diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12157 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999711408040172,
+  "eval_steps": 500,
+  "global_step": 17325,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005771839196559984,
+      "grad_norm": 15.207763525645962,
+      "learning_rate": 9.995959129481038e-06,
+      "loss": 2.6,
+      "step": 10
+    },
+    {
+      "epoch": 0.0011543678393119967,
+      "grad_norm": 12.615616581147215,
+      "learning_rate": 9.990186457311089e-06,
+      "loss": 0.7554,
+      "step": 20
+    },
+    {
+      "epoch": 0.0017315517589679952,
+      "grad_norm": 38.969453309289854,
+      "learning_rate": 9.984413785141142e-06,
+      "loss": 0.6704,
+      "step": 30
+    },
+    {
+      "epoch": 0.0023087356786239935,
+      "grad_norm": 22.432338112146017,
+      "learning_rate": 9.978641112971195e-06,
+      "loss": 0.6122,
+      "step": 40
+    },
+    {
+      "epoch": 0.0028859195982799918,
+      "grad_norm": 6.366544502174357,
+      "learning_rate": 9.972868440801248e-06,
+      "loss": 0.5704,
+      "step": 50
+    },
+    {
+      "epoch": 0.0034631035179359905,
+      "grad_norm": 5.991950420382146,
+      "learning_rate": 9.967095768631301e-06,
+      "loss": 0.5725,
+      "step": 60
+    },
+    {
+      "epoch": 0.004040287437591989,
+      "grad_norm": 27.360506966186097,
+      "learning_rate": 9.961323096461352e-06,
+      "loss": 0.5616,
+      "step": 70
+    },
+    {
+      "epoch": 0.004617471357247987,
+      "grad_norm": 7.6551554884885835,
+      "learning_rate": 9.955550424291405e-06,
+      "loss": 0.5581,
+      "step": 80
+    },
+    {
+      "epoch": 0.005194655276903985,
+      "grad_norm": 5.380738579039347,
+      "learning_rate": 9.949777752121458e-06,
+      "loss": 0.5386,
+      "step": 90
+    },
+    {
+      "epoch": 0.0057718391965599835,
+      "grad_norm": 14.701566820920394,
+      "learning_rate": 9.944005079951511e-06,
+      "loss": 0.5277,
+      "step": 100
+    },
+    {
+      "epoch": 0.006349023116215982,
+      "grad_norm": 24.2926776932226,
+      "learning_rate": 9.938232407781563e-06,
+      "loss": 0.534,
+      "step": 110
+    },
+    {
+      "epoch": 0.006926207035871981,
+      "grad_norm": 18.333541125312422,
+      "learning_rate": 9.932459735611616e-06,
+      "loss": 0.5357,
+      "step": 120
+    },
+    {
+      "epoch": 0.007503390955527979,
+      "grad_norm": 4.413124779621301,
+      "learning_rate": 9.926687063441667e-06,
+      "loss": 0.5189,
+      "step": 130
+    },
+    {
+      "epoch": 0.008080574875183977,
+      "grad_norm": 6.736449877624926,
+      "learning_rate": 9.920914391271722e-06,
+      "loss": 0.5089,
+      "step": 140
+    },
+    {
+      "epoch": 0.008657758794839977,
+      "grad_norm": 7.499085213610092,
+      "learning_rate": 9.915141719101773e-06,
+      "loss": 0.515,
+      "step": 150
+    },
+    {
+      "epoch": 0.009234942714495974,
+      "grad_norm": 7.213309423254868,
+      "learning_rate": 9.909369046931826e-06,
+      "loss": 0.516,
+      "step": 160
+    },
+    {
+      "epoch": 0.009812126634151973,
+      "grad_norm": 6.224375486360128,
+      "learning_rate": 9.903596374761877e-06,
+      "loss": 0.4787,
+      "step": 170
+    },
+    {
+      "epoch": 0.01038931055380797,
+      "grad_norm": 4.666329365397574,
+      "learning_rate": 9.89782370259193e-06,
+      "loss": 0.5457,
+      "step": 180
+    },
+    {
+      "epoch": 0.01096649447346397,
+      "grad_norm": 7.269438501572431,
+      "learning_rate": 9.892051030421983e-06,
+      "loss": 0.4999,
+      "step": 190
+    },
+    {
+      "epoch": 0.011543678393119967,
+      "grad_norm": 3.4282676515510633,
+      "learning_rate": 9.886278358252037e-06,
+      "loss": 0.4979,
+      "step": 200
+    },
+    {
+      "epoch": 0.012120862312775966,
+      "grad_norm": 6.398684245798025,
+      "learning_rate": 9.880505686082088e-06,
+      "loss": 0.5097,
+      "step": 210
+    },
+    {
+      "epoch": 0.012698046232431964,
+      "grad_norm": 9.308090815837401,
+      "learning_rate": 9.874733013912141e-06,
+      "loss": 0.5003,
+      "step": 220
+    },
+    {
+      "epoch": 0.013275230152087963,
+      "grad_norm": 6.572806422530607,
+      "learning_rate": 9.868960341742194e-06,
+      "loss": 0.5151,
+      "step": 230
+    },
+    {
+      "epoch": 0.013852414071743962,
+      "grad_norm": 3.9790215956458206,
+      "learning_rate": 9.863187669572247e-06,
+      "loss": 0.5036,
+      "step": 240
+    },
+    {
+      "epoch": 0.01442959799139996,
+      "grad_norm": 9.129499916264713,
+      "learning_rate": 9.857414997402298e-06,
+      "loss": 0.4993,
+      "step": 250
+    },
+    {
+      "epoch": 0.015006781911055958,
+      "grad_norm": 5.341270073182352,
+      "learning_rate": 9.851642325232351e-06,
+      "loss": 0.5068,
+      "step": 260
+    },
+    {
+      "epoch": 0.015583965830711956,
+      "grad_norm": 8.88171344021306,
+      "learning_rate": 9.845869653062403e-06,
+      "loss": 0.5106,
+      "step": 270
+    },
+    {
+      "epoch": 0.016161149750367955,
+      "grad_norm": 6.256421713727477,
+      "learning_rate": 9.840096980892456e-06,
+      "loss": 0.5012,
+      "step": 280
+    },
+    {
+      "epoch": 0.016738333670023954,
+      "grad_norm": 14.828284045011356,
+      "learning_rate": 9.834324308722509e-06,
+      "loss": 0.4916,
+      "step": 290
+    },
+    {
+      "epoch": 0.017315517589679953,
+      "grad_norm": 13.134769047818303,
+      "learning_rate": 9.828551636552562e-06,
+      "loss": 0.5023,
+      "step": 300
+    },
+    {
+      "epoch": 0.01789270150933595,
+      "grad_norm": 13.132579195124707,
+      "learning_rate": 9.822778964382613e-06,
+      "loss": 0.4979,
+      "step": 310
+    },
+    {
+      "epoch": 0.018469885428991948,
+      "grad_norm": 7.799223760078651,
+      "learning_rate": 9.817006292212666e-06,
+      "loss": 0.5047,
+      "step": 320
+    },
+    {
+      "epoch": 0.019047069348647947,
+      "grad_norm": 9.15638603603503,
+      "learning_rate": 9.811233620042719e-06,
+      "loss": 0.5089,
+      "step": 330
+    },
+    {
+      "epoch": 0.019624253268303946,
+      "grad_norm": 4.959585555459444,
+      "learning_rate": 9.805460947872772e-06,
+      "loss": 0.4952,
+      "step": 340
+    },
+    {
+      "epoch": 0.020201437187959942,
+      "grad_norm": 11.68075308396199,
+      "learning_rate": 9.799688275702823e-06,
+      "loss": 0.4947,
+      "step": 350
+    },
+    {
+      "epoch": 0.02077862110761594,
+      "grad_norm": 5.031742433330389,
+      "learning_rate": 9.793915603532876e-06,
+      "loss": 0.4722,
+      "step": 360
+    },
+    {
+      "epoch": 0.02135580502727194,
+      "grad_norm": 7.396670940592179,
+      "learning_rate": 9.788142931362928e-06,
+      "loss": 0.4784,
+      "step": 370
+    },
+    {
+      "epoch": 0.02193298894692794,
+      "grad_norm": 12.81464819318774,
+      "learning_rate": 9.78237025919298e-06,
+      "loss": 0.5016,
+      "step": 380
+    },
+    {
+      "epoch": 0.02251017286658394,
+      "grad_norm": 6.952612493218752,
+      "learning_rate": 9.776597587023034e-06,
+      "loss": 0.486,
+      "step": 390
+    },
+    {
+      "epoch": 0.023087356786239934,
+      "grad_norm": 3.8994319647851134,
+      "learning_rate": 9.770824914853087e-06,
+      "loss": 0.5066,
+      "step": 400
+    },
+    {
+      "epoch": 0.023664540705895933,
+      "grad_norm": 8.754480087072261,
+      "learning_rate": 9.765052242683138e-06,
+      "loss": 0.477,
+      "step": 410
+    },
+    {
+      "epoch": 0.024241724625551932,
+      "grad_norm": 14.00837557848952,
+      "learning_rate": 9.759279570513191e-06,
+      "loss": 0.4698,
+      "step": 420
+    },
+    {
+      "epoch": 0.02481890854520793,
+      "grad_norm": 17.668700193812345,
+      "learning_rate": 9.753506898343244e-06,
+      "loss": 0.4809,
+      "step": 430
+    },
+    {
+      "epoch": 0.025396092464863927,
+      "grad_norm": 7.562292913968297,
+      "learning_rate": 9.747734226173297e-06,
+      "loss": 0.5051,
+      "step": 440
+    },
+    {
+      "epoch": 0.025973276384519926,
+      "grad_norm": 10.300306620163619,
+      "learning_rate": 9.741961554003348e-06,
+      "loss": 0.5036,
+      "step": 450
+    },
+    {
+      "epoch": 0.026550460304175925,
+      "grad_norm": 10.310969163346487,
+      "learning_rate": 9.736188881833401e-06,
+      "loss": 0.5029,
+      "step": 460
+    },
+    {
+      "epoch": 0.027127644223831925,
+      "grad_norm": 5.858481709305026,
+      "learning_rate": 9.730416209663453e-06,
+      "loss": 0.5021,
+      "step": 470
+    },
+    {
+      "epoch": 0.027704828143487924,
+      "grad_norm": 92.82221880785818,
+      "learning_rate": 9.724643537493506e-06,
+      "loss": 0.487,
+      "step": 480
+    },
+    {
+      "epoch": 0.02828201206314392,
+      "grad_norm": 5.060105328004147,
+      "learning_rate": 9.718870865323559e-06,
+      "loss": 0.5053,
+      "step": 490
+    },
+    {
+      "epoch": 0.02885919598279992,
+      "grad_norm": 5.269210555899195,
+      "learning_rate": 9.713098193153612e-06,
+      "loss": 0.5028,
+      "step": 500
+    },
+    {
+      "epoch": 0.029436379902455918,
+      "grad_norm": 2.809713279481296,
+      "learning_rate": 9.707325520983663e-06,
+      "loss": 0.4789,
+      "step": 510
+    },
+    {
+      "epoch": 0.030013563822111917,
+      "grad_norm": 5.364421325217877,
+      "learning_rate": 9.701552848813716e-06,
+      "loss": 0.488,
+      "step": 520
+    },
+    {
+      "epoch": 0.030590747741767916,
+      "grad_norm": 81.42225421545997,
+      "learning_rate": 9.69578017664377e-06,
+      "loss": 0.4923,
+      "step": 530
+    },
+    {
+      "epoch": 0.03116793166142391,
+      "grad_norm": 5.472841855533951,
+      "learning_rate": 9.690007504473822e-06,
+      "loss": 0.4801,
+      "step": 540
+    },
+    {
+      "epoch": 0.03174511558107991,
+      "grad_norm": 9.314853911550161,
+      "learning_rate": 9.684234832303874e-06,
+      "loss": 0.4889,
+      "step": 550
+    },
+    {
+      "epoch": 0.03232229950073591,
+      "grad_norm": 5.657736368331141,
+      "learning_rate": 9.678462160133927e-06,
+      "loss": 0.468,
+      "step": 560
+    },
+    {
+      "epoch": 0.03289948342039191,
+      "grad_norm": 5.316343227859894,
+      "learning_rate": 9.672689487963978e-06,
+      "loss": 0.4636,
+      "step": 570
+    },
+    {
+      "epoch": 0.03347666734004791,
+      "grad_norm": 6.818051213532348,
+      "learning_rate": 9.666916815794033e-06,
+      "loss": 0.4785,
+      "step": 580
+    },
+    {
+      "epoch": 0.03405385125970391,
+      "grad_norm": 5.515733833721638,
+      "learning_rate": 9.661144143624086e-06,
+      "loss": 0.4676,
+      "step": 590
+    },
+    {
+      "epoch": 0.034631035179359906,
+      "grad_norm": 4.414342893400196,
+      "learning_rate": 9.655371471454137e-06,
+      "loss": 0.4536,
+      "step": 600
+    },
+    {
+      "epoch": 0.0352082190990159,
+      "grad_norm": 5.5253782212489355,
+      "learning_rate": 9.64959879928419e-06,
+      "loss": 0.4831,
+      "step": 610
+    },
+    {
+      "epoch": 0.0357854030186719,
+      "grad_norm": 3.47774426972913,
+      "learning_rate": 9.643826127114241e-06,
+      "loss": 0.4877,
+      "step": 620
+    },
+    {
+      "epoch": 0.0363625869383279,
+      "grad_norm": 7.150841155138223,
+      "learning_rate": 9.638053454944294e-06,
+      "loss": 0.4629,
+      "step": 630
+    },
+    {
+      "epoch": 0.036939770857983896,
+      "grad_norm": 3.7894713504219357,
+      "learning_rate": 9.632280782774347e-06,
+      "loss": 0.4816,
+      "step": 640
+    },
+    {
+      "epoch": 0.037516954777639895,
+      "grad_norm": 5.346386715612294,
+      "learning_rate": 9.6265081106044e-06,
+      "loss": 0.4861,
+      "step": 650
+    },
+    {
+      "epoch": 0.038094138697295894,
+      "grad_norm": 4.822544974681145,
+      "learning_rate": 9.620735438434452e-06,
+      "loss": 0.4756,
+      "step": 660
+    },
+    {
+      "epoch": 0.03867132261695189,
+      "grad_norm": 5.810598113453792,
+      "learning_rate": 9.614962766264505e-06,
+      "loss": 0.4752,
+      "step": 670
+    },
+    {
+      "epoch": 0.03924850653660789,
+      "grad_norm": 19.319543215067025,
+      "learning_rate": 9.609190094094558e-06,
+      "loss": 0.4664,
+      "step": 680
+    },
+    {
+      "epoch": 0.03982569045626389,
+      "grad_norm": 3.6009359491010864,
+      "learning_rate": 9.60341742192461e-06,
+      "loss": 0.481,
+      "step": 690
+    },
+    {
+      "epoch": 0.040402874375919884,
+      "grad_norm": 5.102687170049628,
+      "learning_rate": 9.597644749754662e-06,
+      "loss": 0.4751,
+      "step": 700
+    },
+    {
+      "epoch": 0.04098005829557588,
+      "grad_norm": 6.202922207392348,
+      "learning_rate": 9.591872077584715e-06,
+      "loss": 0.4849,
+      "step": 710
+    },
+    {
+      "epoch": 0.04155724221523188,
+      "grad_norm": 7.8448485511355965,
+      "learning_rate": 9.586099405414766e-06,
+      "loss": 0.4947,
+      "step": 720
+    },
+    {
+      "epoch": 0.04213442613488788,
+      "grad_norm": 3.9340219935160863,
+      "learning_rate": 9.58032673324482e-06,
+      "loss": 0.4948,
+      "step": 730
+    },
+    {
+      "epoch": 0.04271161005454388,
+      "grad_norm": 4.596617570306747,
+      "learning_rate": 9.574554061074873e-06,
+      "loss": 0.4743,
+      "step": 740
+    },
+    {
+      "epoch": 0.04328879397419988,
+      "grad_norm": 2.1114969175976923,
+      "learning_rate": 9.568781388904926e-06,
+      "loss": 0.4717,
+      "step": 750
+    },
+    {
+      "epoch": 0.04386597789385588,
+      "grad_norm": 5.18107474670299,
+      "learning_rate": 9.563008716734977e-06,
+      "loss": 0.468,
+      "step": 760
+    },
+    {
+      "epoch": 0.04444316181351188,
+      "grad_norm": 4.705370463352637,
+      "learning_rate": 9.55723604456503e-06,
+      "loss": 0.4594,
+      "step": 770
+    },
+    {
+      "epoch": 0.04502034573316788,
+      "grad_norm": 4.477204626343746,
+      "learning_rate": 9.551463372395083e-06,
+      "loss": 0.4726,
+      "step": 780
+    },
+    {
+      "epoch": 0.04559752965282387,
+      "grad_norm": 5.11055150918499,
+      "learning_rate": 9.545690700225136e-06,
+      "loss": 0.4878,
+      "step": 790
+    },
+    {
+      "epoch": 0.04617471357247987,
+      "grad_norm": 5.92970242815697,
+      "learning_rate": 9.539918028055187e-06,
+      "loss": 0.4562,
+      "step": 800
+    },
+    {
+      "epoch": 0.04675189749213587,
+      "grad_norm": 32.04814941479638,
+      "learning_rate": 9.53414535588524e-06,
+      "loss": 0.4668,
+      "step": 810
+    },
+    {
+      "epoch": 0.047329081411791867,
+      "grad_norm": 7.483269872312162,
+      "learning_rate": 9.528372683715292e-06,
+      "loss": 0.4593,
+      "step": 820
+    },
+    {
+      "epoch": 0.047906265331447866,
+      "grad_norm": 3.0548203710383026,
+      "learning_rate": 9.522600011545345e-06,
+      "loss": 0.4734,
+      "step": 830
+    },
+    {
+      "epoch": 0.048483449251103865,
+      "grad_norm": 2.846647873568613,
+      "learning_rate": 9.516827339375398e-06,
+      "loss": 0.4583,
+      "step": 840
+    },
+    {
+      "epoch": 0.049060633170759864,
+      "grad_norm": 2.9958931469528753,
+      "learning_rate": 9.51105466720545e-06,
+      "loss": 0.4503,
+      "step": 850
+    },
+    {
+      "epoch": 0.04963781709041586,
+      "grad_norm": 3.669267079128399,
+      "learning_rate": 9.505281995035502e-06,
+      "loss": 0.4543,
+      "step": 860
+    },
+    {
+      "epoch": 0.05021500101007186,
+      "grad_norm": 4.959841634234083,
+      "learning_rate": 9.499509322865555e-06,
+      "loss": 0.4638,
+      "step": 870
+    },
+    {
+      "epoch": 0.050792184929727854,
+      "grad_norm": 6.2400754071583355,
+      "learning_rate": 9.493736650695608e-06,
+      "loss": 0.4675,
+      "step": 880
+    },
+    {
+      "epoch": 0.051369368849383854,
+      "grad_norm": 4.8904554144848325,
+      "learning_rate": 9.487963978525661e-06,
+      "loss": 0.4795,
+      "step": 890
+    },
+    {
+      "epoch": 0.05194655276903985,
+      "grad_norm": 5.718936473226969,
+      "learning_rate": 9.482191306355712e-06,
+      "loss": 0.475,
+      "step": 900
+    },
+    {
+      "epoch": 0.05252373668869585,
+      "grad_norm": 2.813275923923208,
+      "learning_rate": 9.476418634185765e-06,
+      "loss": 0.4588,
+      "step": 910
+    },
+    {
+      "epoch": 0.05310092060835185,
+      "grad_norm": 3.7622872130635825,
+      "learning_rate": 9.470645962015817e-06,
+      "loss": 0.4738,
+      "step": 920
+    },
+    {
+      "epoch": 0.05367810452800785,
+      "grad_norm": 5.251759558774021,
+      "learning_rate": 9.464873289845871e-06,
+      "loss": 0.4548,
+      "step": 930
+    },
+    {
+      "epoch": 0.05425528844766385,
+      "grad_norm": 4.650175199113373,
+      "learning_rate": 9.459100617675923e-06,
+      "loss": 0.469,
+      "step": 940
+    },
+    {
+      "epoch": 0.05483247236731985,
+      "grad_norm": 19.07121359753558,
+      "learning_rate": 9.453327945505976e-06,
+      "loss": 0.4593,
+      "step": 950
+    },
+    {
+      "epoch": 0.05540965628697585,
+      "grad_norm": 4.899239522938927,
+      "learning_rate": 9.447555273336027e-06,
+      "loss": 0.4693,
+      "step": 960
+    },
+    {
+      "epoch": 0.05598684020663185,
+      "grad_norm": 11.615659586114845,
+      "learning_rate": 9.44178260116608e-06,
+      "loss": 0.4447,
+      "step": 970
+    },
+    {
+      "epoch": 0.05656402412628784,
+      "grad_norm": 10.04941518596728,
+      "learning_rate": 9.436009928996133e-06,
+      "loss": 0.4617,
+      "step": 980
+    },
+    {
+      "epoch": 0.05714120804594384,
+      "grad_norm": 8.509458377026059,
+      "learning_rate": 9.430237256826186e-06,
+      "loss": 0.4449,
+      "step": 990
+    },
+    {
+      "epoch": 0.05771839196559984,
+      "grad_norm": 7.280354559581083,
+      "learning_rate": 9.424464584656238e-06,
+      "loss": 0.4473,
+      "step": 1000
+    },
+    {
+      "epoch": 0.058295575885255836,
+      "grad_norm": 6.178315104553298,
+      "learning_rate": 9.41869191248629e-06,
+      "loss": 0.4489,
+      "step": 1010
+    },
+    {
+      "epoch": 0.058872759804911835,
+      "grad_norm": 4.326714170134293,
+      "learning_rate": 9.412919240316344e-06,
+      "loss": 0.4421,
+      "step": 1020
+    },
+    {
+      "epoch": 0.059449943724567834,
+      "grad_norm": 5.121692072940591,
+      "learning_rate": 9.407146568146397e-06,
+      "loss": 0.4474,
+      "step": 1030
+    },
+    {
+      "epoch": 0.060027127644223834,
+      "grad_norm": 9.081808249358685,
+      "learning_rate": 9.401373895976448e-06,
+      "loss": 0.4425,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06060431156387983,
+      "grad_norm": 3.1699426939861644,
+      "learning_rate": 9.395601223806501e-06,
+      "loss": 0.4527,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06118149548353583,
+      "grad_norm": 4.867955044244513,
+      "learning_rate": 9.389828551636552e-06,
+      "loss": 0.4364,
+      "step": 1060
+    },
+    {
+      "epoch": 0.061758679403191824,
+      "grad_norm": 4.757339575426131,
+      "learning_rate": 9.384055879466605e-06,
+      "loss": 0.4744,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06233586332284782,
+      "grad_norm": 7.0039435743401235,
+      "learning_rate": 9.378283207296658e-06,
+      "loss": 0.4542,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06291304724250382,
+      "grad_norm": 3.3067436767056964,
+      "learning_rate": 9.372510535126711e-06,
+      "loss": 0.4434,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06349023116215982,
+      "grad_norm": 5.215962094810896,
+      "learning_rate": 9.366737862956763e-06,
+      "loss": 0.4413,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06406741508181582,
+      "grad_norm": 5.30554157155974,
+      "learning_rate": 9.360965190786816e-06,
+      "loss": 0.4492,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06464459900147182,
+      "grad_norm": 9.282132535227714,
+      "learning_rate": 9.355192518616869e-06,
+      "loss": 0.4552,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06522178292112782,
+      "grad_norm": 4.436227742317887,
+      "learning_rate": 9.349419846446922e-06,
+      "loss": 0.4629,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06579896684078382,
+      "grad_norm": 5.072582901364672,
+      "learning_rate": 9.343647174276975e-06,
+      "loss": 0.4394,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06637615076043982,
+      "grad_norm": 3.6484490415265287,
+      "learning_rate": 9.337874502107026e-06,
+      "loss": 0.4404,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06695333468009582,
+      "grad_norm": 5.179047870445265,
+      "learning_rate": 9.332101829937079e-06,
+      "loss": 0.4401,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06753051859975182,
+      "grad_norm": 5.1113945913805345,
+      "learning_rate": 9.32632915776713e-06,
+      "loss": 0.4642,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06810770251940781,
+      "grad_norm": 13.536539735261888,
+      "learning_rate": 9.320556485597183e-06,
+      "loss": 0.44,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06868488643906381,
+      "grad_norm": 39.38178075944957,
+      "learning_rate": 9.314783813427236e-06,
+      "loss": 0.4401,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06926207035871981,
+      "grad_norm": 5.347846132280397,
+      "learning_rate": 9.30901114125729e-06,
+      "loss": 0.4279,
+      "step": 1200
+    },
+    {
+      "epoch": 0.0698392542783758,
+      "grad_norm": 17.998657342305947,
+      "learning_rate": 9.30323846908734e-06,
+      "loss": 0.4289,
+      "step": 1210
+    },
+    {
+      "epoch": 0.0704164381980318,
+      "grad_norm": 3.352558494444376,
+      "learning_rate": 9.297465796917394e-06,
+      "loss": 0.4345,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0709936221176878,
+      "grad_norm": 5.425024450686575,
+      "learning_rate": 9.291693124747447e-06,
+      "loss": 0.4551,
+      "step": 1230
+    },
+    {
+      "epoch": 0.0715708060373438,
+      "grad_norm": 5.053276208988115,
+      "learning_rate": 9.2859204525775e-06,
+      "loss": 0.455,
+      "step": 1240
+    },
+    {
+      "epoch": 0.0721479899569998,
+      "grad_norm": 4.369475575649534,
+      "learning_rate": 9.280147780407551e-06,
+      "loss": 0.4359,
+      "step": 1250
+    },
+    {
+      "epoch": 0.0727251738766558,
+      "grad_norm": 7.966064775548766,
+      "learning_rate": 9.274375108237604e-06,
+      "loss": 0.4408,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07330235779631179,
+      "grad_norm": 2.9840558262546626,
+      "learning_rate": 9.268602436067656e-06,
+      "loss": 0.445,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07387954171596779,
+      "grad_norm": 6.923266339159757,
+      "learning_rate": 9.26282976389771e-06,
+      "loss": 0.4319,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07445672563562379,
+      "grad_norm": 5.445323456459659,
+      "learning_rate": 9.257057091727762e-06,
+      "loss": 0.4649,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07503390955527979,
+      "grad_norm": 2.258392109239514,
+      "learning_rate": 9.251284419557815e-06,
+      "loss": 0.4435,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07561109347493579,
+      "grad_norm": 6.963934239231957,
+      "learning_rate": 9.245511747387866e-06,
+      "loss": 0.4516,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07618827739459179,
+      "grad_norm": 3.741922936601378,
+      "learning_rate": 9.239739075217919e-06,
+      "loss": 0.4457,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07676546131424779,
+      "grad_norm": 4.697657485962023,
+      "learning_rate": 9.233966403047972e-06,
+      "loss": 0.4585,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07734264523390379,
+      "grad_norm": 2.7800726567106886,
+      "learning_rate": 9.228193730878025e-06,
+      "loss": 0.4303,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07791982915355979,
+      "grad_norm": 2.374574795518818,
+      "learning_rate": 9.222421058708076e-06,
+      "loss": 0.4468,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07849701307321579,
+      "grad_norm": 3.379706889838213,
+      "learning_rate": 9.21664838653813e-06,
+      "loss": 0.4468,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07907419699287178,
+      "grad_norm": 5.02793266796178,
+      "learning_rate": 9.210875714368182e-06,
+      "loss": 0.4541,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07965138091252778,
+      "grad_norm": 5.377287835326619,
+      "learning_rate": 9.205103042198235e-06,
+      "loss": 0.4577,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08022856483218377,
+      "grad_norm": 4.118238939930749,
+      "learning_rate": 9.199330370028287e-06,
+      "loss": 0.4535,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08080574875183977,
+      "grad_norm": 3.6462071059200785,
+      "learning_rate": 9.19355769785834e-06,
+      "loss": 0.452,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08138293267149577,
+      "grad_norm": 2.023709004387077,
+      "learning_rate": 9.187785025688391e-06,
+      "loss": 0.4385,
+      "step": 1410
+    },
+    {
+      "epoch": 0.08196011659115177,
+      "grad_norm": 2.718469268180074,
+      "learning_rate": 9.182012353518444e-06,
+      "loss": 0.4565,
+      "step": 1420
+    },
+    {
+      "epoch": 0.08253730051080777,
+      "grad_norm": 4.494982724398743,
+      "learning_rate": 9.176239681348497e-06,
+      "loss": 0.455,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08311448443046376,
+      "grad_norm": 3.008004247279657,
+      "learning_rate": 9.17046700917855e-06,
+      "loss": 0.4338,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08369166835011976,
+      "grad_norm": 4.471511152653035,
+      "learning_rate": 9.164694337008601e-06,
+      "loss": 0.4498,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08426885226977576,
+      "grad_norm": 41.046308308564996,
+      "learning_rate": 9.158921664838654e-06,
+      "loss": 0.444,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08484603618943176,
+      "grad_norm": 2.2817565591543087,
+      "learning_rate": 9.153148992668707e-06,
+      "loss": 0.4524,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08542322010908776,
+      "grad_norm": 2.7552178530343974,
+      "learning_rate": 9.14737632049876e-06,
+      "loss": 0.4395,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08600040402874376,
+      "grad_norm": 7.350119226751549,
+      "learning_rate": 9.141603648328812e-06,
+      "loss": 0.4439,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08657758794839976,
+      "grad_norm": 4.927329771744563,
+      "learning_rate": 9.135830976158865e-06,
+      "loss": 0.4435,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08715477186805576,
+      "grad_norm": 3.2061592885446433,
+      "learning_rate": 9.130058303988916e-06,
+      "loss": 0.4551,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08773195578771176,
+      "grad_norm": 2.81815402264405,
+      "learning_rate": 9.12428563181897e-06,
+      "loss": 0.4541,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08830913970736776,
+      "grad_norm": 1.8995196764329627,
+      "learning_rate": 9.118512959649022e-06,
+      "loss": 0.426,
+      "step": 1530
+    },
+    {
+      "epoch": 0.08888632362702376,
+      "grad_norm": 4.57597268577496,
+      "learning_rate": 9.112740287479075e-06,
+      "loss": 0.4388,
+      "step": 1540
+    },
+    {
+      "epoch": 0.08946350754667975,
+      "grad_norm": 4.888253541319005,
+      "learning_rate": 9.106967615309127e-06,
+      "loss": 0.4557,
+      "step": 1550
+    },
+    {
+      "epoch": 0.09004069146633575,
+      "grad_norm": 2.7824915648882853,
+      "learning_rate": 9.10119494313918e-06,
+      "loss": 0.4544,
+      "step": 1560
+    },
+    {
+      "epoch": 0.09061787538599175,
+      "grad_norm": 4.3535000068694645,
+      "learning_rate": 9.095422270969233e-06,
+      "loss": 0.4545,
+      "step": 1570
+    },
+    {
+      "epoch": 0.09119505930564774,
+      "grad_norm": 3.465323902631204,
+      "learning_rate": 9.089649598799286e-06,
+      "loss": 0.4343,
+      "step": 1580
+    },
+    {
+      "epoch": 0.09177224322530374,
+      "grad_norm": 3.685921584283666,
+      "learning_rate": 9.083876926629337e-06,
+      "loss": 0.4571,
+      "step": 1590
+    },
+    {
+      "epoch": 0.09234942714495974,
+      "grad_norm": 3.7270367548310457,
+      "learning_rate": 9.07810425445939e-06,
+      "loss": 0.4368,
+      "step": 1600
+    },
+    {
+      "epoch": 0.09292661106461574,
+      "grad_norm": 4.635316351567143,
+      "learning_rate": 9.072331582289441e-06,
+      "loss": 0.4333,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09350379498427173,
+      "grad_norm": 7.835128472709014,
+      "learning_rate": 9.066558910119494e-06,
+      "loss": 0.4371,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09408097890392773,
+      "grad_norm": 3.497453269659607,
+      "learning_rate": 9.060786237949547e-06,
+      "loss": 0.441,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09465816282358373,
+      "grad_norm": 8.942601742638523,
+      "learning_rate": 9.0550135657796e-06,
+      "loss": 0.4381,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09523534674323973,
+      "grad_norm": 4.589054052495082,
+      "learning_rate": 9.049240893609653e-06,
+      "loss": 0.4382,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09581253066289573,
+      "grad_norm": 5.380659756568862,
+      "learning_rate": 9.043468221439705e-06,
+      "loss": 0.4409,
+      "step": 1660
+    },
+    {
+      "epoch": 0.09638971458255173,
+      "grad_norm": 2.868711750692323,
+      "learning_rate": 9.037695549269758e-06,
+      "loss": 0.4456,
+      "step": 1670
+    },
+    {
+      "epoch": 0.09696689850220773,
+      "grad_norm": 2.5446684388172884,
+      "learning_rate": 9.03192287709981e-06,
+      "loss": 0.4491,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09754408242186373,
+      "grad_norm": 2.258241750087974,
+      "learning_rate": 9.026150204929864e-06,
+      "loss": 0.4366,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09812126634151973,
+      "grad_norm": 2.602005403010541,
+      "learning_rate": 9.020377532759915e-06,
+      "loss": 0.4336,
+      "step": 1700
+    },
+    {
+      "epoch": 0.09869845026117573,
+      "grad_norm": 2.209187153695394,
+      "learning_rate": 9.014604860589968e-06,
+      "loss": 0.4438,
+      "step": 1710
+    },
+    {
+      "epoch": 0.09927563418083173,
+      "grad_norm": 3.0657261742371205,
+      "learning_rate": 9.008832188420021e-06,
+      "loss": 0.4438,
+      "step": 1720
+    },
+    {
+      "epoch": 0.09985281810048773,
+      "grad_norm": 3.199466707600606,
+      "learning_rate": 9.003059516250074e-06,
+      "loss": 0.4315,
+      "step": 1730
+    },
+    {
+      "epoch": 0.10043000202014372,
+      "grad_norm": 2.6315666639919657,
+      "learning_rate": 8.997286844080125e-06,
+      "loss": 0.4309,
+      "step": 1740
+    },
+    {
+      "epoch": 0.10100718593979972,
+      "grad_norm": 2.438482770459306,
+      "learning_rate": 8.991514171910178e-06,
+      "loss": 0.4385,
+      "step": 1750
+    },
+    {
+      "epoch": 0.10158436985945571,
+      "grad_norm": 17.29321542355523,
+      "learning_rate": 8.98574149974023e-06,
+      "loss": 0.4408,
+      "step": 1760
+    },
+    {
+      "epoch": 0.10216155377911171,
+      "grad_norm": 4.982684643457552,
+      "learning_rate": 8.979968827570283e-06,
+      "loss": 0.4298,
+      "step": 1770
+    },
+    {
+      "epoch": 0.10273873769876771,
+      "grad_norm": 6.799624435849738,
+      "learning_rate": 8.974196155400336e-06,
+      "loss": 0.4334,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1033159216184237,
+      "grad_norm": 3.774388255492694,
+      "learning_rate": 8.968423483230389e-06,
+      "loss": 0.4308,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1038931055380797,
+      "grad_norm": 10.033128674066695,
+      "learning_rate": 8.96265081106044e-06,
+      "loss": 0.4341,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1044702894577357,
+      "grad_norm": 2.5219113776584123,
+      "learning_rate": 8.956878138890493e-06,
+      "loss": 0.4362,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1050474733773917,
+      "grad_norm": 2.419786699423789,
+      "learning_rate": 8.951105466720546e-06,
+      "loss": 0.4242,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1056246572970477,
+      "grad_norm": 60.08295959626865,
+      "learning_rate": 8.9453327945506e-06,
+      "loss": 0.4385,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1062018412167037,
+      "grad_norm": 3.096217392102352,
+      "learning_rate": 8.93956012238065e-06,
+      "loss": 0.4241,
+      "step": 1840
+    },
+    {
+      "epoch": 0.1067790251363597,
+      "grad_norm": 2.5832709944291503,
+      "learning_rate": 8.933787450210704e-06,
+      "loss": 0.4361,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1073562090560157,
+      "grad_norm": 2.3859955361814302,
+      "learning_rate": 8.928014778040755e-06,
+      "loss": 0.4213,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1079333929756717,
+      "grad_norm": 3.223331997025027,
+      "learning_rate": 8.922242105870808e-06,
+      "loss": 0.4489,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1085105768953277,
+      "grad_norm": 4.411834387897919,
+      "learning_rate": 8.916469433700861e-06,
+      "loss": 0.4336,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1090877608149837,
+      "grad_norm": 9.832585313890915,
+      "learning_rate": 8.910696761530914e-06,
+      "loss": 0.4441,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1096649447346397,
+      "grad_norm": 5.365037688190915,
+      "learning_rate": 8.904924089360965e-06,
+      "loss": 0.4367,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1102421286542957,
+      "grad_norm": 3.819456675382363,
+      "learning_rate": 8.899151417191018e-06,
+      "loss": 0.4272,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1108193125739517,
+      "grad_norm": 3.036048623046518,
+      "learning_rate": 8.893378745021071e-06,
+      "loss": 0.4338,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1113964964936077,
+      "grad_norm": 3.2508150432730196,
+      "learning_rate": 8.887606072851124e-06,
+      "loss": 0.4469,
+      "step": 1930
+    },
+    {
+      "epoch": 0.1119736804132637,
+      "grad_norm": 2.9078235731412305,
+      "learning_rate": 8.881833400681176e-06,
+      "loss": 0.4452,
+      "step": 1940
+    },
+    {
+      "epoch": 0.11255086433291968,
+      "grad_norm": 2.4501912449235648,
+      "learning_rate": 8.876060728511229e-06,
+      "loss": 0.4246,
+      "step": 1950
+    },
+    {
+      "epoch": 0.11312804825257568,
+      "grad_norm": 6.242529869364432,
+      "learning_rate": 8.87028805634128e-06,
+      "loss": 0.4459,
+      "step": 1960
+    },
+    {
+      "epoch": 0.11370523217223168,
+      "grad_norm": 25.285733450369438,
+      "learning_rate": 8.864515384171333e-06,
+      "loss": 0.4377,
+      "step": 1970
+    },
+    {
+      "epoch": 0.11428241609188768,
+      "grad_norm": 3.177411911875137,
+      "learning_rate": 8.858742712001386e-06,
+      "loss": 0.4197,
+      "step": 1980
+    },
+    {
+      "epoch": 0.11485960001154367,
+      "grad_norm": 3.117021809901308,
+      "learning_rate": 8.852970039831439e-06,
+      "loss": 0.4147,
+      "step": 1990
+    },
+    {
+      "epoch": 0.11543678393119967,
+      "grad_norm": 3.919723512318409,
+      "learning_rate": 8.84719736766149e-06,
+      "loss": 0.4414,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11601396785085567,
+      "grad_norm": 2.9342199285272796,
+      "learning_rate": 8.841424695491543e-06,
+      "loss": 0.427,
+      "step": 2010
+    },
+    {
+      "epoch": 0.11659115177051167,
+      "grad_norm": 1.918258053028722,
+      "learning_rate": 8.835652023321596e-06,
+      "loss": 0.4381,
+      "step": 2020
+    },
+    {
+      "epoch": 0.11716833569016767,
+      "grad_norm": 1.7751748211927454,
+      "learning_rate": 8.82987935115165e-06,
+      "loss": 0.4228,
+      "step": 2030
+    },
+    {
+      "epoch": 0.11774551960982367,
+      "grad_norm": 2.2012455830534385,
