+      "grad_norm": 1.1150609254837036,
+      "learning_rate": 3.3256881370383183e-06,
+      "loss": 0.5865,
+      "step": 9180
+    },
+    {
+      "epoch": 0.9495763587518082,
+      "grad_norm": 0.8272204995155334,
+      "learning_rate": 3.1937496030732536e-06,
+      "loss": 0.7075,
+      "step": 9190
+    },
+    {
+      "epoch": 0.9506096300888613,
+      "grad_norm": 0.9530496597290039,
+      "learning_rate": 3.0644646591577174e-06,
+      "loss": 0.5166,
+      "step": 9200
+    },
+    {
+      "epoch": 0.9516429014259145,
+      "grad_norm": 0.7508150935173035,
+      "learning_rate": 2.9378346953264633e-06,
+      "loss": 0.5569,
+      "step": 9210
+    },
+    {
+      "epoch": 0.9526761727629676,
+      "grad_norm": 1.5538454055786133,
+      "learning_rate": 2.8138610730684686e-06,
+      "loss": 0.6117,
+      "step": 9220
+    },
+    {
+      "epoch": 0.9537094441000207,
+      "grad_norm": 1.6959922313690186,
+      "learning_rate": 2.692545125312612e-06,
+      "loss": 0.5843,
+      "step": 9230
+    },
+    {
+      "epoch": 0.9547427154370738,
+      "grad_norm": 1.9940367937088013,
+      "learning_rate": 2.573888156413212e-06,
+      "loss": 0.6021,
+      "step": 9240
+    },
+    {
+      "epoch": 0.9557759867741269,
+      "grad_norm": 1.3337304592132568,
+      "learning_rate": 2.4578914421359288e-06,
+      "loss": 0.525,
+      "step": 9250
+    },
+    {
+      "epoch": 0.95680925811118,
+      "grad_norm": 0.8987991213798523,
+      "learning_rate": 2.344556229644218e-06,
+      "loss": 0.6227,
+      "step": 9260
+    },
+    {
+      "epoch": 0.9578425294482331,
+      "grad_norm": 0.8351725339889526,
+      "learning_rate": 2.233883737485731e-06,
+      "loss": 0.5019,
+      "step": 9270
+    },
+    {
+      "epoch": 0.9588758007852862,
+      "grad_norm": 0.8255094289779663,
+      "learning_rate": 2.1258751555794654e-06,
+      "loss": 0.6226,
+      "step": 9280
+    },
+    {
+      "epoch": 0.9599090721223393,
+      "grad_norm": 0.882256269454956,
+      "learning_rate": 2.020531645202689e-06,
+      "loss": 0.4691,
+      "step": 9290
+    },
+    {
+      "epoch": 0.9609423434593924,
+      "grad_norm": 1.3614681959152222,
+      "learning_rate": 1.9178543389786475e-06,
+      "loss": 0.6947,
+      "step": 9300
+    },
+    {
+      "epoch": 0.9619756147964456,
+      "grad_norm": 1.0351027250289917,
+      "learning_rate": 1.8178443408642386e-06,
+      "loss": 0.5705,
+      "step": 9310
+    },
+    {
+      "epoch": 0.9630088861334987,
+      "grad_norm": 1.4347681999206543,
+      "learning_rate": 1.7205027261383565e-06,
+      "loss": 0.659,
+      "step": 9320
+    },
+    {
+      "epoch": 0.9640421574705518,
+      "grad_norm": 0.7792657017707825,
+      "learning_rate": 1.625830541390122e-06,
+      "loss": 0.5753,
+      "step": 9330
+    },
+    {
+      "epoch": 0.9650754288076049,
+      "grad_norm": 2.364046812057495,
+      "learning_rate": 1.5338288045076697e-06,
+      "loss": 0.626,
+      "step": 9340
+    },
+    {
+      "epoch": 0.966108700144658,
+      "grad_norm": 0.6518005728721619,
+      "learning_rate": 1.4444985046673798e-06,
+      "loss": 0.5726,