+      "learning_rate": 8.8241066789817e-06,
+      "loss": 0.4209,
+      "step": 2040
+    },
+    {
+      "epoch": 0.11832270352947967,
+      "grad_norm": 2.8237417214584544,
+      "learning_rate": 8.818334006811754e-06,
+      "loss": 0.4177,
+      "step": 2050
+    },
+    {
+      "epoch": 0.11889988744913567,
+      "grad_norm": 4.6392306678295805,
+      "learning_rate": 8.812561334641805e-06,
+      "loss": 0.4087,
+      "step": 2060
+    },
+    {
+      "epoch": 0.11947707136879167,
+      "grad_norm": 2.735063304359889,
+      "learning_rate": 8.80678866247186e-06,
+      "loss": 0.4272,
+      "step": 2070
+    },
+    {
+      "epoch": 0.12005425528844767,
+      "grad_norm": 3.0299802933466657,
+      "learning_rate": 8.801015990301911e-06,
+      "loss": 0.4371,
+      "step": 2080
+    },
+    {
+      "epoch": 0.12063143920810367,
+      "grad_norm": 5.124807390829953,
+      "learning_rate": 8.795243318131964e-06,
+      "loss": 0.4165,
+      "step": 2090
+    },
+    {
+      "epoch": 0.12120862312775967,
+      "grad_norm": 5.1154449505887385,
+      "learning_rate": 8.789470645962016e-06,
+      "loss": 0.4346,
+      "step": 2100
+    },
+    {
+      "epoch": 0.12178580704741566,
+      "grad_norm": 2.203947517060104,
+      "learning_rate": 8.783697973792069e-06,
+      "loss": 0.4225,
+      "step": 2110
+    },
+    {
+      "epoch": 0.12236299096707166,
+      "grad_norm": 2.1610133845450683,
+      "learning_rate": 8.777925301622122e-06,
+      "loss": 0.4121,
+      "step": 2120
+    },
+    {
+      "epoch": 0.12294017488672765,
+      "grad_norm": 13.953365114471524,
+      "learning_rate": 8.772152629452175e-06,
+      "loss": 0.4304,
+      "step": 2130
+    },
+    {
+      "epoch": 0.12351735880638365,
+      "grad_norm": 8.125499458140256,
+      "learning_rate": 8.766379957282226e-06,
+      "loss": 0.4351,
+      "step": 2140
+    },
+    {
+      "epoch": 0.12409454272603965,
+      "grad_norm": 10.554186301118584,
+      "learning_rate": 8.760607285112279e-06,
+      "loss": 0.4266,
+      "step": 2150
+    },
+    {
+      "epoch": 0.12467172664569565,
+      "grad_norm": 4.270037728186706,
+      "learning_rate": 8.754834612942332e-06,
+      "loss": 0.438,
+      "step": 2160
+    },
+    {
+      "epoch": 0.12524891056535165,
+      "grad_norm": 8.246386750659413,
+      "learning_rate": 8.749061940772385e-06,
+      "loss": 0.4283,
+      "step": 2170
+    },
+    {
+      "epoch": 0.12582609448500764,
+      "grad_norm": 3.506946209841679,
+      "learning_rate": 8.743289268602438e-06,
+      "loss": 0.4106,
+      "step": 2180
+    },
+    {
+      "epoch": 0.12640327840466364,
+      "grad_norm": 6.297671982703174,
+      "learning_rate": 8.73751659643249e-06,
+      "loss": 0.4329,
+      "step": 2190
+    },
+    {
+      "epoch": 0.12698046232431964,
+      "grad_norm": 3.18530097096491,
+      "learning_rate": 8.731743924262542e-06,
+      "loss": 0.4185,
+      "step": 2200
+    },
+    {
+      "epoch": 0.12755764624397564,
+      "grad_norm": 8.542498824615011,
+      "learning_rate": 8.725971252092594e-06,
+      "loss": 0.4255,
+      "step": 2210
+    },
+    {
+      "epoch": 0.12813483016363164,
+      "grad_norm": 4.533535099713348,
+      "learning_rate": 8.720198579922647e-06,
+      "loss": 0.423,
+      "step": 2220
+    },
+    {
+      "epoch": 0.12871201408328764,
+      "grad_norm": 3.889530684586912,
+      "learning_rate": 8.7144259077527e-06,
+      "loss": 0.4235,
+      "step": 2230
+    },
+    {
+      "epoch": 0.12928919800294364,
+      "grad_norm": 3.6486853451512777,
+      "learning_rate": 8.708653235582753e-06,
+      "loss": 0.4211,
+      "step": 2240
+    },
+    {
+      "epoch": 0.12986638192259964,
+      "grad_norm": 4.615918028596828,
+      "learning_rate": 8.702880563412804e-06,
+      "loss": 0.4088,
+      "step": 2250
+    },
+    {
+      "epoch": 0.13044356584225564,
+      "grad_norm": 3.850082033721987,
+      "learning_rate": 8.697107891242857e-06,
+      "loss": 0.4282,
+      "step": 2260
+    },
+    {
+      "epoch": 0.13102074976191164,
+      "grad_norm": 14.518871261878948,
+      "learning_rate": 8.69133521907291e-06,
+      "loss": 0.431,
+      "step": 2270
+    },
+    {
+      "epoch": 0.13159793368156764,
+      "grad_norm": 3.5577521941063797,
+      "learning_rate": 8.685562546902963e-06,
+      "loss": 0.4269,
+      "step": 2280
+    },
+    {
+      "epoch": 0.13217511760122364,
+      "grad_norm": 4.319284164265623,
+      "learning_rate": 8.679789874733014e-06,
+      "loss": 0.4151,
+      "step": 2290
+    },
+    {
+      "epoch": 0.13275230152087963,
+      "grad_norm": 5.95977091224079,
+      "learning_rate": 8.674017202563068e-06,
+      "loss": 0.434,
+      "step": 2300
+    },
+    {
+      "epoch": 0.13332948544053563,
+      "grad_norm": 5.717322004077494,
+      "learning_rate": 8.668244530393119e-06,
+      "loss": 0.412,
+      "step": 2310
+    },
+    {
+      "epoch": 0.13390666936019163,
+      "grad_norm": 4.1757793177667315,
+      "learning_rate": 8.662471858223172e-06,
+      "loss": 0.4154,
+      "step": 2320
+    },
+    {
+      "epoch": 0.13448385327984763,
+      "grad_norm": 2.4222311786523822,
+      "learning_rate": 8.656699186053225e-06,
+      "loss": 0.4075,
+      "step": 2330
+    },
+    {
+      "epoch": 0.13506103719950363,
+      "grad_norm": 4.366879694604865,
+      "learning_rate": 8.650926513883278e-06,
+      "loss": 0.4212,
+      "step": 2340
+    },
+    {
+      "epoch": 0.13563822111915963,
+      "grad_norm": 4.573113527280892,
+      "learning_rate": 8.64515384171333e-06,
+      "loss": 0.4235,
+      "step": 2350
+    },
+    {
+      "epoch": 0.13621540503881563,
+      "grad_norm": 2.7183246240824612,
+      "learning_rate": 8.639381169543382e-06,
+      "loss": 0.4067,
+      "step": 2360
+    },
+    {
+      "epoch": 0.13679258895847163,
+      "grad_norm": 2.858161612405957,
+      "learning_rate": 8.633608497373435e-06,
+      "loss": 0.404,
+      "step": 2370
+    },
+    {
+      "epoch": 0.13736977287812763,
+      "grad_norm": 4.083715288704242,
+      "learning_rate": 8.627835825203488e-06,
+      "loss": 0.4063,
+      "step": 2380
+    },
+    {
+      "epoch": 0.13794695679778363,
+      "grad_norm": 5.0777931108973595,
+      "learning_rate": 8.62206315303354e-06,
+      "loss": 0.422,
+      "step": 2390
+    },
+    {
+      "epoch": 0.13852414071743963,
+      "grad_norm": 7.507110618573909,
+      "learning_rate": 8.616290480863593e-06,
+      "loss": 0.421,
+      "step": 2400
+    },
+    {
+      "epoch": 0.1391013246370956,
+      "grad_norm": 5.833761997884338,
+      "learning_rate": 8.610517808693644e-06,
+      "loss": 0.4149,
+      "step": 2410
+    },
+    {
+      "epoch": 0.1396785085567516,
+      "grad_norm": 2.9291582156343523,
+      "learning_rate": 8.604745136523697e-06,
+      "loss": 0.4223,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1402556924764076,
+      "grad_norm": 3.8527608900197245,
+      "learning_rate": 8.59897246435375e-06,
+      "loss": 0.4381,
+      "step": 2430
+    },
+    {
+      "epoch": 0.1408328763960636,
+      "grad_norm": 5.009308450797531,
+      "learning_rate": 8.593199792183803e-06,
+      "loss": 0.4265,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1414100603157196,
+      "grad_norm": 4.305682778167611,
+      "learning_rate": 8.587427120013854e-06,
+      "loss": 0.422,
+      "step": 2450
+    },
+    {
+      "epoch": 0.1419872442353756,
+      "grad_norm": 4.119426729774013,
+      "learning_rate": 8.581654447843907e-06,
+      "loss": 0.4079,
+      "step": 2460
+    },
+    {
+      "epoch": 0.1425644281550316,
+      "grad_norm": 6.546817474930748,
+      "learning_rate": 8.57588177567396e-06,
+      "loss": 0.42,
+      "step": 2470
+    },
+    {
+      "epoch": 0.1431416120746876,
+      "grad_norm": 3.668016204374388,
+      "learning_rate": 8.570109103504013e-06,
+      "loss": 0.4323,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1437187959943436,
+      "grad_norm": 4.577088984604785,
+      "learning_rate": 8.564336431334065e-06,
+      "loss": 0.412,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1442959799139996,
+      "grad_norm": 5.57584596543327,
+      "learning_rate": 8.558563759164118e-06,
+      "loss": 0.4263,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1448731638336556,
+      "grad_norm": 2.695769154250156,
+      "learning_rate": 8.55279108699417e-06,
+      "loss": 0.4107,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1454503477533116,
+      "grad_norm": 13.95443843109925,
+      "learning_rate": 8.547018414824224e-06,
+      "loss": 0.4213,
+      "step": 2520
+    },
+    {
+      "epoch": 0.1460275316729676,
+      "grad_norm": 7.136056655011844,
+      "learning_rate": 8.541245742654275e-06,
+      "loss": 0.4274,
+      "step": 2530
+    },
+    {
+      "epoch": 0.14660471559262359,
+      "grad_norm": 6.544753723165297,
+      "learning_rate": 8.535473070484328e-06,
+      "loss": 0.414,
+      "step": 2540
+    },
+    {
+      "epoch": 0.14718189951227958,
+      "grad_norm": 5.397598920832979,
+      "learning_rate": 8.52970039831438e-06,
+      "loss": 0.4204,
+      "step": 2550
+    },
+    {
+      "epoch": 0.14775908343193558,
+      "grad_norm": 4.756636071432368,
+      "learning_rate": 8.523927726144432e-06,
+      "loss": 0.4124,
+      "step": 2560
+    },
+    {
+      "epoch": 0.14833626735159158,
+      "grad_norm": 14.516917457987418,
+      "learning_rate": 8.518155053974486e-06,
+      "loss": 0.4287,
+      "step": 2570
+    },
+    {
+      "epoch": 0.14891345127124758,
+      "grad_norm": 8.042684749735274,
+      "learning_rate": 8.512382381804539e-06,
+      "loss": 0.399,
+      "step": 2580
+    },
+    {
+      "epoch": 0.14949063519090358,
+      "grad_norm": 7.4003473379136775,
+      "learning_rate": 8.50660970963459e-06,
+      "loss": 0.4113,
+      "step": 2590
+    },
+    {
+      "epoch": 0.15006781911055958,
+      "grad_norm": 3.771759855408406,
+      "learning_rate": 8.500837037464643e-06,
+      "loss": 0.4056,
+      "step": 2600
+    },
+    {
+      "epoch": 0.15064500303021558,
+      "grad_norm": 3.6286193403799682,
+      "learning_rate": 8.495064365294696e-06,
+      "loss": 0.4246,
+      "step": 2610
+    },
+    {
+      "epoch": 0.15122218694987158,
+      "grad_norm": 3.863418428458885,
+      "learning_rate": 8.489291693124749e-06,
+      "loss": 0.4085,
+      "step": 2620
+    },
+    {
+      "epoch": 0.15179937086952758,
+      "grad_norm": 4.263367210964064,
+      "learning_rate": 8.4835190209548e-06,
+      "loss": 0.4223,
+      "step": 2630
+    },
+    {
+      "epoch": 0.15237655478918358,
+      "grad_norm": 3.337620340539389,
+      "learning_rate": 8.477746348784853e-06,
+      "loss": 0.4175,
+      "step": 2640
+    },
+    {
+      "epoch": 0.15295373870883958,
+      "grad_norm": 3.0267657042390788,
+      "learning_rate": 8.471973676614905e-06,
+      "loss": 0.4113,
+      "step": 2650
+    },
+    {
+      "epoch": 0.15353092262849558,
+      "grad_norm": 2.6106010567447893,
+      "learning_rate": 8.466201004444958e-06,
+      "loss": 0.4157,
+      "step": 2660
+    },
+    {
+      "epoch": 0.15410810654815157,
+      "grad_norm": 2.264152062153991,
+      "learning_rate": 8.46042833227501e-06,
+      "loss": 0.4338,
+      "step": 2670
+    },
+    {
+      "epoch": 0.15468529046780757,
+      "grad_norm": 1.9215310401757064,
+      "learning_rate": 8.454655660105064e-06,
+      "loss": 0.4114,
+      "step": 2680
+    },
+    {
+      "epoch": 0.15526247438746357,
+      "grad_norm": 2.615154051211967,
+      "learning_rate": 8.448882987935117e-06,
+      "loss": 0.4216,
+      "step": 2690
+    },
+    {
+      "epoch": 0.15583965830711957,
+      "grad_norm": 2.491915544143726,
+      "learning_rate": 8.443110315765168e-06,
+      "loss": 0.411,
+      "step": 2700
+    },
+    {
+      "epoch": 0.15641684222677557,
+      "grad_norm": 10.894040933901527,
+      "learning_rate": 8.437337643595221e-06,
+      "loss": 0.43,
+      "step": 2710
+    },
+    {
+      "epoch": 0.15699402614643157,
+      "grad_norm": 7.186343591522965,
+      "learning_rate": 8.431564971425274e-06,
+      "loss": 0.4119,
+      "step": 2720
+    },
+    {
+      "epoch": 0.15757121006608757,
+      "grad_norm": 11.15516113836552,
+      "learning_rate": 8.425792299255327e-06,
+      "loss": 0.4206,
+      "step": 2730
+    },
+    {
+      "epoch": 0.15814839398574357,
+      "grad_norm": 3.3849670049000626,
+      "learning_rate": 8.420019627085378e-06,
+      "loss": 0.4098,
+      "step": 2740
+    },
+    {
+      "epoch": 0.15872557790539957,
+      "grad_norm": 2.3478329302680176,
+      "learning_rate": 8.414246954915431e-06,
+      "loss": 0.4358,
+      "step": 2750
+    },
+    {
+      "epoch": 0.15930276182505557,
+      "grad_norm": 5.368826342998146,
+      "learning_rate": 8.408474282745483e-06,
+      "loss": 0.4114,
+      "step": 2760
+    },
+    {
+      "epoch": 0.15987994574471157,
+      "grad_norm": 2.778388386877281,
+      "learning_rate": 8.402701610575536e-06,
+      "loss": 0.4149,
+      "step": 2770
+    },
+    {
+      "epoch": 0.16045712966436754,
+      "grad_norm": 6.269723327157733,
+      "learning_rate": 8.396928938405589e-06,
+      "loss": 0.4059,
+      "step": 2780
+    },
+    {
+      "epoch": 0.16103431358402354,
+      "grad_norm": 3.51009855789622,
+      "learning_rate": 8.391156266235642e-06,
+      "loss": 0.4073,
+      "step": 2790
+    },
+    {
+      "epoch": 0.16161149750367954,
+      "grad_norm": 2.84216423707538,
+      "learning_rate": 8.385383594065693e-06,
+      "loss": 0.4249,
+      "step": 2800
+    },
+    {
+      "epoch": 0.16218868142333553,
+      "grad_norm": 4.458775881028344,
+      "learning_rate": 8.379610921895746e-06,
+      "loss": 0.4248,
+      "step": 2810
+    },
+    {
+      "epoch": 0.16276586534299153,
+      "grad_norm": 4.7717011271745875,
+      "learning_rate": 8.3738382497258e-06,
+      "loss": 0.4031,
+      "step": 2820
+    },
+    {
+      "epoch": 0.16334304926264753,
+      "grad_norm": 2.339306133151702,
+      "learning_rate": 8.368065577555852e-06,
+      "loss": 0.4089,
+      "step": 2830
+    },
+    {
+      "epoch": 0.16392023318230353,
+      "grad_norm": 2.242140863987944,
+      "learning_rate": 8.362292905385904e-06,
+      "loss": 0.4183,
+      "step": 2840
+    },
+    {
+      "epoch": 0.16449741710195953,
+      "grad_norm": 2.50085114867505,
+      "learning_rate": 8.356520233215957e-06,
+      "loss": 0.4209,
+      "step": 2850
+    },
+    {
+      "epoch": 0.16507460102161553,
+      "grad_norm": 2.426762382327375,
+      "learning_rate": 8.350747561046008e-06,
+      "loss": 0.4009,
+      "step": 2860
+    },
+    {
+      "epoch": 0.16565178494127153,
+      "grad_norm": 3.535990085619889,
+      "learning_rate": 8.344974888876063e-06,
+      "loss": 0.4145,
+      "step": 2870
+    },
+    {
+      "epoch": 0.16622896886092753,
+      "grad_norm": 1.9434196504603485,
+      "learning_rate": 8.339202216706114e-06,
+      "loss": 0.4029,
+      "step": 2880
+    },
+    {
+      "epoch": 0.16680615278058353,
+      "grad_norm": 3.647575465250371,
+      "learning_rate": 8.333429544536167e-06,
+      "loss": 0.4165,
+      "step": 2890
+    },
+    {
+      "epoch": 0.16738333670023953,
+      "grad_norm": 3.21796325121302,
+      "learning_rate": 8.327656872366218e-06,
+      "loss": 0.419,
+      "step": 2900
+    },
+    {
+      "epoch": 0.16796052061989553,
+      "grad_norm": 2.49775247695972,
+      "learning_rate": 8.321884200196271e-06,
+      "loss": 0.4179,
+      "step": 2910
+    },
+    {
+      "epoch": 0.16853770453955152,
+      "grad_norm": 1.992198339323461,
+      "learning_rate": 8.316111528026324e-06,
+      "loss": 0.4062,
+      "step": 2920
+    },
+    {
+      "epoch": 0.16911488845920752,
+      "grad_norm": 2.82923364634973,
+      "learning_rate": 8.310338855856377e-06,
+      "loss": 0.3997,
+      "step": 2930
+    },
+    {
+      "epoch": 0.16969207237886352,
+      "grad_norm": 2.414622342801528,
+      "learning_rate": 8.304566183686429e-06,
+      "loss": 0.3978,
+      "step": 2940
+    },
+    {
+      "epoch": 0.17026925629851952,
+      "grad_norm": 3.500929543134566,
+      "learning_rate": 8.298793511516482e-06,
+      "loss": 0.4214,
+      "step": 2950
+    },
+    {
+      "epoch": 0.17084644021817552,
+      "grad_norm": 4.533770301559384,
+      "learning_rate": 8.293020839346535e-06,
+      "loss": 0.4003,
+      "step": 2960
+    },
+    {
+      "epoch": 0.17142362413783152,
+      "grad_norm": 3.463538477874354,
+      "learning_rate": 8.287248167176588e-06,
+      "loss": 0.4049,
+      "step": 2970
+    },
+    {
+      "epoch": 0.17200080805748752,
+      "grad_norm": 2.6643523986040023,
+      "learning_rate": 8.281475495006639e-06,
+      "loss": 0.4157,
+      "step": 2980
+    },
+    {
+      "epoch": 0.17257799197714352,
+      "grad_norm": 5.1202436242415175,
+      "learning_rate": 8.275702822836692e-06,
+      "loss": 0.4119,
+      "step": 2990
+    },
+    {
+      "epoch": 0.17315517589679952,
+      "grad_norm": 4.375642688680107,
+      "learning_rate": 8.269930150666743e-06,
+      "loss": 0.4241,
+      "step": 3000
+    },
+    {
+      "epoch": 0.17373235981645552,
+      "grad_norm": 13.184415266130015,
+      "learning_rate": 8.264157478496796e-06,
+      "loss": 0.3931,
+      "step": 3010
+    },
+    {
+      "epoch": 0.17430954373611152,
+      "grad_norm": 3.5119329436961872,
+      "learning_rate": 8.25838480632685e-06,
+      "loss": 0.4332,
+      "step": 3020
+    },
+    {
+      "epoch": 0.17488672765576752,
+      "grad_norm": 2.057788212146213,
+      "learning_rate": 8.252612134156902e-06,
+      "loss": 0.4026,
+      "step": 3030
+    },
+    {
+      "epoch": 0.17546391157542351,
+      "grad_norm": 2.4025914280540293,
+      "learning_rate": 8.246839461986954e-06,
+      "loss": 0.4297,
+      "step": 3040
+    },
+    {
+      "epoch": 0.1760410954950795,
+      "grad_norm": 2.3273476646183187,
+      "learning_rate": 8.241066789817007e-06,
+      "loss": 0.4256,
+      "step": 3050
+    },
+    {
+      "epoch": 0.1766182794147355,
+      "grad_norm": 3.968034409940849,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.3911,
+      "step": 3060
+    },
+    {
+      "epoch": 0.1771954633343915,
+      "grad_norm": 3.5987187085577648,
+      "learning_rate": 8.229521445477113e-06,
+      "loss": 0.4137,
+      "step": 3070
+    },
+    {
+      "epoch": 0.1777726472540475,
+      "grad_norm": 1.6685524653169845,
+      "learning_rate": 8.223748773307164e-06,
+      "loss": 0.4115,
+      "step": 3080
+    },
+    {
+      "epoch": 0.1783498311737035,
+      "grad_norm": 15.206856747451255,
+      "learning_rate": 8.217976101137217e-06,
+      "loss": 0.3993,
+      "step": 3090
+    },
+    {
+      "epoch": 0.1789270150933595,
+      "grad_norm": 3.2248196017828685,
+      "learning_rate": 8.212203428967268e-06,
+      "loss": 0.4263,
+      "step": 3100
+    },
+    {
+      "epoch": 0.1795041990130155,
+      "grad_norm": 5.06230398614513,
+      "learning_rate": 8.206430756797322e-06,
+      "loss": 0.4241,
+      "step": 3110
+    },
+    {
+      "epoch": 0.1800813829326715,
+      "grad_norm": 24.83780832952265,
+      "learning_rate": 8.200658084627375e-06,
+      "loss": 0.4139,
+      "step": 3120
+    },
+    {
+      "epoch": 0.1806585668523275,
+      "grad_norm": 2.8321722632943094,
+      "learning_rate": 8.194885412457428e-06,
+      "loss": 0.4233,
+      "step": 3130
+    },
+    {
+      "epoch": 0.1812357507719835,
+      "grad_norm": 3.0260534676681727,
+      "learning_rate": 8.189112740287479e-06,
+      "loss": 0.4142,
+      "step": 3140
+    },
+    {
+      "epoch": 0.18181293469163948,
+      "grad_norm": 4.68099176672326,
+      "learning_rate": 8.183340068117532e-06,
+      "loss": 0.4118,
+      "step": 3150
+    },
+    {
+      "epoch": 0.18239011861129548,
+      "grad_norm": 11.310308197618612,
+      "learning_rate": 8.177567395947585e-06,
+      "loss": 0.415,
+      "step": 3160
+    },
+    {
+      "epoch": 0.18296730253095148,
+      "grad_norm": 4.572818458115979,
+      "learning_rate": 8.171794723777638e-06,
+      "loss": 0.4015,
+      "step": 3170
+    },
+    {
+      "epoch": 0.18354448645060747,
+      "grad_norm": 6.450421868687323,
+      "learning_rate": 8.16602205160769e-06,
+      "loss": 0.4263,
+      "step": 3180
+    },
+    {
+      "epoch": 0.18412167037026347,
+      "grad_norm": 3.4296934325228263,
+      "learning_rate": 8.160249379437742e-06,
+      "loss": 0.4124,
+      "step": 3190
+    },
+    {
+      "epoch": 0.18469885428991947,
+      "grad_norm": 9.959813392971029,
+      "learning_rate": 8.154476707267794e-06,
+      "loss": 0.4015,
+      "step": 3200
+    },
+    {
+      "epoch": 0.18527603820957547,
+      "grad_norm": 3.3029855110594695,
+      "learning_rate": 8.148704035097847e-06,
+      "loss": 0.4009,
+      "step": 3210
+    },
+    {
+      "epoch": 0.18585322212923147,
+      "grad_norm": 2.8097447076161273,
+      "learning_rate": 8.142931362927901e-06,
+      "loss": 0.4002,
+      "step": 3220
+    },
+    {
+      "epoch": 0.18643040604888747,
+      "grad_norm": 34.98571611934199,
+      "learning_rate": 8.137158690757953e-06,
+      "loss": 0.405,
+      "step": 3230
+    },
+    {
+      "epoch": 0.18700758996854347,
+      "grad_norm": 9.148426684066074,
+      "learning_rate": 8.131386018588006e-06,
+      "loss": 0.4044,
+      "step": 3240
+    },
+    {
+      "epoch": 0.18758477388819947,
+      "grad_norm": 4.173548086962396,
+      "learning_rate": 8.125613346418057e-06,
+      "loss": 0.4217,
+      "step": 3250
+    },
+    {
+      "epoch": 0.18816195780785547,
+      "grad_norm": 53.37411486094812,
+      "learning_rate": 8.11984067424811e-06,
+      "loss": 0.4138,
+      "step": 3260
+    },
+    {
+      "epoch": 0.18873914172751147,
+      "grad_norm": 2.3480379802050733,
+      "learning_rate": 8.114068002078163e-06,
+      "loss": 0.4033,
+      "step": 3270
+    },
+    {
+      "epoch": 0.18931632564716747,
+      "grad_norm": 2.221291345204337,
+      "learning_rate": 8.108295329908216e-06,
+      "loss": 0.3914,
+      "step": 3280
+    },
+    {
+      "epoch": 0.18989350956682347,
+      "grad_norm": 4.990540206022043,
+      "learning_rate": 8.102522657738267e-06,
+      "loss": 0.3985,
+      "step": 3290
+    },
+    {
+      "epoch": 0.19047069348647946,
+      "grad_norm": 4.025051683089965,
+      "learning_rate": 8.09674998556832e-06,
+      "loss": 0.415,
+      "step": 3300
+    },
+    {
+      "epoch": 0.19104787740613546,
+      "grad_norm": 2.5251837331957607,
+      "learning_rate": 8.090977313398373e-06,
+      "loss": 0.4133,
+      "step": 3310
+    },
+    {
+      "epoch": 0.19162506132579146,
+      "grad_norm": 2.380391015882718,
+      "learning_rate": 8.085204641228426e-06,
+      "loss": 0.3926,
+      "step": 3320
+    },
+    {
+      "epoch": 0.19220224524544746,
+      "grad_norm": 4.41017165770354,
+      "learning_rate": 8.079431969058478e-06,
+      "loss": 0.4211,
+      "step": 3330
+    },
+    {
+      "epoch": 0.19277942916510346,
+      "grad_norm": 2.3063890276035846,
+      "learning_rate": 8.07365929688853e-06,
+      "loss": 0.4138,
+      "step": 3340
+    },
+    {
+      "epoch": 0.19335661308475946,
+      "grad_norm": 4.271620366325266,
+      "learning_rate": 8.067886624718582e-06,
+      "loss": 0.4093,
+      "step": 3350
+    },
+    {
+      "epoch": 0.19393379700441546,
+      "grad_norm": 32.70807377417423,
+      "learning_rate": 8.062113952548635e-06,
+      "loss": 0.4035,
+      "step": 3360
+    },
+    {
+      "epoch": 0.19451098092407146,
+      "grad_norm": 2.4018179935448276,
+      "learning_rate": 8.056341280378688e-06,
+      "loss": 0.4126,
+      "step": 3370
+    },
+    {
+      "epoch": 0.19508816484372746,
+      "grad_norm": 3.1209658711505552,
+      "learning_rate": 8.050568608208741e-06,
+      "loss": 0.4073,
+      "step": 3380
+    },
+    {
+      "epoch": 0.19566534876338346,
+      "grad_norm": 3.6351629831980024,
+      "learning_rate": 8.044795936038793e-06,
+      "loss": 0.4041,
+      "step": 3390
+    },
+    {
+      "epoch": 0.19624253268303946,
+      "grad_norm": 3.652368336229329,
+      "learning_rate": 8.039023263868846e-06,
+      "loss": 0.4112,
+      "step": 3400
+    },
+    {
+      "epoch": 0.19681971660269545,
+      "grad_norm": 2.984404633116569,
+      "learning_rate": 8.033250591698899e-06,
+      "loss": 0.4134,
+      "step": 3410
+    },
+    {
+      "epoch": 0.19739690052235145,
+      "grad_norm": 2.4526774932392024,
+      "learning_rate": 8.027477919528952e-06,
+      "loss": 0.4011,
+      "step": 3420
+    },
+    {
+      "epoch": 0.19797408444200745,
+      "grad_norm": 4.571072305725846,
+      "learning_rate": 8.021705247359003e-06,
+      "loss": 0.4097,
+      "step": 3430
+    },
+    {
+      "epoch": 0.19855126836166345,
+      "grad_norm": 2.6003306321869895,
+      "learning_rate": 8.015932575189056e-06,
+      "loss": 0.406,
+      "step": 3440
+    },
+    {
+      "epoch": 0.19912845228131945,
+      "grad_norm": 4.171122512202554,
+      "learning_rate": 8.010159903019107e-06,
+      "loss": 0.4083,
+      "step": 3450
+    },
+    {
+      "epoch": 0.19970563620097545,
+      "grad_norm": 3.2250910237655006,
+      "learning_rate": 8.00438723084916e-06,
+      "loss": 0.415,
+      "step": 3460
+    },
+    {
+      "epoch": 0.20028282012063145,
+      "grad_norm": 4.871724300699273,
+      "learning_rate": 7.998614558679213e-06,
+      "loss": 0.4296,
+      "step": 3470
+    },
+    {
+      "epoch": 0.20086000404028745,
+      "grad_norm": 2.9738579217002887,
+      "learning_rate": 7.992841886509266e-06,
+      "loss": 0.4109,
+      "step": 3480
+    },
+    {
+      "epoch": 0.20143718795994345,
+      "grad_norm": 5.488557624180365,
+      "learning_rate": 7.987069214339318e-06,
+      "loss": 0.4203,
+      "step": 3490
+    },
+    {
+      "epoch": 0.20201437187959945,
+      "grad_norm": 6.30148557707432,
+      "learning_rate": 7.98129654216937e-06,
+      "loss": 0.4166,
+      "step": 3500
+    },
+    {
+      "epoch": 0.20259155579925545,
+      "grad_norm": 4.159946548536443,
+      "learning_rate": 7.975523869999424e-06,
+      "loss": 0.4126,
+      "step": 3510
+    },
+    {
+      "epoch": 0.20316873971891142,
+      "grad_norm": 15.334889223997754,
+      "learning_rate": 7.969751197829477e-06,
+      "loss": 0.3947,
+      "step": 3520
+    },
+    {
+      "epoch": 0.20374592363856742,
+      "grad_norm": 3.025285897677573,
+      "learning_rate": 7.963978525659528e-06,
+      "loss": 0.403,
+      "step": 3530
+    },
+    {
+      "epoch": 0.20432310755822342,
+      "grad_norm": 4.262956488693212,
+      "learning_rate": 7.958205853489581e-06,
+      "loss": 0.4066,
+      "step": 3540
+    },
+    {
+      "epoch": 0.20490029147787941,
+      "grad_norm": 2.5780055567960143,
+      "learning_rate": 7.952433181319632e-06,
+      "loss": 0.4032,
+      "step": 3550
+    },
+    {
+      "epoch": 0.20547747539753541,
+      "grad_norm": 4.956342345823864,
+      "learning_rate": 7.946660509149685e-06,
+      "loss": 0.4022,
+      "step": 3560
+    },
+    {
+      "epoch": 0.2060546593171914,
+      "grad_norm": 3.89966793053596,
+      "learning_rate": 7.940887836979738e-06,
+      "loss": 0.4059,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2066318432368474,
+      "grad_norm": 2.598383914711163,
+      "learning_rate": 7.935115164809791e-06,
+      "loss": 0.3988,
+      "step": 3580
+    },
+    {
+      "epoch": 0.2072090271565034,
+      "grad_norm": 3.673402619499888,
+      "learning_rate": 7.929342492639843e-06,
+      "loss": 0.4117,
+      "step": 3590
+    },
+    {
+      "epoch": 0.2077862110761594,
+      "grad_norm": 5.7308049694271235,
+      "learning_rate": 7.923569820469896e-06,
+      "loss": 0.4008,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2083633949958154,
+      "grad_norm": 7.935773567013425,
+      "learning_rate": 7.917797148299949e-06,
+      "loss": 0.3877,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2089405789154714,
+      "grad_norm": 6.009048915243287,
+      "learning_rate": 7.912024476130002e-06,
+      "loss": 0.4065,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2095177628351274,
+      "grad_norm": 2.95995624359623,
+      "learning_rate": 7.906251803960053e-06,
+      "loss": 0.401,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2100949467547834,
+      "grad_norm": 3.1037746592103255,
+      "learning_rate": 7.900479131790106e-06,
+      "loss": 0.4013,
+      "step": 3640
+    },
+    {
+      "epoch": 0.2106721306744394,
+      "grad_norm": 6.029705929267264,
+      "learning_rate": 7.894706459620158e-06,
+      "loss": 0.4085,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2112493145940954,
+      "grad_norm": 5.81081103172759,
+      "learning_rate": 7.888933787450212e-06,
+      "loss": 0.4062,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2118264985137514,
+      "grad_norm": 6.6915581684773935,
+      "learning_rate": 7.883161115280264e-06,
+      "loss": 0.4146,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2124036824334074,
+      "grad_norm": 4.483451936143915,
+      "learning_rate": 7.877388443110317e-06,
+      "loss": 0.3964,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2129808663530634,
+      "grad_norm": 4.847478023953505,
+      "learning_rate": 7.871615770940368e-06,
+      "loss": 0.3968,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2135580502727194,
+      "grad_norm": 4.734123259756348,
+      "learning_rate": 7.865843098770421e-06,
+      "loss": 0.411,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2141352341923754,
+      "grad_norm": 67.93482554004888,
+      "learning_rate": 7.860070426600474e-06,
+      "loss": 0.4171,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2147124181120314,
+      "grad_norm": 4.427723281232474,
+      "learning_rate": 7.854297754430527e-06,
+      "loss": 0.3901,
+      "step": 3720
+    },
+    {
+      "epoch": 0.2152896020316874,
+      "grad_norm": 5.970970375519135,
+      "learning_rate": 7.848525082260578e-06,
+      "loss": 0.392,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2158667859513434,
+      "grad_norm": 3.364616367304386,
+      "learning_rate": 7.842752410090631e-06,
+      "loss": 0.4262,
+      "step": 3740
+    },
+    {
+      "epoch": 0.2164439698709994,
+      "grad_norm": 14.709875574791262,
+      "learning_rate": 7.836979737920684e-06,
+      "loss": 0.4189,
+      "step": 3750
+    },
+    {
+      "epoch": 0.2170211537906554,
+      "grad_norm": 34.85728037006575,
+      "learning_rate": 7.831207065750737e-06,
+      "loss": 0.4281,
+      "step": 3760
+    },
+    {
+      "epoch": 0.2175983377103114,
+      "grad_norm": 3.8349995808560835,
+      "learning_rate": 7.82543439358079e-06,
+      "loss": 0.4197,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2181755216299674,
+      "grad_norm": 9.826012532840847,
+      "learning_rate": 7.819661721410842e-06,
+      "loss": 0.411,
+      "step": 3780
+    },
+    {
+      "epoch": 0.2187527055496234,
+      "grad_norm": 5.148779185873613,
+      "learning_rate": 7.813889049240895e-06,
+      "loss": 0.408,
+      "step": 3790
+    },
+    {
+      "epoch": 0.2193298894692794,
+      "grad_norm": 6.034690169401536,
+      "learning_rate": 7.808116377070946e-06,
+      "loss": 0.389,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2199070733889354,
+      "grad_norm": 7.98093273312519,
+      "learning_rate": 7.802343704900999e-06,
+      "loss": 0.4087,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2204842573085914,
+      "grad_norm": 20.795705430196968,
+      "learning_rate": 7.796571032731052e-06,
+      "loss": 0.4164,
+      "step": 3820
+    },
+    {
+      "epoch": 0.2210614412282474,
+      "grad_norm": 4.546394166434191,
+      "learning_rate": 7.790798360561105e-06,
+      "loss": 0.4198,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2216386251479034,
+      "grad_norm": 2.8073314189395457,
+      "learning_rate": 7.785025688391156e-06,
+      "loss": 0.4201,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2222158090675594,
+      "grad_norm": 24.430024193972454,
+      "learning_rate": 7.77925301622121e-06,
+      "loss": 0.3916,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2227929929872154,
+      "grad_norm": 3.7314691801815525,
+      "learning_rate": 7.773480344051262e-06,
+      "loss": 0.4034,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2233701769068714,
+      "grad_norm": 4.463974565740604,
+      "learning_rate": 7.767707671881316e-06,
+      "loss": 0.4177,
+      "step": 3870
+    },
+    {
+      "epoch": 0.2239473608265274,
+      "grad_norm": 3.935819483400635,
+      "learning_rate": 7.761934999711367e-06,
+      "loss": 0.4186,
+      "step": 3880
+    },
+    {
+      "epoch": 0.22452454474618336,
+      "grad_norm": 2.7465902804726343,
+      "learning_rate": 7.75616232754142e-06,
+      "loss": 0.4355,
+      "step": 3890
+    },
+    {
+      "epoch": 0.22510172866583936,
+      "grad_norm": 2.59329896630652,
+      "learning_rate": 7.750389655371471e-06,
+      "loss": 0.4139,
+      "step": 3900
+    },
+    {
+      "epoch": 0.22567891258549536,
+      "grad_norm": 2.8540002686300214,
+      "learning_rate": 7.744616983201524e-06,
+      "loss": 0.4062,
+      "step": 3910
+    },
+    {
+      "epoch": 0.22625609650515136,
+      "grad_norm": 6.48349060337823,
+      "learning_rate": 7.738844311031577e-06,
+      "loss": 0.4002,
+      "step": 3920
+    },
+    {
+      "epoch": 0.22683328042480735,
+      "grad_norm": 4.668616924928805,
+      "learning_rate": 7.73307163886163e-06,
+      "loss": 0.4041,
+      "step": 3930
+    },
+    {
+      "epoch": 0.22741046434446335,
+      "grad_norm": 6.539752215401261,
+      "learning_rate": 7.727298966691682e-06,
+      "loss": 0.3994,
+      "step": 3940
+    },
+    {
+      "epoch": 0.22798764826411935,
+      "grad_norm": 3.3595541678536156,
+      "learning_rate": 7.721526294521735e-06,
+      "loss": 0.3995,
+      "step": 3950
+    },
+    {
+      "epoch": 0.22856483218377535,
+      "grad_norm": 2.8441727594832886,
+      "learning_rate": 7.715753622351788e-06,
+      "loss": 0.4,
+      "step": 3960
+    },
+    {
+      "epoch": 0.22914201610343135,
+      "grad_norm": 2.849353128300574,
+      "learning_rate": 7.70998095018184e-06,
+      "loss": 0.4048,
+      "step": 3970
+    },
+    {
+      "epoch": 0.22971920002308735,
+      "grad_norm": 4.881214366450045,
+      "learning_rate": 7.704208278011892e-06,
+      "loss": 0.4054,
+      "step": 3980
+    },
+    {
+      "epoch": 0.23029638394274335,
+      "grad_norm": 15.46486836243805,
+      "learning_rate": 7.698435605841945e-06,
+      "loss": 0.3923,
+      "step": 3990
+    },
+    {
+      "epoch": 0.23087356786239935,
+      "grad_norm": 3.847585477525563,
+      "learning_rate": 7.692662933671996e-06,
+      "loss": 0.393,
+      "step": 4000
+    },
+    {
+      "epoch": 0.23145075178205535,
+      "grad_norm": 38.3399434241166,
+      "learning_rate": 7.686890261502051e-06,
+      "loss": 0.3926,
+      "step": 4010
+    },
+    {
+      "epoch": 0.23202793570171135,
+      "grad_norm": 2.6580858390650466,
+      "learning_rate": 7.681117589332102e-06,
+      "loss": 0.3948,
+      "step": 4020
+    },
+    {
+      "epoch": 0.23260511962136735,
+      "grad_norm": 4.466615319467433,
+      "learning_rate": 7.675344917162155e-06,
+      "loss": 0.402,
+      "step": 4030
+    },
+    {
+      "epoch": 0.23318230354102334,
+      "grad_norm": 6.7414838396917505,
+      "learning_rate": 7.669572244992207e-06,
+      "loss": 0.3904,
+      "step": 4040
+    },
+    {
+      "epoch": 0.23375948746067934,
+      "grad_norm": 4.03997703346681,
+      "learning_rate": 7.66379957282226e-06,
+      "loss": 0.3938,
+      "step": 4050
+    },
+    {
+      "epoch": 0.23433667138033534,
+      "grad_norm": 4.258742526842458,
+      "learning_rate": 7.658026900652313e-06,
+      "loss": 0.3998,
+      "step": 4060
+    },
+    {
+      "epoch": 0.23491385529999134,
+      "grad_norm": 5.100923170543839,
+      "learning_rate": 7.652254228482366e-06,
+      "loss": 0.3968,
+      "step": 4070
+    },
+    {
+      "epoch": 0.23549103921964734,
+      "grad_norm": 8.332021271422962,
+      "learning_rate": 7.646481556312417e-06,
+      "loss": 0.4017,
+      "step": 4080
+    },
+    {
+      "epoch": 0.23606822313930334,
+      "grad_norm": 5.001487618559788,
+      "learning_rate": 7.64070888414247e-06,
+      "loss": 0.3929,
+      "step": 4090
+    },
+    {
+      "epoch": 0.23664540705895934,
+      "grad_norm": 6.605470741420995,
+      "learning_rate": 7.634936211972523e-06,
+      "loss": 0.3995,
+      "step": 4100
+    },
+    {
+      "epoch": 0.23722259097861534,
+      "grad_norm": 4.352594377363156,
+      "learning_rate": 7.629163539802575e-06,
+      "loss": 0.4008,
+      "step": 4110
+    },
+    {
+      "epoch": 0.23779977489827134,
+      "grad_norm": 8.143604964743357,
+      "learning_rate": 7.6233908676326275e-06,
+      "loss": 0.3987,
+      "step": 4120
+    },
+    {
+      "epoch": 0.23837695881792734,
+      "grad_norm": 3.9869800783007427,
+      "learning_rate": 7.6176181954626805e-06,
+      "loss": 0.3874,
+      "step": 4130
+    },
+    {
+      "epoch": 0.23895414273758334,
+      "grad_norm": 3.4272782347037207,
+      "learning_rate": 7.611845523292733e-06,
+      "loss": 0.3869,
+      "step": 4140
+    },
+    {
+      "epoch": 0.23953132665723934,
+      "grad_norm": 2.964033904850173,
+      "learning_rate": 7.606072851122786e-06,
+      "loss": 0.3962,
+      "step": 4150
+    },
+    {
+      "epoch": 0.24010851057689533,
+      "grad_norm": 55.68370603421082,
+      "learning_rate": 7.600300178952838e-06,
+      "loss": 0.3981,
+      "step": 4160
+    },
+    {
+      "epoch": 0.24068569449655133,
+      "grad_norm": 4.402682573412378,
+      "learning_rate": 7.594527506782891e-06,
+      "loss": 0.4152,
+      "step": 4170
+    },
+    {
+      "epoch": 0.24126287841620733,
+      "grad_norm": 2.8760794787596997,
+      "learning_rate": 7.588754834612942e-06,
+      "loss": 0.3881,
+      "step": 4180
+    },
+    {
+      "epoch": 0.24184006233586333,
+      "grad_norm": 2.0239240283122575,
+      "learning_rate": 7.582982162442995e-06,
+      "loss": 0.4023,
+      "step": 4190
+    },
+    {
+      "epoch": 0.24241724625551933,
+      "grad_norm": 2.5930742840295986,
+      "learning_rate": 7.577209490273047e-06,
+      "loss": 0.4146,
+      "step": 4200
+    },
+    {
+      "epoch": 0.24299443017517533,
+      "grad_norm": 12.25806910583576,
+      "learning_rate": 7.5714368181031e-06,
+      "loss": 0.4268,
+      "step": 4210
+    },
+    {
+      "epoch": 0.24357161409483133,
+      "grad_norm": 9.121606156996025,
+      "learning_rate": 7.565664145933153e-06,
+      "loss": 0.3926,
+      "step": 4220
+    },
+    {
+      "epoch": 0.24414879801448733,
+      "grad_norm": 9.12741140098973,
+      "learning_rate": 7.559891473763206e-06,
+      "loss": 0.3908,
+      "step": 4230
+    },
+    {
+      "epoch": 0.24472598193414333,
+      "grad_norm": 2.988966863298224,
+      "learning_rate": 7.554118801593258e-06,
+      "loss": 0.4077,
+      "step": 4240
+    },
+    {
+      "epoch": 0.24530316585379933,
+      "grad_norm": 4.755614024652895,
+      "learning_rate": 7.548346129423311e-06,
+      "loss": 0.3923,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2458803497734553,
+      "grad_norm": 3.286123151497483,
+      "learning_rate": 7.542573457253363e-06,
+      "loss": 0.3842,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2464575336931113,
+      "grad_norm": 24.75179268960897,
+      "learning_rate": 7.536800785083416e-06,
+      "loss": 0.4348,
+      "step": 4270
+    },
+    {
+      "epoch": 0.2470347176127673,
+      "grad_norm": 2.1639276020923064,
+      "learning_rate": 7.531028112913469e-06,
+      "loss": 0.4064,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2476119015324233,
+      "grad_norm": 1.9508786971501029,
+      "learning_rate": 7.525255440743521e-06,
+      "loss": 0.4085,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2481890854520793,
+      "grad_norm": 1.8580731020746826,
+      "learning_rate": 7.519482768573574e-06,
+      "loss": 0.4111,
+      "step": 4300
+    },
+    {
+      "epoch": 0.2487662693717353,
+      "grad_norm": 2.6928452549457407,
+      "learning_rate": 7.5137100964036255e-06,
+      "loss": 0.4115,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2493434532913913,
+      "grad_norm": 2.054083128359146,
+      "learning_rate": 7.5079374242336786e-06,
+      "loss": 0.4121,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2499206372110473,
+      "grad_norm": 2.6316000209561663,
+      "learning_rate": 7.502164752063731e-06,
+      "loss": 0.4186,
+      "step": 4330
+    },
+    {
+      "epoch": 0.2504978211307033,
+      "grad_norm": 1.8771833797620388,
+      "learning_rate": 7.496392079893784e-06,
+      "loss": 0.4001,
+      "step": 4340
+    },
+    {
+      "epoch": 0.2510750050503593,
+      "grad_norm": 9.221358255158016,
+      "learning_rate": 7.490619407723836e-06,
+      "loss": 0.4177,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2516521889700153,
+      "grad_norm": 1.8539444504213582,
+      "learning_rate": 7.484846735553889e-06,
+      "loss": 0.4055,
+      "step": 4360
+    },
+    {
+      "epoch": 0.2522293728896713,
+      "grad_norm": 2.284022523834839,
+      "learning_rate": 7.479074063383941e-06,
+      "loss": 0.4228,
+      "step": 4370
+    },
+    {
+      "epoch": 0.2528065568093273,
+      "grad_norm": 1.8613795057409426,
+      "learning_rate": 7.473301391213994e-06,
+      "loss": 0.3886,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2533837407289833,
+      "grad_norm": 3.7779942998503855,
+      "learning_rate": 7.467528719044046e-06,
+      "loss": 0.4158,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2539609246486393,
+      "grad_norm": 3.7840065627156365,
+      "learning_rate": 7.461756046874099e-06,
+      "loss": 0.4187,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2545381085682953,
+      "grad_norm": 5.099586666363089,
+      "learning_rate": 7.455983374704151e-06,
+      "loss": 0.4098,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2551152924879513,
+      "grad_norm": 2.6870891278248337,
+      "learning_rate": 7.450210702534204e-06,
+      "loss": 0.3986,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2556924764076073,
+      "grad_norm": 2.7248868412027583,
+      "learning_rate": 7.444438030364256e-06,
+      "loss": 0.4038,
+      "step": 4430
+    },
+    {
+      "epoch": 0.2562696603272633,
+      "grad_norm": 3.714403433710303,
+      "learning_rate": 7.438665358194309e-06,
+      "loss": 0.4045,
+      "step": 4440
+    },
+    {
+      "epoch": 0.2568468442469193,
+      "grad_norm": 2.9240340589059644,
+      "learning_rate": 7.432892686024361e-06,
+      "loss": 0.3911,
+      "step": 4450
+    },
+    {
+      "epoch": 0.2574240281665753,
+      "grad_norm": 4.331854988527969,
+      "learning_rate": 7.427120013854414e-06,
+      "loss": 0.394,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2580012120862313,
+      "grad_norm": 6.84340943547103,
+      "learning_rate": 7.421347341684466e-06,
+      "loss": 0.3884,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2585783960058873,
+      "grad_norm": 26.71610826157837,
+      "learning_rate": 7.415574669514519e-06,
+      "loss": 0.4126,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2591555799255433,
+      "grad_norm": 3.691538028091923,
+      "learning_rate": 7.4098019973445714e-06,
+      "loss": 0.3997,
+      "step": 4490
+    },
+    {
+      "epoch": 0.2597327638451993,
+      "grad_norm": 3.6466304992174527,
+      "learning_rate": 7.4040293251746245e-06,
+      "loss": 0.387,
+      "step": 4500
+    },
+    {
+      "epoch": 0.2603099477648553,
+      "grad_norm": 3.069337367045407,
+      "learning_rate": 7.398256653004677e-06,
+      "loss": 0.3925,
+      "step": 4510
+    },
+    {
+      "epoch": 0.2608871316845113,
+      "grad_norm": 17.7941278304272,
+      "learning_rate": 7.39248398083473e-06,
+      "loss": 0.3904,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2614643156041673,
+      "grad_norm": 4.010302222583594,
+      "learning_rate": 7.386711308664781e-06,
+      "loss": 0.4,
+      "step": 4530
+    },
+    {
+      "epoch": 0.2620414995238233,
+      "grad_norm": 2.891669458141575,
+      "learning_rate": 7.380938636494834e-06,
+      "loss": 0.3916,
+      "step": 4540
+    },
+    {
+      "epoch": 0.2626186834434793,
+      "grad_norm": 3.5781925847929736,
+      "learning_rate": 7.375165964324886e-06,
+      "loss": 0.3901,
+      "step": 4550
+    },
+    {
+      "epoch": 0.26319586736313527,
+      "grad_norm": 2.8217409668695814,
+      "learning_rate": 7.369393292154939e-06,
+      "loss": 0.41,
+      "step": 4560
+    },
+    {
+      "epoch": 0.26377305128279127,
+      "grad_norm": 4.326569507185014,
+      "learning_rate": 7.363620619984991e-06,
+      "loss": 0.4091,
+      "step": 4570
+    },
+    {
+      "epoch": 0.26435023520244727,
+      "grad_norm": 5.515247686980751,
+      "learning_rate": 7.357847947815044e-06,
+      "loss": 0.4203,
+      "step": 4580
+    },
+    {
+      "epoch": 0.26492741912210327,
+      "grad_norm": 6.561612435080219,
+      "learning_rate": 7.3520752756450966e-06,
+      "loss": 0.3951,
+      "step": 4590
+    },
+    {
+      "epoch": 0.26550460304175927,
+      "grad_norm": 2.241772546310698,
+      "learning_rate": 7.34630260347515e-06,
+      "loss": 0.3985,
+      "step": 4600
+    },
+    {
+      "epoch": 0.26608178696141527,
+      "grad_norm": 2.524827575292416,
+      "learning_rate": 7.340529931305202e-06,
+      "loss": 0.3981,
+      "step": 4610
+    },
+    {
+      "epoch": 0.26665897088107127,
+      "grad_norm": 2.4686565154848106,
+      "learning_rate": 7.334757259135255e-06,
+      "loss": 0.3948,
+      "step": 4620
+    },
+    {
+      "epoch": 0.26723615480072727,
+      "grad_norm": 5.055286394352697,
+      "learning_rate": 7.328984586965306e-06,
+      "loss": 0.3966,
+      "step": 4630
+    },
+    {
+      "epoch": 0.26781333872038326,
+      "grad_norm": 2.4105713306719023,
+      "learning_rate": 7.323211914795359e-06,
+      "loss": 0.3896,
+      "step": 4640
+    },
+    {
+      "epoch": 0.26839052264003926,
+      "grad_norm": 8.884358381031186,
+      "learning_rate": 7.317439242625411e-06,
+      "loss": 0.4043,
+      "step": 4650
+    },
+    {
+      "epoch": 0.26896770655969526,
+      "grad_norm": 3.3786577911171465,
+      "learning_rate": 7.311666570455464e-06,
+      "loss": 0.4087,
+      "step": 4660
+    },
+    {
+      "epoch": 0.26954489047935126,
+      "grad_norm": 2.241166004757874,
+      "learning_rate": 7.3058938982855165e-06,
+      "loss": 0.3961,
+      "step": 4670
+    },
+    {
+      "epoch": 0.27012207439900726,
+      "grad_norm": 3.7398767727731195,
+      "learning_rate": 7.3001212261155695e-06,
+      "loss": 0.4025,
+      "step": 4680
+    },
+    {
+      "epoch": 0.27069925831866326,
+      "grad_norm": 2.4627601331961024,
+      "learning_rate": 7.294348553945622e-06,
+      "loss": 0.3967,
+      "step": 4690
+    },
+    {
+      "epoch": 0.27127644223831926,
+      "grad_norm": 2.9400014965222243,
+      "learning_rate": 7.288575881775675e-06,
+      "loss": 0.398,
+      "step": 4700
+    },
+    {
+      "epoch": 0.27185362615797526,
+      "grad_norm": 2.371642161881622,
+      "learning_rate": 7.282803209605727e-06,
+      "loss": 0.3878,
+      "step": 4710
+    },
+    {
+      "epoch": 0.27243081007763126,
+      "grad_norm": 2.1217448647861943,
+      "learning_rate": 7.27703053743578e-06,
+      "loss": 0.3961,
+      "step": 4720
+    },
+    {
+      "epoch": 0.27300799399728726,
+      "grad_norm": 6.480519075871927,
+      "learning_rate": 7.271257865265832e-06,
+      "loss": 0.3904,
+      "step": 4730
+    },
+    {
+      "epoch": 0.27358517791694326,
+      "grad_norm": 3.7782562668503292,
+      "learning_rate": 7.265485193095885e-06,
+      "loss": 0.4015,
+      "step": 4740
+    },
+    {
+      "epoch": 0.27416236183659926,
+      "grad_norm": 27.34719190524476,
+      "learning_rate": 7.259712520925936e-06,
+      "loss": 0.3995,
+      "step": 4750
+    },
+    {
+      "epoch": 0.27473954575625525,
+      "grad_norm": 2.5984426690848044,
+      "learning_rate": 7.2539398487559894e-06,
+      "loss": 0.4091,
+      "step": 4760
+    },
+    {
+      "epoch": 0.27531672967591125,
+      "grad_norm": 2.269931164816007,
+      "learning_rate": 7.248167176586042e-06,
+      "loss": 0.391,
+      "step": 4770
+    },
+    {
+      "epoch": 0.27589391359556725,
+      "grad_norm": 2.0949435472109443,
+      "learning_rate": 7.242394504416095e-06,
+      "loss": 0.3867,
+      "step": 4780
+    },
+    {
+      "epoch": 0.27647109751522325,
+      "grad_norm": 2.1688865736563794,
+      "learning_rate": 7.236621832246147e-06,
+      "loss": 0.3819,
+      "step": 4790
+    },
+    {
+      "epoch": 0.27704828143487925,
+      "grad_norm": 3.6275115123885744,
+      "learning_rate": 7.2308491600762e-06,
+      "loss": 0.403,
+      "step": 4800
+    },
+    {
+      "epoch": 0.27762546535453525,
+      "grad_norm": 2.7044630613298204,
+      "learning_rate": 7.225076487906253e-06,
+      "loss": 0.401,
+      "step": 4810
+    },
+    {
+      "epoch": 0.2782026492741912,
+      "grad_norm": 3.6256795573786853,
+      "learning_rate": 7.219303815736305e-06,
+      "loss": 0.4008,
+      "step": 4820
+    },
+    {
+      "epoch": 0.2787798331938472,
+      "grad_norm": 2.3560879474365595,
+      "learning_rate": 7.213531143566358e-06,
+      "loss": 0.382,
+      "step": 4830
+    },
+    {
+      "epoch": 0.2793570171135032,
+      "grad_norm": 6.363609259389832,
+      "learning_rate": 7.20775847139641e-06,
+      "loss": 0.3889,
+      "step": 4840
+    },
+    {
+      "epoch": 0.2799342010331592,
+      "grad_norm": 2.447343796783594,
+      "learning_rate": 7.201985799226463e-06,
+      "loss": 0.3819,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2805113849528152,
+      "grad_norm": 4.454942195776334,
+      "learning_rate": 7.1962131270565146e-06,
+      "loss": 0.4029,
+      "step": 4860
+    },
+    {
+      "epoch": 0.2810885688724712,
+      "grad_norm": 4.7341175135353675,
+      "learning_rate": 7.1904404548865684e-06,
+      "loss": 0.3962,
+      "step": 4870
+    },
+    {
+      "epoch": 0.2816657527921272,
+      "grad_norm": 1.8389081163765115,
+      "learning_rate": 7.18466778271662e-06,
+      "loss": 0.3962,
+      "step": 4880
+    },
+    {
+      "epoch": 0.2822429367117832,
+      "grad_norm": 3.2296499793817612,
+      "learning_rate": 7.178895110546673e-06,
+      "loss": 0.3942,
+      "step": 4890
+    },
+    {
+      "epoch": 0.2828201206314392,
+      "grad_norm": 3.947947823306894,
+      "learning_rate": 7.173122438376725e-06,
+      "loss": 0.3976,
+      "step": 4900
+    },
+    {
+      "epoch": 0.2833973045510952,
+      "grad_norm": 3.562193655967395,
+      "learning_rate": 7.167349766206778e-06,
+      "loss": 0.3876,
+      "step": 4910
+    },
+    {
+      "epoch": 0.2839744884707512,
+      "grad_norm": 2.0441973593521086,
+      "learning_rate": 7.16157709403683e-06,
+      "loss": 0.404,
+      "step": 4920
+    },
+    {
+      "epoch": 0.2845516723904072,
+      "grad_norm": 2.3917393046670075,
+      "learning_rate": 7.155804421866883e-06,
+      "loss": 0.4169,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2851288563100632,
+      "grad_norm": 2.328855705516118,
+      "learning_rate": 7.150031749696935e-06,
+      "loss": 0.3886,
+      "step": 4940
+    },
+    {
+      "epoch": 0.2857060402297192,
+      "grad_norm": 5.150276479313902,
+      "learning_rate": 7.144259077526988e-06,
+      "loss": 0.401,
+      "step": 4950
+    },
+    {
+      "epoch": 0.2862832241493752,
+      "grad_norm": 2.4042125393360907,
+      "learning_rate": 7.1384864053570405e-06,
+      "loss": 0.4003,
+      "step": 4960
+    },
+    {
+      "epoch": 0.2868604080690312,
+      "grad_norm": 2.8838052574781257,
+      "learning_rate": 7.1327137331870936e-06,
+      "loss": 0.3992,
+      "step": 4970
+    },
+    {
+      "epoch": 0.2874375919886872,
+      "grad_norm": 3.3196900990562646,
+      "learning_rate": 7.126941061017145e-06,
+      "loss": 0.4025,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2880147759083432,
+      "grad_norm": 5.299768426314854,
+      "learning_rate": 7.121168388847198e-06,
+      "loss": 0.3994,
+      "step": 4990
+    },
+    {
+      "epoch": 0.2885919598279992,
+      "grad_norm": 27.899807155688983,
+      "learning_rate": 7.11539571667725e-06,
+      "loss": 0.402,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2891691437476552,
+      "grad_norm": 2.2901385928116484,
+      "learning_rate": 7.109623044507303e-06,
+      "loss": 0.3936,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2897463276673112,
+      "grad_norm": 5.293699045227353,
+      "learning_rate": 7.103850372337355e-06,
+      "loss": 0.3977,
+      "step": 5020
+    },
+    {
+      "epoch": 0.2903235115869672,
+      "grad_norm": 2.2638394900698624,
+      "learning_rate": 7.098077700167408e-06,
+      "loss": 0.3932,
+      "step": 5030
+    },
+    {
+      "epoch": 0.2909006955066232,
+      "grad_norm": 2.2021953017040317,
+      "learning_rate": 7.0923050279974605e-06,
+      "loss": 0.4063,
+      "step": 5040
+    },
+    {
+      "epoch": 0.2914778794262792,
+      "grad_norm": 3.0086020205487363,
+      "learning_rate": 7.0865323558275135e-06,
+      "loss": 0.3948,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2920550633459352,
+      "grad_norm": 2.1426519571708615,
+      "learning_rate": 7.080759683657566e-06,
+      "loss": 0.3967,
+      "step": 5060
+    },
+    {
+      "epoch": 0.29263224726559117,
+      "grad_norm": 2.396350457849943,
+      "learning_rate": 7.074987011487619e-06,
+      "loss": 0.3699,
+      "step": 5070
+    },
+    {
+      "epoch": 0.29320943118524717,
+      "grad_norm": 48.09761207732283,
+      "learning_rate": 7.06921433931767e-06,
+      "loss": 0.3783,
+      "step": 5080
+    },
+    {
+      "epoch": 0.29378661510490317,
+      "grad_norm": 1.9858233601346664,
+      "learning_rate": 7.063441667147724e-06,
+      "loss": 0.3885,
+      "step": 5090
+    },
+    {
+      "epoch": 0.29436379902455917,
+      "grad_norm": 1.8720378976655367,
+      "learning_rate": 7.057668994977775e-06,
+      "loss": 0.408,
+      "step": 5100
+    },
+    {
+      "epoch": 0.29494098294421517,
+      "grad_norm": 1.9939149039664568,
+      "learning_rate": 7.051896322807828e-06,
+      "loss": 0.3954,
+      "step": 5110
+    },
+    {
+      "epoch": 0.29551816686387117,
+      "grad_norm": 1.654582540560213,
+      "learning_rate": 7.04612365063788e-06,
+      "loss": 0.3914,
+      "step": 5120
+    },
+    {
+      "epoch": 0.29609535078352717,
+      "grad_norm": 3.06484715205326,
+      "learning_rate": 7.040350978467933e-06,
+      "loss": 0.3941,
+      "step": 5130
+    },
+    {
+      "epoch": 0.29667253470318317,
+      "grad_norm": 4.663889722159032,
+      "learning_rate": 7.034578306297986e-06,
+      "loss": 0.3864,
+      "step": 5140
+    },
+    {
+      "epoch": 0.29724971862283917,
+      "grad_norm": 7.729000917336516,
+      "learning_rate": 7.028805634128039e-06,
+      "loss": 0.383,
+      "step": 5150
+    },
+    {
+      "epoch": 0.29782690254249516,
+      "grad_norm": 3.2930128906393357,
+      "learning_rate": 7.023032961958091e-06,
+      "loss": 0.4141,
+      "step": 5160
+    },
+    {
+      "epoch": 0.29840408646215116,
+      "grad_norm": 2.410906648107205,
+      "learning_rate": 7.017260289788144e-06,
+      "loss": 0.3954,
+      "step": 5170
+    },
+    {
+      "epoch": 0.29898127038180716,
+      "grad_norm": 2.334082494973446,
+      "learning_rate": 7.011487617618196e-06,
+      "loss": 0.3946,
+      "step": 5180
+    },
+    {
+      "epoch": 0.29955845430146316,
+      "grad_norm": 2.5351433587502568,
+      "learning_rate": 7.005714945448249e-06,
+      "loss": 0.3793,
+      "step": 5190
+    },
+    {
+      "epoch": 0.30013563822111916,
+      "grad_norm": 4.293591510370919,
+      "learning_rate": 6.9999422732783e-06,
+      "loss": 0.3932,
+      "step": 5200
+    },
+    {
+      "epoch": 0.30071282214077516,
+      "grad_norm": 2.7812977150432032,
+      "learning_rate": 6.994169601108353e-06,
+      "loss": 0.4031,
+      "step": 5210
+    },
+    {
+      "epoch": 0.30129000606043116,
+      "grad_norm": 2.6445327211143392,
+      "learning_rate": 6.9883969289384055e-06,
+      "loss": 0.3971,
+      "step": 5220
+    },
+    {
+      "epoch": 0.30186718998008716,
+      "grad_norm": 3.667055707656093,
+      "learning_rate": 6.9826242567684585e-06,
+      "loss": 0.3863,
+      "step": 5230
+    },
+    {
+      "epoch": 0.30244437389974316,
+      "grad_norm": 7.251793531417223,
+      "learning_rate": 6.976851584598511e-06,
+      "loss": 0.3957,
+      "step": 5240
+    },
+    {
+      "epoch": 0.30302155781939916,
+      "grad_norm": 2.3512832022952415,
+      "learning_rate": 6.971078912428564e-06,
+      "loss": 0.3795,
+      "step": 5250
+    },
+    {
+      "epoch": 0.30359874173905516,
+      "grad_norm": 3.8467854833379103,
+      "learning_rate": 6.965306240258616e-06,
+      "loss": 0.4022,
+      "step": 5260
+    },
+    {
+      "epoch": 0.30417592565871115,
+      "grad_norm": 3.452869510119873,
+      "learning_rate": 6.959533568088669e-06,
+      "loss": 0.367,
+      "step": 5270
+    },
+    {
+      "epoch": 0.30475310957836715,
+      "grad_norm": 1.7928454544936765,
+      "learning_rate": 6.953760895918721e-06,
+      "loss": 0.3835,
+      "step": 5280
+    },
+    {
+      "epoch": 0.30533029349802315,
+      "grad_norm": 3.3141188092724057,
+      "learning_rate": 6.947988223748774e-06,
+      "loss": 0.3841,
+      "step": 5290
+    },
+    {
+      "epoch": 0.30590747741767915,
+      "grad_norm": 7.1200801655892905,
+      "learning_rate": 6.942215551578826e-06,
+      "loss": 0.3851,
+      "step": 5300
+    },
+    {
+      "epoch": 0.30648466133733515,
+      "grad_norm": 2.6078657985349563,
+      "learning_rate": 6.936442879408879e-06,
+      "loss": 0.3851,
+      "step": 5310
+    },
+    {
+      "epoch": 0.30706184525699115,
+      "grad_norm": 3.8742908966217873,
+      "learning_rate": 6.930670207238931e-06,
+      "loss": 0.3657,
+      "step": 5320
+    },
+    {
+      "epoch": 0.30763902917664715,
+      "grad_norm": 2.0698861633639885,
+      "learning_rate": 6.924897535068984e-06,
+      "loss": 0.3911,
+      "step": 5330
+    },
+    {
+      "epoch": 0.30821621309630315,
+      "grad_norm": 2.5491051422292412,
+      "learning_rate": 6.919124862899037e-06,
+      "loss": 0.3765,
+      "step": 5340
+    },
+    {
+      "epoch": 0.30879339701595915,
+      "grad_norm": 3.734949482545124,
+      "learning_rate": 6.913352190729089e-06,
+      "loss": 0.3895,
+      "step": 5350
+    },
+    {
+      "epoch": 0.30937058093561515,
+      "grad_norm": 3.6071556886180356,
+      "learning_rate": 6.907579518559142e-06,
+      "loss": 0.3855,
+      "step": 5360
+    },
+    {
+      "epoch": 0.30994776485527115,
+      "grad_norm": 2.048967073465003,
+      "learning_rate": 6.901806846389194e-06,
+      "loss": 0.3955,
+      "step": 5370
+    },
+    {
+      "epoch": 0.31052494877492715,
+      "grad_norm": 3.6739243647918016,
+      "learning_rate": 6.896034174219247e-06,
+      "loss": 0.3886,
+      "step": 5380
+    },
+    {
+      "epoch": 0.31110213269458314,
+      "grad_norm": 6.86781835267949,
+      "learning_rate": 6.890261502049299e-06,
+      "loss": 0.3865,
+      "step": 5390
+    },
+    {
+      "epoch": 0.31167931661423914,
+      "grad_norm": 2.3848433309003445,
+      "learning_rate": 6.884488829879352e-06,
+      "loss": 0.382,
+      "step": 5400
+    },
+    {
+      "epoch": 0.31225650053389514,
+      "grad_norm": 4.973233732358959,
+      "learning_rate": 6.8787161577094044e-06,
+      "loss": 0.3902,
+      "step": 5410
+    },
+    {
+      "epoch": 0.31283368445355114,
+      "grad_norm": 4.047034417439155,
+      "learning_rate": 6.8729434855394575e-06,
+      "loss": 0.3848,
+      "step": 5420
+    },
+    {
+      "epoch": 0.31341086837320714,
+      "grad_norm": 2.0502326573281464,
+      "learning_rate": 6.867170813369509e-06,
+      "loss": 0.3775,
+      "step": 5430
+    },
+    {
+      "epoch": 0.31398805229286314,
+      "grad_norm": 2.3429460144586747,
+      "learning_rate": 6.861398141199563e-06,
+      "loss": 0.3854,
+      "step": 5440
+    },
+    {
+      "epoch": 0.31456523621251914,
+      "grad_norm": 2.366729120218485,
+      "learning_rate": 6.855625469029614e-06,
+      "loss": 0.392,
+      "step": 5450
+    },
+    {
+      "epoch": 0.31514242013217514,
+      "grad_norm": 2.846824703607735,
+      "learning_rate": 6.849852796859667e-06,
+      "loss": 0.3899,
+      "step": 5460
+    },
+    {
+      "epoch": 0.31571960405183114,
+      "grad_norm": 4.059470208038008,
+      "learning_rate": 6.844080124689719e-06,
+      "loss": 0.3824,
+      "step": 5470
+    },
+    {
+      "epoch": 0.31629678797148714,
+      "grad_norm": 3.060450364883157,
+      "learning_rate": 6.838307452519772e-06,
+      "loss": 0.386,
+      "step": 5480
+    },
+    {
+      "epoch": 0.31687397189114314,
+      "grad_norm": 6.0419702046843735,
+      "learning_rate": 6.832534780349824e-06,
+      "loss": 0.3989,
+      "step": 5490
+    },
+    {
+      "epoch": 0.31745115581079913,
+      "grad_norm": 148.73139965795136,
+      "learning_rate": 6.826762108179877e-06,
+      "loss": 0.3972,
+      "step": 5500
+    },
+    {
+      "epoch": 0.31802833973045513,
+      "grad_norm": 4.457525994355414,
+      "learning_rate": 6.8209894360099296e-06,
+      "loss": 0.3917,
+      "step": 5510
+    },
+    {
+      "epoch": 0.31860552365011113,
+      "grad_norm": 2.172441867365423,
+      "learning_rate": 6.815216763839983e-06,
+      "loss": 0.3877,
+      "step": 5520
+    },
+    {
+      "epoch": 0.31918270756976713,
+      "grad_norm": 3.287076044270403,
+      "learning_rate": 6.809444091670035e-06,
+      "loss": 0.3697,
+      "step": 5530
+    },
+    {
+      "epoch": 0.31975989148942313,
+      "grad_norm": 2.0889311701748747,
+      "learning_rate": 6.803671419500088e-06,
+      "loss": 0.3744,
+      "step": 5540
+    },
+    {
+      "epoch": 0.32033707540907913,
+      "grad_norm": 3.2176355512083616,
+      "learning_rate": 6.797898747330139e-06,
+      "loss": 0.3936,
+      "step": 5550
+    },
+    {
+      "epoch": 0.3209142593287351,
+      "grad_norm": 2.16842859391982,
+      "learning_rate": 6.792126075160192e-06,
+      "loss": 0.3945,
+      "step": 5560
+    },
+    {
+      "epoch": 0.3214914432483911,
+      "grad_norm": 3.8670603387604134,
+      "learning_rate": 6.786353402990244e-06,
+      "loss": 0.3883,
+      "step": 5570
+    },
+    {
+      "epoch": 0.3220686271680471,
+      "grad_norm": 16.096137517550215,
+      "learning_rate": 6.780580730820297e-06,
+      "loss": 0.3982,
+      "step": 5580
+    },
+    {
+      "epoch": 0.32264581108770307,
+      "grad_norm": 3.253563357226522,
+      "learning_rate": 6.7748080586503495e-06,
+      "loss": 0.3971,
+      "step": 5590
+    },
+    {
+      "epoch": 0.32322299500735907,
+      "grad_norm": 5.849409027392066,
+      "learning_rate": 6.7690353864804025e-06,
+      "loss": 0.3966,
+      "step": 5600
+    },
+    {
+      "epoch": 0.32380017892701507,
+      "grad_norm": 4.446139033898797,
+      "learning_rate": 6.763262714310455e-06,
+      "loss": 0.3871,
+      "step": 5610
+    },
+    {
+      "epoch": 0.32437736284667107,
+      "grad_norm": 2.4319593332542953,
+      "learning_rate": 6.757490042140508e-06,
+      "loss": 0.381,
+      "step": 5620
+    },
+    {
+      "epoch": 0.32495454676632707,
+      "grad_norm": 4.734141112466146,
+      "learning_rate": 6.75171736997056e-06,
+      "loss": 0.3688,
+      "step": 5630
+    },
+    {
+      "epoch": 0.32553173068598307,
+      "grad_norm": 3.0227214340364386,
+      "learning_rate": 6.745944697800613e-06,
+      "loss": 0.3939,
+      "step": 5640
+    },
+    {
+      "epoch": 0.32610891460563907,
+      "grad_norm": 6.9528804097069274,
+      "learning_rate": 6.740172025630664e-06,
+      "loss": 0.3859,
+      "step": 5650
+    },
+    {
+      "epoch": 0.32668609852529507,
+      "grad_norm": 2.4438240264660527,
+      "learning_rate": 6.734399353460718e-06,
+      "loss": 0.3929,
+      "step": 5660
+    },
+    {
+      "epoch": 0.32726328244495106,
+      "grad_norm": 7.001401722934106,
+      "learning_rate": 6.728626681290769e-06,
+      "loss": 0.3878,
+      "step": 5670
+    },
+    {
+      "epoch": 0.32784046636460706,
+      "grad_norm": 4.631134068889566,
+      "learning_rate": 6.7228540091208224e-06,
+      "loss": 0.3886,
+      "step": 5680
+    },
+    {
+      "epoch": 0.32841765028426306,
+      "grad_norm": 3.0416584434332274,
+      "learning_rate": 6.717081336950875e-06,
+      "loss": 0.3732,
+      "step": 5690
+    },
+    {
+      "epoch": 0.32899483420391906,
+      "grad_norm": 3.189118603447828,
+      "learning_rate": 6.711308664780928e-06,
+      "loss": 0.3788,
+      "step": 5700
+    },
+    {
+      "epoch": 0.32957201812357506,
+      "grad_norm": 4.412271751435158,
+      "learning_rate": 6.70553599261098e-06,
+      "loss": 0.3903,
+      "step": 5710
+    },
+    {
+      "epoch": 0.33014920204323106,
+      "grad_norm": 5.781124605717443,
+      "learning_rate": 6.699763320441033e-06,
+      "loss": 0.3742,
+      "step": 5720
+    },
+    {
+      "epoch": 0.33072638596288706,
+      "grad_norm": 2.38240681303682,
+      "learning_rate": 6.693990648271085e-06,
+      "loss": 0.3846,
+      "step": 5730
+    },
+    {
+      "epoch": 0.33130356988254306,
+      "grad_norm": 5.627940078972001,
+      "learning_rate": 6.688217976101138e-06,
+      "loss": 0.383,
+      "step": 5740
+    },
+    {
+      "epoch": 0.33188075380219906,
+      "grad_norm": 2.5562400913295695,
+      "learning_rate": 6.68244530393119e-06,
+      "loss": 0.388,
+      "step": 5750
+    },
+    {
+      "epoch": 0.33245793772185506,
+      "grad_norm": 2.009018555010131,
+      "learning_rate": 6.676672631761243e-06,
+      "loss": 0.3863,
+      "step": 5760
+    },
+    {
+      "epoch": 0.33303512164151106,
+      "grad_norm": 2.6584190178994223,
+      "learning_rate": 6.6708999595912945e-06,
+      "loss": 0.382,
+      "step": 5770
+    },
+    {
+      "epoch": 0.33361230556116706,
+      "grad_norm": 1.6637756869209672,
+      "learning_rate": 6.6651272874213476e-06,
+      "loss": 0.384,
+      "step": 5780
+    },
+    {
+      "epoch": 0.33418948948082305,
+      "grad_norm": 2.3195781804624174,
+      "learning_rate": 6.6593546152514e-06,
+      "loss": 0.3766,
+      "step": 5790
+    },
+    {
+      "epoch": 0.33476667340047905,
+      "grad_norm": 3.760084084073311,
+      "learning_rate": 6.653581943081453e-06,
+      "loss": 0.4035,
+      "step": 5800
+    },
+    {
+      "epoch": 0.33534385732013505,
+      "grad_norm": 2.1527295119213607,
+      "learning_rate": 6.647809270911505e-06,
+      "loss": 0.3746,
+      "step": 5810
+    },
+    {
+      "epoch": 0.33592104123979105,
+      "grad_norm": 2.50518271064483,
+      "learning_rate": 6.642036598741558e-06,
+      "loss": 0.3788,
+      "step": 5820
+    },
+    {
+      "epoch": 0.33649822515944705,
+      "grad_norm": 3.4692947058918895,
+      "learning_rate": 6.63626392657161e-06,
+      "loss": 0.3893,
+      "step": 5830
+    },
+    {
+      "epoch": 0.33707540907910305,
+      "grad_norm": 2.512775623033029,
+      "learning_rate": 6.630491254401663e-06,
+      "loss": 0.3719,
+      "step": 5840
+    },
+    {
+      "epoch": 0.33765259299875905,
+      "grad_norm": 1.9666671304858914,
+      "learning_rate": 6.624718582231716e-06,
+      "loss": 0.3785,
+      "step": 5850
+    },
+    {
+      "epoch": 0.33822977691841505,
+      "grad_norm": 2.724605715859374,
+      "learning_rate": 6.618945910061768e-06,
+      "loss": 0.4004,
+      "step": 5860
+    },
+    {
+      "epoch": 0.33880696083807105,
+      "grad_norm": 2.7489712656132044,
+      "learning_rate": 6.613173237891821e-06,
+      "loss": 0.3771,
+      "step": 5870
+    },
+    {
+      "epoch": 0.33938414475772705,
+      "grad_norm": 2.599485813176942,
+      "learning_rate": 6.6074005657218735e-06,
+      "loss": 0.4003,
+      "step": 5880
+    },
+    {
+      "epoch": 0.33996132867738305,
+      "grad_norm": 2.746831225729797,
+      "learning_rate": 6.6016278935519266e-06,
+      "loss": 0.3699,
+      "step": 5890
+    },
+    {
+      "epoch": 0.34053851259703904,
+      "grad_norm": 10.192368895005096,
+      "learning_rate": 6.595855221381978e-06,
+      "loss": 0.3825,
+      "step": 5900
+    },
+    {
+      "epoch": 0.34111569651669504,
+      "grad_norm": 2.3362074486231785,
+      "learning_rate": 6.590082549212031e-06,
+      "loss": 0.3816,
+      "step": 5910
+    },
+    {
+      "epoch": 0.34169288043635104,
+      "grad_norm": 3.2505345689597194,
+      "learning_rate": 6.584309877042083e-06,
+      "loss": 0.4147,
+      "step": 5920
+    },
+    {
+      "epoch": 0.34227006435600704,
+      "grad_norm": 2.771813125028045,
+      "learning_rate": 6.578537204872136e-06,
+      "loss": 0.3822,
+      "step": 5930
+    },
+    {
+      "epoch": 0.34284724827566304,
+      "grad_norm": 2.600143172725041,
+      "learning_rate": 6.572764532702188e-06,
+      "loss": 0.3718,
+      "step": 5940
+    },
+    {
+      "epoch": 0.34342443219531904,
+      "grad_norm": 3.904138501463535,
+      "learning_rate": 6.566991860532241e-06,
+      "loss": 0.3886,
+      "step": 5950
+    },
+    {
+      "epoch": 0.34400161611497504,
+      "grad_norm": 5.0319871188012515,
+      "learning_rate": 6.5612191883622935e-06,
+      "loss": 0.3877,
+      "step": 5960
+    },
+    {
+      "epoch": 0.34457880003463104,
+      "grad_norm": 13.625359588311552,
+      "learning_rate": 6.5554465161923465e-06,
+      "loss": 0.3687,
+      "step": 5970
+    },
+    {
+      "epoch": 0.34515598395428704,
+      "grad_norm": 2.920042053925227,
+      "learning_rate": 6.549673844022399e-06,
+      "loss": 0.3859,
+      "step": 5980
+    },
+    {
+      "epoch": 0.34573316787394304,
+      "grad_norm": 3.5277767201369223,
+      "learning_rate": 6.543901171852452e-06,
+      "loss": 0.3789,
+      "step": 5990
+    },
+    {
+      "epoch": 0.34631035179359904,
+      "grad_norm": 2.7899995578571426,
+      "learning_rate": 6.538128499682503e-06,
+      "loss": 0.3668,
+      "step": 6000
+    },
+    {
+      "epoch": 0.34688753571325504,
+      "grad_norm": 5.64453785605591,
+      "learning_rate": 6.532355827512557e-06,
+      "loss": 0.3824,
+      "step": 6010
+    },
+    {
+      "epoch": 0.34746471963291103,
+      "grad_norm": 3.440174338768187,
+      "learning_rate": 6.526583155342608e-06,
+      "loss": 0.3844,
+      "step": 6020
+    },
+    {
+      "epoch": 0.34804190355256703,
+      "grad_norm": 3.9486486049020635,
+      "learning_rate": 6.520810483172661e-06,
+      "loss": 0.3897,
+      "step": 6030
+    },
+    {
+      "epoch": 0.34861908747222303,
+      "grad_norm": 7.253991235141298,
+      "learning_rate": 6.515037811002713e-06,
+      "loss": 0.3677,
+      "step": 6040
+    },
+    {
+      "epoch": 0.34919627139187903,
+      "grad_norm": 7.656685618930045,
+      "learning_rate": 6.509265138832766e-06,
+      "loss": 0.388,
+      "step": 6050
+    },
+    {
+      "epoch": 0.34977345531153503,
+      "grad_norm": 2.213875264653562,
+      "learning_rate": 6.503492466662819e-06,
+      "loss": 0.3851,
+      "step": 6060
+    },
+    {
+      "epoch": 0.35035063923119103,
+      "grad_norm": 5.474678610165808,
+      "learning_rate": 6.497719794492872e-06,
+      "loss": 0.3777,
+      "step": 6070
+    },
+    {
+      "epoch": 0.35092782315084703,
+      "grad_norm": 4.538506198098333,
+      "learning_rate": 6.491947122322924e-06,
+      "loss": 0.3606,
+      "step": 6080
+    },
+    {
+      "epoch": 0.35150500707050303,
+      "grad_norm": 2.7367624612828627,
+      "learning_rate": 6.486174450152977e-06,
+      "loss": 0.3824,
+      "step": 6090
+    },
+    {
+      "epoch": 0.352082190990159,
+      "grad_norm": 3.9600996597048432,
+      "learning_rate": 6.480401777983029e-06,
+      "loss": 0.3865,
+      "step": 6100
+    },
+    {
+      "epoch": 0.352659374909815,
+      "grad_norm": 6.138903729575434,
+      "learning_rate": 6.474629105813082e-06,
+      "loss": 0.3836,
+      "step": 6110
+    },
+    {
+      "epoch": 0.353236558829471,
+      "grad_norm": 3.1110082739843676,