+      "step": 9350
+    },
+    {
+      "epoch": 0.9671419714817111,
+      "grad_norm": 0.8066127896308899,
+      "learning_rate": 1.3578406023229962e-06,
+      "loss": 0.6068,
+      "step": 9360
+    },
+    {
+      "epoch": 0.9681752428187642,
+      "grad_norm": 0.918552815914154,
+      "learning_rate": 1.2738560291954415e-06,
+      "loss": 0.5854,
+      "step": 9370
+    },
+    {
+      "epoch": 0.9692085141558173,
+      "grad_norm": 0.840164840221405,
+      "learning_rate": 1.1925456882627417e-06,
+      "loss": 0.5582,
+      "step": 9380
+    },
+    {
+      "epoch": 0.9702417854928704,
+      "grad_norm": 1.3730981349945068,
+      "learning_rate": 1.113910453750394e-06,
+      "loss": 0.6132,
+      "step": 9390
+    },
+    {
+      "epoch": 0.9712750568299235,
+      "grad_norm": 0.9575563669204712,
+      "learning_rate": 1.0379511711219313e-06,
+      "loss": 0.552,
+      "step": 9400
+    },
+    {
+      "epoch": 0.9723083281669767,
+      "grad_norm": 1.423699975013733,
+      "learning_rate": 9.646686570697061e-07,
+      "loss": 0.6279,
+      "step": 9410
+    },
+    {
+      "epoch": 0.9733415995040298,
+      "grad_norm": 1.0341185331344604,
+      "learning_rate": 8.9406369950637e-07,
+      "loss": 0.6182,
+      "step": 9420
+    },
+    {
+      "epoch": 0.9743748708410829,
+      "grad_norm": 0.9466649889945984,
+      "learning_rate": 8.261370575561866e-07,
+      "loss": 0.6104,
+      "step": 9430
+    },
+    {
+      "epoch": 0.975408142178136,
+      "grad_norm": 1.4443756341934204,
+      "learning_rate": 7.608894615468709e-07,
+      "loss": 0.6099,
+      "step": 9440
+    },
+    {
+      "epoch": 0.9764414135151891,
+      "grad_norm": 1.057411551475525,
+      "learning_rate": 6.983216130019288e-07,
+      "loss": 0.621,
+      "step": 9450
+    },
+    {
+      "epoch": 0.9774746848522422,
+      "grad_norm": 1.0595217943191528,
+      "learning_rate": 6.384341846329134e-07,
+      "loss": 0.6057,
+      "step": 9460
+    },
+    {
+      "epoch": 0.9785079561892953,
+      "grad_norm": 1.049188494682312,
+      "learning_rate": 5.812278203322918e-07,
+      "loss": 0.5873,
+      "step": 9470
+    },
+    {
+      "epoch": 0.9795412275263484,
+      "grad_norm": 1.212692379951477,
+      "learning_rate": 5.267031351664786e-07,
+      "loss": 0.6586,
+      "step": 9480
+    },
+    {
+      "epoch": 0.9805744988634015,
+      "grad_norm": 0.6824125647544861,
+      "learning_rate": 4.7486071536925745e-07,
+      "loss": 0.5165,
+      "step": 9490
+    },
+    {
+      "epoch": 0.9816077702004546,
+      "grad_norm": 0.6535896062850952,
+      "learning_rate": 4.257011183354809e-07,
+      "loss": 0.568,
+      "step": 9500
+    },
+    {
+      "epoch": 0.9826410415375078,
+      "grad_norm": 1.9366480112075806,
+      "learning_rate": 3.792248726150471e-07,
+      "loss": 0.5708,
+      "step": 9510
+    },
+    {
+      "epoch": 0.9836743128745609,
+      "grad_norm": 0.8946526646614075,
+      "learning_rate": 3.354324779071827e-07,
+      "loss": 0.499,
+      "step": 9520
+    },
+    {
+      "epoch": 0.984707584211614,
+      "grad_norm": 1.2145295143127441,