+      "learning_rate": 6.468856433643133e-06,
+      "loss": 0.3941,
+      "step": 6120
+    },
+    {
+      "epoch": 0.353813742749127,
+      "grad_norm": 3.0125957250660997,
+      "learning_rate": 6.463083761473186e-06,
+      "loss": 0.3907,
+      "step": 6130
+    },
+    {
+      "epoch": 0.354390926668783,
+      "grad_norm": 5.777332948819984,
+      "learning_rate": 6.4573110893032385e-06,
+      "loss": 0.3926,
+      "step": 6140
+    },
+    {
+      "epoch": 0.354968110588439,
+      "grad_norm": 10.69646287431515,
+      "learning_rate": 6.4515384171332915e-06,
+      "loss": 0.3892,
+      "step": 6150
+    },
+    {
+      "epoch": 0.355545294508095,
+      "grad_norm": 5.389336302495197,
+      "learning_rate": 6.445765744963344e-06,
+      "loss": 0.3957,
+      "step": 6160
+    },
+    {
+      "epoch": 0.356122478427751,
+      "grad_norm": 12.802235491479053,
+      "learning_rate": 6.439993072793397e-06,
+      "loss": 0.3742,
+      "step": 6170
+    },
+    {
+      "epoch": 0.356699662347407,
+      "grad_norm": 4.196338750242119,
+      "learning_rate": 6.434220400623449e-06,
+      "loss": 0.3878,
+      "step": 6180
+    },
+    {
+      "epoch": 0.357276846267063,
+      "grad_norm": 3.7684375534000276,
+      "learning_rate": 6.428447728453502e-06,
+      "loss": 0.3724,
+      "step": 6190
+    },
+    {
+      "epoch": 0.357854030186719,
+      "grad_norm": 2.4825477710744446,
+      "learning_rate": 6.422675056283554e-06,
+      "loss": 0.3665,
+      "step": 6200
+    },
+    {
+      "epoch": 0.358431214106375,
+      "grad_norm": 2.5273547043428244,
+      "learning_rate": 6.416902384113607e-06,
+      "loss": 0.3682,
+      "step": 6210
+    },
+    {
+      "epoch": 0.359008398026031,
+      "grad_norm": 3.3691141387535453,
+      "learning_rate": 6.4111297119436584e-06,
+      "loss": 0.3884,
+      "step": 6220
+    },
+    {
+      "epoch": 0.359585581945687,
+      "grad_norm": 3.986041227799815,
+      "learning_rate": 6.405357039773712e-06,
+      "loss": 0.3793,
+      "step": 6230
+    },
+    {
+      "epoch": 0.360162765865343,
+      "grad_norm": 4.388692532796717,
+      "learning_rate": 6.399584367603764e-06,
+      "loss": 0.3826,
+      "step": 6240
+    },
+    {
+      "epoch": 0.360739949784999,
+      "grad_norm": 5.61124433419293,
+      "learning_rate": 6.393811695433817e-06,
+      "loss": 0.3725,
+      "step": 6250
+    },
+    {
+      "epoch": 0.361317133704655,
+      "grad_norm": 5.79217310796387,
+      "learning_rate": 6.388039023263869e-06,
+      "loss": 0.3803,
+      "step": 6260
+    },
+    {
+      "epoch": 0.361894317624311,
+      "grad_norm": 3.3092133128777017,
+      "learning_rate": 6.382266351093922e-06,
+      "loss": 0.3727,
+      "step": 6270
+    },
+    {
+      "epoch": 0.362471501543967,
+      "grad_norm": 2.6436967571480308,
+      "learning_rate": 6.376493678923974e-06,
+      "loss": 0.3869,
+      "step": 6280
+    },
+    {
+      "epoch": 0.363048685463623,
+      "grad_norm": 4.870192599092706,
+      "learning_rate": 6.370721006754027e-06,
+      "loss": 0.3711,
+      "step": 6290
+    },
+    {
+      "epoch": 0.36362586938327895,
+      "grad_norm": 6.412850489358521,
+      "learning_rate": 6.364948334584079e-06,
+      "loss": 0.3778,
+      "step": 6300
+    },
+    {
+      "epoch": 0.36420305330293495,
+      "grad_norm": 6.723734658950526,
+      "learning_rate": 6.359175662414132e-06,
+      "loss": 0.375,
+      "step": 6310
+    },
+    {
+      "epoch": 0.36478023722259095,
+      "grad_norm": 2.9855811704461916,
+      "learning_rate": 6.353402990244184e-06,
+      "loss": 0.3968,
+      "step": 6320
+    },
+    {
+      "epoch": 0.36535742114224695,
+      "grad_norm": 4.253318369577758,
+      "learning_rate": 6.3476303180742374e-06,
+      "loss": 0.3835,
+      "step": 6330
+    },
+    {
+      "epoch": 0.36593460506190295,
+      "grad_norm": 3.350507256204714,
+      "learning_rate": 6.341857645904289e-06,
+      "loss": 0.3934,
+      "step": 6340
+    },
+    {
+      "epoch": 0.36651178898155895,
+      "grad_norm": 7.6596745260829024,
+      "learning_rate": 6.336084973734342e-06,
+      "loss": 0.3695,
+      "step": 6350
+    },
+    {
+      "epoch": 0.36708897290121495,
+      "grad_norm": 5.170501025053992,
+      "learning_rate": 6.330312301564394e-06,
+      "loss": 0.3822,
+      "step": 6360
+    },
+    {
+      "epoch": 0.36766615682087095,
+      "grad_norm": 3.3572394366722342,
+      "learning_rate": 6.324539629394447e-06,
+      "loss": 0.3756,
+      "step": 6370
+    },
+    {
+      "epoch": 0.36824334074052695,
+      "grad_norm": 4.222113472184697,
+      "learning_rate": 6.3187669572245e-06,
+      "loss": 0.3934,
+      "step": 6380
+    },
+    {
+      "epoch": 0.36882052466018295,
+      "grad_norm": 4.239041230078147,
+      "learning_rate": 6.312994285054552e-06,
+      "loss": 0.3841,
+      "step": 6390
+    },
+    {
+      "epoch": 0.36939770857983895,
+      "grad_norm": 3.8859530952931425,
+      "learning_rate": 6.307221612884605e-06,
+      "loss": 0.3747,
+      "step": 6400
+    },
+    {
+      "epoch": 0.36997489249949495,
+      "grad_norm": 5.0107950962438315,
+      "learning_rate": 6.301448940714657e-06,
+      "loss": 0.3781,
+      "step": 6410
+    },
+    {
+      "epoch": 0.37055207641915094,
+      "grad_norm": 19.339302178387896,
+      "learning_rate": 6.29567626854471e-06,
+      "loss": 0.3804,
+      "step": 6420
+    },
+    {
+      "epoch": 0.37112926033880694,
+      "grad_norm": 103.6902739739912,
+      "learning_rate": 6.2899035963747626e-06,
+      "loss": 0.3614,
+      "step": 6430
+    },
+    {
+      "epoch": 0.37170644425846294,
+      "grad_norm": 9.795833830078546,
+      "learning_rate": 6.284130924204816e-06,
+      "loss": 0.3806,
+      "step": 6440
+    },
+    {
+      "epoch": 0.37228362817811894,
+      "grad_norm": 2.835184422290757,
+      "learning_rate": 6.278358252034868e-06,
+      "loss": 0.3635,
+      "step": 6450
+    },
+    {
+      "epoch": 0.37286081209777494,
+      "grad_norm": 17.867434742010555,
+      "learning_rate": 6.272585579864921e-06,
+      "loss": 0.3825,
+      "step": 6460
+    },
+    {
+      "epoch": 0.37343799601743094,
+      "grad_norm": 3.5064803159499554,
+      "learning_rate": 6.266812907694972e-06,
+      "loss": 0.3777,
+      "step": 6470
+    },
+    {
+      "epoch": 0.37401517993708694,
+      "grad_norm": 6.18556265562049,
+      "learning_rate": 6.261040235525025e-06,
+      "loss": 0.354,
+      "step": 6480
+    },
+    {
+      "epoch": 0.37459236385674294,
+      "grad_norm": 5.3643329344286,
+      "learning_rate": 6.255267563355077e-06,
+      "loss": 0.3653,
+      "step": 6490
+    },
+    {
+      "epoch": 0.37516954777639894,
+      "grad_norm": 4.370066633666132,
+      "learning_rate": 6.24949489118513e-06,
+      "loss": 0.3799,
+      "step": 6500
+    },
+    {
+      "epoch": 0.37574673169605494,
+      "grad_norm": 3.802802160469247,
+      "learning_rate": 6.2437222190151825e-06,
+      "loss": 0.3771,
+      "step": 6510
+    },
+    {
+      "epoch": 0.37632391561571094,
+      "grad_norm": 2.999312565631662,
+      "learning_rate": 6.2379495468452355e-06,
+      "loss": 0.3761,
+      "step": 6520
+    },
+    {
+      "epoch": 0.37690109953536693,
+      "grad_norm": 7.852310497644898,
+      "learning_rate": 6.232176874675288e-06,
+      "loss": 0.3823,
+      "step": 6530
+    },
+    {
+      "epoch": 0.37747828345502293,
+      "grad_norm": 4.876630434197547,
+      "learning_rate": 6.226404202505341e-06,
+      "loss": 0.3704,
+      "step": 6540
+    },
+    {
+      "epoch": 0.37805546737467893,
+      "grad_norm": 4.989568751322678,
+      "learning_rate": 6.220631530335393e-06,
+      "loss": 0.3836,
+      "step": 6550
+    },
+    {
+      "epoch": 0.37863265129433493,
+      "grad_norm": 3.7548796873491135,
+      "learning_rate": 6.214858858165446e-06,
+      "loss": 0.3759,
+      "step": 6560
+    },
+    {
+      "epoch": 0.37920983521399093,
+      "grad_norm": 2.7376006597130513,
+      "learning_rate": 6.209086185995497e-06,
+      "loss": 0.3825,
+      "step": 6570
+    },
+    {
+      "epoch": 0.37978701913364693,
+      "grad_norm": 3.835955921955649,
+      "learning_rate": 6.203313513825551e-06,
+      "loss": 0.3892,
+      "step": 6580
+    },
+    {
+      "epoch": 0.38036420305330293,
+      "grad_norm": 2.728138816573778,
+      "learning_rate": 6.197540841655602e-06,
+      "loss": 0.3806,
+      "step": 6590
+    },
+    {
+      "epoch": 0.38094138697295893,
+      "grad_norm": 6.331854012865428,
+      "learning_rate": 6.1917681694856554e-06,
+      "loss": 0.3849,
+      "step": 6600
+    },
+    {
+      "epoch": 0.3815185708926149,
+      "grad_norm": 2.7712160688455394,
+      "learning_rate": 6.185995497315708e-06,
+      "loss": 0.366,
+      "step": 6610
+    },
+    {
+      "epoch": 0.3820957548122709,
+      "grad_norm": 2.2697691277132406,
+      "learning_rate": 6.180222825145761e-06,
+      "loss": 0.3625,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3826729387319269,
+      "grad_norm": 4.51763189851551,
+      "learning_rate": 6.174450152975813e-06,
+      "loss": 0.3745,
+      "step": 6630
+    },
+    {
+      "epoch": 0.3832501226515829,
+      "grad_norm": 22.00920716007038,
+      "learning_rate": 6.168677480805866e-06,
+      "loss": 0.379,
+      "step": 6640
+    },
+    {
+      "epoch": 0.3838273065712389,
+      "grad_norm": 8.50487988964264,
+      "learning_rate": 6.162904808635918e-06,
+      "loss": 0.3798,
+      "step": 6650
+    },
+    {
+      "epoch": 0.3844044904908949,
+      "grad_norm": 2.756542308650349,
+      "learning_rate": 6.157132136465971e-06,
+      "loss": 0.3777,
+      "step": 6660
+    },
+    {
+      "epoch": 0.3849816744105509,
+      "grad_norm": 1.8967629666152492,
+      "learning_rate": 6.151359464296023e-06,
+      "loss": 0.3536,
+      "step": 6670
+    },
+    {
+      "epoch": 0.3855588583302069,
+      "grad_norm": 2.208647530669507,
+      "learning_rate": 6.145586792126076e-06,
+      "loss": 0.3757,
+      "step": 6680
+    },
+    {
+      "epoch": 0.3861360422498629,
+      "grad_norm": 3.18500818944882,
+      "learning_rate": 6.1398141199561275e-06,
+      "loss": 0.381,
+      "step": 6690
+    },
+    {
+      "epoch": 0.3867132261695189,
+      "grad_norm": 3.7319272107204267,
+      "learning_rate": 6.1340414477861806e-06,
+      "loss": 0.3895,
+      "step": 6700
+    },
+    {
+      "epoch": 0.3872904100891749,
+      "grad_norm": 6.233879379077169,
+      "learning_rate": 6.128268775616233e-06,
+      "loss": 0.3931,
+      "step": 6710
+    },
+    {
+      "epoch": 0.3878675940088309,
+      "grad_norm": 2.5172058960090147,
+      "learning_rate": 6.122496103446286e-06,
+      "loss": 0.3696,
+      "step": 6720
+    },
+    {
+      "epoch": 0.3884447779284869,
+      "grad_norm": 2.4821687545852544,
+      "learning_rate": 6.116723431276338e-06,
+      "loss": 0.3783,
+      "step": 6730
+    },
+    {
+      "epoch": 0.3890219618481429,
+      "grad_norm": 2.5811379708324984,
+      "learning_rate": 6.110950759106391e-06,
+      "loss": 0.3883,
+      "step": 6740
+    },
+    {
+      "epoch": 0.3895991457677989,
+      "grad_norm": 4.606721510393016,
+      "learning_rate": 6.105178086936443e-06,
+      "loss": 0.364,
+      "step": 6750
+    },
+    {
+      "epoch": 0.3901763296874549,
+      "grad_norm": 5.353229433626119,
+      "learning_rate": 6.099405414766496e-06,
+      "loss": 0.3882,
+      "step": 6760
+    },
+    {
+      "epoch": 0.3907535136071109,
+      "grad_norm": 2.3516345262109617,
+      "learning_rate": 6.093632742596548e-06,
+      "loss": 0.3788,
+      "step": 6770
+    },
+    {
+      "epoch": 0.3913306975267669,
+      "grad_norm": 11.487680253286674,
+      "learning_rate": 6.087860070426601e-06,
+      "loss": 0.3889,
+      "step": 6780
+    },
+    {
+      "epoch": 0.3919078814464229,
+      "grad_norm": 3.18290202413646,
+      "learning_rate": 6.082087398256653e-06,
+      "loss": 0.3607,
+      "step": 6790
+    },
+    {
+      "epoch": 0.3924850653660789,
+      "grad_norm": 2.7380986355865917,
+      "learning_rate": 6.0763147260867065e-06,
+      "loss": 0.3809,
+      "step": 6800
+    },
+    {
+      "epoch": 0.3930622492857349,
+      "grad_norm": 2.985403565819371,
+      "learning_rate": 6.070542053916758e-06,
+      "loss": 0.3785,
+      "step": 6810
+    },
+    {
+      "epoch": 0.3936394332053909,
+      "grad_norm": 7.257850197480963,
+      "learning_rate": 6.064769381746811e-06,
+      "loss": 0.3813,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3942166171250469,
+      "grad_norm": 2.651981631840983,
+      "learning_rate": 6.058996709576863e-06,
+      "loss": 0.3911,
+      "step": 6830
+    },
+    {
+      "epoch": 0.3947938010447029,
+      "grad_norm": 3.007540853480136,
+      "learning_rate": 6.053224037406916e-06,
+      "loss": 0.3787,
+      "step": 6840
+    },
+    {
+      "epoch": 0.3953709849643589,
+      "grad_norm": 4.967113215124695,
+      "learning_rate": 6.047451365236968e-06,
+      "loss": 0.3729,
+      "step": 6850
+    },
+    {
+      "epoch": 0.3959481688840149,
+      "grad_norm": 2.4113519734571627,
+      "learning_rate": 6.041678693067021e-06,
+      "loss": 0.3576,
+      "step": 6860
+    },
+    {
+      "epoch": 0.3965253528036709,
+      "grad_norm": 1.5215990778439656,
+      "learning_rate": 6.0359060208970734e-06,
+      "loss": 0.3813,
+      "step": 6870
+    },
+    {
+      "epoch": 0.3971025367233269,
+      "grad_norm": 1.9980571139407164,
+      "learning_rate": 6.0301333487271265e-06,
+      "loss": 0.3764,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3976797206429829,
+      "grad_norm": 3.851850368869639,
+      "learning_rate": 6.024360676557179e-06,
+      "loss": 0.3793,
+      "step": 6890
+    },
+    {
+      "epoch": 0.3982569045626389,
+      "grad_norm": 2.819413612633571,
+      "learning_rate": 6.018588004387232e-06,
+      "loss": 0.3915,
+      "step": 6900
+    },
+    {
+      "epoch": 0.3988340884822949,
+      "grad_norm": 2.2801532893497733,
+      "learning_rate": 6.012815332217285e-06,
+      "loss": 0.3927,
+      "step": 6910
+    },
+    {
+      "epoch": 0.3994112724019509,
+      "grad_norm": 3.0396536780138734,
+      "learning_rate": 6.007042660047336e-06,
+      "loss": 0.3787,
+      "step": 6920
+    },
+    {
+      "epoch": 0.3999884563216069,
+      "grad_norm": 1.9908365824878806,
+      "learning_rate": 6.00126998787739e-06,
+      "loss": 0.3751,
+      "step": 6930
+    },
+    {
+      "epoch": 0.4005656402412629,
+      "grad_norm": 2.5287281468598817,
+      "learning_rate": 5.995497315707441e-06,
+      "loss": 0.3739,
+      "step": 6940
+    },
+    {
+      "epoch": 0.4011428241609189,
+      "grad_norm": 2.3877649628077404,
+      "learning_rate": 5.989724643537494e-06,
+      "loss": 0.3931,
+      "step": 6950
+    },
+    {
+      "epoch": 0.4017200080805749,
+      "grad_norm": 1.751274625854989,
+      "learning_rate": 5.983951971367546e-06,
+      "loss": 0.3725,
+      "step": 6960
+    },
+    {
+      "epoch": 0.4022971920002309,
+      "grad_norm": 7.097265852789204,
+      "learning_rate": 5.978179299197599e-06,
+      "loss": 0.3771,
+      "step": 6970
+    },
+    {
+      "epoch": 0.4028743759198869,
+      "grad_norm": 1.5751898011207688,
+      "learning_rate": 5.972406627027652e-06,
+      "loss": 0.3804,
+      "step": 6980
+    },
+    {
+      "epoch": 0.4034515598395429,
+      "grad_norm": 3.487612778253862,
+      "learning_rate": 5.966633954857705e-06,
+      "loss": 0.3677,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4040287437591989,
+      "grad_norm": 2.622057972311098,
+      "learning_rate": 5.960861282687757e-06,
+      "loss": 0.3778,
+      "step": 7000
+    },
+    {
+      "epoch": 0.4046059276788549,
+      "grad_norm": 2.7368469799858532,
+      "learning_rate": 5.95508861051781e-06,
+      "loss": 0.3768,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4051831115985109,
+      "grad_norm": 1.6133398127083427,
+      "learning_rate": 5.949315938347862e-06,
+      "loss": 0.3799,
+      "step": 7020
+    },
+    {
+      "epoch": 0.4057602955181669,
+      "grad_norm": 3.191334805976918,
+      "learning_rate": 5.943543266177915e-06,
+      "loss": 0.3813,
+      "step": 7030
+    },
+    {
+      "epoch": 0.40633747943782283,
+      "grad_norm": 2.8991810624406784,
+      "learning_rate": 5.937770594007966e-06,
+      "loss": 0.376,
+      "step": 7040
+    },
+    {
+      "epoch": 0.40691466335747883,
+      "grad_norm": 2.0785390805202684,
+      "learning_rate": 5.931997921838019e-06,
+      "loss": 0.3729,
+      "step": 7050
+    },
+    {
+      "epoch": 0.40749184727713483,
+      "grad_norm": 1.9512094562324862,
+      "learning_rate": 5.9262252496680715e-06,
+      "loss": 0.3732,
+      "step": 7060
+    },
+    {
+      "epoch": 0.40806903119679083,
+      "grad_norm": 3.3176725840902206,
+      "learning_rate": 5.9204525774981245e-06,
+      "loss": 0.3874,
+      "step": 7070
+    },
+    {
+      "epoch": 0.40864621511644683,
+      "grad_norm": 3.000837724994079,
+      "learning_rate": 5.914679905328177e-06,
+      "loss": 0.3745,
+      "step": 7080
+    },
+    {
+      "epoch": 0.40922339903610283,
+      "grad_norm": 1.8158962267665133,
+      "learning_rate": 5.90890723315823e-06,
+      "loss": 0.3756,
+      "step": 7090
+    },
+    {
+      "epoch": 0.40980058295575883,
+      "grad_norm": 2.324389501252935,
+      "learning_rate": 5.903134560988282e-06,
+      "loss": 0.3886,
+      "step": 7100
+    },
+    {
+      "epoch": 0.41037776687541483,
+      "grad_norm": 2.894571332845524,
+      "learning_rate": 5.897361888818335e-06,
+      "loss": 0.3869,
+      "step": 7110
+    },
+    {
+      "epoch": 0.41095495079507083,
+      "grad_norm": 2.629677485680801,
+      "learning_rate": 5.891589216648387e-06,
+      "loss": 0.3615,
+      "step": 7120
+    },
+    {
+      "epoch": 0.4115321347147268,
+      "grad_norm": 13.759434005163566,
+      "learning_rate": 5.88581654447844e-06,
+      "loss": 0.3678,
+      "step": 7130
+    },
+    {
+      "epoch": 0.4121093186343828,
+      "grad_norm": 3.187249272214218,
+      "learning_rate": 5.8800438723084915e-06,
+      "loss": 0.3496,
+      "step": 7140
+    },
+    {
+      "epoch": 0.4126865025540388,
+      "grad_norm": 25.01828326406148,
+      "learning_rate": 5.874271200138545e-06,
+      "loss": 0.3831,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4132636864736948,
+      "grad_norm": 4.28899556920541,
+      "learning_rate": 5.868498527968597e-06,
+      "loss": 0.3847,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4138408703933508,
+      "grad_norm": 5.017592395582479,
+      "learning_rate": 5.86272585579865e-06,
+      "loss": 0.3528,
+      "step": 7170
+    },
+    {
+      "epoch": 0.4144180543130068,
+      "grad_norm": 3.6138133944499686,
+      "learning_rate": 5.856953183628702e-06,
+      "loss": 0.3615,
+      "step": 7180
+    },
+    {
+      "epoch": 0.4149952382326628,
+      "grad_norm": 11.345281193048963,
+      "learning_rate": 5.851180511458755e-06,
+      "loss": 0.36,
+      "step": 7190
+    },
+    {
+      "epoch": 0.4155724221523188,
+      "grad_norm": 4.1575514029124525,
+      "learning_rate": 5.845407839288807e-06,
+      "loss": 0.3707,
+      "step": 7200
+    },
+    {
+      "epoch": 0.4161496060719748,
+      "grad_norm": 5.184879687211155,
+      "learning_rate": 5.83963516711886e-06,
+      "loss": 0.3584,
+      "step": 7210
+    },
+    {
+      "epoch": 0.4167267899916308,
+      "grad_norm": 3.6353294525038256,
+      "learning_rate": 5.833862494948912e-06,
+      "loss": 0.3922,
+      "step": 7220
+    },
+    {
+      "epoch": 0.4173039739112868,
+      "grad_norm": 10.083912587939164,
+      "learning_rate": 5.828089822778965e-06,
+      "loss": 0.358,
+      "step": 7230
+    },
+    {
+      "epoch": 0.4178811578309428,
+      "grad_norm": 3.795430776940293,
+      "learning_rate": 5.822317150609017e-06,
+      "loss": 0.3584,
+      "step": 7240
+    },
+    {
+      "epoch": 0.4184583417505988,
+      "grad_norm": 2.735432424886805,
+      "learning_rate": 5.8165444784390704e-06,
+      "loss": 0.3628,
+      "step": 7250
+    },
+    {
+      "epoch": 0.4190355256702548,
+      "grad_norm": 3.7538394849350034,
+      "learning_rate": 5.810771806269122e-06,
+      "loss": 0.3808,
+      "step": 7260
+    },
+    {
+      "epoch": 0.4196127095899108,
+      "grad_norm": 3.486146744000872,
+      "learning_rate": 5.804999134099175e-06,
+      "loss": 0.3677,
+      "step": 7270
+    },
+    {
+      "epoch": 0.4201898935095668,
+      "grad_norm": 6.482596192596544,
+      "learning_rate": 5.799226461929227e-06,
+      "loss": 0.3715,
+      "step": 7280
+    },
+    {
+      "epoch": 0.4207670774292228,
+      "grad_norm": 15.287481532081374,
+      "learning_rate": 5.79345378975928e-06,
+      "loss": 0.3607,
+      "step": 7290
+    },
+    {
+      "epoch": 0.4213442613488788,
+      "grad_norm": 5.756011268210783,
+      "learning_rate": 5.787681117589332e-06,
+      "loss": 0.3829,
+      "step": 7300
+    },
+    {
+      "epoch": 0.4219214452685348,
+      "grad_norm": 5.238271188240731,
+      "learning_rate": 5.781908445419385e-06,
+      "loss": 0.3685,
+      "step": 7310
+    },
+    {
+      "epoch": 0.4224986291881908,
+      "grad_norm": 5.072523904302979,
+      "learning_rate": 5.776135773249437e-06,
+      "loss": 0.3785,
+      "step": 7320
+    },
+    {
+      "epoch": 0.4230758131078468,
+      "grad_norm": 2.7230926250144494,
+      "learning_rate": 5.77036310107949e-06,
+      "loss": 0.3586,
+      "step": 7330
+    },
+    {
+      "epoch": 0.4236529970275028,
+      "grad_norm": 3.1651643016202438,
+      "learning_rate": 5.7645904289095425e-06,
+      "loss": 0.3675,
+      "step": 7340
+    },
+    {
+      "epoch": 0.4242301809471588,
+      "grad_norm": 5.575569273909336,
+      "learning_rate": 5.7588177567395956e-06,
+      "loss": 0.3659,
+      "step": 7350
+    },
+    {
+      "epoch": 0.4248073648668148,
+      "grad_norm": 3.4372405530276686,
+      "learning_rate": 5.753045084569647e-06,
+      "loss": 0.3562,
+      "step": 7360
+    },
+    {
+      "epoch": 0.4253845487864708,
+      "grad_norm": 3.1380962366302203,
+      "learning_rate": 5.747272412399701e-06,
+      "loss": 0.3665,
+      "step": 7370
+    },
+    {
+      "epoch": 0.4259617327061268,
+      "grad_norm": 4.195020514299469,
+      "learning_rate": 5.741499740229752e-06,
+      "loss": 0.3834,
+      "step": 7380
+    },
+    {
+      "epoch": 0.4265389166257828,
+      "grad_norm": 2.5201571788814103,
+      "learning_rate": 5.735727068059805e-06,
+      "loss": 0.3598,
+      "step": 7390
+    },
+    {
+      "epoch": 0.4271161005454388,
+      "grad_norm": 7.757366621212017,
+      "learning_rate": 5.729954395889857e-06,
+      "loss": 0.365,
+      "step": 7400
+    },
+    {
+      "epoch": 0.4276932844650948,
+      "grad_norm": 3.6863947123217438,
+      "learning_rate": 5.72418172371991e-06,
+      "loss": 0.3605,
+      "step": 7410
+    },
+    {
+      "epoch": 0.4282704683847508,
+      "grad_norm": 2.713386138832286,
+      "learning_rate": 5.7184090515499625e-06,
+      "loss": 0.3692,
+      "step": 7420
+    },
+    {
+      "epoch": 0.4288476523044068,
+      "grad_norm": 4.061486235134526,
+      "learning_rate": 5.7126363793800155e-06,
+      "loss": 0.3615,
+      "step": 7430
+    },
+    {
+      "epoch": 0.4294248362240628,
+      "grad_norm": 3.171386095616653,
+      "learning_rate": 5.7068637072100685e-06,
+      "loss": 0.3742,
+      "step": 7440
+    },
+    {
+      "epoch": 0.4300020201437188,
+      "grad_norm": 3.0632786743675173,
+      "learning_rate": 5.701091035040121e-06,
+      "loss": 0.3669,
+      "step": 7450
+    },
+    {
+      "epoch": 0.4305792040633748,
+      "grad_norm": 5.9823682538619884,
+      "learning_rate": 5.695318362870174e-06,
+      "loss": 0.3625,
+      "step": 7460
+    },
+    {
+      "epoch": 0.4311563879830308,
+      "grad_norm": 2.587936253733615,
+      "learning_rate": 5.689545690700226e-06,
+      "loss": 0.3654,
+      "step": 7470
+    },
+    {
+      "epoch": 0.4317335719026868,
+      "grad_norm": 3.3311960193145507,
+      "learning_rate": 5.683773018530279e-06,
+      "loss": 0.3849,
+      "step": 7480
+    },
+    {
+      "epoch": 0.4323107558223428,
+      "grad_norm": 7.2504266943512885,
+      "learning_rate": 5.67800034636033e-06,
+      "loss": 0.383,
+      "step": 7490
+    },
+    {
+      "epoch": 0.4328879397419988,
+      "grad_norm": 2.565927845188403,
+      "learning_rate": 5.672227674190383e-06,
+      "loss": 0.3689,
+      "step": 7500
+    },
+    {
+      "epoch": 0.4334651236616548,
+      "grad_norm": 2.596436540237007,
+      "learning_rate": 5.6664550020204354e-06,
+      "loss": 0.3783,
+      "step": 7510
+    },
+    {
+      "epoch": 0.4340423075813108,
+      "grad_norm": 3.2568921956880397,
+      "learning_rate": 5.6606823298504884e-06,
+      "loss": 0.3622,
+      "step": 7520
+    },
+    {
+      "epoch": 0.4346194915009668,
+      "grad_norm": 3.534180092969136,
+      "learning_rate": 5.654909657680541e-06,
+      "loss": 0.374,
+      "step": 7530
+    },
+    {
+      "epoch": 0.4351966754206228,
+      "grad_norm": 2.320956209280894,
+      "learning_rate": 5.649136985510594e-06,
+      "loss": 0.3618,
+      "step": 7540
+    },
+    {
+      "epoch": 0.4357738593402788,
+      "grad_norm": 2.7986565358175732,
+      "learning_rate": 5.643364313340646e-06,
+      "loss": 0.376,
+      "step": 7550
+    },
+    {
+      "epoch": 0.4363510432599348,
+      "grad_norm": 4.3640055595975955,
+      "learning_rate": 5.637591641170699e-06,
+      "loss": 0.359,
+      "step": 7560
+    },
+    {
+      "epoch": 0.4369282271795908,
+      "grad_norm": 2.4295878318318893,
+      "learning_rate": 5.631818969000751e-06,
+      "loss": 0.3837,
+      "step": 7570
+    },
+    {
+      "epoch": 0.4375054110992468,
+      "grad_norm": 2.5358015892610304,
+      "learning_rate": 5.626046296830804e-06,
+      "loss": 0.3824,
+      "step": 7580
+    },
+    {
+      "epoch": 0.4380825950189028,
+      "grad_norm": 2.732560193932699,
+      "learning_rate": 5.620273624660856e-06,
+      "loss": 0.3657,
+      "step": 7590
+    },
+    {
+      "epoch": 0.4386597789385588,
+      "grad_norm": 4.150107259821488,
+      "learning_rate": 5.614500952490909e-06,
+      "loss": 0.3805,
+      "step": 7600
+    },
+    {
+      "epoch": 0.4392369628582148,
+      "grad_norm": 6.027002919837396,
+      "learning_rate": 5.6087282803209605e-06,
+      "loss": 0.3733,
+      "step": 7610
+    },
+    {
+      "epoch": 0.4398141467778708,
+      "grad_norm": 4.383047686001244,
+      "learning_rate": 5.6029556081510136e-06,
+      "loss": 0.3798,
+      "step": 7620
+    },
+    {
+      "epoch": 0.4403913306975268,
+      "grad_norm": 3.183548428631444,
+      "learning_rate": 5.597182935981066e-06,
+      "loss": 0.3704,
+      "step": 7630
+    },
+    {
+      "epoch": 0.4409685146171828,
+      "grad_norm": 3.2995502847867364,
+      "learning_rate": 5.591410263811119e-06,
+      "loss": 0.3868,
+      "step": 7640
+    },
+    {
+      "epoch": 0.4415456985368388,
+      "grad_norm": 2.9302522543070384,
+      "learning_rate": 5.585637591641171e-06,
+      "loss": 0.3719,
+      "step": 7650
+    },
+    {
+      "epoch": 0.4421228824564948,
+      "grad_norm": 3.699827330713927,
+      "learning_rate": 5.579864919471224e-06,
+      "loss": 0.3843,
+      "step": 7660
+    },
+    {
+      "epoch": 0.4427000663761508,
+      "grad_norm": 10.131740140492866,
+      "learning_rate": 5.574092247301276e-06,
+      "loss": 0.3741,
+      "step": 7670
+    },
+    {
+      "epoch": 0.4432772502958068,
+      "grad_norm": 2.53184828015941,
+      "learning_rate": 5.568319575131329e-06,
+      "loss": 0.3679,
+      "step": 7680
+    },
+    {
+      "epoch": 0.4438544342154628,
+      "grad_norm": 2.735336367396379,
+      "learning_rate": 5.562546902961381e-06,
+      "loss": 0.3794,
+      "step": 7690
+    },
+    {
+      "epoch": 0.4444316181351188,
+      "grad_norm": 3.118950089045635,
+      "learning_rate": 5.556774230791434e-06,
+      "loss": 0.3749,
+      "step": 7700
+    },
+    {
+      "epoch": 0.4450088020547748,
+      "grad_norm": 2.0345152708541736,
+      "learning_rate": 5.551001558621486e-06,
+      "loss": 0.3797,
+      "step": 7710
+    },
+    {
+      "epoch": 0.4455859859744308,
+      "grad_norm": 2.216729946357023,
+      "learning_rate": 5.5452288864515395e-06,
+      "loss": 0.3704,
+      "step": 7720
+    },
+    {
+      "epoch": 0.4461631698940868,
+      "grad_norm": 11.605317744790039,
+      "learning_rate": 5.539456214281591e-06,
+      "loss": 0.3826,
+      "step": 7730
+    },
+    {
+      "epoch": 0.4467403538137428,
+      "grad_norm": 5.555060033849291,
+      "learning_rate": 5.533683542111644e-06,
+      "loss": 0.3768,
+      "step": 7740
+    },
+    {
+      "epoch": 0.4473175377333988,
+      "grad_norm": 2.5005756710793507,
+      "learning_rate": 5.527910869941696e-06,
+      "loss": 0.3667,
+      "step": 7750
+    },
+    {
+      "epoch": 0.4478947216530548,
+      "grad_norm": 4.5009505264076655,
+      "learning_rate": 5.522138197771749e-06,
+      "loss": 0.3773,
+      "step": 7760
+    },
+    {
+      "epoch": 0.44847190557271077,
+      "grad_norm": 2.6271424623008945,
+      "learning_rate": 5.516365525601801e-06,
+      "loss": 0.3737,
+      "step": 7770
+    },
+    {
+      "epoch": 0.4490490894923667,
+      "grad_norm": 12.12839502479119,
+      "learning_rate": 5.510592853431854e-06,
+      "loss": 0.3573,
+      "step": 7780
+    },
+    {
+      "epoch": 0.4496262734120227,
+      "grad_norm": 4.194839597547066,
+      "learning_rate": 5.5048201812619064e-06,
+      "loss": 0.3774,
+      "step": 7790
+    },
+    {
+      "epoch": 0.4502034573316787,
+      "grad_norm": 2.1887367164733016,
+      "learning_rate": 5.4990475090919595e-06,
+      "loss": 0.381,
+      "step": 7800
+    },
+    {
+      "epoch": 0.4507806412513347,
+      "grad_norm": 3.1886528624855925,
+      "learning_rate": 5.493274836922012e-06,
+      "loss": 0.37,
+      "step": 7810
+    },
+    {
+      "epoch": 0.4513578251709907,
+      "grad_norm": 2.4110545743480527,
+      "learning_rate": 5.487502164752065e-06,
+      "loss": 0.3604,
+      "step": 7820
+    },
+    {
+      "epoch": 0.4519350090906467,
+      "grad_norm": 2.9847050808092166,
+      "learning_rate": 5.481729492582116e-06,
+      "loss": 0.3675,
+      "step": 7830
+    },
+    {
+      "epoch": 0.4525121930103027,
+      "grad_norm": 5.885118240316819,
+      "learning_rate": 5.475956820412169e-06,
+      "loss": 0.3856,
+      "step": 7840
+    },
+    {
+      "epoch": 0.4530893769299587,
+      "grad_norm": 5.303575867358966,
+      "learning_rate": 5.470184148242221e-06,
+      "loss": 0.3607,
+      "step": 7850
+    },
+    {
+      "epoch": 0.4536665608496147,
+      "grad_norm": 7.1549036006295035,
+      "learning_rate": 5.464411476072274e-06,
+      "loss": 0.3699,
+      "step": 7860
+    },
+    {
+      "epoch": 0.4542437447692707,
+      "grad_norm": 1.8926865310221554,
+      "learning_rate": 5.458638803902326e-06,
+      "loss": 0.3676,
+      "step": 7870
+    },
+    {
+      "epoch": 0.4548209286889267,
+      "grad_norm": 4.130403133399794,
+      "learning_rate": 5.452866131732379e-06,
+      "loss": 0.3624,
+      "step": 7880
+    },
+    {
+      "epoch": 0.4553981126085827,
+      "grad_norm": 3.4241484954051677,
+      "learning_rate": 5.4470934595624316e-06,
+      "loss": 0.3652,
+      "step": 7890
+    },
+    {
+      "epoch": 0.4559752965282387,
+      "grad_norm": 7.8305369558715725,
+      "learning_rate": 5.441320787392485e-06,
+      "loss": 0.35,
+      "step": 7900
+    },
+    {
+      "epoch": 0.4565524804478947,
+      "grad_norm": 7.1372161394964575,
+      "learning_rate": 5.435548115222537e-06,
+      "loss": 0.3742,
+      "step": 7910
+    },
+    {
+      "epoch": 0.4571296643675507,
+      "grad_norm": 20.273585832785447,
+      "learning_rate": 5.42977544305259e-06,
+      "loss": 0.3769,
+      "step": 7920
+    },
+    {
+      "epoch": 0.4577068482872067,
+      "grad_norm": 2.794498898766565,
+      "learning_rate": 5.424002770882641e-06,
+      "loss": 0.384,
+      "step": 7930
+    },
+    {
+      "epoch": 0.4582840322068627,
+      "grad_norm": 7.066469891261649,
+      "learning_rate": 5.418230098712695e-06,
+      "loss": 0.3795,
+      "step": 7940
+    },
+    {
+      "epoch": 0.4588612161265187,
+      "grad_norm": 2.8353456413911737,
+      "learning_rate": 5.412457426542746e-06,
+      "loss": 0.3643,
+      "step": 7950
+    },
+    {
+      "epoch": 0.4594384000461747,
+      "grad_norm": 3.0383570429357345,
+      "learning_rate": 5.406684754372799e-06,
+      "loss": 0.3718,
+      "step": 7960
+    },
+    {
+      "epoch": 0.4600155839658307,
+      "grad_norm": 3.6164323938018734,
+      "learning_rate": 5.400912082202852e-06,
+      "loss": 0.3849,
+      "step": 7970
+    },
+    {
+      "epoch": 0.4605927678854867,
+      "grad_norm": 2.7123845726783262,
+      "learning_rate": 5.3951394100329045e-06,
+      "loss": 0.3683,
+      "step": 7980
+    },
+    {
+      "epoch": 0.4611699518051427,
+      "grad_norm": 2.312599361300853,
+      "learning_rate": 5.3893667378629575e-06,
+      "loss": 0.3798,
+      "step": 7990
+    },
+    {
+      "epoch": 0.4617471357247987,
+      "grad_norm": 3.251330933463336,
+      "learning_rate": 5.38359406569301e-06,
+      "loss": 0.3587,
+      "step": 8000
+    },
+    {
+      "epoch": 0.4623243196444547,
+      "grad_norm": 6.596375932856641,
+      "learning_rate": 5.377821393523063e-06,
+      "loss": 0.3802,
+      "step": 8010
+    },
+    {
+      "epoch": 0.4629015035641107,
+      "grad_norm": 1.8050467998180781,
+      "learning_rate": 5.372048721353115e-06,
+      "loss": 0.3712,
+      "step": 8020
+    },
+    {
+      "epoch": 0.4634786874837667,
+      "grad_norm": 5.472845808317412,
+      "learning_rate": 5.366276049183168e-06,
+      "loss": 0.377,
+      "step": 8030
+    },
+    {
+      "epoch": 0.4640558714034227,
+      "grad_norm": 21.192833022577837,
+      "learning_rate": 5.36050337701322e-06,
+      "loss": 0.3745,
+      "step": 8040
+    },
+    {
+      "epoch": 0.4646330553230787,
+      "grad_norm": 2.3954727021255677,
+      "learning_rate": 5.354730704843273e-06,
+      "loss": 0.368,
+      "step": 8050
+    },
+    {
+      "epoch": 0.4652102392427347,
+      "grad_norm": 3.6751726559767652,
+      "learning_rate": 5.3489580326733245e-06,
+      "loss": 0.3803,
+      "step": 8060
+    },
+    {
+      "epoch": 0.4657874231623907,
+      "grad_norm": 2.6299220528922316,
+      "learning_rate": 5.3431853605033775e-06,
+      "loss": 0.3655,
+      "step": 8070
+    },
+    {
+      "epoch": 0.4663646070820467,
+      "grad_norm": 4.147182980327485,
+      "learning_rate": 5.33741268833343e-06,
+      "loss": 0.376,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4669417910017027,
+      "grad_norm": 2.3318408925884526,
+      "learning_rate": 5.331640016163483e-06,
+      "loss": 0.3793,
+      "step": 8090
+    },
+    {
+      "epoch": 0.4675189749213587,
+      "grad_norm": 2.5630808791681106,
+      "learning_rate": 5.325867343993535e-06,
+      "loss": 0.3841,
+      "step": 8100
+    },
+    {
+      "epoch": 0.4680961588410147,
+      "grad_norm": 16.561602524939726,
+      "learning_rate": 5.320094671823588e-06,
+      "loss": 0.3628,
+      "step": 8110
+    },
+    {
+      "epoch": 0.4686733427606707,
+      "grad_norm": 2.655435817667697,
+      "learning_rate": 5.31432199965364e-06,
+      "loss": 0.3827,
+      "step": 8120
+    },
+    {
+      "epoch": 0.4692505266803267,
+      "grad_norm": 2.082608101455672,
+      "learning_rate": 5.308549327483693e-06,
+      "loss": 0.3726,
+      "step": 8130
+    },
+    {
+      "epoch": 0.4698277105999827,
+      "grad_norm": 7.450725164048278,
+      "learning_rate": 5.302776655313745e-06,
+      "loss": 0.3607,
+      "step": 8140
+    },
+    {
+      "epoch": 0.4704048945196387,
+      "grad_norm": 3.727260702005544,
+      "learning_rate": 5.297003983143798e-06,
+      "loss": 0.3666,
+      "step": 8150
+    },
+    {
+      "epoch": 0.4709820784392947,
+      "grad_norm": 2.9372607445086816,
+      "learning_rate": 5.29123131097385e-06,
+      "loss": 0.3638,
+      "step": 8160
+    },
+    {
+      "epoch": 0.4715592623589507,
+      "grad_norm": 3.011083825137573,
+      "learning_rate": 5.2854586388039034e-06,
+      "loss": 0.3711,
+      "step": 8170
+    },
+    {
+      "epoch": 0.4721364462786067,
+      "grad_norm": 5.58881805540413,
+      "learning_rate": 5.279685966633955e-06,
+      "loss": 0.3737,
+      "step": 8180
+    },
+    {
+      "epoch": 0.4727136301982627,
+      "grad_norm": 1.9307643066917193,
+      "learning_rate": 5.273913294464008e-06,
+      "loss": 0.3633,
+      "step": 8190
+    },
+    {
+      "epoch": 0.4732908141179187,
+      "grad_norm": 2.246158235550298,
+      "learning_rate": 5.26814062229406e-06,
+      "loss": 0.3811,
+      "step": 8200
+    },
+    {
+      "epoch": 0.4738679980375747,
+      "grad_norm": 5.378095626076901,
+      "learning_rate": 5.262367950124113e-06,
+      "loss": 0.3772,
+      "step": 8210
+    },
+    {
+      "epoch": 0.4744451819572307,
+      "grad_norm": 3.594146830036725,
+      "learning_rate": 5.256595277954165e-06,
+      "loss": 0.3689,
+      "step": 8220
+    },
+    {
+      "epoch": 0.4750223658768867,
+      "grad_norm": 1.880161955339062,
+      "learning_rate": 5.250822605784218e-06,
+      "loss": 0.3682,
+      "step": 8230
+    },
+    {
+      "epoch": 0.4755995497965427,
+      "grad_norm": 3.655512057019781,
+      "learning_rate": 5.24504993361427e-06,
+      "loss": 0.3601,
+      "step": 8240
+    },
+    {
+      "epoch": 0.4761767337161987,
+      "grad_norm": 4.5822643890603345,
+      "learning_rate": 5.239277261444323e-06,
+      "loss": 0.3799,
+      "step": 8250
+    },
+    {
+      "epoch": 0.4767539176358547,
+      "grad_norm": 10.062266012222976,
+      "learning_rate": 5.2335045892743755e-06,
+      "loss": 0.3651,
+      "step": 8260
+    },
+    {
+      "epoch": 0.4773311015555107,
+      "grad_norm": 7.820363388846345,
+      "learning_rate": 5.2277319171044286e-06,
+      "loss": 0.3629,
+      "step": 8270
+    },
+    {
+      "epoch": 0.47790828547516667,
+      "grad_norm": 3.04998159803599,
+      "learning_rate": 5.22195924493448e-06,
+      "loss": 0.3693,
+      "step": 8280
+    },
+    {
+      "epoch": 0.47848546939482267,
+      "grad_norm": 4.5635676614995475,
+      "learning_rate": 5.216186572764533e-06,
+      "loss": 0.384,
+      "step": 8290
+    },
+    {
+      "epoch": 0.47906265331447867,
+      "grad_norm": 2.1736842754902708,
+      "learning_rate": 5.210413900594585e-06,
+      "loss": 0.363,
+      "step": 8300
+    },
+    {
+      "epoch": 0.47963983723413467,
+      "grad_norm": 2.5418326594021887,
+      "learning_rate": 5.204641228424638e-06,
+      "loss": 0.3697,
+      "step": 8310
+    },
+    {
+      "epoch": 0.48021702115379067,
+      "grad_norm": 2.692093940326692,
+      "learning_rate": 5.19886855625469e-06,
+      "loss": 0.3691,
+      "step": 8320
+    },
+    {
+      "epoch": 0.48079420507344667,
+      "grad_norm": 2.4052933532982816,
+      "learning_rate": 5.193095884084743e-06,
+      "loss": 0.3506,
+      "step": 8330
+    },
+    {
+      "epoch": 0.48137138899310267,
+      "grad_norm": 6.620567354797733,
+      "learning_rate": 5.1873232119147955e-06,
+      "loss": 0.3838,
+      "step": 8340
+    },
+    {
+      "epoch": 0.48194857291275867,
+      "grad_norm": 2.9255268074292555,
+      "learning_rate": 5.1815505397448485e-06,
+      "loss": 0.3694,
+      "step": 8350
+    },
+    {
+      "epoch": 0.48252575683241467,
+      "grad_norm": 13.37602989555681,
+      "learning_rate": 5.175777867574901e-06,
+      "loss": 0.3752,
+      "step": 8360
+    },
+    {
+      "epoch": 0.48310294075207066,
+      "grad_norm": 3.511794563236054,
+      "learning_rate": 5.170005195404954e-06,
+      "loss": 0.3741,
+      "step": 8370
+    },
+    {
+      "epoch": 0.48368012467172666,
+      "grad_norm": 4.7758963928181375,
+      "learning_rate": 5.164232523235006e-06,
+      "loss": 0.369,
+      "step": 8380
+    },
+    {
+      "epoch": 0.48425730859138266,
+      "grad_norm": 3.8240515141083002,
+      "learning_rate": 5.158459851065059e-06,
+      "loss": 0.374,
+      "step": 8390
+    },
+    {
+      "epoch": 0.48483449251103866,
+      "grad_norm": 8.100590461673363,
+      "learning_rate": 5.15268717889511e-06,
+      "loss": 0.3563,
+      "step": 8400
+    },
+    {
+      "epoch": 0.48541167643069466,
+      "grad_norm": 4.117224392256427,
+      "learning_rate": 5.146914506725163e-06,
+      "loss": 0.3626,
+      "step": 8410
+    },
+    {
+      "epoch": 0.48598886035035066,
+      "grad_norm": 3.4662871354919904,
+      "learning_rate": 5.141141834555215e-06,
+      "loss": 0.3848,
+      "step": 8420
+    },
+    {
+      "epoch": 0.48656604427000666,
+      "grad_norm": 4.959524985987204,
+      "learning_rate": 5.1353691623852684e-06,
+      "loss": 0.3649,
+      "step": 8430
+    },
+    {
+      "epoch": 0.48714322818966266,
+      "grad_norm": 5.172300477902163,
+      "learning_rate": 5.129596490215321e-06,
+      "loss": 0.3732,
+      "step": 8440
+    },
+    {
+      "epoch": 0.48772041210931866,
+      "grad_norm": 2.996587870660032,
+      "learning_rate": 5.123823818045374e-06,
+      "loss": 0.3544,
+      "step": 8450
+    },
+    {
+      "epoch": 0.48829759602897466,
+      "grad_norm": 5.228719957469869,
+      "learning_rate": 5.118051145875426e-06,
+      "loss": 0.3623,
+      "step": 8460
+    },
+    {
+      "epoch": 0.48887477994863066,
+      "grad_norm": 7.078111368668544,
+      "learning_rate": 5.112278473705479e-06,
+      "loss": 0.3521,
+      "step": 8470
+    },
+    {
+      "epoch": 0.48945196386828665,
+      "grad_norm": 4.585249570356133,
+      "learning_rate": 5.106505801535531e-06,
+      "loss": 0.3602,
+      "step": 8480
+    },
+    {
+      "epoch": 0.49002914778794265,
+      "grad_norm": 3.769909535272591,
+      "learning_rate": 5.100733129365584e-06,
+      "loss": 0.3642,
+      "step": 8490
+    },
+    {
+      "epoch": 0.49060633170759865,
+      "grad_norm": 5.843171007267111,
+      "learning_rate": 5.094960457195637e-06,
+      "loss": 0.3844,
+      "step": 8500
+    },
+    {
+      "epoch": 0.49118351562725465,
+      "grad_norm": 5.011620359523228,
+      "learning_rate": 5.089187785025688e-06,
+      "loss": 0.3748,
+      "step": 8510
+    },
+    {
+      "epoch": 0.4917606995469106,
+      "grad_norm": 3.1629027771513667,
+      "learning_rate": 5.083415112855742e-06,
+      "loss": 0.3756,
+      "step": 8520
+    },
+    {
+      "epoch": 0.4923378834665666,
+      "grad_norm": 3.4287778879420583,
+      "learning_rate": 5.0776424406857936e-06,
+      "loss": 0.3654,
+      "step": 8530
+    },
+    {
+      "epoch": 0.4929150673862226,
+      "grad_norm": 2.3995913819961077,
+      "learning_rate": 5.0718697685158466e-06,
+      "loss": 0.3497,
+      "step": 8540
+    },
+    {
+      "epoch": 0.4934922513058786,
+      "grad_norm": 7.033069082305426,
+      "learning_rate": 5.066097096345899e-06,
+      "loss": 0.3679,
+      "step": 8550
+    },
+    {
+      "epoch": 0.4940694352255346,
+      "grad_norm": 9.94984806585782,
+      "learning_rate": 5.060324424175952e-06,
+      "loss": 0.3528,
+      "step": 8560
+    },
+    {
+      "epoch": 0.4946466191451906,
+      "grad_norm": 3.544058073278801,
+      "learning_rate": 5.054551752006004e-06,
+      "loss": 0.3578,
+      "step": 8570
+    },
+    {
+      "epoch": 0.4952238030648466,
+      "grad_norm": 2.5120291222524176,
+      "learning_rate": 5.048779079836057e-06,
+      "loss": 0.3548,
+      "step": 8580
+    },
+    {
+      "epoch": 0.4958009869845026,
+      "grad_norm": 3.3522580875814887,
+      "learning_rate": 5.043006407666109e-06,
+      "loss": 0.3606,
+      "step": 8590
+    },
+    {
+      "epoch": 0.4963781709041586,
+      "grad_norm": 2.412854685811211,
+      "learning_rate": 5.037233735496162e-06,
+      "loss": 0.3605,
+      "step": 8600
+    },
+    {
+      "epoch": 0.4969553548238146,
+      "grad_norm": 2.9209835829832613,
+      "learning_rate": 5.031461063326214e-06,
+      "loss": 0.3566,
+      "step": 8610
+    },
+    {
+      "epoch": 0.4975325387434706,
+      "grad_norm": 2.6580624391224568,
+      "learning_rate": 5.025688391156267e-06,
+      "loss": 0.3806,
+      "step": 8620
+    },
+    {
+      "epoch": 0.4981097226631266,
+      "grad_norm": 2.9252114550534567,
+      "learning_rate": 5.019915718986319e-06,
+      "loss": 0.3596,
+      "step": 8630
+    },
+    {
+      "epoch": 0.4986869065827826,
+      "grad_norm": 2.432042802537537,
+      "learning_rate": 5.014143046816372e-06,
+      "loss": 0.3849,
+      "step": 8640
+    },
+    {
+      "epoch": 0.4992640905024386,
+      "grad_norm": 14.195133545125286,
+      "learning_rate": 5.008370374646424e-06,
+      "loss": 0.3602,
+      "step": 8650
+    },
+    {
+      "epoch": 0.4998412744220946,
+      "grad_norm": 88.84290269765289,
+      "learning_rate": 5.002597702476477e-06,
+      "loss": 0.3642,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5004184583417506,
+      "grad_norm": 2.935701463863382,
+      "learning_rate": 4.996825030306529e-06,
+      "loss": 0.3792,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5009956422614066,
+      "grad_norm": 3.1279829631060396,
+      "learning_rate": 4.991052358136581e-06,
+      "loss": 0.3693,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5015728261810626,
+      "grad_norm": 10.193114728366995,
+      "learning_rate": 4.985279685966634e-06,
+      "loss": 0.3597,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5021500101007186,
+      "grad_norm": 2.4687234809350107,
+      "learning_rate": 4.9795070137966864e-06,
+      "loss": 0.3574,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5027271940203746,
+      "grad_norm": 5.072249220362736,
+      "learning_rate": 4.9737343416267395e-06,
+      "loss": 0.3494,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5033043779400306,
+      "grad_norm": 2.491352620261712,
+      "learning_rate": 4.967961669456792e-06,
+      "loss": 0.3567,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5038815618596866,
+      "grad_norm": 4.883169609666169,
+      "learning_rate": 4.962188997286845e-06,
+      "loss": 0.3563,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5044587457793426,
+      "grad_norm": 3.9469009793644623,
+      "learning_rate": 4.956416325116897e-06,
+      "loss": 0.3669,
+      "step": 8740
+    },
+    {
+      "epoch": 0.5050359296989986,
+      "grad_norm": 5.325339903718785,
+      "learning_rate": 4.95064365294695e-06,
+      "loss": 0.3479,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5056131136186546,
+      "grad_norm": 2.924190713741744,
+      "learning_rate": 4.944870980777002e-06,
+      "loss": 0.3596,
+      "step": 8760
+    },
+    {
+      "epoch": 0.5061902975383106,
+      "grad_norm": 29.94785140992617,
+      "learning_rate": 4.939098308607055e-06,
+      "loss": 0.3482,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5067674814579666,
+      "grad_norm": 4.600461351076627,
+      "learning_rate": 4.933325636437107e-06,
+      "loss": 0.3676,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5073446653776226,
+      "grad_norm": 3.298079688278539,
+      "learning_rate": 4.92755296426716e-06,
+      "loss": 0.3849,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5079218492972786,
+      "grad_norm": 3.3537368346024903,
+      "learning_rate": 4.921780292097212e-06,
+      "loss": 0.354,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5084990332169346,
+      "grad_norm": 4.562710088882002,
+      "learning_rate": 4.9160076199272646e-06,
+      "loss": 0.3762,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5090762171365906,
+      "grad_norm": 3.600922103480094,
+      "learning_rate": 4.910234947757318e-06,
+      "loss": 0.3709,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5096534010562466,
+      "grad_norm": 5.507110655352795,
+      "learning_rate": 4.90446227558737e-06,
+      "loss": 0.35,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5102305849759026,
+      "grad_norm": 14.876481920624023,
+      "learning_rate": 4.898689603417423e-06,
+      "loss": 0.3655,
+      "step": 8840
+    },
+    {
+      "epoch": 0.5108077688955586,
+      "grad_norm": 6.040943909655418,
+      "learning_rate": 4.892916931247475e-06,
+      "loss": 0.3537,
+      "step": 8850
+    },
+    {
+      "epoch": 0.5113849528152146,
+      "grad_norm": 4.784882455474531,
+      "learning_rate": 4.887144259077527e-06,
+      "loss": 0.365,
+      "step": 8860
+    },
+    {
+      "epoch": 0.5119621367348706,
+      "grad_norm": 3.5957944832471864,
+      "learning_rate": 4.88137158690758e-06,
+      "loss": 0.3649,
+      "step": 8870
+    },
+    {
+      "epoch": 0.5125393206545266,
+      "grad_norm": 7.656707006499249,
+      "learning_rate": 4.875598914737632e-06,
+      "loss": 0.3678,
+      "step": 8880
+    },
+    {
+      "epoch": 0.5131165045741826,
+      "grad_norm": 3.6610216347500666,
+      "learning_rate": 4.869826242567685e-06,
+      "loss": 0.3659,
+      "step": 8890
+    },
+    {
+      "epoch": 0.5136936884938386,
+      "grad_norm": 4.181649664719206,
+      "learning_rate": 4.8640535703977375e-06,
+      "loss": 0.358,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5142708724134946,
+      "grad_norm": 1.8612925513884986,
+      "learning_rate": 4.8582808982277905e-06,
+      "loss": 0.3508,
+      "step": 8910
+    },
+    {
+      "epoch": 0.5148480563331506,
+      "grad_norm": 5.0292268546846195,
+      "learning_rate": 4.852508226057843e-06,
+      "loss": 0.3567,
+      "step": 8920
+    },
+    {
+      "epoch": 0.5154252402528066,
+      "grad_norm": 5.77083593828813,
+      "learning_rate": 4.846735553887895e-06,
+      "loss": 0.3706,
+      "step": 8930
+    },
+    {
+      "epoch": 0.5160024241724626,
+      "grad_norm": 3.1575366683166264,
+      "learning_rate": 4.840962881717948e-06,
+      "loss": 0.3603,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5165796080921186,
+      "grad_norm": 3.0092615460602357,
+      "learning_rate": 4.835190209548e-06,
+      "loss": 0.3567,
+      "step": 8950
+    },
+    {
+      "epoch": 0.5171567920117746,
+      "grad_norm": 5.424773899652038,
+      "learning_rate": 4.829417537378053e-06,
+      "loss": 0.3569,
+      "step": 8960
+    },
+    {
+      "epoch": 0.5177339759314306,
+      "grad_norm": 4.0199024195308075,
+      "learning_rate": 4.823644865208105e-06,
+      "loss": 0.376,
+      "step": 8970
+    },
+    {
+      "epoch": 0.5183111598510866,
+      "grad_norm": 8.92239001104145,
+      "learning_rate": 4.8178721930381575e-06,
+      "loss": 0.3572,
+      "step": 8980
+    },
+    {
+      "epoch": 0.5188883437707426,
+      "grad_norm": 10.702196683213096,
+      "learning_rate": 4.8120995208682105e-06,
+      "loss": 0.3651,
+      "step": 8990
+    },
+    {
+      "epoch": 0.5194655276903986,
+      "grad_norm": 4.470162551927128,
+      "learning_rate": 4.806326848698263e-06,
+      "loss": 0.3654,
+      "step": 9000
+    },
+    {
+      "epoch": 0.5200427116100546,
+      "grad_norm": 2.1821708334970737,
+      "learning_rate": 4.800554176528316e-06,
+      "loss": 0.3516,
+      "step": 9010
+    },
+    {
+      "epoch": 0.5206198955297106,
+      "grad_norm": 5.347973377285275,
+      "learning_rate": 4.794781504358368e-06,
+      "loss": 0.3623,
+      "step": 9020
+    },
+    {
+      "epoch": 0.5211970794493666,
+      "grad_norm": 4.623834142826691,
+      "learning_rate": 4.78900883218842e-06,
+      "loss": 0.3727,
+      "step": 9030
+    },
+    {
+      "epoch": 0.5217742633690226,
+      "grad_norm": 5.255808460960779,
+      "learning_rate": 4.783236160018473e-06,
+      "loss": 0.3729,
+      "step": 9040
+    },
+    {
+      "epoch": 0.5223514472886785,
+      "grad_norm": 3.1254534476864215,
+      "learning_rate": 4.777463487848525e-06,
+      "loss": 0.3679,
+      "step": 9050
+    },
+    {
+      "epoch": 0.5229286312083345,
+      "grad_norm": 9.646051621259671,
+      "learning_rate": 4.771690815678578e-06,
+      "loss": 0.3834,
+      "step": 9060
+    },
+    {
+      "epoch": 0.5235058151279905,
+      "grad_norm": 4.490473334084667,
+      "learning_rate": 4.76591814350863e-06,
+      "loss": 0.3656,
+      "step": 9070
+    },
+    {
+      "epoch": 0.5240829990476465,
+      "grad_norm": 2.607385547296513,
+      "learning_rate": 4.760145471338683e-06,
+      "loss": 0.3861,
+      "step": 9080
+    },
+    {
+      "epoch": 0.5246601829673025,
+      "grad_norm": 3.7754579602738136,
+      "learning_rate": 4.754372799168736e-06,
+      "loss": 0.3506,
+      "step": 9090
+    },
+    {
+      "epoch": 0.5252373668869585,
+      "grad_norm": 2.8550942295775896,
+      "learning_rate": 4.748600126998788e-06,
+      "loss": 0.3567,
+      "step": 9100
+    },
+    {
+      "epoch": 0.5258145508066145,
+      "grad_norm": 1.7112590538881849,
+      "learning_rate": 4.742827454828841e-06,
+      "loss": 0.3604,
+      "step": 9110
+    },
+    {
+      "epoch": 0.5263917347262705,
+      "grad_norm": 2.7066239753400585,
+      "learning_rate": 4.737054782658893e-06,
+      "loss": 0.36,
+      "step": 9120
+    },
+    {
+      "epoch": 0.5269689186459265,
+      "grad_norm": 5.758530747558061,
+      "learning_rate": 4.731282110488946e-06,
+      "loss": 0.3634,
+      "step": 9130
+    },
+    {
+      "epoch": 0.5275461025655825,
+      "grad_norm": 210.77018196053547,
+      "learning_rate": 4.725509438318998e-06,
+      "loss": 0.3722,
+      "step": 9140
+    },
+    {
+      "epoch": 0.5281232864852385,
+      "grad_norm": 2.300368854843036,
+      "learning_rate": 4.71973676614905e-06,
+      "loss": 0.3729,
+      "step": 9150
+    },
+    {
+      "epoch": 0.5287004704048945,
+      "grad_norm": 2.210540304841504,
+      "learning_rate": 4.713964093979103e-06,
+      "loss": 0.3796,
+      "step": 9160
+    },
+    {
+      "epoch": 0.5292776543245505,
+      "grad_norm": 2.2634685628485145,
+      "learning_rate": 4.7081914218091555e-06,
+      "loss": 0.356,
+      "step": 9170
+    },
+    {
+      "epoch": 0.5298548382442065,
+      "grad_norm": 3.1055746156661614,
+      "learning_rate": 4.7024187496392085e-06,
+      "loss": 0.3582,
+      "step": 9180
+    },
+    {
+      "epoch": 0.5304320221638625,
+      "grad_norm": 3.654168505440511,
+      "learning_rate": 4.696646077469261e-06,
+      "loss": 0.3607,
+      "step": 9190
+    },
+    {
+      "epoch": 0.5310092060835185,
+      "grad_norm": 3.0537549885812347,
+      "learning_rate": 4.690873405299313e-06,
+      "loss": 0.3767,
+      "step": 9200
+    },
+    {
+      "epoch": 0.5315863900031745,
+      "grad_norm": 1.9675795559729068,
+      "learning_rate": 4.685100733129366e-06,
+      "loss": 0.3657,
+      "step": 9210
+    },
+    {
+      "epoch": 0.5321635739228305,
+      "grad_norm": 2.0115039115011606,
+      "learning_rate": 4.679328060959418e-06,
+      "loss": 0.3601,
+      "step": 9220
+    },
+    {
+      "epoch": 0.5327407578424865,
+      "grad_norm": 5.152045183089219,
+      "learning_rate": 4.673555388789471e-06,
+      "loss": 0.3874,
+      "step": 9230
+    },
+    {
+      "epoch": 0.5333179417621425,
+      "grad_norm": 3.394195501156254,
+      "learning_rate": 4.667782716619523e-06,
+      "loss": 0.3687,
+      "step": 9240
+    },
+    {
+      "epoch": 0.5338951256817985,
+      "grad_norm": 3.733950998673983,
+      "learning_rate": 4.6620100444495755e-06,
+      "loss": 0.3812,
+      "step": 9250
+    },
+    {
+      "epoch": 0.5344723096014545,
+      "grad_norm": 2.172335452644072,
+      "learning_rate": 4.6562373722796285e-06,
+      "loss": 0.3719,
+      "step": 9260
+    },
+    {
+      "epoch": 0.5350494935211105,
+      "grad_norm": 1.8466321499683245,
+      "learning_rate": 4.650464700109681e-06,
+      "loss": 0.3563,
+      "step": 9270
+    },
+    {
+      "epoch": 0.5356266774407665,
+      "grad_norm": 2.411259261337378,
+      "learning_rate": 4.644692027939734e-06,
+      "loss": 0.3652,
+      "step": 9280
+    },
+    {
+      "epoch": 0.5362038613604225,
+      "grad_norm": 2.525799052658632,
+      "learning_rate": 4.638919355769787e-06,
+      "loss": 0.357,
+      "step": 9290
+    },
+    {
+      "epoch": 0.5367810452800785,
+      "grad_norm": 4.322100164387683,
+      "learning_rate": 4.633146683599839e-06,
+      "loss": 0.344,
+      "step": 9300
+    },
+    {
+      "epoch": 0.5373582291997345,
+      "grad_norm": 2.9720413934133267,
+      "learning_rate": 4.627374011429892e-06,
+      "loss": 0.3591,
+      "step": 9310
+    },
+    {
+      "epoch": 0.5379354131193905,
+      "grad_norm": 3.236889177938413,
+      "learning_rate": 4.621601339259944e-06,
+      "loss": 0.3698,
+      "step": 9320
+    },
+    {
+      "epoch": 0.5385125970390465,
+      "grad_norm": 4.173966589558208,
+      "learning_rate": 4.615828667089996e-06,
+      "loss": 0.3687,
+      "step": 9330
+    },
+    {
+      "epoch": 0.5390897809587025,
+      "grad_norm": 3.3050072649704387,
+      "learning_rate": 4.610055994920049e-06,
+      "loss": 0.3559,
+      "step": 9340
+    },
+    {
+      "epoch": 0.5396669648783585,
+      "grad_norm": 2.2463012820732904,
+      "learning_rate": 4.6042833227501014e-06,
+      "loss": 0.3585,
+      "step": 9350
+    },
+    {
+      "epoch": 0.5402441487980145,
+      "grad_norm": 2.3061353512132357,
+      "learning_rate": 4.5985106505801544e-06,
+      "loss": 0.3666,
+      "step": 9360
+    },
+    {
+      "epoch": 0.5408213327176705,
+      "grad_norm": 2.8767755309606393,
+      "learning_rate": 4.592737978410207e-06,
+      "loss": 0.3656,
+      "step": 9370
+    },
+    {
+      "epoch": 0.5413985166373265,
+      "grad_norm": 3.5885590608959603,
+      "learning_rate": 4.586965306240259e-06,
+      "loss": 0.3619,
+      "step": 9380
+    },
+    {
+      "epoch": 0.5419757005569825,
+      "grad_norm": 3.012510637436092,
+      "learning_rate": 4.581192634070312e-06,
+      "loss": 0.3679,
+      "step": 9390
+    },
+    {
+      "epoch": 0.5425528844766385,
+      "grad_norm": 2.074304355176205,
+      "learning_rate": 4.575419961900364e-06,
+      "loss": 0.3558,
+      "step": 9400
+    },
+    {
+      "epoch": 0.5431300683962945,
+      "grad_norm": 2.410649696384616,
+      "learning_rate": 4.569647289730417e-06,
+      "loss": 0.3667,
+      "step": 9410
+    },
+    {
+      "epoch": 0.5437072523159505,
+      "grad_norm": 6.990964309593162,
+      "learning_rate": 4.563874617560469e-06,
+      "loss": 0.3544,
+      "step": 9420
+    },
+    {
+      "epoch": 0.5442844362356065,
+      "grad_norm": 1.6679890672242221,
+      "learning_rate": 4.558101945390521e-06,
+      "loss": 0.3635,
+      "step": 9430
+    },
+    {
+      "epoch": 0.5448616201552625,
+      "grad_norm": 3.5900096135866177,
+      "learning_rate": 4.552329273220574e-06,
+      "loss": 0.3548,
+      "step": 9440
+    },
+    {
+      "epoch": 0.5454388040749185,
+      "grad_norm": 2.8054899052225655,
+      "learning_rate": 4.5465566010506266e-06,
+      "loss": 0.3573,
+      "step": 9450
+    },
+    {
+      "epoch": 0.5460159879945745,
+      "grad_norm": 2.7792687957236315,
+      "learning_rate": 4.5407839288806796e-06,
+      "loss": 0.3677,
+      "step": 9460
+    },
+    {
+      "epoch": 0.5465931719142305,
+      "grad_norm": 2.2614901204437636,
+      "learning_rate": 4.535011256710732e-06,
+      "loss": 0.3643,
+      "step": 9470
+    },
+    {
+      "epoch": 0.5471703558338865,
+      "grad_norm": 2.062841207589413,
+      "learning_rate": 4.529238584540785e-06,
+      "loss": 0.3592,
+      "step": 9480
+    },
+    {
+      "epoch": 0.5477475397535425,
+      "grad_norm": 8.866919074207425,
+      "learning_rate": 4.523465912370837e-06,
+      "loss": 0.3552,
+      "step": 9490
+    },
+    {
+      "epoch": 0.5483247236731985,
+      "grad_norm": 2.5645835670129618,
+      "learning_rate": 4.517693240200889e-06,
+      "loss": 0.3498,
+      "step": 9500
+    },
+    {
+      "epoch": 0.5489019075928545,
+      "grad_norm": 3.1741970680335716,
+      "learning_rate": 4.511920568030942e-06,
+      "loss": 0.3682,
+      "step": 9510
+    },
+    {
+      "epoch": 0.5494790915125105,
+      "grad_norm": 2.0747984580427765,
+      "learning_rate": 4.506147895860994e-06,
+      "loss": 0.3646,
+      "step": 9520
+    },
+    {
+      "epoch": 0.5500562754321665,
+      "grad_norm": 2.3586125635742654,
+      "learning_rate": 4.500375223691047e-06,
+      "loss": 0.355,
+      "step": 9530
+    },
+    {
+      "epoch": 0.5506334593518225,
+      "grad_norm": 3.8182790115085927,
+      "learning_rate": 4.4946025515210995e-06,
+      "loss": 0.3528,
+      "step": 9540
+    },
+    {
+      "epoch": 0.5512106432714785,
+      "grad_norm": 2.623243257421812,
+      "learning_rate": 4.488829879351152e-06,
+      "loss": 0.3551,
+      "step": 9550
+    },
+    {
+      "epoch": 0.5517878271911345,
+      "grad_norm": 2.775469371839904,
+      "learning_rate": 4.483057207181205e-06,
+      "loss": 0.3556,
+      "step": 9560
+    },
+    {
+      "epoch": 0.5523650111107905,
+      "grad_norm": 2.345529859871698,
+      "learning_rate": 4.477284535011257e-06,
+      "loss": 0.3702,
+      "step": 9570
+    },
+    {
+      "epoch": 0.5529421950304465,
+      "grad_norm": 5.265133567547254,
+      "learning_rate": 4.47151186284131e-06,
+      "loss": 0.3617,
+      "step": 9580
+    },
+    {
+      "epoch": 0.5535193789501025,
+      "grad_norm": 3.1602517841252897,
+      "learning_rate": 4.465739190671362e-06,
+      "loss": 0.3524,
+      "step": 9590
+    },
+    {
+      "epoch": 0.5540965628697585,
+      "grad_norm": 1.8906279954560556,
+      "learning_rate": 4.459966518501414e-06,
+      "loss": 0.3584,
+      "step": 9600
+    },
+    {
+      "epoch": 0.5546737467894145,
+      "grad_norm": 3.4743194872868117,
+      "learning_rate": 4.454193846331467e-06,
+      "loss": 0.3531,
+      "step": 9610
+    },
+    {
+      "epoch": 0.5552509307090705,
+      "grad_norm": 2.9951619722989578,
+      "learning_rate": 4.4484211741615194e-06,
+      "loss": 0.3525,
+      "step": 9620
+    },
+    {
+      "epoch": 0.5558281146287264,
+      "grad_norm": 5.694532155563467,
+      "learning_rate": 4.4426485019915725e-06,
+      "loss": 0.3568,
+      "step": 9630
+    },
+    {
+      "epoch": 0.5564052985483824,
+      "grad_norm": 2.0095159465987398,
+      "learning_rate": 4.436875829821625e-06,
+      "loss": 0.3586,
+      "step": 9640
+    },
+    {
+      "epoch": 0.5569824824680384,
+      "grad_norm": 2.703106746299796,
+      "learning_rate": 4.431103157651677e-06,
+      "loss": 0.3586,
+      "step": 9650
+    },
+    {
+      "epoch": 0.5575596663876944,
+      "grad_norm": 4.366070322479832,
+      "learning_rate": 4.42533048548173e-06,
+      "loss": 0.3628,
+      "step": 9660
+    },
+    {
+      "epoch": 0.5581368503073504,
+      "grad_norm": 5.959735845454689,
+      "learning_rate": 4.419557813311782e-06,
+      "loss": 0.3453,
+      "step": 9670
+    },
+    {
+      "epoch": 0.5587140342270064,
+      "grad_norm": 3.887262505871754,
+      "learning_rate": 4.413785141141835e-06,
+      "loss": 0.3579,
+      "step": 9680
+    },
+    {
+      "epoch": 0.5592912181466624,
+      "grad_norm": 3.5847585820260273,
+      "learning_rate": 4.408012468971887e-06,
+      "loss": 0.353,
+      "step": 9690
+    },
+    {
+      "epoch": 0.5598684020663184,
+      "grad_norm": 11.183076093226834,
+      "learning_rate": 4.40223979680194e-06,
+      "loss": 0.354,
+      "step": 9700
+    },
+    {
+      "epoch": 0.5604455859859744,
+      "grad_norm": 8.751616032322445,
+      "learning_rate": 4.396467124631992e-06,
+      "loss": 0.358,
+      "step": 9710
+    },
+    {
+      "epoch": 0.5610227699056304,
+      "grad_norm": 8.759188394381018,
+      "learning_rate": 4.3906944524620446e-06,
+      "loss": 0.3477,
+      "step": 9720
+    },
+    {
+      "epoch": 0.5615999538252864,
+      "grad_norm": 3.185742989121124,
+      "learning_rate": 4.3849217802920976e-06,
+      "loss": 0.3527,
+      "step": 9730
+    },
+    {
+      "epoch": 0.5621771377449424,
+      "grad_norm": 2.6034839783878847,
+      "learning_rate": 4.37914910812215e-06,
+      "loss": 0.3614,
+      "step": 9740
+    },
+    {
+      "epoch": 0.5627543216645984,
+      "grad_norm": 3.8176039303217943,
+      "learning_rate": 4.373376435952203e-06,
+      "loss": 0.3629,
+      "step": 9750
+    },
+    {
+      "epoch": 0.5633315055842544,
+      "grad_norm": 2.892391389713988,
+      "learning_rate": 4.367603763782255e-06,
+      "loss": 0.3543,
+      "step": 9760
+    },
+    {
+      "epoch": 0.5639086895039104,
+      "grad_norm": 3.756297203577958,
+      "learning_rate": 4.361831091612307e-06,
+      "loss": 0.3618,
+      "step": 9770
+    },
+    {
+      "epoch": 0.5644858734235664,
+      "grad_norm": 5.678261576873807,
+      "learning_rate": 4.35605841944236e-06,
+      "loss": 0.3568,
+      "step": 9780
+    },
+    {
+      "epoch": 0.5650630573432224,
+      "grad_norm": 8.441638312518547,
+      "learning_rate": 4.350285747272412e-06,
+      "loss": 0.3729,
+      "step": 9790
+    },
+    {
+      "epoch": 0.5656402412628784,
+      "grad_norm": 11.963283236920255,
+      "learning_rate": 4.344513075102465e-06,
+      "loss": 0.3515,
+      "step": 9800
+    },
+    {
+      "epoch": 0.5662174251825344,
+      "grad_norm": 5.191809552146641,
+      "learning_rate": 4.338740402932518e-06,
+      "loss": 0.3555,
+      "step": 9810
+    },
+    {
+      "epoch": 0.5667946091021904,
+      "grad_norm": 4.402544511071244,
+      "learning_rate": 4.3329677307625705e-06,
+      "loss": 0.3593,
+      "step": 9820
+    },
+    {
+      "epoch": 0.5673717930218464,
+      "grad_norm": 3.877667010250195,
+      "learning_rate": 4.327195058592623e-06,
+      "loss": 0.3671,
+      "step": 9830
+    },
+    {
+      "epoch": 0.5679489769415024,
+      "grad_norm": 3.6190457236390907,
+      "learning_rate": 4.321422386422676e-06,
+      "loss": 0.3821,
+      "step": 9840
+    },
+    {
+      "epoch": 0.5685261608611584,
+      "grad_norm": 6.877296061159591,
+      "learning_rate": 4.315649714252728e-06,
+      "loss": 0.3603,
+      "step": 9850
+    },
+    {
+      "epoch": 0.5691033447808144,
+      "grad_norm": 4.590765318427339,
+      "learning_rate": 4.309877042082781e-06,
+      "loss": 0.3713,
+      "step": 9860
+    },
+    {
+      "epoch": 0.5696805287004704,
+      "grad_norm": 4.474831184892119,
+      "learning_rate": 4.304104369912833e-06,
+      "loss": 0.3475,
+      "step": 9870
+    },
+    {
+      "epoch": 0.5702577126201264,
+      "grad_norm": 6.348172468881283,
+      "learning_rate": 4.298331697742886e-06,
+      "loss": 0.3498,
+      "step": 9880
+    },
+    {
+      "epoch": 0.5708348965397824,
+      "grad_norm": 3.2161248922526906,
+      "learning_rate": 4.292559025572938e-06,
+      "loss": 0.3548,
+      "step": 9890
+    },
+    {
+      "epoch": 0.5714120804594384,
+      "grad_norm": 3.6712025528595476,
+      "learning_rate": 4.2867863534029905e-06,
+      "loss": 0.3698,
+      "step": 9900
+    },
+    {
+      "epoch": 0.5719892643790944,
+      "grad_norm": 6.372020331564792,
+      "learning_rate": 4.2810136812330435e-06,
+      "loss": 0.3774,
+      "step": 9910
+    },
+    {
+      "epoch": 0.5725664482987504,
+      "grad_norm": 7.0760072190976055,
+      "learning_rate": 4.275241009063096e-06,
+      "loss": 0.366,
+      "step": 9920
+    },
+    {
+      "epoch": 0.5731436322184064,
+      "grad_norm": 4.330191932801956,
+      "learning_rate": 4.269468336893149e-06,
+      "loss": 0.3659,
+      "step": 9930
+    },
+    {
+      "epoch": 0.5737208161380624,
+      "grad_norm": 3.7388442828506183,
+      "learning_rate": 4.263695664723201e-06,
+      "loss": 0.3585,
+      "step": 9940
+    },
+    {
+      "epoch": 0.5742980000577184,
+      "grad_norm": 3.788666515258982,
+      "learning_rate": 4.257922992553253e-06,
+      "loss": 0.3624,
+      "step": 9950
+    },
+    {
+      "epoch": 0.5748751839773744,
+      "grad_norm": 3.882574363808373,
+      "learning_rate": 4.252150320383306e-06,
+      "loss": 0.359,
+      "step": 9960
+    },
+    {
+      "epoch": 0.5754523678970304,
+      "grad_norm": 3.3860959596594764,
+      "learning_rate": 4.246377648213358e-06,
+      "loss": 0.3753,
+      "step": 9970
+    },
+    {
+      "epoch": 0.5760295518166864,
+      "grad_norm": 3.0843751033026936,
+      "learning_rate": 4.240604976043411e-06,
+      "loss": 0.3623,
+      "step": 9980
+    },
+    {
+      "epoch": 0.5766067357363424,
+      "grad_norm": 4.2670578544960165,
+      "learning_rate": 4.234832303873463e-06,
+      "loss": 0.3583,
+      "step": 9990
+    },
+    {
+      "epoch": 0.5771839196559984,
+      "grad_norm": 4.252817625823887,
+      "learning_rate": 4.229059631703516e-06,
+      "loss": 0.3576,
+      "step": 10000
+    },
+    {
+      "epoch": 0.5777611035756544,
+      "grad_norm": 3.407625305823021,
+      "learning_rate": 4.223286959533569e-06,
+      "loss": 0.3552,
+      "step": 10010
+    },
+    {
+      "epoch": 0.5783382874953104,
+      "grad_norm": 3.1679099104052484,
+      "learning_rate": 4.217514287363621e-06,
+      "loss": 0.3738,
+      "step": 10020
+    },
+    {
+      "epoch": 0.5789154714149664,
+      "grad_norm": 3.6787102703539443,
+      "learning_rate": 4.211741615193674e-06,
+      "loss": 0.3578,
+      "step": 10030
+    },
+    {
+      "epoch": 0.5794926553346224,
+      "grad_norm": 5.851925555703329,
+      "learning_rate": 4.205968943023726e-06,
+      "loss": 0.3419,
+      "step": 10040
+    },
+    {
+      "epoch": 0.5800698392542784,
+      "grad_norm": 4.329132073061233,
+      "learning_rate": 4.200196270853779e-06,
+      "loss": 0.3643,
+      "step": 10050
+    },
+    {
+      "epoch": 0.5806470231739344,
+      "grad_norm": 5.997643376925449,
+      "learning_rate": 4.194423598683831e-06,
+      "loss": 0.377,
+      "step": 10060
+    },
+    {
+      "epoch": 0.5812242070935904,
+      "grad_norm": 2.847934197964713,
+      "learning_rate": 4.188650926513883e-06,
+      "loss": 0.3637,
+      "step": 10070
+    },
+    {
+      "epoch": 0.5818013910132463,
+      "grad_norm": 5.914457578502053,
+      "learning_rate": 4.182878254343936e-06,
+      "loss": 0.3674,
+      "step": 10080
+    },
+    {
+      "epoch": 0.5823785749329023,
+      "grad_norm": 2.717688952876076,
+      "learning_rate": 4.1771055821739885e-06,
+      "loss": 0.3445,
+      "step": 10090
+    },
+    {
+      "epoch": 0.5829557588525583,
+      "grad_norm": 4.506246357458738,
+      "learning_rate": 4.1713329100040415e-06,
+      "loss": 0.3596,
+      "step": 10100
+    },
+    {
+      "epoch": 0.5835329427722143,
+      "grad_norm": 4.390097302175211,
+      "learning_rate": 4.165560237834094e-06,
+      "loss": 0.3502,
+      "step": 10110
+    },
+    {
+      "epoch": 0.5841101266918703,
+      "grad_norm": 61.470288646964526,
+      "learning_rate": 4.159787565664146e-06,
+      "loss": 0.364,
+      "step": 10120
+    },
+    {
+      "epoch": 0.5846873106115263,
+      "grad_norm": 4.249154244474576,
+      "learning_rate": 4.154014893494199e-06,
+      "loss": 0.3494,
+      "step": 10130
+    },
+    {
+      "epoch": 0.5852644945311823,
+      "grad_norm": 2.6108964866695956,
+      "learning_rate": 4.148242221324251e-06,
+      "loss": 0.3417,
+      "step": 10140
+    },
+    {
+      "epoch": 0.5858416784508383,
+      "grad_norm": 3.655089863468255,
+      "learning_rate": 4.142469549154304e-06,
+      "loss": 0.3576,
+      "step": 10150
+    },
+    {
+      "epoch": 0.5864188623704943,
+      "grad_norm": 19.653637032520724,
+      "learning_rate": 4.136696876984356e-06,
+      "loss": 0.3603,
+      "step": 10160
+    },
+    {
+      "epoch": 0.5869960462901503,
+      "grad_norm": 4.623567506823282,
+      "learning_rate": 4.1309242048144085e-06,
+      "loss": 0.3486,
+      "step": 10170
+    },
+    {
+      "epoch": 0.5875732302098063,
+      "grad_norm": 5.14547645262892,
+      "learning_rate": 4.1251515326444615e-06,
+      "loss": 0.3591,
+      "step": 10180
+    },
+    {
+      "epoch": 0.5881504141294623,
+      "grad_norm": 20.59966690800242,
+      "learning_rate": 4.119378860474514e-06,
+      "loss": 0.3473,
+      "step": 10190
+    },
+    {
+      "epoch": 0.5887275980491183,
+      "grad_norm": 3.3637726076066796,
+      "learning_rate": 4.113606188304567e-06,
+      "loss": 0.3706,
+      "step": 10200
+    },
+    {