+      "learning_rate": 2.9432440505522406e-07,
+      "loss": 0.6079,
+      "step": 9530
+    },
+    {
+      "epoch": 0.9857408555486671,
+      "grad_norm": 0.948176383972168,
+      "learning_rate": 2.559010960413444e-07,
+      "loss": 0.5659,
+      "step": 9540
+    },
+    {
+      "epoch": 0.9867741268857202,
+      "grad_norm": 0.8348559141159058,
+      "learning_rate": 2.201629639819458e-07,
+      "loss": 0.6056,
+      "step": 9550
+    },
+    {
+      "epoch": 0.9878073982227733,
+      "grad_norm": 1.4996442794799805,
+      "learning_rate": 1.87110393123191e-07,
+      "loss": 0.521,
+      "step": 9560
+    },
+    {
+      "epoch": 0.9888406695598264,
+      "grad_norm": 0.999030590057373,
+      "learning_rate": 1.567437388368398e-07,
+      "loss": 0.6173,
+      "step": 9570
+    },
+    {
+      "epoch": 0.9898739408968795,
+      "grad_norm": 2.4354379177093506,
+      "learning_rate": 1.2906332761647433e-07,
+      "loss": 0.6978,
+      "step": 9580
+    },
+    {
+      "epoch": 0.9909072122339326,
+      "grad_norm": 1.067911148071289,
+      "learning_rate": 1.0406945707391868e-07,
+      "loss": 0.5733,
+      "step": 9590
+    },
+    {
+      "epoch": 0.9919404835709857,
+      "grad_norm": 0.8200865387916565,
+      "learning_rate": 8.176239593610246e-08,
+      "loss": 0.602,
+      "step": 9600
+    },
+    {
+      "epoch": 0.9929737549080389,
+      "grad_norm": 0.7436734437942505,
+      "learning_rate": 6.214238404214645e-08,
+      "loss": 0.6507,
+      "step": 9610
+    },
+    {
+      "epoch": 0.994007026245092,
+      "grad_norm": 1.0331168174743652,
+      "learning_rate": 4.520963234083686e-08,
+      "loss": 0.6224,
+      "step": 9620
+    },
+    {
+      "epoch": 0.9950402975821451,
+      "grad_norm": 1.3714491128921509,
+      "learning_rate": 3.09643228882106e-08,
+      "loss": 0.6278,
+      "step": 9630
+    },
+    {
+      "epoch": 0.9960735689191982,
+      "grad_norm": 1.478288173675537,
+      "learning_rate": 1.940660884577894e-08,
+      "loss": 0.6508,
+      "step": 9640
+    },
+    {
+      "epoch": 0.9971068402562513,
+      "grad_norm": 1.7626447677612305,
+      "learning_rate": 1.053661447877885e-08,
+      "loss": 0.5287,
+      "step": 9650
+    },
+    {
+      "epoch": 0.9981401115933044,
+      "grad_norm": 1.4299039840698242,
+      "learning_rate": 4.354435154813042e-09,
+      "loss": 0.6035,
+      "step": 9660
+    },
+    {
+      "epoch": 0.9991733829303575,
+      "grad_norm": 0.8753514885902405,
+      "learning_rate": 8.601373429339976e-10,
+      "loss": 0.6028,
+      "step": 9670
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 2.729660987854004,
+      "eval_runtime": 5409.1379,
+      "eval_samples_per_second": 9.159,
+      "eval_steps_per_second": 0.286,
+      "step": 9678
+    },
+    {
+      "epoch": 1.0,
+      "step": 9678,
+      "total_flos": 0.0,
+      "train_loss": 0.9993063533914609,
+      "train_runtime": 82593.2232,
+      "train_samples_per_second": 7.499,
+      "train_steps_per_second": 0.117
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 9678,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}