+      "epoch": 0.5893047819687743,
+      "grad_norm": 4.061688985881421,
+      "learning_rate": 4.107833516134619e-06,
+      "loss": 0.3694,
+      "step": 10210
+    },
+    {
+      "epoch": 0.5898819658884303,
+      "grad_norm": 4.941411464322626,
+      "learning_rate": 4.102060843964671e-06,
+      "loss": 0.3551,
+      "step": 10220
+    },
+    {
+      "epoch": 0.5904591498080863,
+      "grad_norm": 4.631050291252514,
+      "learning_rate": 4.096288171794724e-06,
+      "loss": 0.3669,
+      "step": 10230
+    },
+    {
+      "epoch": 0.5910363337277423,
+      "grad_norm": 5.833635533863073,
+      "learning_rate": 4.090515499624776e-06,
+      "loss": 0.354,
+      "step": 10240
+    },
+    {
+      "epoch": 0.5916135176473983,
+      "grad_norm": 17.060908158433886,
+      "learning_rate": 4.084742827454829e-06,
+      "loss": 0.355,
+      "step": 10250
+    },
+    {
+      "epoch": 0.5921907015670543,
+      "grad_norm": 4.438473043902829,
+      "learning_rate": 4.078970155284881e-06,
+      "loss": 0.3559,
+      "step": 10260
+    },
+    {
+      "epoch": 0.5927678854867103,
+      "grad_norm": 4.029441903808598,
+      "learning_rate": 4.0731974831149344e-06,
+      "loss": 0.3456,
+      "step": 10270
+    },
+    {
+      "epoch": 0.5933450694063663,
+      "grad_norm": 4.894018752470052,
+      "learning_rate": 4.067424810944987e-06,
+      "loss": 0.342,
+      "step": 10280
+    },
+    {
+      "epoch": 0.5939222533260223,
+      "grad_norm": 2.403763740670601,
+      "learning_rate": 4.061652138775039e-06,
+      "loss": 0.3653,
+      "step": 10290
+    },
+    {
+      "epoch": 0.5944994372456783,
+      "grad_norm": 4.283731997785048,
+      "learning_rate": 4.055879466605092e-06,
+      "loss": 0.3668,
+      "step": 10300
+    },
+    {
+      "epoch": 0.5950766211653343,
+      "grad_norm": 3.021670467218982,
+      "learning_rate": 4.050106794435144e-06,
+      "loss": 0.3554,
+      "step": 10310
+    },
+    {
+      "epoch": 0.5956538050849903,
+      "grad_norm": 39.77509257463927,
+      "learning_rate": 4.044334122265197e-06,
+      "loss": 0.373,
+      "step": 10320
+    },
+    {
+      "epoch": 0.5962309890046463,
+      "grad_norm": 7.734517576459877,
+      "learning_rate": 4.03856145009525e-06,
+      "loss": 0.3545,
+      "step": 10330
+    },
+    {
+      "epoch": 0.5968081729243023,
+      "grad_norm": 12.164334264143207,
+      "learning_rate": 4.032788777925302e-06,
+      "loss": 0.3593,
+      "step": 10340
+    },
+    {
+      "epoch": 0.5973853568439583,
+      "grad_norm": 6.962481288991228,
+      "learning_rate": 4.027016105755354e-06,
+      "loss": 0.3424,
+      "step": 10350
+    },
+    {
+      "epoch": 0.5979625407636143,
+      "grad_norm": 3.265508104782265,
+      "learning_rate": 4.021243433585407e-06,
+      "loss": 0.3401,
+      "step": 10360
+    },
+    {
+      "epoch": 0.5985397246832703,
+      "grad_norm": 3.169522843364254,
+      "learning_rate": 4.0154707614154596e-06,
+      "loss": 0.3534,
+      "step": 10370
+    },
+    {
+      "epoch": 0.5991169086029263,
+      "grad_norm": 3.2365860904520662,
+      "learning_rate": 4.0096980892455126e-06,
+      "loss": 0.3476,
+      "step": 10380
+    },
+    {
+      "epoch": 0.5996940925225823,
+      "grad_norm": 22.34179305900222,
+      "learning_rate": 4.003925417075565e-06,
+      "loss": 0.3615,
+      "step": 10390
+    },
+    {
+      "epoch": 0.6002712764422383,
+      "grad_norm": 3.6824711768745235,
+      "learning_rate": 3.998152744905617e-06,
+      "loss": 0.3598,
+      "step": 10400
+    },
+    {
+      "epoch": 0.6008484603618943,
+      "grad_norm": 2.244270658998301,
+      "learning_rate": 3.99238007273567e-06,
+      "loss": 0.3602,
+      "step": 10410
+    },
+    {
+      "epoch": 0.6014256442815503,
+      "grad_norm": 2.944264669766013,
+      "learning_rate": 3.986607400565722e-06,
+      "loss": 0.3557,
+      "step": 10420
+    },
+    {
+      "epoch": 0.6020028282012063,
+      "grad_norm": 3.6813564246612893,
+      "learning_rate": 3.980834728395775e-06,
+      "loss": 0.3426,
+      "step": 10430
+    },
+    {
+      "epoch": 0.6025800121208623,
+      "grad_norm": 2.4642472523118193,
+      "learning_rate": 3.975062056225827e-06,
+      "loss": 0.3498,
+      "step": 10440
+    },
+    {
+      "epoch": 0.6031571960405183,
+      "grad_norm": 2.612111770933025,
+      "learning_rate": 3.96928938405588e-06,
+      "loss": 0.3738,
+      "step": 10450
+    },
+    {
+      "epoch": 0.6037343799601743,
+      "grad_norm": 4.989290119921459,
+      "learning_rate": 3.9635167118859325e-06,
+      "loss": 0.3437,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6043115638798303,
+      "grad_norm": 3.191111659552641,
+      "learning_rate": 3.957744039715985e-06,
+      "loss": 0.3513,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6048887477994863,
+      "grad_norm": 6.289514020679802,
+      "learning_rate": 3.951971367546038e-06,
+      "loss": 0.3617,
+      "step": 10480
+    },
+    {
+      "epoch": 0.6054659317191423,
+      "grad_norm": 4.369159045847553,
+      "learning_rate": 3.94619869537609e-06,
+      "loss": 0.3622,
+      "step": 10490
+    },
+    {
+      "epoch": 0.6060431156387983,
+      "grad_norm": 4.272747590300094,
+      "learning_rate": 3.940426023206143e-06,
+      "loss": 0.3679,
+      "step": 10500
+    },
+    {
+      "epoch": 0.6066202995584543,
+      "grad_norm": 6.298266009612924,
+      "learning_rate": 3.934653351036195e-06,
+      "loss": 0.3605,
+      "step": 10510
+    },
+    {
+      "epoch": 0.6071974834781103,
+      "grad_norm": 2.853912711053667,
+      "learning_rate": 3.928880678866247e-06,
+      "loss": 0.3643,
+      "step": 10520
+    },
+    {
+      "epoch": 0.6077746673977663,
+      "grad_norm": 4.905191894605176,
+      "learning_rate": 3.9231080066963e-06,
+      "loss": 0.3566,
+      "step": 10530
+    },
+    {
+      "epoch": 0.6083518513174223,
+      "grad_norm": 3.7179571361092307,
+      "learning_rate": 3.9173353345263524e-06,
+      "loss": 0.3534,
+      "step": 10540
+    },
+    {
+      "epoch": 0.6089290352370783,
+      "grad_norm": 4.720816418264325,
+      "learning_rate": 3.9115626623564055e-06,
+      "loss": 0.3445,
+      "step": 10550
+    },
+    {
+      "epoch": 0.6095062191567343,
+      "grad_norm": 6.0994643547541045,
+      "learning_rate": 3.905789990186458e-06,
+      "loss": 0.3447,
+      "step": 10560
+    },
+    {
+      "epoch": 0.6100834030763903,
+      "grad_norm": 7.143461522640564,
+      "learning_rate": 3.90001731801651e-06,
+      "loss": 0.3486,
+      "step": 10570
+    },
+    {
+      "epoch": 0.6106605869960463,
+      "grad_norm": 3.5865672738484515,
+      "learning_rate": 3.894244645846563e-06,
+      "loss": 0.3543,
+      "step": 10580
+    },
+    {
+      "epoch": 0.6112377709157023,
+      "grad_norm": 3.44671753994167,
+      "learning_rate": 3.888471973676615e-06,
+      "loss": 0.339,
+      "step": 10590
+    },
+    {
+      "epoch": 0.6118149548353583,
+      "grad_norm": 4.037111129069171,
+      "learning_rate": 3.882699301506668e-06,
+      "loss": 0.3542,
+      "step": 10600
+    },
+    {
+      "epoch": 0.6123921387550143,
+      "grad_norm": 2.5068462700876752,
+      "learning_rate": 3.87692662933672e-06,
+      "loss": 0.3612,
+      "step": 10610
+    },
+    {
+      "epoch": 0.6129693226746703,
+      "grad_norm": 2.69916896955261,
+      "learning_rate": 3.871153957166772e-06,
+      "loss": 0.3552,
+      "step": 10620
+    },
+    {
+      "epoch": 0.6135465065943263,
+      "grad_norm": 2.12828690128291,
+      "learning_rate": 3.865381284996825e-06,
+      "loss": 0.3464,
+      "step": 10630
+    },
+    {
+      "epoch": 0.6141236905139823,
+      "grad_norm": 2.4651478648163754,
+      "learning_rate": 3.8596086128268776e-06,
+      "loss": 0.3714,
+      "step": 10640
+    },
+    {
+      "epoch": 0.6147008744336383,
+      "grad_norm": 4.709377859928187,
+      "learning_rate": 3.853835940656931e-06,
+      "loss": 0.3462,
+      "step": 10650
+    },
+    {
+      "epoch": 0.6152780583532943,
+      "grad_norm": 14.878118519317356,
+      "learning_rate": 3.848063268486983e-06,
+      "loss": 0.355,
+      "step": 10660
+    },
+    {
+      "epoch": 0.6158552422729503,
+      "grad_norm": 7.0013585156765314,
+      "learning_rate": 3.842290596317036e-06,
+      "loss": 0.3548,
+      "step": 10670
+    },
+    {
+      "epoch": 0.6164324261926063,
+      "grad_norm": 4.781645383408167,
+      "learning_rate": 3.836517924147088e-06,
+      "loss": 0.3544,
+      "step": 10680
+    },
+    {
+      "epoch": 0.6170096101122623,
+      "grad_norm": 2.5404756093298695,
+      "learning_rate": 3.83074525197714e-06,
+      "loss": 0.3534,
+      "step": 10690
+    },
+    {
+      "epoch": 0.6175867940319183,
+      "grad_norm": 8.95780042415011,
+      "learning_rate": 3.824972579807193e-06,
+      "loss": 0.3636,
+      "step": 10700
+    },
+    {
+      "epoch": 0.6181639779515743,
+      "grad_norm": 4.989641662422552,
+      "learning_rate": 3.819199907637245e-06,
+      "loss": 0.3673,
+      "step": 10710
+    },
+    {
+      "epoch": 0.6187411618712303,
+      "grad_norm": 3.0860360298250096,
+      "learning_rate": 3.813427235467298e-06,
+      "loss": 0.3454,
+      "step": 10720
+    },
+    {
+      "epoch": 0.6193183457908863,
+      "grad_norm": 5.75583484430528,
+      "learning_rate": 3.8076545632973505e-06,
+      "loss": 0.3598,
+      "step": 10730
+    },
+    {
+      "epoch": 0.6198955297105423,
+      "grad_norm": 2.949832246732552,
+      "learning_rate": 3.801881891127403e-06,
+      "loss": 0.3553,
+      "step": 10740
+    },
+    {
+      "epoch": 0.6204727136301983,
+      "grad_norm": 5.670166660495844,
+      "learning_rate": 3.7961092189574557e-06,
+      "loss": 0.3626,
+      "step": 10750
+    },
+    {
+      "epoch": 0.6210498975498543,
+      "grad_norm": 3.972108943307402,
+      "learning_rate": 3.790336546787508e-06,
+      "loss": 0.3473,
+      "step": 10760
+    },
+    {
+      "epoch": 0.6216270814695103,
+      "grad_norm": 6.588272622863319,
+      "learning_rate": 3.7845638746175605e-06,
+      "loss": 0.3504,
+      "step": 10770
+    },
+    {
+      "epoch": 0.6222042653891663,
+      "grad_norm": 10.062313562042537,
+      "learning_rate": 3.778791202447613e-06,
+      "loss": 0.348,
+      "step": 10780
+    },
+    {
+      "epoch": 0.6227814493088223,
+      "grad_norm": 2.393822027910724,
+      "learning_rate": 3.7730185302776657e-06,
+      "loss": 0.332,
+      "step": 10790
+    },
+    {
+      "epoch": 0.6233586332284783,
+      "grad_norm": 3.726451963447983,
+      "learning_rate": 3.7672458581077183e-06,
+      "loss": 0.3544,
+      "step": 10800
+    },
+    {
+      "epoch": 0.6239358171481343,
+      "grad_norm": 3.0713208436951405,
+      "learning_rate": 3.761473185937771e-06,
+      "loss": 0.3569,
+      "step": 10810
+    },
+    {
+      "epoch": 0.6245130010677903,
+      "grad_norm": 8.506793083475245,
+      "learning_rate": 3.755700513767823e-06,
+      "loss": 0.3498,
+      "step": 10820
+    },
+    {
+      "epoch": 0.6250901849874463,
+      "grad_norm": 5.0436471202946205,
+      "learning_rate": 3.7499278415978756e-06,
+      "loss": 0.3539,
+      "step": 10830
+    },
+    {
+      "epoch": 0.6256673689071023,
+      "grad_norm": 3.0241635545445957,
+      "learning_rate": 3.7441551694279282e-06,
+      "loss": 0.3623,
+      "step": 10840
+    },
+    {
+      "epoch": 0.6262445528267583,
+      "grad_norm": 2.4939286036772703,
+      "learning_rate": 3.7383824972579812e-06,
+      "loss": 0.3578,
+      "step": 10850
+    },
+    {
+      "epoch": 0.6268217367464143,
+      "grad_norm": 3.3678230813106373,
+      "learning_rate": 3.732609825088034e-06,
+      "loss": 0.3568,
+      "step": 10860
+    },
+    {
+      "epoch": 0.6273989206660703,
+      "grad_norm": 3.1630251340425795,
+      "learning_rate": 3.7268371529180864e-06,
+      "loss": 0.3488,
+      "step": 10870
+    },
+    {
+      "epoch": 0.6279761045857263,
+      "grad_norm": 4.715668356612278,
+      "learning_rate": 3.721064480748139e-06,
+      "loss": 0.3487,
+      "step": 10880
+    },
+    {
+      "epoch": 0.6285532885053823,
+      "grad_norm": 3.430089287433023,
+      "learning_rate": 3.7152918085781912e-06,
+      "loss": 0.3676,
+      "step": 10890
+    },
+    {
+      "epoch": 0.6291304724250383,
+      "grad_norm": 2.9339747157831546,
+      "learning_rate": 3.709519136408244e-06,
+      "loss": 0.3507,
+      "step": 10900
+    },
+    {
+      "epoch": 0.6297076563446943,
+      "grad_norm": 4.2204605393920485,
+      "learning_rate": 3.7037464642382964e-06,
+      "loss": 0.3446,
+      "step": 10910
+    },
+    {
+      "epoch": 0.6302848402643503,
+      "grad_norm": 2.5323029057405764,
+      "learning_rate": 3.697973792068349e-06,
+      "loss": 0.3507,
+      "step": 10920
+    },
+    {
+      "epoch": 0.6308620241840063,
+      "grad_norm": 1.959685524861653,
+      "learning_rate": 3.6922011198984016e-06,
+      "loss": 0.3551,
+      "step": 10930
+    },
+    {
+      "epoch": 0.6314392081036623,
+      "grad_norm": 2.2201928530131085,
+      "learning_rate": 3.6864284477284538e-06,
+      "loss": 0.3476,
+      "step": 10940
+    },
+    {
+      "epoch": 0.6320163920233183,
+      "grad_norm": 2.5294366254069645,
+      "learning_rate": 3.6806557755585064e-06,
+      "loss": 0.3502,
+      "step": 10950
+    },
+    {
+      "epoch": 0.6325935759429743,
+      "grad_norm": 2.5929823326561103,
+      "learning_rate": 3.674883103388559e-06,
+      "loss": 0.3477,
+      "step": 10960
+    },
+    {
+      "epoch": 0.6331707598626303,
+      "grad_norm": 3.0643397308226903,
+      "learning_rate": 3.6691104312186116e-06,
+      "loss": 0.3511,
+      "step": 10970
+    },
+    {
+      "epoch": 0.6337479437822863,
+      "grad_norm": 3.725143554468828,
+      "learning_rate": 3.663337759048664e-06,
+      "loss": 0.3597,
+      "step": 10980
+    },
+    {
+      "epoch": 0.6343251277019423,
+      "grad_norm": 2.332988363561149,
+      "learning_rate": 3.6575650868787168e-06,
+      "loss": 0.3547,
+      "step": 10990
+    },
+    {
+      "epoch": 0.6349023116215983,
+      "grad_norm": 4.338506151135665,
+      "learning_rate": 3.651792414708769e-06,
+      "loss": 0.3621,
+      "step": 11000
+    },
+    {
+      "epoch": 0.6354794955412543,
+      "grad_norm": 5.853920773449472,
+      "learning_rate": 3.6460197425388215e-06,
+      "loss": 0.3467,
+      "step": 11010
+    },
+    {
+      "epoch": 0.6360566794609103,
+      "grad_norm": 2.9801395957721217,
+      "learning_rate": 3.640247070368874e-06,
+      "loss": 0.3533,
+      "step": 11020
+    },
+    {
+      "epoch": 0.6366338633805663,
+      "grad_norm": 5.428993197624115,
+      "learning_rate": 3.6344743981989267e-06,
+      "loss": 0.3477,
+      "step": 11030
+    },
+    {
+      "epoch": 0.6372110473002223,
+      "grad_norm": 2.1575911965605914,
+      "learning_rate": 3.6287017260289793e-06,
+      "loss": 0.3463,
+      "step": 11040
+    },
+    {
+      "epoch": 0.6377882312198783,
+      "grad_norm": 3.3210877709918982,
+      "learning_rate": 3.622929053859032e-06,
+      "loss": 0.3546,
+      "step": 11050
+    },
+    {
+      "epoch": 0.6383654151395343,
+      "grad_norm": 2.686843207231148,
+      "learning_rate": 3.617156381689084e-06,
+      "loss": 0.3518,
+      "step": 11060
+    },
+    {
+      "epoch": 0.6389425990591903,
+      "grad_norm": 5.280345153851947,
+      "learning_rate": 3.6113837095191367e-06,
+      "loss": 0.3579,
+      "step": 11070
+    },
+    {
+      "epoch": 0.6395197829788463,
+      "grad_norm": 5.403871542937742,
+      "learning_rate": 3.6056110373491893e-06,
+      "loss": 0.3489,
+      "step": 11080
+    },
+    {
+      "epoch": 0.6400969668985023,
+      "grad_norm": 2.9735701887326833,
+      "learning_rate": 3.599838365179242e-06,
+      "loss": 0.3502,
+      "step": 11090
+    },
+    {
+      "epoch": 0.6406741508181583,
+      "grad_norm": 2.6891252705595368,
+      "learning_rate": 3.5940656930092945e-06,
+      "loss": 0.3612,
+      "step": 11100
+    },
+    {
+      "epoch": 0.6412513347378141,
+      "grad_norm": 4.660072834904341,
+      "learning_rate": 3.5882930208393467e-06,
+      "loss": 0.3408,
+      "step": 11110
+    },
+    {
+      "epoch": 0.6418285186574701,
+      "grad_norm": 8.181225664129359,
+      "learning_rate": 3.5825203486693993e-06,
+      "loss": 0.3466,
+      "step": 11120
+    },
+    {
+      "epoch": 0.6424057025771261,
+      "grad_norm": 3.8250794191372943,
+      "learning_rate": 3.576747676499452e-06,
+      "loss": 0.3428,
+      "step": 11130
+    },
+    {
+      "epoch": 0.6429828864967821,
+      "grad_norm": 2.5770691997974975,
+      "learning_rate": 3.5709750043295044e-06,
+      "loss": 0.3617,
+      "step": 11140
+    },
+    {
+      "epoch": 0.6435600704164381,
+      "grad_norm": 3.518147076569533,
+      "learning_rate": 3.565202332159557e-06,
+      "loss": 0.3598,
+      "step": 11150
+    },
+    {
+      "epoch": 0.6441372543360941,
+      "grad_norm": 2.3978173360258332,
+      "learning_rate": 3.5594296599896096e-06,
+      "loss": 0.3486,
+      "step": 11160
+    },
+    {
+      "epoch": 0.6447144382557501,
+      "grad_norm": 4.963521259349147,
+      "learning_rate": 3.553656987819662e-06,
+      "loss": 0.3465,
+      "step": 11170
+    },
+    {
+      "epoch": 0.6452916221754061,
+      "grad_norm": 3.7768523495827,
+      "learning_rate": 3.5478843156497144e-06,
+      "loss": 0.3507,
+      "step": 11180
+    },
+    {
+      "epoch": 0.6458688060950621,
+      "grad_norm": 1.9990854220435814,
+      "learning_rate": 3.542111643479767e-06,
+      "loss": 0.3485,
+      "step": 11190
+    },
+    {
+      "epoch": 0.6464459900147181,
+      "grad_norm": 2.434785484655442,
+      "learning_rate": 3.5363389713098196e-06,
+      "loss": 0.3414,
+      "step": 11200
+    },
+    {
+      "epoch": 0.6470231739343741,
+      "grad_norm": 3.764273559187499,
+      "learning_rate": 3.530566299139872e-06,
+      "loss": 0.3475,
+      "step": 11210
+    },
+    {
+      "epoch": 0.6476003578540301,
+      "grad_norm": 3.063611287505477,
+      "learning_rate": 3.5247936269699244e-06,
+      "loss": 0.3487,
+      "step": 11220
+    },
+    {
+      "epoch": 0.6481775417736861,
+      "grad_norm": 2.7961532657594357,
+      "learning_rate": 3.519020954799977e-06,
+      "loss": 0.3591,
+      "step": 11230
+    },
+    {
+      "epoch": 0.6487547256933421,
+      "grad_norm": 5.675631273424128,
+      "learning_rate": 3.5132482826300296e-06,
+      "loss": 0.3387,
+      "step": 11240
+    },
+    {
+      "epoch": 0.6493319096129981,
+      "grad_norm": 4.038281786465871,
+      "learning_rate": 3.507475610460082e-06,
+      "loss": 0.3544,
+      "step": 11250
+    },
+    {
+      "epoch": 0.6499090935326541,
+      "grad_norm": 19.461560362822837,
+      "learning_rate": 3.5017029382901348e-06,
+      "loss": 0.3612,
+      "step": 11260
+    },
+    {
+      "epoch": 0.6504862774523101,
+      "grad_norm": 2.9170938007747838,
+      "learning_rate": 3.4959302661201874e-06,
+      "loss": 0.3633,
+      "step": 11270
+    },
+    {
+      "epoch": 0.6510634613719661,
+      "grad_norm": 2.100392402638713,
+      "learning_rate": 3.4901575939502395e-06,
+      "loss": 0.3628,
+      "step": 11280
+    },
+    {
+      "epoch": 0.6516406452916221,
+      "grad_norm": 7.466900360838518,
+      "learning_rate": 3.484384921780292e-06,
+      "loss": 0.3544,
+      "step": 11290
+    },
+    {
+      "epoch": 0.6522178292112781,
+      "grad_norm": 2.3522582138412984,
+      "learning_rate": 3.4786122496103447e-06,
+      "loss": 0.3768,
+      "step": 11300
+    },
+    {
+      "epoch": 0.6527950131309341,
+      "grad_norm": 2.0677131586462556,
+      "learning_rate": 3.4728395774403973e-06,
+      "loss": 0.3514,
+      "step": 11310
+    },
+    {
+      "epoch": 0.6533721970505901,
+      "grad_norm": 3.1804108497752943,
+      "learning_rate": 3.46706690527045e-06,
+      "loss": 0.3506,
+      "step": 11320
+    },
+    {
+      "epoch": 0.6539493809702461,
+      "grad_norm": 5.396390003664786,
+      "learning_rate": 3.461294233100502e-06,
+      "loss": 0.3539,
+      "step": 11330
+    },
+    {
+      "epoch": 0.6545265648899021,
+      "grad_norm": 5.1304776342645235,
+      "learning_rate": 3.4555215609305547e-06,
+      "loss": 0.3578,
+      "step": 11340
+    },
+    {
+      "epoch": 0.6551037488095581,
+      "grad_norm": 2.625555244563686,
+      "learning_rate": 3.4497488887606073e-06,
+      "loss": 0.3344,
+      "step": 11350
+    },
+    {
+      "epoch": 0.6556809327292141,
+      "grad_norm": 2.765166389820208,
+      "learning_rate": 3.44397621659066e-06,
+      "loss": 0.3467,
+      "step": 11360
+    },
+    {
+      "epoch": 0.6562581166488701,
+      "grad_norm": 3.1348291760774556,
+      "learning_rate": 3.4382035444207125e-06,
+      "loss": 0.348,
+      "step": 11370
+    },
+    {
+      "epoch": 0.6568353005685261,
+      "grad_norm": 2.6149817919704486,
+      "learning_rate": 3.4324308722507655e-06,
+      "loss": 0.3581,
+      "step": 11380
+    },
+    {
+      "epoch": 0.6574124844881821,
+      "grad_norm": 2.4071247325320084,
+      "learning_rate": 3.426658200080818e-06,
+      "loss": 0.3425,
+      "step": 11390
+    },
+    {
+      "epoch": 0.6579896684078381,
+      "grad_norm": 6.966272376463285,
+      "learning_rate": 3.4208855279108703e-06,
+      "loss": 0.3648,
+      "step": 11400
+    },
+    {
+      "epoch": 0.6585668523274941,
+      "grad_norm": 2.410341785080001,
+      "learning_rate": 3.415112855740923e-06,
+      "loss": 0.3518,
+      "step": 11410
+    },
+    {
+      "epoch": 0.6591440362471501,
+      "grad_norm": 4.052333811947672,
+      "learning_rate": 3.4093401835709755e-06,
+      "loss": 0.3538,
+      "step": 11420
+    },
+    {
+      "epoch": 0.6597212201668061,
+      "grad_norm": 5.112675962153542,
+      "learning_rate": 3.403567511401028e-06,
+      "loss": 0.3462,
+      "step": 11430
+    },
+    {
+      "epoch": 0.6602984040864621,
+      "grad_norm": 2.0711228360250873,
+      "learning_rate": 3.3977948392310807e-06,
+      "loss": 0.3523,
+      "step": 11440
+    },
+    {
+      "epoch": 0.6608755880061181,
+      "grad_norm": 3.0723705312379677,
+      "learning_rate": 3.3920221670611333e-06,
+      "loss": 0.3538,
+      "step": 11450
+    },
+    {
+      "epoch": 0.6614527719257741,
+      "grad_norm": 7.22672591716136,
+      "learning_rate": 3.3862494948911854e-06,
+      "loss": 0.3508,
+      "step": 11460
+    },
+    {
+      "epoch": 0.6620299558454301,
+      "grad_norm": 3.4674616674239447,
+      "learning_rate": 3.380476822721238e-06,
+      "loss": 0.3488,
+      "step": 11470
+    },
+    {
+      "epoch": 0.6626071397650861,
+      "grad_norm": 3.0526763692239602,
+      "learning_rate": 3.3747041505512906e-06,
+      "loss": 0.3413,
+      "step": 11480
+    },
+    {
+      "epoch": 0.6631843236847421,
+      "grad_norm": 8.92552529404141,
+      "learning_rate": 3.3689314783813432e-06,
+      "loss": 0.3559,
+      "step": 11490
+    },
+    {
+      "epoch": 0.6637615076043981,
+      "grad_norm": 7.126184068548845,
+      "learning_rate": 3.363158806211396e-06,
+      "loss": 0.3431,
+      "step": 11500
+    },
+    {
+      "epoch": 0.6643386915240541,
+      "grad_norm": 3.6795312021204993,
+      "learning_rate": 3.357386134041448e-06,
+      "loss": 0.3644,
+      "step": 11510
+    },
+    {
+      "epoch": 0.6649158754437101,
+      "grad_norm": 3.1394487426454765,
+      "learning_rate": 3.3516134618715006e-06,
+      "loss": 0.355,
+      "step": 11520
+    },
+    {
+      "epoch": 0.6654930593633661,
+      "grad_norm": 3.115800444710574,
+      "learning_rate": 3.345840789701553e-06,
+      "loss": 0.3595,
+      "step": 11530
+    },
+    {
+      "epoch": 0.6660702432830221,
+      "grad_norm": 3.04464377321414,
+      "learning_rate": 3.3400681175316058e-06,
+      "loss": 0.3474,
+      "step": 11540
+    },
+    {
+      "epoch": 0.6666474272026781,
+      "grad_norm": 11.414003410738056,
+      "learning_rate": 3.3342954453616584e-06,
+      "loss": 0.344,
+      "step": 11550
+    },
+    {
+      "epoch": 0.6672246111223341,
+      "grad_norm": 7.794881460371124,
+      "learning_rate": 3.328522773191711e-06,
+      "loss": 0.3489,
+      "step": 11560
+    },
+    {
+      "epoch": 0.6678017950419901,
+      "grad_norm": 3.606879802040075,
+      "learning_rate": 3.322750101021763e-06,
+      "loss": 0.3542,
+      "step": 11570
+    },
+    {
+      "epoch": 0.6683789789616461,
+      "grad_norm": 3.271233323948874,
+      "learning_rate": 3.3169774288518158e-06,
+      "loss": 0.3572,
+      "step": 11580
+    },
+    {
+      "epoch": 0.6689561628813021,
+      "grad_norm": 5.312528784803595,
+      "learning_rate": 3.3112047566818683e-06,
+      "loss": 0.344,
+      "step": 11590
+    },
+    {
+      "epoch": 0.6695333468009581,
+      "grad_norm": 4.414037045732359,
+      "learning_rate": 3.305432084511921e-06,
+      "loss": 0.3757,
+      "step": 11600
+    },
+    {
+      "epoch": 0.6701105307206141,
+      "grad_norm": 4.420990727422642,
+      "learning_rate": 3.2996594123419735e-06,
+      "loss": 0.3361,
+      "step": 11610
+    },
+    {
+      "epoch": 0.6706877146402701,
+      "grad_norm": 28.79897728988663,
+      "learning_rate": 3.2938867401720257e-06,
+      "loss": 0.3549,
+      "step": 11620
+    },
+    {
+      "epoch": 0.6712648985599261,
+      "grad_norm": 5.938341510395738,
+      "learning_rate": 3.2881140680020783e-06,
+      "loss": 0.3528,
+      "step": 11630
+    },
+    {
+      "epoch": 0.6718420824795821,
+      "grad_norm": 7.625014010774144,
+      "learning_rate": 3.282341395832131e-06,
+      "loss": 0.3569,
+      "step": 11640
+    },
+    {
+      "epoch": 0.6724192663992381,
+      "grad_norm": 2.627772129382934,
+      "learning_rate": 3.2765687236621835e-06,
+      "loss": 0.341,
+      "step": 11650
+    },
+    {
+      "epoch": 0.6729964503188941,
+      "grad_norm": 6.6462109495436765,
+      "learning_rate": 3.270796051492236e-06,
+      "loss": 0.3432,
+      "step": 11660
+    },
+    {
+      "epoch": 0.6735736342385501,
+      "grad_norm": 4.140894146799749,
+      "learning_rate": 3.2650233793222887e-06,
+      "loss": 0.35,
+      "step": 11670
+    },
+    {
+      "epoch": 0.6741508181582061,
+      "grad_norm": 7.023320528238819,
+      "learning_rate": 3.259250707152341e-06,
+      "loss": 0.3483,
+      "step": 11680
+    },
+    {
+      "epoch": 0.6747280020778621,
+      "grad_norm": 3.56371765942958,
+      "learning_rate": 3.2534780349823935e-06,
+      "loss": 0.3456,
+      "step": 11690
+    },
+    {
+      "epoch": 0.6753051859975181,
+      "grad_norm": 4.693701665628699,
+      "learning_rate": 3.247705362812446e-06,
+      "loss": 0.3556,
+      "step": 11700
+    },
+    {
+      "epoch": 0.6758823699171741,
+      "grad_norm": 4.8769232133317955,
+      "learning_rate": 3.2419326906424987e-06,
+      "loss": 0.3406,
+      "step": 11710
+    },
+    {
+      "epoch": 0.6764595538368301,
+      "grad_norm": 14.213756604351863,
+      "learning_rate": 3.2361600184725513e-06,
+      "loss": 0.3422,
+      "step": 11720
+    },
+    {
+      "epoch": 0.6770367377564861,
+      "grad_norm": 4.009649954155962,
+      "learning_rate": 3.2303873463026034e-06,
+      "loss": 0.3449,
+      "step": 11730
+    },
+    {
+      "epoch": 0.6776139216761421,
+      "grad_norm": 3.538273145119479,
+      "learning_rate": 3.224614674132656e-06,
+      "loss": 0.3457,
+      "step": 11740
+    },
+    {
+      "epoch": 0.6781911055957981,
+      "grad_norm": 3.0329790960952026,
+      "learning_rate": 3.2188420019627086e-06,
+      "loss": 0.3479,
+      "step": 11750
+    },
+    {
+      "epoch": 0.6787682895154541,
+      "grad_norm": 8.21491132526687,
+      "learning_rate": 3.2130693297927612e-06,
+      "loss": 0.3577,
+      "step": 11760
+    },
+    {
+      "epoch": 0.6793454734351101,
+      "grad_norm": 4.981836103874383,
+      "learning_rate": 3.207296657622814e-06,
+      "loss": 0.355,
+      "step": 11770
+    },
+    {
+      "epoch": 0.6799226573547661,
+      "grad_norm": 5.1547852515451975,
+      "learning_rate": 3.2015239854528664e-06,
+      "loss": 0.3392,
+      "step": 11780
+    },
+    {
+      "epoch": 0.6804998412744221,
+      "grad_norm": 9.811331834930291,
+      "learning_rate": 3.1957513132829186e-06,
+      "loss": 0.3545,
+      "step": 11790
+    },
+    {
+      "epoch": 0.6810770251940781,
+      "grad_norm": 4.083480395202693,
+      "learning_rate": 3.189978641112971e-06,
+      "loss": 0.3575,
+      "step": 11800
+    },
+    {
+      "epoch": 0.6816542091137341,
+      "grad_norm": 4.764988431769556,
+      "learning_rate": 3.184205968943024e-06,
+      "loss": 0.3372,
+      "step": 11810
+    },
+    {
+      "epoch": 0.6822313930333901,
+      "grad_norm": 6.794798971465098,
+      "learning_rate": 3.1784332967730764e-06,
+      "loss": 0.3539,
+      "step": 11820
+    },
+    {
+      "epoch": 0.6828085769530461,
+      "grad_norm": 7.111672589883507,
+      "learning_rate": 3.172660624603129e-06,
+      "loss": 0.3544,
+      "step": 11830
+    },
+    {
+      "epoch": 0.6833857608727021,
+      "grad_norm": 24.124344459801147,
+      "learning_rate": 3.166887952433181e-06,
+      "loss": 0.3602,
+      "step": 11840
+    },
+    {
+      "epoch": 0.6839629447923581,
+      "grad_norm": 12.424128663469016,
+      "learning_rate": 3.1611152802632338e-06,
+      "loss": 0.3441,
+      "step": 11850
+    },
+    {
+      "epoch": 0.6845401287120141,
+      "grad_norm": 3.903963843774075,
+      "learning_rate": 3.1553426080932864e-06,
+      "loss": 0.3572,
+      "step": 11860
+    },
+    {
+      "epoch": 0.6851173126316701,
+      "grad_norm": 7.5954443349430525,
+      "learning_rate": 3.149569935923339e-06,
+      "loss": 0.3459,
+      "step": 11870
+    },
+    {
+      "epoch": 0.6856944965513261,
+      "grad_norm": 5.147769651660261,
+      "learning_rate": 3.1437972637533915e-06,
+      "loss": 0.3443,
+      "step": 11880
+    },
+    {
+      "epoch": 0.6862716804709821,
+      "grad_norm": 8.531813464808828,
+      "learning_rate": 3.138024591583444e-06,
+      "loss": 0.3427,
+      "step": 11890
+    },
+    {
+      "epoch": 0.6868488643906381,
+      "grad_norm": 9.203737344622347,
+      "learning_rate": 3.1322519194134963e-06,
+      "loss": 0.3466,
+      "step": 11900
+    },
+    {
+      "epoch": 0.6874260483102941,
+      "grad_norm": 4.1913039016792055,
+      "learning_rate": 3.1264792472435498e-06,
+      "loss": 0.3667,
+      "step": 11910
+    },
+    {
+      "epoch": 0.6880032322299501,
+      "grad_norm": 4.984073528450747,
+      "learning_rate": 3.120706575073602e-06,
+      "loss": 0.3417,
+      "step": 11920
+    },
+    {
+      "epoch": 0.6885804161496061,
+      "grad_norm": 4.3895825229927725,
+      "learning_rate": 3.1149339029036545e-06,
+      "loss": 0.3341,
+      "step": 11930
+    },
+    {
+      "epoch": 0.6891576000692621,
+      "grad_norm": 5.323675472371107,
+      "learning_rate": 3.109161230733707e-06,
+      "loss": 0.3341,
+      "step": 11940
+    },
+    {
+      "epoch": 0.6897347839889181,
+      "grad_norm": 4.263152619263457,
+      "learning_rate": 3.1033885585637597e-06,
+      "loss": 0.3463,
+      "step": 11950
+    },
+    {
+      "epoch": 0.6903119679085741,
+      "grad_norm": 6.116151443830828,
+      "learning_rate": 3.0976158863938123e-06,
+      "loss": 0.3603,
+      "step": 11960
+    },
+    {
+      "epoch": 0.6908891518282301,
+      "grad_norm": 10.96676225065857,
+      "learning_rate": 3.0918432142238645e-06,
+      "loss": 0.3317,
+      "step": 11970
+    },
+    {
+      "epoch": 0.6914663357478861,
+      "grad_norm": 5.204338558399774,
+      "learning_rate": 3.086070542053917e-06,
+      "loss": 0.3432,
+      "step": 11980
+    },
+    {
+      "epoch": 0.6920435196675421,
+      "grad_norm": 3.481643749818502,
+      "learning_rate": 3.0802978698839697e-06,
+      "loss": 0.3343,
+      "step": 11990
+    },
+    {
+      "epoch": 0.6926207035871981,
+      "grad_norm": 5.422983882620972,
+      "learning_rate": 3.0745251977140223e-06,
+      "loss": 0.3444,
+      "step": 12000
+    },
+    {
+      "epoch": 0.6931978875068541,
+      "grad_norm": 3.293726670681602,
+      "learning_rate": 3.068752525544075e-06,
+      "loss": 0.3536,
+      "step": 12010
+    },
+    {
+      "epoch": 0.6937750714265101,
+      "grad_norm": 11.162281611641948,
+      "learning_rate": 3.0629798533741275e-06,
+      "loss": 0.3461,
+      "step": 12020
+    },
+    {
+      "epoch": 0.6943522553461661,
+      "grad_norm": 4.538803769431588,
+      "learning_rate": 3.0572071812041797e-06,
+      "loss": 0.3404,
+      "step": 12030
+    },
+    {
+      "epoch": 0.6949294392658221,
+      "grad_norm": 7.297266576912264,
+      "learning_rate": 3.0514345090342323e-06,
+      "loss": 0.3399,
+      "step": 12040
+    },
+    {
+      "epoch": 0.6955066231854781,
+      "grad_norm": 4.498082885030529,
+      "learning_rate": 3.045661836864285e-06,
+      "loss": 0.3572,
+      "step": 12050
+    },
+    {
+      "epoch": 0.6960838071051341,
+      "grad_norm": 6.71445000313715,
+      "learning_rate": 3.0398891646943374e-06,
+      "loss": 0.3456,
+      "step": 12060
+    },
+    {
+      "epoch": 0.6966609910247901,
+      "grad_norm": 4.130838744263147,
+      "learning_rate": 3.03411649252439e-06,
+      "loss": 0.3382,
+      "step": 12070
+    },
+    {
+      "epoch": 0.6972381749444461,
+      "grad_norm": 2.959998512168581,
+      "learning_rate": 3.0283438203544422e-06,
+      "loss": 0.3441,
+      "step": 12080
+    },
+    {
+      "epoch": 0.6978153588641021,
+      "grad_norm": 8.68519096842326,
+      "learning_rate": 3.022571148184495e-06,
+      "loss": 0.3536,
+      "step": 12090
+    },
+    {
+      "epoch": 0.6983925427837581,
+      "grad_norm": 6.068123748807202,
+      "learning_rate": 3.0167984760145474e-06,
+      "loss": 0.3336,
+      "step": 12100
+    },
+    {
+      "epoch": 0.6989697267034141,
+      "grad_norm": 13.720643945389472,
+      "learning_rate": 3.0110258038446e-06,
+      "loss": 0.3453,
+      "step": 12110
+    },
+    {
+      "epoch": 0.6995469106230701,
+      "grad_norm": 5.2345993949285115,
+      "learning_rate": 3.0052531316746526e-06,
+      "loss": 0.3514,
+      "step": 12120
+    },
+    {
+      "epoch": 0.7001240945427261,
+      "grad_norm": 5.995140869193482,
+      "learning_rate": 2.999480459504705e-06,
+      "loss": 0.3465,
+      "step": 12130
+    },
+    {
+      "epoch": 0.7007012784623821,
+      "grad_norm": 5.534508344244959,
+      "learning_rate": 2.9937077873347574e-06,
+      "loss": 0.3436,
+      "step": 12140
+    },
+    {
+      "epoch": 0.7012784623820381,
+      "grad_norm": 3.837652212059965,
+      "learning_rate": 2.98793511516481e-06,
+      "loss": 0.347,
+      "step": 12150
+    },
+    {
+      "epoch": 0.7018556463016941,
+      "grad_norm": 32.35178311186503,
+      "learning_rate": 2.9821624429948626e-06,
+      "loss": 0.3543,
+      "step": 12160
+    },
+    {
+      "epoch": 0.7024328302213501,
+      "grad_norm": 4.1065566840261125,
+      "learning_rate": 2.976389770824915e-06,
+      "loss": 0.3421,
+      "step": 12170
+    },
+    {
+      "epoch": 0.7030100141410061,
+      "grad_norm": 6.144592670774153,
+      "learning_rate": 2.9706170986549678e-06,
+      "loss": 0.3466,
+      "step": 12180
+    },
+    {
+      "epoch": 0.7035871980606621,
+      "grad_norm": 5.684740566371751,
+      "learning_rate": 2.96484442648502e-06,
+      "loss": 0.349,
+      "step": 12190
+    },
+    {
+      "epoch": 0.704164381980318,
+      "grad_norm": 5.234589933221641,
+      "learning_rate": 2.9590717543150725e-06,
+      "loss": 0.3221,
+      "step": 12200
+    },
+    {
+      "epoch": 0.704741565899974,
+      "grad_norm": 5.9090591405034205,
+      "learning_rate": 2.953299082145125e-06,
+      "loss": 0.3461,
+      "step": 12210
+    },
+    {
+      "epoch": 0.70531874981963,
+      "grad_norm": 5.161621785333446,
+      "learning_rate": 2.9475264099751777e-06,
+      "loss": 0.3343,
+      "step": 12220
+    },
+    {
+      "epoch": 0.705895933739286,
+      "grad_norm": 7.454875045097898,
+      "learning_rate": 2.9417537378052303e-06,
+      "loss": 0.339,
+      "step": 12230
+    },
+    {
+      "epoch": 0.706473117658942,
+      "grad_norm": 3.3533355785311936,
+      "learning_rate": 2.935981065635283e-06,
+      "loss": 0.3226,
+      "step": 12240
+    },
+    {
+      "epoch": 0.707050301578598,
+      "grad_norm": 3.3526244501016507,
+      "learning_rate": 2.930208393465335e-06,
+      "loss": 0.3525,
+      "step": 12250
+    },
+    {
+      "epoch": 0.707627485498254,
+      "grad_norm": 46.63609588889749,
+      "learning_rate": 2.9244357212953877e-06,
+      "loss": 0.3388,
+      "step": 12260
+    },
+    {
+      "epoch": 0.70820466941791,
+      "grad_norm": 6.343222491694745,
+      "learning_rate": 2.9186630491254403e-06,
+      "loss": 0.3458,
+      "step": 12270
+    },
+    {
+      "epoch": 0.708781853337566,
+      "grad_norm": 7.406012410848603,
+      "learning_rate": 2.912890376955493e-06,
+      "loss": 0.3523,
+      "step": 12280
+    },
+    {
+      "epoch": 0.709359037257222,
+      "grad_norm": 4.391956311756113,
+      "learning_rate": 2.9071177047855455e-06,
+      "loss": 0.3462,
+      "step": 12290
+    },
+    {
+      "epoch": 0.709936221176878,
+      "grad_norm": 7.365023031476813,
+      "learning_rate": 2.9013450326155977e-06,
+      "loss": 0.3507,
+      "step": 12300
+    },
+    {
+      "epoch": 0.710513405096534,
+      "grad_norm": 8.355149496371373,
+      "learning_rate": 2.8955723604456503e-06,
+      "loss": 0.3516,
+      "step": 12310
+    },
+    {
+      "epoch": 0.71109058901619,
+      "grad_norm": 4.12597154889129,
+      "learning_rate": 2.889799688275703e-06,
+      "loss": 0.3387,
+      "step": 12320
+    },
+    {
+      "epoch": 0.711667772935846,
+      "grad_norm": 22.783055812157006,
+      "learning_rate": 2.8840270161057555e-06,
+      "loss": 0.3505,
+      "step": 12330
+    },
+    {
+      "epoch": 0.712244956855502,
+      "grad_norm": 6.598877289364409,
+      "learning_rate": 2.878254343935808e-06,
+      "loss": 0.3511,
+      "step": 12340
+    },
+    {
+      "epoch": 0.712822140775158,
+      "grad_norm": 6.658331270547365,
+      "learning_rate": 2.8724816717658606e-06,
+      "loss": 0.3472,
+      "step": 12350
+    },
+    {
+      "epoch": 0.713399324694814,
+      "grad_norm": 4.249593904236529,
+      "learning_rate": 2.866708999595913e-06,
+      "loss": 0.3462,
+      "step": 12360
+    },
+    {
+      "epoch": 0.71397650861447,
+      "grad_norm": 4.820541412534286,
+      "learning_rate": 2.8609363274259654e-06,
+      "loss": 0.3327,
+      "step": 12370
+    },
+    {
+      "epoch": 0.714553692534126,
+      "grad_norm": 29.19756399575411,
+      "learning_rate": 2.855163655256018e-06,
+      "loss": 0.3423,
+      "step": 12380
+    },
+    {
+      "epoch": 0.715130876453782,
+      "grad_norm": 2.5170693039148695,
+      "learning_rate": 2.8493909830860706e-06,
+      "loss": 0.3507,
+      "step": 12390
+    },
+    {
+      "epoch": 0.715708060373438,
+      "grad_norm": 4.848573397975011,
+      "learning_rate": 2.843618310916123e-06,
+      "loss": 0.3614,
+      "step": 12400
+    },
+    {
+      "epoch": 0.716285244293094,
+      "grad_norm": 8.083798553592858,
+      "learning_rate": 2.8378456387461754e-06,
+      "loss": 0.3444,
+      "step": 12410
+    },
+    {
+      "epoch": 0.71686242821275,
+      "grad_norm": 6.013516986737268,
+      "learning_rate": 2.832072966576228e-06,
+      "loss": 0.3411,
+      "step": 12420
+    },
+    {
+      "epoch": 0.717439612132406,
+      "grad_norm": 58.40825846546123,
+      "learning_rate": 2.826300294406281e-06,
+      "loss": 0.3639,
+      "step": 12430
+    },
+    {
+      "epoch": 0.718016796052062,
+      "grad_norm": 3.486306193338465,
+      "learning_rate": 2.8205276222363336e-06,
+      "loss": 0.3508,
+      "step": 12440
+    },
+    {
+      "epoch": 0.718593979971718,
+      "grad_norm": 11.117203715725822,
+      "learning_rate": 2.814754950066386e-06,
+      "loss": 0.3441,
+      "step": 12450
+    },
+    {
+      "epoch": 0.719171163891374,
+      "grad_norm": 6.959568782940119,
+      "learning_rate": 2.808982277896439e-06,
+      "loss": 0.355,
+      "step": 12460
+    },
+    {
+      "epoch": 0.71974834781103,
+      "grad_norm": 4.77575740578489,
+      "learning_rate": 2.8032096057264914e-06,
+      "loss": 0.3634,
+      "step": 12470
+    },
+    {
+      "epoch": 0.720325531730686,
+      "grad_norm": 12.450518772632652,
+      "learning_rate": 2.797436933556544e-06,
+      "loss": 0.3454,
+      "step": 12480
+    },
+    {
+      "epoch": 0.720902715650342,
+      "grad_norm": 4.606989122142115,
+      "learning_rate": 2.791664261386596e-06,
+      "loss": 0.338,
+      "step": 12490
+    },
+    {
+      "epoch": 0.721479899569998,
+      "grad_norm": 4.456459932783003,
+      "learning_rate": 2.7858915892166488e-06,
+      "loss": 0.3474,
+      "step": 12500
+    },
+    {
+      "epoch": 0.722057083489654,
+      "grad_norm": 4.348978440263518,
+      "learning_rate": 2.7801189170467014e-06,
+      "loss": 0.3488,
+      "step": 12510
+    },
+    {
+      "epoch": 0.72263426740931,
+      "grad_norm": 3.220553815408838,
+      "learning_rate": 2.774346244876754e-06,
+      "loss": 0.3372,
+      "step": 12520
+    },
+    {
+      "epoch": 0.723211451328966,
+      "grad_norm": 6.662149536726665,
+      "learning_rate": 2.7685735727068065e-06,
+      "loss": 0.3487,
+      "step": 12530
+    },
+    {
+      "epoch": 0.723788635248622,
+      "grad_norm": 4.489888078931735,
+      "learning_rate": 2.7628009005368587e-06,
+      "loss": 0.3589,
+      "step": 12540
+    },
+    {
+      "epoch": 0.724365819168278,
+      "grad_norm": 3.911221624507832,
+      "learning_rate": 2.7570282283669113e-06,
+      "loss": 0.3364,
+      "step": 12550
+    },
+    {
+      "epoch": 0.724943003087934,
+      "grad_norm": 4.397535770990173,
+      "learning_rate": 2.751255556196964e-06,
+      "loss": 0.3468,
+      "step": 12560
+    },
+    {
+      "epoch": 0.72552018700759,
+      "grad_norm": 4.138162758687126,
+      "learning_rate": 2.7454828840270165e-06,
+      "loss": 0.351,
+      "step": 12570
+    },
+    {
+      "epoch": 0.726097370927246,
+      "grad_norm": 3.6393806674007396,
+      "learning_rate": 2.739710211857069e-06,
+      "loss": 0.3439,
+      "step": 12580
+    },
+    {
+      "epoch": 0.7266745548469019,
+      "grad_norm": 10.959014317841367,
+      "learning_rate": 2.7339375396871217e-06,
+      "loss": 0.3466,
+      "step": 12590
+    },
+    {
+      "epoch": 0.7272517387665579,
+      "grad_norm": 8.5956598139555,
+      "learning_rate": 2.728164867517174e-06,
+      "loss": 0.3474,
+      "step": 12600
+    },
+    {
+      "epoch": 0.7278289226862139,
+      "grad_norm": 8.199922135307672,
+      "learning_rate": 2.7223921953472265e-06,
+      "loss": 0.3334,
+      "step": 12610
+    },
+    {
+      "epoch": 0.7284061066058699,
+      "grad_norm": 3.4555204304735563,
+      "learning_rate": 2.716619523177279e-06,
+      "loss": 0.3437,
+      "step": 12620
+    },
+    {
+      "epoch": 0.7289832905255259,
+      "grad_norm": 4.838169983017387,
+      "learning_rate": 2.7108468510073317e-06,
+      "loss": 0.3444,
+      "step": 12630
+    },
+    {
+      "epoch": 0.7295604744451819,
+      "grad_norm": 3.294894158340646,
+      "learning_rate": 2.7050741788373843e-06,
+      "loss": 0.345,
+      "step": 12640
+    },
+    {
+      "epoch": 0.7301376583648379,
+      "grad_norm": 3.886142779724859,
+      "learning_rate": 2.6993015066674364e-06,
+      "loss": 0.34,
+      "step": 12650
+    },
+    {
+      "epoch": 0.7307148422844939,
+      "grad_norm": 3.1827555908179814,
+      "learning_rate": 2.693528834497489e-06,
+      "loss": 0.3383,
+      "step": 12660
+    },
+    {
+      "epoch": 0.7312920262041499,
+      "grad_norm": 20.97526772421675,
+      "learning_rate": 2.6877561623275416e-06,
+      "loss": 0.3416,
+      "step": 12670
+    },
+    {
+      "epoch": 0.7318692101238059,
+      "grad_norm": 3.474528368008189,
+      "learning_rate": 2.6819834901575942e-06,
+      "loss": 0.3359,
+      "step": 12680
+    },
+    {
+      "epoch": 0.7324463940434619,
+      "grad_norm": 11.34967364860296,
+      "learning_rate": 2.676210817987647e-06,
+      "loss": 0.3581,
+      "step": 12690
+    },
+    {
+      "epoch": 0.7330235779631179,
+      "grad_norm": 3.495369176137086,
+      "learning_rate": 2.6704381458176994e-06,
+      "loss": 0.3369,
+      "step": 12700
+    },
+    {
+      "epoch": 0.7336007618827739,
+      "grad_norm": 4.744362874719428,
+      "learning_rate": 2.6646654736477516e-06,
+      "loss": 0.3502,
+      "step": 12710
+    },
+    {
+      "epoch": 0.7341779458024299,
+      "grad_norm": 2.554299735229823,
+      "learning_rate": 2.658892801477804e-06,
+      "loss": 0.3547,
+      "step": 12720
+    },
+    {
+      "epoch": 0.7347551297220859,
+      "grad_norm": 4.114324789692039,
+      "learning_rate": 2.653120129307857e-06,
+      "loss": 0.3406,
+      "step": 12730
+    },
+    {
+      "epoch": 0.7353323136417419,
+      "grad_norm": 5.99465461001967,
+      "learning_rate": 2.6473474571379094e-06,
+      "loss": 0.3568,
+      "step": 12740
+    },
+    {
+      "epoch": 0.7359094975613979,
+      "grad_norm": 5.08850145360636,
+      "learning_rate": 2.641574784967962e-06,
+      "loss": 0.3437,
+      "step": 12750
+    },
+    {
+      "epoch": 0.7364866814810539,
+      "grad_norm": 2.901320860603511,
+      "learning_rate": 2.635802112798014e-06,
+      "loss": 0.334,
+      "step": 12760
+    },
+    {
+      "epoch": 0.7370638654007099,
+      "grad_norm": 7.740697658957906,
+      "learning_rate": 2.6300294406280668e-06,
+      "loss": 0.3361,
+      "step": 12770
+    },
+    {
+      "epoch": 0.7376410493203659,
+      "grad_norm": 6.245600892762995,
+      "learning_rate": 2.6242567684581194e-06,
+      "loss": 0.3308,
+      "step": 12780
+    },
+    {
+      "epoch": 0.7382182332400219,
+      "grad_norm": 5.308126429822944,
+      "learning_rate": 2.618484096288172e-06,
+      "loss": 0.3527,
+      "step": 12790
+    },
+    {
+      "epoch": 0.7387954171596779,
+      "grad_norm": 4.212291401570202,
+      "learning_rate": 2.6127114241182245e-06,
+      "loss": 0.3409,
+      "step": 12800
+    },
+    {
+      "epoch": 0.7393726010793339,
+      "grad_norm": 3.7060833092802556,
+      "learning_rate": 2.606938751948277e-06,
+      "loss": 0.337,
+      "step": 12810
+    },
+    {
+      "epoch": 0.7399497849989899,
+      "grad_norm": 13.303775807909831,
+      "learning_rate": 2.6011660797783293e-06,
+      "loss": 0.3521,
+      "step": 12820
+    },
+    {
+      "epoch": 0.7405269689186459,
+      "grad_norm": 5.855236173414753,
+      "learning_rate": 2.595393407608382e-06,
+      "loss": 0.344,
+      "step": 12830
+    },
+    {
+      "epoch": 0.7411041528383019,
+      "grad_norm": 2.862607162292994,
+      "learning_rate": 2.5896207354384345e-06,
+      "loss": 0.3386,
+      "step": 12840
+    },
+    {
+      "epoch": 0.7416813367579579,
+      "grad_norm": 3.190707127525178,
+      "learning_rate": 2.583848063268487e-06,
+      "loss": 0.3451,
+      "step": 12850
+    },
+    {
+      "epoch": 0.7422585206776139,
+      "grad_norm": 7.479465788278306,
+      "learning_rate": 2.5780753910985397e-06,
+      "loss": 0.3418,
+      "step": 12860
+    },
+    {
+      "epoch": 0.7428357045972699,
+      "grad_norm": 3.9314452580240795,
+      "learning_rate": 2.572302718928592e-06,
+      "loss": 0.3558,
+      "step": 12870
+    },
+    {
+      "epoch": 0.7434128885169259,
+      "grad_norm": 2.586918485101635,
+      "learning_rate": 2.5665300467586445e-06,
+      "loss": 0.3521,
+      "step": 12880
+    },
+    {
+      "epoch": 0.7439900724365819,
+      "grad_norm": 5.285658124727487,
+      "learning_rate": 2.560757374588697e-06,
+      "loss": 0.3467,
+      "step": 12890
+    },
+    {
+      "epoch": 0.7445672563562379,
+      "grad_norm": 16.29975276837285,
+      "learning_rate": 2.5549847024187497e-06,
+      "loss": 0.338,
+      "step": 12900
+    },
+    {
+      "epoch": 0.7451444402758939,
+      "grad_norm": 5.18800369782506,
+      "learning_rate": 2.5492120302488023e-06,
+      "loss": 0.3423,
+      "step": 12910
+    },
+    {
+      "epoch": 0.7457216241955499,
+      "grad_norm": 2.7621346083831737,
+      "learning_rate": 2.543439358078855e-06,
+      "loss": 0.3474,
+      "step": 12920
+    },
+    {
+      "epoch": 0.7462988081152059,
+      "grad_norm": 9.136730427994907,
+      "learning_rate": 2.537666685908907e-06,
+      "loss": 0.3599,
+      "step": 12930
+    },
+    {
+      "epoch": 0.7468759920348619,
+      "grad_norm": 4.494380679642752,
+      "learning_rate": 2.5318940137389596e-06,
+      "loss": 0.3509,
+      "step": 12940
+    },
+    {
+      "epoch": 0.7474531759545179,
+      "grad_norm": 3.8278742483492554,
+      "learning_rate": 2.5261213415690122e-06,
+      "loss": 0.3419,
+      "step": 12950
+    },
+    {
+      "epoch": 0.7480303598741739,
+      "grad_norm": 13.52933963754357,
+      "learning_rate": 2.5203486693990653e-06,
+      "loss": 0.3255,
+      "step": 12960
+    },
+    {
+      "epoch": 0.7486075437938299,
+      "grad_norm": 4.13788172504123,
+      "learning_rate": 2.514575997229118e-06,
+      "loss": 0.3446,
+      "step": 12970
+    },
+    {
+      "epoch": 0.7491847277134859,
+      "grad_norm": 6.859750860335564,
+      "learning_rate": 2.5088033250591704e-06,
+      "loss": 0.3586,
+      "step": 12980
+    },
+    {
+      "epoch": 0.7497619116331419,
+      "grad_norm": 9.235304443739427,
+      "learning_rate": 2.503030652889223e-06,
+      "loss": 0.3393,
+      "step": 12990
+    },
+    {
+      "epoch": 0.7503390955527979,
+      "grad_norm": 5.829087177972612,
+      "learning_rate": 2.4972579807192752e-06,
+      "loss": 0.3518,
+      "step": 13000
+    },
+    {
+      "epoch": 0.7509162794724539,
+      "grad_norm": 4.209497926424704,
+      "learning_rate": 2.491485308549328e-06,
+      "loss": 0.3392,
+      "step": 13010
+    },
+    {
+      "epoch": 0.7514934633921099,
+      "grad_norm": 7.292629657573057,
+      "learning_rate": 2.48571263637938e-06,
+      "loss": 0.3253,
+      "step": 13020
+    },
+    {
+      "epoch": 0.7520706473117659,
+      "grad_norm": 2.647834968655611,
+      "learning_rate": 2.4799399642094326e-06,
+      "loss": 0.3304,
+      "step": 13030
+    },
+    {
+      "epoch": 0.7526478312314219,
+      "grad_norm": 3.2625519939417638,
+      "learning_rate": 2.474167292039485e-06,
+      "loss": 0.3389,
+      "step": 13040
+    },
+    {
+      "epoch": 0.7532250151510779,
+      "grad_norm": 11.316624355429083,
+      "learning_rate": 2.4683946198695378e-06,
+      "loss": 0.3369,
+      "step": 13050
+    },
+    {
+      "epoch": 0.7538021990707339,
+      "grad_norm": 9.673257149513491,
+      "learning_rate": 2.4626219476995904e-06,
+      "loss": 0.3334,
+      "step": 13060
+    },
+    {
+      "epoch": 0.7543793829903899,
+      "grad_norm": 4.342384289505803,
+      "learning_rate": 2.4568492755296426e-06,
+      "loss": 0.3464,
+      "step": 13070
+    },
+    {
+      "epoch": 0.7549565669100459,
+      "grad_norm": 11.18476020427808,
+      "learning_rate": 2.451076603359695e-06,
+      "loss": 0.3427,
+      "step": 13080
+    },
+    {
+      "epoch": 0.7555337508297019,
+      "grad_norm": 3.7726393590227274,
+      "learning_rate": 2.445303931189748e-06,
+      "loss": 0.334,
+      "step": 13090
+    },
+    {
+      "epoch": 0.7561109347493579,
+      "grad_norm": 12.714136611436194,
+      "learning_rate": 2.4395312590198008e-06,
+      "loss": 0.3268,
+      "step": 13100
+    },
+    {
+      "epoch": 0.7566881186690139,
+      "grad_norm": 2.1269070192971307,
+      "learning_rate": 2.433758586849853e-06,
+      "loss": 0.3441,
+      "step": 13110
+    },
+    {
+      "epoch": 0.7572653025886699,
+      "grad_norm": 6.054989937046523,
+      "learning_rate": 2.4279859146799055e-06,
+      "loss": 0.3428,
+      "step": 13120
+    },
+    {
+      "epoch": 0.7578424865083259,
+      "grad_norm": 4.270071372155915,
+      "learning_rate": 2.422213242509958e-06,
+      "loss": 0.3539,
+      "step": 13130
+    },
+    {
+      "epoch": 0.7584196704279819,
+      "grad_norm": 3.1183499805433814,
+      "learning_rate": 2.4164405703400107e-06,
+      "loss": 0.3347,
+      "step": 13140
+    },
+    {
+      "epoch": 0.7589968543476379,
+      "grad_norm": 8.120134607992474,
+      "learning_rate": 2.4106678981700633e-06,
+      "loss": 0.3513,
+      "step": 13150
+    },
+    {
+      "epoch": 0.7595740382672939,
+      "grad_norm": 4.275672737070938,
+      "learning_rate": 2.4048952260001155e-06,
+      "loss": 0.3333,
+      "step": 13160
+    },
+    {
+      "epoch": 0.7601512221869499,
+      "grad_norm": 4.309409121092216,
+      "learning_rate": 2.399122553830168e-06,
+      "loss": 0.342,
+      "step": 13170
+    },
+    {
+      "epoch": 0.7607284061066059,
+      "grad_norm": 14.975672326288917,
+      "learning_rate": 2.3933498816602207e-06,
+      "loss": 0.3491,
+      "step": 13180
+    },
+    {
+      "epoch": 0.7613055900262619,
+      "grad_norm": 3.3391093670284984,
+      "learning_rate": 2.3875772094902733e-06,
+      "loss": 0.3377,
+      "step": 13190
+    },
+    {
+      "epoch": 0.7618827739459179,
+      "grad_norm": 9.934238745041949,
+      "learning_rate": 2.381804537320326e-06,
+      "loss": 0.3316,
+      "step": 13200
+    },
+    {
+      "epoch": 0.7624599578655739,
+      "grad_norm": 3.323343149796346,
+      "learning_rate": 2.3760318651503785e-06,
+      "loss": 0.3217,
+      "step": 13210
+    },
+    {
+      "epoch": 0.7630371417852299,
+      "grad_norm": 5.084030929299541,
+      "learning_rate": 2.3702591929804307e-06,
+      "loss": 0.3582,
+      "step": 13220
+    },
+    {
+      "epoch": 0.7636143257048859,
+      "grad_norm": 7.012741879467996,
+      "learning_rate": 2.3644865208104833e-06,
+      "loss": 0.3433,
+      "step": 13230
+    },
+    {
+      "epoch": 0.7641915096245419,
+      "grad_norm": 4.5407340579502025,
+      "learning_rate": 2.358713848640536e-06,
+      "loss": 0.3388,
+      "step": 13240
+    },
+    {
+      "epoch": 0.7647686935441979,
+      "grad_norm": 4.194705290127997,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.3312,
+      "step": 13250
+    },
+    {
+      "epoch": 0.7653458774638539,
+      "grad_norm": 3.7140413786909203,
+      "learning_rate": 2.347168504300641e-06,
+      "loss": 0.3496,
+      "step": 13260
+    },
+    {
+      "epoch": 0.7659230613835099,
+      "grad_norm": 3.257080159852974,
+      "learning_rate": 2.3413958321306936e-06,
+      "loss": 0.3366,
+      "step": 13270
+    },
+    {
+      "epoch": 0.7665002453031659,
+      "grad_norm": 5.234732296881502,
+      "learning_rate": 2.335623159960746e-06,
+      "loss": 0.3427,
+      "step": 13280
+    },
+    {
+      "epoch": 0.7670774292228218,
+      "grad_norm": 6.870037441960529,
+      "learning_rate": 2.3298504877907984e-06,
+      "loss": 0.3354,
+      "step": 13290
+    },
+    {
+      "epoch": 0.7676546131424778,
+      "grad_norm": 9.876776601961089,
+      "learning_rate": 2.324077815620851e-06,
+      "loss": 0.3315,
+      "step": 13300
+    },
+    {
+      "epoch": 0.7682317970621338,
+      "grad_norm": 5.912982961911957,
+      "learning_rate": 2.3183051434509036e-06,
+      "loss": 0.3485,
+      "step": 13310
+    },
+    {
+      "epoch": 0.7688089809817898,
+      "grad_norm": 4.832716437104076,
+      "learning_rate": 2.312532471280956e-06,
+      "loss": 0.3348,
+      "step": 13320
+    },
+    {
+      "epoch": 0.7693861649014458,
+      "grad_norm": 5.098375721379664,
+      "learning_rate": 2.3067597991110084e-06,
+      "loss": 0.3484,
+      "step": 13330
+    },
+    {
+      "epoch": 0.7699633488211018,
+      "grad_norm": 4.63656346157866,
+      "learning_rate": 2.300987126941061e-06,
+      "loss": 0.3366,
+      "step": 13340
+    },
+    {
+      "epoch": 0.7705405327407578,
+      "grad_norm": 6.588466951994697,
+      "learning_rate": 2.2952144547711136e-06,
+      "loss": 0.3457,
+      "step": 13350
+    },
+    {
+      "epoch": 0.7711177166604138,
+      "grad_norm": 4.824094255832894,
+      "learning_rate": 2.2894417826011666e-06,
+      "loss": 0.341,
+      "step": 13360
+    },
+    {
+      "epoch": 0.7716949005800698,
+      "grad_norm": 4.637581573877111,
+      "learning_rate": 2.2836691104312188e-06,
+      "loss": 0.3554,
+      "step": 13370
+    },
+    {
+      "epoch": 0.7722720844997258,
+      "grad_norm": 6.056831078560241,
+      "learning_rate": 2.2778964382612714e-06,
+      "loss": 0.3186,
+      "step": 13380
+    },
+    {
+      "epoch": 0.7728492684193818,
+      "grad_norm": 45.47481090136826,
+      "learning_rate": 2.272123766091324e-06,
+      "loss": 0.3299,
+      "step": 13390
+    },
+    {
+      "epoch": 0.7734264523390378,
+      "grad_norm": 22.884051964592864,
+      "learning_rate": 2.2663510939213766e-06,
+      "loss": 0.3392,
+      "step": 13400
+    },
+    {
+      "epoch": 0.7740036362586938,
+      "grad_norm": 6.492791637464672,
+      "learning_rate": 2.260578421751429e-06,
+      "loss": 0.3436,
+      "step": 13410
+    },
+    {
+      "epoch": 0.7745808201783498,
+      "grad_norm": 5.820865415832969,
+      "learning_rate": 2.2548057495814813e-06,
+      "loss": 0.3347,
+      "step": 13420
+    },
+    {
+      "epoch": 0.7751580040980058,
+      "grad_norm": 3.6093618399609406,
+      "learning_rate": 2.249033077411534e-06,
+      "loss": 0.3457,
+      "step": 13430
+    },
+    {
+      "epoch": 0.7757351880176618,
+      "grad_norm": 3.8630308240200977,
+      "learning_rate": 2.2432604052415865e-06,
+      "loss": 0.345,
+      "step": 13440
+    },
+    {
+      "epoch": 0.7763123719373178,
+      "grad_norm": 5.534539731109275,
+      "learning_rate": 2.237487733071639e-06,
+      "loss": 0.3383,
+      "step": 13450
+    },
+    {
+      "epoch": 0.7768895558569738,
+      "grad_norm": 9.50423287415909,
+      "learning_rate": 2.2317150609016917e-06,
+      "loss": 0.3528,
+      "step": 13460
+    },
+    {
+      "epoch": 0.7774667397766298,
+      "grad_norm": 11.597704923923128,
+      "learning_rate": 2.2259423887317443e-06,
+      "loss": 0.3482,
+      "step": 13470
+    },
+    {
+      "epoch": 0.7780439236962858,
+      "grad_norm": 4.724944252593318,
+      "learning_rate": 2.2201697165617965e-06,
+      "loss": 0.341,
+      "step": 13480
+    },
+    {
+      "epoch": 0.7786211076159418,
+      "grad_norm": 4.741588520697521,
+      "learning_rate": 2.214397044391849e-06,
+      "loss": 0.3443,
+      "step": 13490
+    },
+    {
+      "epoch": 0.7791982915355978,
+      "grad_norm": 4.723745867093743,
+      "learning_rate": 2.2086243722219017e-06,
+      "loss": 0.3468,
+      "step": 13500
+    },
+    {
+      "epoch": 0.7797754754552538,
+      "grad_norm": 4.302802386626385,
+      "learning_rate": 2.2028517000519543e-06,
+      "loss": 0.3534,
+      "step": 13510
+    },
+    {
+      "epoch": 0.7803526593749098,
+      "grad_norm": 3.6957459272751385,
+      "learning_rate": 2.197079027882007e-06,
+      "loss": 0.3438,
+      "step": 13520
+    },
+    {
+      "epoch": 0.7809298432945658,
+      "grad_norm": 9.02964602805584,
+      "learning_rate": 2.191306355712059e-06,
+      "loss": 0.3507,
+      "step": 13530
+    },
+    {
+      "epoch": 0.7815070272142218,
+      "grad_norm": 4.584120438677978,
+      "learning_rate": 2.1855336835421117e-06,
+      "loss": 0.3392,
+      "step": 13540
+    },
+    {
+      "epoch": 0.7820842111338778,
+      "grad_norm": 3.634370316749477,
+      "learning_rate": 2.1797610113721642e-06,
+      "loss": 0.3434,
+      "step": 13550
+    },
+    {
+      "epoch": 0.7826613950535338,
+      "grad_norm": 7.974537610574205,
+      "learning_rate": 2.173988339202217e-06,
+      "loss": 0.3488,
+      "step": 13560
+    },
+    {
+      "epoch": 0.7832385789731898,
+      "grad_norm": 15.500270051358006,
+      "learning_rate": 2.1682156670322694e-06,
+      "loss": 0.3319,
+      "step": 13570
+    },
+    {
+      "epoch": 0.7838157628928458,
+      "grad_norm": 3.6214865336598288,
+      "learning_rate": 2.162442994862322e-06,
+      "loss": 0.3431,
+      "step": 13580
+    },
+    {
+      "epoch": 0.7843929468125018,
+      "grad_norm": 4.251663962722991,
+      "learning_rate": 2.1566703226923742e-06,
+      "loss": 0.339,
+      "step": 13590
+    },
+    {
+      "epoch": 0.7849701307321578,
+      "grad_norm": 5.8477391728665875,
+      "learning_rate": 2.150897650522427e-06,
+      "loss": 0.3425,
+      "step": 13600
+    },
+    {
+      "epoch": 0.7855473146518138,
+      "grad_norm": 12.661348588140084,
+      "learning_rate": 2.1451249783524794e-06,
+      "loss": 0.3385,
+      "step": 13610
+    },
+    {
+      "epoch": 0.7861244985714698,
+      "grad_norm": 3.153875030872274,
+      "learning_rate": 2.139352306182532e-06,
+      "loss": 0.3429,
+      "step": 13620
+    },
+    {
+      "epoch": 0.7867016824911258,
+      "grad_norm": 7.53800643811605,
+      "learning_rate": 2.1335796340125846e-06,
+      "loss": 0.3492,
+      "step": 13630
+    },
+    {
+      "epoch": 0.7872788664107818,
+      "grad_norm": 4.143985301762935,
+      "learning_rate": 2.127806961842637e-06,
+      "loss": 0.3361,
+      "step": 13640
+    },
+    {
+      "epoch": 0.7878560503304378,
+      "grad_norm": 3.005269949918135,
+      "learning_rate": 2.12203428967269e-06,
+      "loss": 0.3405,
+      "step": 13650
+    },
+    {
+      "epoch": 0.7884332342500938,
+      "grad_norm": 3.018587955955484,
+      "learning_rate": 2.1162616175027424e-06,
+      "loss": 0.3447,
+      "step": 13660
+    },
+    {
+      "epoch": 0.7890104181697498,
+      "grad_norm": 3.002589774112856,
+      "learning_rate": 2.110488945332795e-06,
+      "loss": 0.3253,
+      "step": 13670
+    },
+    {
+      "epoch": 0.7895876020894058,
+      "grad_norm": 3.85103590680487,
+      "learning_rate": 2.104716273162847e-06,
+      "loss": 0.3395,
+      "step": 13680
+    },
+    {
+      "epoch": 0.7901647860090618,
+      "grad_norm": 4.040110821545416,
+      "learning_rate": 2.0989436009928998e-06,
+      "loss": 0.3439,
+      "step": 13690
+    },
+    {
+      "epoch": 0.7907419699287178,
+      "grad_norm": 8.395833969271274,
+      "learning_rate": 2.0931709288229524e-06,
+      "loss": 0.3334,
+      "step": 13700
+    },
+    {
+      "epoch": 0.7913191538483738,
+      "grad_norm": 8.808629294549078,
+      "learning_rate": 2.087398256653005e-06,
+      "loss": 0.3516,
+      "step": 13710
+    },
+    {
+      "epoch": 0.7918963377680298,
+      "grad_norm": 5.183013748264493,
+      "learning_rate": 2.0816255844830576e-06,
+      "loss": 0.3468,
+      "step": 13720
+    },
+    {
+      "epoch": 0.7924735216876858,
+      "grad_norm": 4.651858039147579,
+      "learning_rate": 2.0758529123131097e-06,
+      "loss": 0.3383,
+      "step": 13730
+    },
+    {
+      "epoch": 0.7930507056073418,
+      "grad_norm": 5.147274649825693,
+      "learning_rate": 2.0700802401431623e-06,
+      "loss": 0.3435,
+      "step": 13740
+    },
+    {
+      "epoch": 0.7936278895269978,
+      "grad_norm": 3.3157064357282646,
+      "learning_rate": 2.064307567973215e-06,
+      "loss": 0.3374,
+      "step": 13750
+    },
+    {
+      "epoch": 0.7942050734466538,
+      "grad_norm": 5.359738899671234,
+      "learning_rate": 2.0585348958032675e-06,
+      "loss": 0.3519,
+      "step": 13760
+    },
+    {
+      "epoch": 0.7947822573663098,
+      "grad_norm": 3.232416306077255,
+      "learning_rate": 2.05276222363332e-06,
+      "loss": 0.3466,
+      "step": 13770
+    },
+    {
+      "epoch": 0.7953594412859658,
+      "grad_norm": 5.634484175482586,
+      "learning_rate": 2.0469895514633727e-06,
+      "loss": 0.3435,
+      "step": 13780
+    },
+    {
+      "epoch": 0.7959366252056218,
+      "grad_norm": 4.812177943413768,
+      "learning_rate": 2.041216879293425e-06,
+      "loss": 0.3394,
+      "step": 13790
+    },
+    {
+      "epoch": 0.7965138091252778,
+      "grad_norm": 4.18690284048581,
+      "learning_rate": 2.0354442071234775e-06,
+      "loss": 0.339,
+      "step": 13800
+    },
+    {
+      "epoch": 0.7970909930449338,
+      "grad_norm": 2.6302278323472024,
+      "learning_rate": 2.02967153495353e-06,
+      "loss": 0.3211,
+      "step": 13810
+    },
+    {
+      "epoch": 0.7976681769645898,
+      "grad_norm": 3.7677609162356567,
+      "learning_rate": 2.0238988627835827e-06,
+      "loss": 0.3459,
+      "step": 13820
+    },
+    {
+      "epoch": 0.7982453608842458,
+      "grad_norm": 3.877790721238228,
+      "learning_rate": 2.0181261906136353e-06,
+      "loss": 0.3527,
+      "step": 13830
+    },
+    {
+      "epoch": 0.7988225448039018,
+      "grad_norm": 3.3882391753576897,
+      "learning_rate": 2.0123535184436874e-06,
+      "loss": 0.3418,
+      "step": 13840
+    },
+    {
+      "epoch": 0.7993997287235578,
+      "grad_norm": 8.275704148522843,
+      "learning_rate": 2.00658084627374e-06,
+      "loss": 0.3584,
+      "step": 13850
+    },
+    {
+      "epoch": 0.7999769126432138,
+      "grad_norm": 4.07120977222577,
+      "learning_rate": 2.0008081741037926e-06,
+      "loss": 0.3369,
+      "step": 13860
+    },
+    {
+      "epoch": 0.8005540965628698,
+      "grad_norm": 6.796016189984939,
+      "learning_rate": 1.9950355019338452e-06,
+      "loss": 0.3359,
+      "step": 13870
+    },
+    {
+      "epoch": 0.8011312804825258,
+      "grad_norm": 4.70618096154267,
+      "learning_rate": 1.989262829763898e-06,
+      "loss": 0.3383,
+      "step": 13880
+    },
+    {
+      "epoch": 0.8017084644021818,
+      "grad_norm": 9.622472004428674,
+      "learning_rate": 1.9834901575939504e-06,
+      "loss": 0.3492,
+      "step": 13890
+    },
+    {
+      "epoch": 0.8022856483218378,
+      "grad_norm": 20.930166870041855,
+      "learning_rate": 1.977717485424003e-06,
+      "loss": 0.3343,
+      "step": 13900
+    },
+    {
+      "epoch": 0.8028628322414938,
+      "grad_norm": 10.85418413095689,
+      "learning_rate": 1.9719448132540556e-06,
+      "loss": 0.3399,
+      "step": 13910
+    },
+    {
+      "epoch": 0.8034400161611498,
+      "grad_norm": 3.9293112562677353,
+      "learning_rate": 1.9661721410841082e-06,
+      "loss": 0.363,
+      "step": 13920
+    },
+    {
+      "epoch": 0.8040172000808058,
+      "grad_norm": 10.773909045381199,
+      "learning_rate": 1.960399468914161e-06,
+      "loss": 0.3427,
+      "step": 13930
+    },
+    {
+      "epoch": 0.8045943840004618,
+      "grad_norm": 4.932078250676689,
+      "learning_rate": 1.954626796744213e-06,
+      "loss": 0.3353,
+      "step": 13940
+    },
+    {
+      "epoch": 0.8051715679201178,
+      "grad_norm": 5.376321864962683,
+      "learning_rate": 1.9488541245742656e-06,
+      "loss": 0.3502,
+      "step": 13950
+    },
+    {
+      "epoch": 0.8057487518397738,
+      "grad_norm": 4.014977881261309,
+      "learning_rate": 1.943081452404318e-06,
+      "loss": 0.3169,
+      "step": 13960
+    },
+    {
+      "epoch": 0.8063259357594298,
+      "grad_norm": 8.651359182721874,
+      "learning_rate": 1.9373087802343708e-06,
+      "loss": 0.3349,
+      "step": 13970
+    },
+    {
+      "epoch": 0.8069031196790858,
+      "grad_norm": 4.683539873403375,
+      "learning_rate": 1.9315361080644234e-06,
+      "loss": 0.3366,
+      "step": 13980
+    },
+    {
+      "epoch": 0.8074803035987418,
+      "grad_norm": 7.755677266741849,
+      "learning_rate": 1.9257634358944756e-06,
+      "loss": 0.3377,
+      "step": 13990
+    },
+    {
+      "epoch": 0.8080574875183978,
+      "grad_norm": 14.533972050557356,
+      "learning_rate": 1.919990763724528e-06,
+      "loss": 0.3337,
+      "step": 14000
+    },
+    {
+      "epoch": 0.8086346714380538,
+      "grad_norm": 3.5287179981983052,
+      "learning_rate": 1.9142180915545807e-06,
+      "loss": 0.3371,
+      "step": 14010
+    },
+    {
+      "epoch": 0.8092118553577098,
+      "grad_norm": 3.6897961566375588,
+      "learning_rate": 1.9084454193846333e-06,
+      "loss": 0.3437,
+      "step": 14020
+    },
+    {
+      "epoch": 0.8097890392773658,
+      "grad_norm": 4.275916003879041,
+      "learning_rate": 1.9026727472146857e-06,
+      "loss": 0.3505,
+      "step": 14030
+    },
+    {
+      "epoch": 0.8103662231970218,
+      "grad_norm": 5.853581087992128,
+      "learning_rate": 1.8969000750447383e-06,
+      "loss": 0.3447,
+      "step": 14040
+    },
+    {
+      "epoch": 0.8109434071166778,
+      "grad_norm": 4.456219664367194,
+      "learning_rate": 1.891127402874791e-06,
+      "loss": 0.3405,
+      "step": 14050
+    },
+    {
+      "epoch": 0.8115205910363338,
+      "grad_norm": 3.609665854590679,
+      "learning_rate": 1.8853547307048433e-06,
+      "loss": 0.3539,
+      "step": 14060
+    },
+    {
+      "epoch": 0.8120977749559897,
+      "grad_norm": 8.81894657017933,
+      "learning_rate": 1.879582058534896e-06,
+      "loss": 0.3434,
+      "step": 14070
+    },
+    {
+      "epoch": 0.8126749588756457,
+      "grad_norm": 73.62257234022623,
+      "learning_rate": 1.8738093863649485e-06,
+      "loss": 0.3362,
+      "step": 14080
+    },
+    {
+      "epoch": 0.8132521427953017,
+      "grad_norm": 4.843976388543906,
+      "learning_rate": 1.8680367141950009e-06,
+      "loss": 0.3522,
+      "step": 14090
+    },
+    {
+      "epoch": 0.8138293267149577,
+      "grad_norm": 6.3647153836539925,
+      "learning_rate": 1.8622640420250535e-06,
+      "loss": 0.3351,
+      "step": 14100
+    },
+    {
+      "epoch": 0.8144065106346137,
+      "grad_norm": 15.067604508734842,
+      "learning_rate": 1.8564913698551059e-06,
+      "loss": 0.341,
+      "step": 14110
+    },
+    {
+      "epoch": 0.8149836945542697,
+      "grad_norm": 2.837118277081176,
+      "learning_rate": 1.8507186976851585e-06,
+      "loss": 0.3433,
+      "step": 14120
+    },
+    {
+      "epoch": 0.8155608784739257,
+      "grad_norm": 3.8243460158503337,
+      "learning_rate": 1.844946025515211e-06,
+      "loss": 0.3433,
+      "step": 14130
+    },
+    {
+      "epoch": 0.8161380623935817,
+      "grad_norm": 5.476403512789154,
+      "learning_rate": 1.8391733533452635e-06,
+      "loss": 0.3507,
+      "step": 14140
+    },
+    {
+      "epoch": 0.8167152463132377,
+      "grad_norm": 7.503101973214223,
+      "learning_rate": 1.8334006811753163e-06,
+      "loss": 0.3426,
+      "step": 14150
+    },
+    {
+      "epoch": 0.8172924302328937,
+      "grad_norm": 13.62010751923973,
+      "learning_rate": 1.8276280090053689e-06,
+      "loss": 0.3467,
+      "step": 14160
+    },
+    {
+      "epoch": 0.8178696141525497,
+      "grad_norm": 5.792511670187818,
+      "learning_rate": 1.8218553368354215e-06,
+      "loss": 0.3618,
+      "step": 14170
+    },
+    {
+      "epoch": 0.8184467980722057,
+      "grad_norm": 3.268214743983259,
+      "learning_rate": 1.8160826646654738e-06,
+      "loss": 0.3383,
+      "step": 14180
+    },
+    {
+      "epoch": 0.8190239819918617,
+      "grad_norm": 4.2120443389536,
+      "learning_rate": 1.8103099924955264e-06,
+      "loss": 0.3444,
+      "step": 14190
+    },
+    {
+      "epoch": 0.8196011659115177,
+      "grad_norm": 5.497732760033444,
+      "learning_rate": 1.804537320325579e-06,
+      "loss": 0.3321,
+      "step": 14200
+    },
+    {
+      "epoch": 0.8201783498311737,
+      "grad_norm": 3.752608960335712,
+      "learning_rate": 1.7987646481556314e-06,
+      "loss": 0.34,
+      "step": 14210
+    },
+    {
+      "epoch": 0.8207555337508297,
+      "grad_norm": 5.61044709148666,
+      "learning_rate": 1.792991975985684e-06,
+      "loss": 0.3477,
+      "step": 14220
+    },
+    {
+      "epoch": 0.8213327176704857,
+      "grad_norm": 3.2289347743222,
+      "learning_rate": 1.7872193038157364e-06,
+      "loss": 0.3354,
+      "step": 14230
+    },
+    {
+      "epoch": 0.8219099015901417,
+      "grad_norm": 4.132295269856051,
+      "learning_rate": 1.781446631645789e-06,
+      "loss": 0.3454,
+      "step": 14240
+    },
+    {
+      "epoch": 0.8224870855097977,
+      "grad_norm": 8.94256483462556,
+      "learning_rate": 1.7756739594758416e-06,
+      "loss": 0.3437,
+      "step": 14250
+    },
+    {
+      "epoch": 0.8230642694294537,
+      "grad_norm": 6.100086954024239,
+      "learning_rate": 1.769901287305894e-06,
+      "loss": 0.3358,
+      "step": 14260
+    },
+    {
+      "epoch": 0.8236414533491097,
+      "grad_norm": 4.7378243458253255,
+      "learning_rate": 1.7641286151359466e-06,
+      "loss": 0.3357,
+      "step": 14270
+    },
+    {
+      "epoch": 0.8242186372687657,
+      "grad_norm": 4.402684079414389,
+      "learning_rate": 1.7583559429659992e-06,
+      "loss": 0.346,
+      "step": 14280
+    },
+    {
+      "epoch": 0.8247958211884217,
+      "grad_norm": 7.94048495060501,
+      "learning_rate": 1.7525832707960516e-06,
+      "loss": 0.3334,
+      "step": 14290
+    },
+    {
+      "epoch": 0.8253730051080777,
+      "grad_norm": 4.248624676291536,
+      "learning_rate": 1.7468105986261042e-06,
+      "loss": 0.3348,
+      "step": 14300
+    },
+    {
+      "epoch": 0.8259501890277336,
+      "grad_norm": 5.863500294171999,
+      "learning_rate": 1.7410379264561568e-06,
+      "loss": 0.3545,
+      "step": 14310
+    },
+    {
+      "epoch": 0.8265273729473896,
+      "grad_norm": 5.462151225452732,
+      "learning_rate": 1.7352652542862091e-06,
+      "loss": 0.3324,
+      "step": 14320
+    },
+    {
+      "epoch": 0.8271045568670456,
+      "grad_norm": 2.757724626485063,
+      "learning_rate": 1.7294925821162617e-06,
+      "loss": 0.3411,
+      "step": 14330
+    },
+    {
+      "epoch": 0.8276817407867016,
+      "grad_norm": 2.5548734181876456,
+      "learning_rate": 1.7237199099463141e-06,
+      "loss": 0.3278,
+      "step": 14340
+    },
+    {
+      "epoch": 0.8282589247063576,
+      "grad_norm": 4.871936623866052,
+      "learning_rate": 1.7179472377763667e-06,
+      "loss": 0.3306,
+      "step": 14350
+    },
+    {
+      "epoch": 0.8288361086260136,
+      "grad_norm": 3.10101382610628,
+      "learning_rate": 1.7121745656064193e-06,
+      "loss": 0.3387,
+      "step": 14360
+    },
+    {
+      "epoch": 0.8294132925456696,
+      "grad_norm": 10.575051742858964,
+      "learning_rate": 1.7064018934364717e-06,
+      "loss": 0.3428,
+      "step": 14370
+    },
+    {
+      "epoch": 0.8299904764653256,
+      "grad_norm": 4.367916255458309,
+      "learning_rate": 1.7006292212665243e-06,
+      "loss": 0.3397,
+      "step": 14380
+    },
+    {
+      "epoch": 0.8305676603849816,
+      "grad_norm": 6.616840660969775,
+      "learning_rate": 1.694856549096577e-06,
+      "loss": 0.331,
+      "step": 14390
+    },
+    {
+      "epoch": 0.8311448443046376,
+      "grad_norm": 12.71189460753895,
+      "learning_rate": 1.6890838769266293e-06,
+      "loss": 0.3352,
+      "step": 14400
+    },
+    {
+      "epoch": 0.8317220282242936,
+      "grad_norm": 8.817998813310439,
+      "learning_rate": 1.683311204756682e-06,
+      "loss": 0.3425,
+      "step": 14410
+    },
+    {
+      "epoch": 0.8322992121439496,
+      "grad_norm": 4.766835462814113,
+      "learning_rate": 1.6775385325867347e-06,
+      "loss": 0.3398,
+      "step": 14420
+    },
+    {
+      "epoch": 0.8328763960636056,
+      "grad_norm": 4.169057753174133,
+      "learning_rate": 1.6717658604167873e-06,
+      "loss": 0.3294,
+      "step": 14430
+    },
+    {
+      "epoch": 0.8334535799832616,
+      "grad_norm": 5.150260088289917,
+      "learning_rate": 1.6659931882468397e-06,
+      "loss": 0.3273,
+      "step": 14440
+    },
+    {
+      "epoch": 0.8340307639029176,
+      "grad_norm": 5.053684619580007,
+      "learning_rate": 1.6602205160768923e-06,
+      "loss": 0.3642,
+      "step": 14450
+    },
+    {
+      "epoch": 0.8346079478225736,
+      "grad_norm": 4.068134996065944,
+      "learning_rate": 1.6544478439069447e-06,
+      "loss": 0.3441,
+      "step": 14460
+    },
+    {
+      "epoch": 0.8351851317422296,
+      "grad_norm": 4.004964659229183,
+      "learning_rate": 1.6486751717369972e-06,
+      "loss": 0.3408,
+      "step": 14470
+    },
+    {
+      "epoch": 0.8357623156618856,
+      "grad_norm": 4.02559591240971,
+      "learning_rate": 1.6429024995670498e-06,
+      "loss": 0.3385,
+      "step": 14480
+    },
+    {
+      "epoch": 0.8363394995815416,
+      "grad_norm": 5.168203382157519,
+      "learning_rate": 1.6371298273971022e-06,
+      "loss": 0.3424,
+      "step": 14490
+    },
+    {
+      "epoch": 0.8369166835011976,
+      "grad_norm": 6.3562693056901285,
+      "learning_rate": 1.6313571552271548e-06,
+      "loss": 0.3432,
+      "step": 14500
+    },
+    {
+      "epoch": 0.8374938674208536,
+      "grad_norm": 7.105132387517162,
+      "learning_rate": 1.6255844830572074e-06,
+      "loss": 0.3385,
+      "step": 14510
+    },
+    {
+      "epoch": 0.8380710513405096,
+      "grad_norm": 5.167657366774912,
+      "learning_rate": 1.6198118108872598e-06,
+      "loss": 0.3357,
+      "step": 14520
+    },
+    {
+      "epoch": 0.8386482352601656,
+      "grad_norm": 5.322779082569645,
+      "learning_rate": 1.6140391387173124e-06,
+      "loss": 0.351,
+      "step": 14530
+    },
+    {
+      "epoch": 0.8392254191798216,
+      "grad_norm": 5.555960049234855,
+      "learning_rate": 1.608266466547365e-06,
+      "loss": 0.3427,
+      "step": 14540
+    },
+    {
+      "epoch": 0.8398026030994776,
+      "grad_norm": 14.30097637805443,
+      "learning_rate": 1.6024937943774174e-06,
+      "loss": 0.3268,
+      "step": 14550
+    },
+    {
+      "epoch": 0.8403797870191336,
+      "grad_norm": 5.930941981679822,
+      "learning_rate": 1.59672112220747e-06,
+      "loss": 0.3378,
+      "step": 14560
+    },
+    {
+      "epoch": 0.8409569709387896,
+      "grad_norm": 6.41989158034055,
+      "learning_rate": 1.5909484500375224e-06,
+      "loss": 0.3425,
+      "step": 14570
+    },
+    {
+      "epoch": 0.8415341548584456,
+      "grad_norm": 5.309996017042821,
+      "learning_rate": 1.585175777867575e-06,
+      "loss": 0.347,
+      "step": 14580
+    },
+    {
+      "epoch": 0.8421113387781016,
+      "grad_norm": 5.483533161522683,
+      "learning_rate": 1.5794031056976276e-06,
+      "loss": 0.3292,
+      "step": 14590
+    },
+    {
+      "epoch": 0.8426885226977576,
+      "grad_norm": 5.692527937195376,
+      "learning_rate": 1.57363043352768e-06,
+      "loss": 0.3377,
+      "step": 14600
+    },
+    {
+      "epoch": 0.8432657066174136,
+      "grad_norm": 4.521114906145438,
+      "learning_rate": 1.5678577613577325e-06,
+      "loss": 0.33,
+      "step": 14610
+    },
+    {
+      "epoch": 0.8438428905370696,
+      "grad_norm": 5.387502803416387,
+      "learning_rate": 1.5620850891877851e-06,
+      "loss": 0.3418,
+      "step": 14620
+    },
+    {
+      "epoch": 0.8444200744567256,
+      "grad_norm": 3.5934061953432783,
+      "learning_rate": 1.5563124170178375e-06,
+      "loss": 0.3386,
+      "step": 14630
+    },
+    {
+      "epoch": 0.8449972583763816,
+      "grad_norm": 11.194143114734654,
+      "learning_rate": 1.5505397448478901e-06,
+      "loss": 0.345,
+      "step": 14640
+    },
+    {
+      "epoch": 0.8455744422960376,
+      "grad_norm": 10.094088344395155,
+      "learning_rate": 1.5447670726779427e-06,
+      "loss": 0.3526,
+      "step": 14650
+    },
+    {
+      "epoch": 0.8461516262156936,
+      "grad_norm": 3.4477998903369538,
+      "learning_rate": 1.5389944005079951e-06,
+      "loss": 0.3373,
+      "step": 14660
+    },
+    {
+      "epoch": 0.8467288101353496,
+      "grad_norm": 7.047707090908251,
+      "learning_rate": 1.533221728338048e-06,
+      "loss": 0.3403,
+      "step": 14670
+    },
+    {
+      "epoch": 0.8473059940550056,
+      "grad_norm": 2.8576436370742893,
+      "learning_rate": 1.5274490561681005e-06,
+      "loss": 0.3453,
+      "step": 14680
+    },
+    {
+      "epoch": 0.8478831779746616,
+      "grad_norm": 12.938468126614902,
+      "learning_rate": 1.521676383998153e-06,
+      "loss": 0.336,
+      "step": 14690
+    },
+    {
+      "epoch": 0.8484603618943176,
+      "grad_norm": 5.011512190053106,
+      "learning_rate": 1.5159037118282055e-06,
+      "loss": 0.3424,
+      "step": 14700
+    },
+    {
+      "epoch": 0.8490375458139736,
+      "grad_norm": 5.063829942230378,
+      "learning_rate": 1.510131039658258e-06,
+      "loss": 0.346,
+      "step": 14710
+    },
+    {
+      "epoch": 0.8496147297336296,
+      "grad_norm": 6.679043646426254,
+      "learning_rate": 1.5043583674883105e-06,
+      "loss": 0.3422,
+      "step": 14720
+    },
+    {
+      "epoch": 0.8501919136532856,
+      "grad_norm": 5.800788583586723,
+      "learning_rate": 1.498585695318363e-06,
+      "loss": 0.3421,
+      "step": 14730
+    },
+    {
+      "epoch": 0.8507690975729416,
+      "grad_norm": 3.968739599812534,
+      "learning_rate": 1.4928130231484157e-06,
+      "loss": 0.3395,
+      "step": 14740
+    },
+    {
+      "epoch": 0.8513462814925976,
+      "grad_norm": 7.880056504602316,
+      "learning_rate": 1.487040350978468e-06,
+      "loss": 0.3326,
+      "step": 14750
+    },
+    {
+      "epoch": 0.8519234654122536,
+      "grad_norm": 5.685089944033708,
+      "learning_rate": 1.4812676788085207e-06,
+      "loss": 0.3356,
+      "step": 14760
+    },
+    {
+      "epoch": 0.8525006493319096,
+      "grad_norm": 4.795808126193561,
+      "learning_rate": 1.475495006638573e-06,
+      "loss": 0.316,
+      "step": 14770
+    },
+    {
+      "epoch": 0.8530778332515656,
+      "grad_norm": 6.693555623774071,
+      "learning_rate": 1.4697223344686256e-06,
+      "loss": 0.3416,
+      "step": 14780
+    },
+    {
+      "epoch": 0.8536550171712216,
+      "grad_norm": 5.752237694272574,
+      "learning_rate": 1.4639496622986782e-06,
+      "loss": 0.3348,
+      "step": 14790
+    },
+    {
+      "epoch": 0.8542322010908776,
+      "grad_norm": 7.135726397784308,
+      "learning_rate": 1.4581769901287306e-06,
+      "loss": 0.3378,
+      "step": 14800
+    },
+    {
+      "epoch": 0.8548093850105336,
+      "grad_norm": 3.2833774570611234,
+      "learning_rate": 1.4524043179587832e-06,
+      "loss": 0.3504,
+      "step": 14810
+    },
+    {
+      "epoch": 0.8553865689301896,
+      "grad_norm": 6.256462721047408,
+      "learning_rate": 1.4466316457888358e-06,
+      "loss": 0.349,
+      "step": 14820
+    },
+    {
+      "epoch": 0.8559637528498456,
+      "grad_norm": 6.040295806596955,
+      "learning_rate": 1.4408589736188882e-06,
+      "loss": 0.3208,
+      "step": 14830
+    },
+    {
+      "epoch": 0.8565409367695016,
+      "grad_norm": 4.729301211824621,
+      "learning_rate": 1.4350863014489408e-06,
+      "loss": 0.3277,
+      "step": 14840
+    },
+    {
+      "epoch": 0.8571181206891576,
+      "grad_norm": 4.313779706679082,
+      "learning_rate": 1.4293136292789934e-06,
+      "loss": 0.3306,
+      "step": 14850
+    },
+    {
+      "epoch": 0.8576953046088136,
+      "grad_norm": 11.790043476255672,
+      "learning_rate": 1.4235409571090458e-06,
+      "loss": 0.3391,
+      "step": 14860
+    },
+    {
+      "epoch": 0.8582724885284696,
+      "grad_norm": 7.642639050872643,
+      "learning_rate": 1.4177682849390984e-06,
+      "loss": 0.3388,
+      "step": 14870
+    },
+    {
+      "epoch": 0.8588496724481256,
+      "grad_norm": 7.260077362208394,
+      "learning_rate": 1.4119956127691508e-06,
+      "loss": 0.3422,
+      "step": 14880
+    },
+    {
+      "epoch": 0.8594268563677816,
+      "grad_norm": 4.754130590048299,
+      "learning_rate": 1.4062229405992034e-06,
+      "loss": 0.3436,
+      "step": 14890
+    },
+    {
+      "epoch": 0.8600040402874376,
+      "grad_norm": 6.225554657816755,
+      "learning_rate": 1.400450268429256e-06,
+      "loss": 0.3352,
+      "step": 14900
+    },
+    {
+      "epoch": 0.8605812242070936,
+      "grad_norm": 9.369073827925245,
+      "learning_rate": 1.3946775962593083e-06,
+      "loss": 0.3552,
+      "step": 14910
+    },
+    {
+      "epoch": 0.8611584081267496,
+      "grad_norm": 22.876915271022913,
+      "learning_rate": 1.388904924089361e-06,
+      "loss": 0.3492,
+      "step": 14920
+    },
+    {
+      "epoch": 0.8617355920464056,
+      "grad_norm": 3.2981136451706132,
+      "learning_rate": 1.3831322519194135e-06,
+      "loss": 0.3328,
+      "step": 14930
+    },
+    {
+      "epoch": 0.8623127759660616,
+      "grad_norm": 15.61626227774467,
+      "learning_rate": 1.3773595797494663e-06,
+      "loss": 0.337,
+      "step": 14940
+    },
+    {
+      "epoch": 0.8628899598857176,
+      "grad_norm": 3.605130100716397,
+      "learning_rate": 1.3715869075795187e-06,
+      "loss": 0.3385,
+      "step": 14950
+    },
+    {
+      "epoch": 0.8634671438053736,
+      "grad_norm": 12.284387392936685,
+      "learning_rate": 1.3658142354095713e-06,
+      "loss": 0.338,
+      "step": 14960
+    },
+    {
+      "epoch": 0.8640443277250296,
+      "grad_norm": 3.278014138295641,
+      "learning_rate": 1.360041563239624e-06,
+      "loss": 0.3337,
+      "step": 14970
+    },
+    {
+      "epoch": 0.8646215116446856,
+      "grad_norm": 6.104850658183799,
+      "learning_rate": 1.3542688910696763e-06,
+      "loss": 0.3276,
+      "step": 14980
+    },
+    {
+      "epoch": 0.8651986955643416,
+      "grad_norm": 5.279235884437097,
+      "learning_rate": 1.348496218899729e-06,
+      "loss": 0.3289,
+      "step": 14990
+    },
+    {
+      "epoch": 0.8657758794839976,
+      "grad_norm": 8.449467479597608,
+      "learning_rate": 1.3427235467297813e-06,
+      "loss": 0.3314,
+      "step": 15000
+    },
+    {
+      "epoch": 0.8663530634036536,
+      "grad_norm": 5.665683889458834,
+      "learning_rate": 1.3369508745598339e-06,
+      "loss": 0.3416,
+      "step": 15010
+    },
+    {
+      "epoch": 0.8669302473233096,
+      "grad_norm": 3.9335607560335735,
+      "learning_rate": 1.3311782023898865e-06,
+      "loss": 0.3525,
+      "step": 15020
+    },
+    {
+      "epoch": 0.8675074312429656,
+      "grad_norm": 2.894014459167942,
+      "learning_rate": 1.3254055302199389e-06,
+      "loss": 0.3328,
+      "step": 15030
+    },
+    {
+      "epoch": 0.8680846151626216,
+      "grad_norm": 65.75688988910291,
+      "learning_rate": 1.3196328580499915e-06,
+      "loss": 0.331,
+      "step": 15040
+    },
+    {
+      "epoch": 0.8686617990822776,
+      "grad_norm": 61.00701606633841,
+      "learning_rate": 1.313860185880044e-06,
+      "loss": 0.3407,
+      "step": 15050
+    },
+    {
+      "epoch": 0.8692389830019336,
+      "grad_norm": 4.364812051967769,
+      "learning_rate": 1.3080875137100965e-06,
+      "loss": 0.335,
+      "step": 15060
+    },
+    {
+      "epoch": 0.8698161669215896,
+      "grad_norm": 24.662511548298095,
+      "learning_rate": 1.302314841540149e-06,
+      "loss": 0.3348,
+      "step": 15070
+    },
+    {
+      "epoch": 0.8703933508412456,
+      "grad_norm": 4.160466497716753,
+      "learning_rate": 1.2965421693702016e-06,
+      "loss": 0.3252,
+      "step": 15080
+    },
+    {
+      "epoch": 0.8709705347609016,
+      "grad_norm": 6.134539334165056,
+      "learning_rate": 1.290769497200254e-06,
+      "loss": 0.343,
+      "step": 15090
+    },
+    {
+      "epoch": 0.8715477186805576,
+      "grad_norm": 7.228713295937482,
+      "learning_rate": 1.2849968250303066e-06,
+      "loss": 0.3338,
+      "step": 15100
+    },
+    {
+      "epoch": 0.8721249026002136,
+      "grad_norm": 4.918201123965152,
+      "learning_rate": 1.279224152860359e-06,
+      "loss": 0.3527,
+      "step": 15110
+    },
+    {
+      "epoch": 0.8727020865198696,
+      "grad_norm": 11.079488033226095,
+      "learning_rate": 1.2734514806904116e-06,
+      "loss": 0.3485,
+      "step": 15120
+    },
+    {
+      "epoch": 0.8732792704395256,
+      "grad_norm": 4.928609183215457,
+      "learning_rate": 1.2676788085204642e-06,
+      "loss": 0.3354,
+      "step": 15130
+    },
+    {
+      "epoch": 0.8738564543591816,
+      "grad_norm": 6.740385950730952,
+      "learning_rate": 1.2619061363505166e-06,
+      "loss": 0.333,
+      "step": 15140
+    },
+    {
+      "epoch": 0.8744336382788376,
+      "grad_norm": 4.67490770106929,
+      "learning_rate": 1.2561334641805692e-06,
+      "loss": 0.331,
+      "step": 15150
+    },
+    {
+      "epoch": 0.8750108221984936,
+      "grad_norm": 13.018881009854521,
+      "learning_rate": 1.2503607920106218e-06,
+      "loss": 0.3418,
+      "step": 15160
+    },
+    {
+      "epoch": 0.8755880061181496,
+      "grad_norm": 5.447550954730696,
+      "learning_rate": 1.2445881198406744e-06,
+      "loss": 0.3372,
+      "step": 15170
+    },
+    {
+      "epoch": 0.8761651900378056,
+      "grad_norm": 3.273410656701021,
+      "learning_rate": 1.238815447670727e-06,
+      "loss": 0.3199,
+      "step": 15180
+    },
+    {
+      "epoch": 0.8767423739574616,
+      "grad_norm": 4.277523732895546,
+      "learning_rate": 1.2330427755007794e-06,
+      "loss": 0.341,
+      "step": 15190
+    },
+    {
+      "epoch": 0.8773195578771176,
+      "grad_norm": 15.742782882783079,
+      "learning_rate": 1.227270103330832e-06,
+      "loss": 0.341,
+      "step": 15200
+    },
+    {
+      "epoch": 0.8778967417967736,
+      "grad_norm": 5.131013300289669,
+      "learning_rate": 1.2214974311608844e-06,
+      "loss": 0.3303,
+      "step": 15210
+    },
+    {
+      "epoch": 0.8784739257164296,
+      "grad_norm": 12.866336037106171,
+      "learning_rate": 1.215724758990937e-06,
+      "loss": 0.3442,
+      "step": 15220
+    },
+    {
+      "epoch": 0.8790511096360856,
+      "grad_norm": 5.127710193300859,
+      "learning_rate": 1.2099520868209895e-06,
+      "loss": 0.352,
+      "step": 15230
+    },
+    {
+      "epoch": 0.8796282935557416,
+      "grad_norm": 7.005612361267027,
+      "learning_rate": 1.204179414651042e-06,
+      "loss": 0.3376,
+      "step": 15240
+    },
+    {
+      "epoch": 0.8802054774753976,
+      "grad_norm": 19.994629507428854,
+      "learning_rate": 1.1984067424810945e-06,
+      "loss": 0.3366,
+      "step": 15250
+    },
+    {
+      "epoch": 0.8807826613950536,
+      "grad_norm": 19.289162587657003,
+      "learning_rate": 1.1926340703111471e-06,
+      "loss": 0.3339,
+      "step": 15260
+    },
+    {
+      "epoch": 0.8813598453147096,
+      "grad_norm": 5.780176197085242,
+      "learning_rate": 1.1868613981411997e-06,
+      "loss": 0.3328,
+      "step": 15270
+    },
+    {
+      "epoch": 0.8819370292343656,
+      "grad_norm": 5.322625710215796,
+      "learning_rate": 1.1810887259712523e-06,
+      "loss": 0.3364,
+      "step": 15280
+    },
+    {
+      "epoch": 0.8825142131540216,
+      "grad_norm": 4.160665784312188,
+      "learning_rate": 1.1753160538013047e-06,
+      "loss": 0.3298,
+      "step": 15290
+    },
+    {
+      "epoch": 0.8830913970736776,
+      "grad_norm": 39.76071118055878,
+      "learning_rate": 1.1695433816313573e-06,
+      "loss": 0.348,
+      "step": 15300
+    },
+    {
+      "epoch": 0.8836685809933336,
+      "grad_norm": 3.5659508030627958,
+      "learning_rate": 1.16377070946141e-06,
+      "loss": 0.3271,
+      "step": 15310
+    },
+    {
+      "epoch": 0.8842457649129896,
+      "grad_norm": 4.436695102429374,
+      "learning_rate": 1.1579980372914623e-06,
+      "loss": 0.3407,
+      "step": 15320
+    },
+    {
+      "epoch": 0.8848229488326456,
+      "grad_norm": 11.403167564876071,
+      "learning_rate": 1.1522253651215149e-06,
+      "loss": 0.3383,
+      "step": 15330
+    },
+    {
+      "epoch": 0.8854001327523016,
+      "grad_norm": 4.535641723893359,
+      "learning_rate": 1.1464526929515673e-06,
+      "loss": 0.3328,
+      "step": 15340
+    },
+    {
+      "epoch": 0.8859773166719576,
+      "grad_norm": 7.663265877666311,
+      "learning_rate": 1.1406800207816199e-06,
+      "loss": 0.3365,
+      "step": 15350
+    },
+    {
+      "epoch": 0.8865545005916136,
+      "grad_norm": 7.820505381715719,
+      "learning_rate": 1.1349073486116725e-06,
+      "loss": 0.3432,
+      "step": 15360
+    },
+    {
+      "epoch": 0.8871316845112696,
+      "grad_norm": 6.006599271526383,
+      "learning_rate": 1.1291346764417248e-06,
+      "loss": 0.3487,
+      "step": 15370
+    },
+    {
+      "epoch": 0.8877088684309256,
+      "grad_norm": 11.881435307105928,
+      "learning_rate": 1.1233620042717774e-06,
+      "loss": 0.3332,
+      "step": 15380
+    },
+    {
+      "epoch": 0.8882860523505816,
+      "grad_norm": 3.4819561819318103,
+      "learning_rate": 1.11758933210183e-06,
+      "loss": 0.335,
+      "step": 15390
+    },
+    {
+      "epoch": 0.8888632362702376,
+      "grad_norm": 6.893650513052578,
+      "learning_rate": 1.1118166599318826e-06,
+      "loss": 0.3493,
+      "step": 15400
+    },
+    {
+      "epoch": 0.8894404201898936,
+      "grad_norm": 6.925493159604081,
+      "learning_rate": 1.1060439877619352e-06,
+      "loss": 0.3448,
+      "step": 15410
+    },
+    {
+      "epoch": 0.8900176041095496,
+      "grad_norm": 5.5200268397262775,
+      "learning_rate": 1.1002713155919876e-06,
+      "loss": 0.3256,
+      "step": 15420
+    },
+    {
+      "epoch": 0.8905947880292056,
+      "grad_norm": 7.638890582405007,
+      "learning_rate": 1.0944986434220402e-06,
+      "loss": 0.3327,
+      "step": 15430
+    },
+    {
+      "epoch": 0.8911719719488616,
+      "grad_norm": 4.585636430659814,
+      "learning_rate": 1.0887259712520926e-06,
+      "loss": 0.3386,
+      "step": 15440
+    },
+    {
+      "epoch": 0.8917491558685176,
+      "grad_norm": 7.147252901738225,
+      "learning_rate": 1.0829532990821452e-06,
+      "loss": 0.3402,
+      "step": 15450
+    },
+    {
+      "epoch": 0.8923263397881736,
+      "grad_norm": 5.40597402568476,
+      "learning_rate": 1.0771806269121978e-06,
+      "loss": 0.3385,
+      "step": 15460
+    },
+    {
+      "epoch": 0.8929035237078295,
+      "grad_norm": 4.260919590930795,
+      "learning_rate": 1.0714079547422502e-06,
+      "loss": 0.3232,
+      "step": 15470
+    },
+    {
+      "epoch": 0.8934807076274855,
+      "grad_norm": 9.083532528696407,
+      "learning_rate": 1.0656352825723028e-06,
+      "loss": 0.354,
+      "step": 15480
+    },
+    {
+      "epoch": 0.8940578915471415,
+      "grad_norm": 4.996043459346209,
+      "learning_rate": 1.0598626104023554e-06,
+      "loss": 0.329,
+      "step": 15490
+    },
+    {
+      "epoch": 0.8946350754667975,
+      "grad_norm": 4.117500728065101,
+      "learning_rate": 1.0540899382324078e-06,
+      "loss": 0.3269,
+      "step": 15500
+    },
+    {
+      "epoch": 0.8952122593864535,
+      "grad_norm": 4.9248570764155435,
+      "learning_rate": 1.0483172660624604e-06,
+      "loss": 0.3321,
+      "step": 15510
+    },
+    {
+      "epoch": 0.8957894433061095,
+      "grad_norm": 4.696024737813827,
+      "learning_rate": 1.042544593892513e-06,
+      "loss": 0.3309,
+      "step": 15520
+    },
+    {
+      "epoch": 0.8963666272257655,
+      "grad_norm": 7.733578705149999,
+      "learning_rate": 1.0367719217225656e-06,
+      "loss": 0.3305,
+      "step": 15530
+    },
+    {
+      "epoch": 0.8969438111454215,
+      "grad_norm": 5.394346330597671,
+      "learning_rate": 1.0309992495526181e-06,
+      "loss": 0.3388,
+      "step": 15540
+    },
+    {
+      "epoch": 0.8975209950650774,
+      "grad_norm": 5.311667852016155,
+      "learning_rate": 1.0252265773826705e-06,
+      "loss": 0.3376,
+      "step": 15550
+    },
+    {
+      "epoch": 0.8980981789847334,
+      "grad_norm": 25.908831946398365,
+      "learning_rate": 1.0194539052127231e-06,
+      "loss": 0.3265,
+      "step": 15560
+    },
+    {
+      "epoch": 0.8986753629043894,
+      "grad_norm": 33.665974796607145,
+      "learning_rate": 1.0136812330427755e-06,
+      "loss": 0.3348,
+      "step": 15570
+    },
+    {
+      "epoch": 0.8992525468240454,
+      "grad_norm": 6.825259095726871,
+      "learning_rate": 1.0079085608728281e-06,
+      "loss": 0.3453,
+      "step": 15580
+    },
+    {
+      "epoch": 0.8998297307437014,
+      "grad_norm": 3.1735048623731092,
+      "learning_rate": 1.0021358887028807e-06,
+      "loss": 0.3218,
+      "step": 15590
+    },
+    {
+      "epoch": 0.9004069146633574,
+      "grad_norm": 9.114656814093948,
+      "learning_rate": 9.96363216532933e-07,
+      "loss": 0.3228,
+      "step": 15600
+    },
+    {
+      "epoch": 0.9009840985830134,
+      "grad_norm": 5.87155385500696,
+      "learning_rate": 9.905905443629857e-07,
+      "loss": 0.3309,
+      "step": 15610
+    },
+    {
+      "epoch": 0.9015612825026694,
+      "grad_norm": 14.717991716657272,
+      "learning_rate": 9.848178721930383e-07,
+      "loss": 0.3486,
+      "step": 15620
+    },
+    {
+      "epoch": 0.9021384664223254,
+      "grad_norm": 3.7767948462311067,
+      "learning_rate": 9.790452000230907e-07,
+      "loss": 0.3182,
+      "step": 15630
+    },
+    {
+      "epoch": 0.9027156503419814,
+      "grad_norm": 3.6787944993169006,
+      "learning_rate": 9.732725278531433e-07,
+      "loss": 0.3336,
+      "step": 15640
+    },
+    {
+      "epoch": 0.9032928342616374,
+      "grad_norm": 11.812395475228488,
+      "learning_rate": 9.674998556831959e-07,
+      "loss": 0.3286,
+      "step": 15650
+    },
+    {
+      "epoch": 0.9038700181812934,
+      "grad_norm": 5.186436534605756,
+      "learning_rate": 9.617271835132485e-07,
+      "loss": 0.3275,
+      "step": 15660
+    },
+    {
+      "epoch": 0.9044472021009494,
+      "grad_norm": 4.424154650673984,
+      "learning_rate": 9.559545113433009e-07,
+      "loss": 0.3185,
+      "step": 15670
+    },
+    {
+      "epoch": 0.9050243860206054,
+      "grad_norm": 6.3580022358578105,
+      "learning_rate": 9.501818391733534e-07,
+      "loss": 0.3226,
+      "step": 15680
+    },
+    {
+      "epoch": 0.9056015699402614,
+      "grad_norm": 6.310999676892304,
+      "learning_rate": 9.444091670034059e-07,
+      "loss": 0.3304,
+      "step": 15690
+    },
+    {
+      "epoch": 0.9061787538599174,
+      "grad_norm": 3.7913010315172633,
+      "learning_rate": 9.386364948334585e-07,
+      "loss": 0.3412,
+      "step": 15700
+    },
+    {
+      "epoch": 0.9067559377795734,
+      "grad_norm": 3.67407928861924,
+      "learning_rate": 9.32863822663511e-07,
+      "loss": 0.341,
+      "step": 15710
+    },
+    {
+      "epoch": 0.9073331216992294,
+      "grad_norm": 7.334570042771433,
+      "learning_rate": 9.270911504935635e-07,
+      "loss": 0.3265,
+      "step": 15720
+    },
+    {
+      "epoch": 0.9079103056188854,
+      "grad_norm": 6.456404235720165,
+      "learning_rate": 9.213184783236161e-07,
+      "loss": 0.3444,
+      "step": 15730
+    },
+    {
+      "epoch": 0.9084874895385414,
+      "grad_norm": 4.201612752804453,
+      "learning_rate": 9.155458061536686e-07,
+      "loss": 0.348,
+      "step": 15740
+    },
+    {
+      "epoch": 0.9090646734581974,
+      "grad_norm": 5.071365974662106,
+      "learning_rate": 9.097731339837211e-07,
+      "loss": 0.3153,
+      "step": 15750
+    },
+    {
+      "epoch": 0.9096418573778534,
+      "grad_norm": 5.035641882967374,
+      "learning_rate": 9.040004618137736e-07,
+      "loss": 0.3359,
+      "step": 15760
+    },
+    {
+      "epoch": 0.9102190412975094,
+      "grad_norm": 12.03037859002001,
+      "learning_rate": 8.982277896438262e-07,
+      "loss": 0.3242,
+      "step": 15770
+    },
+    {
+      "epoch": 0.9107962252171654,
+      "grad_norm": 7.140437657447056,
+      "learning_rate": 8.924551174738787e-07,
+      "loss": 0.3333,
+      "step": 15780
+    },
+    {
+      "epoch": 0.9113734091368214,
+      "grad_norm": 8.294728901327792,
+      "learning_rate": 8.866824453039313e-07,
+      "loss": 0.3267,
+      "step": 15790
+    },
+    {
+      "epoch": 0.9119505930564774,
+      "grad_norm": 5.258892430294796,
+      "learning_rate": 8.809097731339839e-07,
+      "loss": 0.3306,
+      "step": 15800
+    },
+    {
+      "epoch": 0.9125277769761334,
+      "grad_norm": 4.56809810586393,
+      "learning_rate": 8.751371009640364e-07,
+      "loss": 0.3354,
+      "step": 15810
+    },
+    {
+      "epoch": 0.9131049608957894,
+      "grad_norm": 8.754714473224658,
+      "learning_rate": 8.693644287940889e-07,
+      "loss": 0.3258,
+      "step": 15820
+    },
+    {
+      "epoch": 0.9136821448154454,
+      "grad_norm": 10.722318127648052,
+      "learning_rate": 8.635917566241415e-07,
+      "loss": 0.3251,
+      "step": 15830
+    },
+    {
+      "epoch": 0.9142593287351014,
+      "grad_norm": 17.100240147200765,
+      "learning_rate": 8.578190844541939e-07,
+      "loss": 0.3243,
+      "step": 15840
+    },
+    {
+      "epoch": 0.9148365126547574,
+      "grad_norm": 6.487613210408211,
+      "learning_rate": 8.520464122842464e-07,
+      "loss": 0.3299,
+      "step": 15850
+    },
+    {
+      "epoch": 0.9154136965744134,
+      "grad_norm": 3.047470063867609,
+      "learning_rate": 8.462737401142989e-07,
+      "loss": 0.3277,
+      "step": 15860
+    },
+    {
+      "epoch": 0.9159908804940694,
+      "grad_norm": 3.3099261534656823,
+      "learning_rate": 8.405010679443515e-07,
+      "loss": 0.3225,
+      "step": 15870
+    },
+    {
+      "epoch": 0.9165680644137254,
+      "grad_norm": 12.904829402744845,
+      "learning_rate": 8.34728395774404e-07,
+      "loss": 0.3456,
+      "step": 15880
+    },
+    {
+      "epoch": 0.9171452483333814,
+      "grad_norm": 3.874918781355711,
+      "learning_rate": 8.289557236044565e-07,
+      "loss": 0.324,
+      "step": 15890
+    },
+    {
+      "epoch": 0.9177224322530374,
+      "grad_norm": 8.856329412411933,
+      "learning_rate": 8.231830514345091e-07,
+      "loss": 0.3336,
+      "step": 15900
+    },
+    {
+      "epoch": 0.9182996161726934,
+      "grad_norm": 8.142861706815804,
+      "learning_rate": 8.174103792645616e-07,
+      "loss": 0.3346,
+      "step": 15910
+    },
+    {
+      "epoch": 0.9188768000923494,
+      "grad_norm": 6.512456297032329,
+      "learning_rate": 8.116377070946142e-07,
+      "loss": 0.3356,
+      "step": 15920
+    },
+    {
+      "epoch": 0.9194539840120054,
+      "grad_norm": 5.20827680094837,
+      "learning_rate": 8.058650349246668e-07,
+      "loss": 0.327,
+      "step": 15930
+    },
+    {
+      "epoch": 0.9200311679316614,
+      "grad_norm": 3.6566931406552166,
+      "learning_rate": 8.000923627547193e-07,
+      "loss": 0.3393,
+      "step": 15940
+    },
+    {
+      "epoch": 0.9206083518513174,
+      "grad_norm": 5.430549059652793,
+      "learning_rate": 7.943196905847718e-07,
+      "loss": 0.336,
+      "step": 15950
+    },
+    {
+      "epoch": 0.9211855357709734,
+      "grad_norm": 11.48641695737308,
+      "learning_rate": 7.885470184148243e-07,
+      "loss": 0.3239,
+      "step": 15960
+    },
+    {
+      "epoch": 0.9217627196906294,
+      "grad_norm": 5.74247518929047,
+      "learning_rate": 7.827743462448769e-07,
+      "loss": 0.3272,
+      "step": 15970
+    },
+    {
+      "epoch": 0.9223399036102854,
+      "grad_norm": 2.9972434937325954,
+      "learning_rate": 7.770016740749293e-07,
+      "loss": 0.3444,
+      "step": 15980
+    },
+    {
+      "epoch": 0.9229170875299414,
+      "grad_norm": 4.3485773330395405,
+      "learning_rate": 7.712290019049818e-07,
+      "loss": 0.3343,
+      "step": 15990
+    },
+    {
+      "epoch": 0.9234942714495974,
+      "grad_norm": 5.8420315281490725,
+      "learning_rate": 7.654563297350344e-07,
+      "loss": 0.3418,
+      "step": 16000
+    },
+    {
+      "epoch": 0.9240714553692534,
+      "grad_norm": 3.3727101894208924,
+      "learning_rate": 7.596836575650869e-07,
+      "loss": 0.3313,
+      "step": 16010
+    },
+    {
+      "epoch": 0.9246486392889094,
+      "grad_norm": 4.3271834892078305,
+      "learning_rate": 7.539109853951394e-07,
+      "loss": 0.3263,
+      "step": 16020
+    },
+    {
+      "epoch": 0.9252258232085654,
+      "grad_norm": 4.114539060448778,
+      "learning_rate": 7.481383132251919e-07,
+      "loss": 0.3494,
+      "step": 16030
+    },
+    {
+      "epoch": 0.9258030071282214,
+      "grad_norm": 3.6454496106451946,
+      "learning_rate": 7.423656410552445e-07,
+      "loss": 0.3287,
+      "step": 16040
+    },
+    {
+      "epoch": 0.9263801910478774,
+      "grad_norm": 3.738978996136776,
+      "learning_rate": 7.365929688852971e-07,
+      "loss": 0.3463,
+      "step": 16050
+    },
+    {
+      "epoch": 0.9269573749675334,
+      "grad_norm": 4.789282815458411,
+      "learning_rate": 7.308202967153497e-07,
+      "loss": 0.3237,
+      "step": 16060
+    },
+    {
+      "epoch": 0.9275345588871894,
+      "grad_norm": 9.761730219338645,
+      "learning_rate": 7.250476245454022e-07,
+      "loss": 0.3469,
+      "step": 16070
+    },
+    {
+      "epoch": 0.9281117428068454,
+      "grad_norm": 3.028208359321862,
+      "learning_rate": 7.192749523754547e-07,
+      "loss": 0.3355,
+      "step": 16080
+    },
+    {
+      "epoch": 0.9286889267265014,
+      "grad_norm": 18.45944225954803,
+      "learning_rate": 7.135022802055072e-07,
+      "loss": 0.3177,
+      "step": 16090
+    },
+    {
+      "epoch": 0.9292661106461574,
+      "grad_norm": 7.253195406338047,
+      "learning_rate": 7.077296080355598e-07,
+      "loss": 0.3439,
+      "step": 16100
+    },
+    {
+      "epoch": 0.9298432945658134,
+      "grad_norm": 4.739735623729803,
+      "learning_rate": 7.019569358656123e-07,
+      "loss": 0.3357,
+      "step": 16110
+    },
+    {
+      "epoch": 0.9304204784854694,
+      "grad_norm": 7.262658023729907,
+      "learning_rate": 6.961842636956648e-07,
+      "loss": 0.3307,
+      "step": 16120
+    },
+    {
+      "epoch": 0.9309976624051254,
+      "grad_norm": 4.3276336459159275,
+      "learning_rate": 6.904115915257172e-07,
+      "loss": 0.3397,
+      "step": 16130
+    },
+    {
+      "epoch": 0.9315748463247814,
+      "grad_norm": 4.773618213493451,
+      "learning_rate": 6.846389193557698e-07,
+      "loss": 0.3265,
+      "step": 16140
+    },
+    {
+      "epoch": 0.9321520302444374,
+      "grad_norm": 3.469713298213091,
+      "learning_rate": 6.788662471858223e-07,
+      "loss": 0.3356,
+      "step": 16150
+    },
+    {
+      "epoch": 0.9327292141640934,
+      "grad_norm": 2.680891067948385,
+      "learning_rate": 6.730935750158748e-07,
+      "loss": 0.3351,
+      "step": 16160
+    },
+    {
+      "epoch": 0.9333063980837494,
+      "grad_norm": 4.795647821005584,
+      "learning_rate": 6.673209028459274e-07,
+      "loss": 0.316,
+      "step": 16170
+    },
+    {
+      "epoch": 0.9338835820034054,
+      "grad_norm": 4.072329465945383,
+      "learning_rate": 6.615482306759799e-07,
+      "loss": 0.3323,
+      "step": 16180
+    },
+    {
+      "epoch": 0.9344607659230614,
+      "grad_norm": 6.139834777308556,
+      "learning_rate": 6.557755585060325e-07,
+      "loss": 0.3389,
+      "step": 16190
+    },
+    {
+      "epoch": 0.9350379498427174,
+      "grad_norm": 6.679020147433282,
+      "learning_rate": 6.500028863360851e-07,
+      "loss": 0.3213,
+      "step": 16200
+    },
+    {
+      "epoch": 0.9356151337623734,
+      "grad_norm": 3.931169778597499,
+      "learning_rate": 6.442302141661376e-07,
+      "loss": 0.3202,
+      "step": 16210
+    },
+    {
+      "epoch": 0.9361923176820294,
+      "grad_norm": 9.793848920416451,
+      "learning_rate": 6.384575419961901e-07,
+      "loss": 0.3312,
+      "step": 16220
+    },
+    {
+      "epoch": 0.9367695016016854,
+      "grad_norm": 3.165076810544466,
+      "learning_rate": 6.326848698262427e-07,
+      "loss": 0.3268,
+      "step": 16230
+    },
+    {
+      "epoch": 0.9373466855213414,
+      "grad_norm": 11.613653214904037,
+      "learning_rate": 6.269121976562952e-07,
+      "loss": 0.3395,
+      "step": 16240
+    },
+    {
+      "epoch": 0.9379238694409974,
+      "grad_norm": 8.287793663837986,
+      "learning_rate": 6.211395254863477e-07,
+      "loss": 0.337,
+      "step": 16250
+    },
+    {
+      "epoch": 0.9385010533606534,
+      "grad_norm": 4.919246593010951,
+      "learning_rate": 6.153668533164002e-07,
+      "loss": 0.3183,
+      "step": 16260
+    },
+    {
+      "epoch": 0.9390782372803094,
+      "grad_norm": 12.710584575769794,
+      "learning_rate": 6.095941811464528e-07,
+      "loss": 0.3371,
+      "step": 16270
+    },
+    {
+      "epoch": 0.9396554211999654,
+      "grad_norm": 13.642759112617313,
+      "learning_rate": 6.038215089765054e-07,
+      "loss": 0.3364,
+      "step": 16280
+    },
+    {
+      "epoch": 0.9402326051196214,
+      "grad_norm": 3.2163528904964567,
+      "learning_rate": 5.980488368065578e-07,
+      "loss": 0.3482,
+      "step": 16290
+    },
+    {
+      "epoch": 0.9408097890392774,
+      "grad_norm": 6.3234687340970845,
+      "learning_rate": 5.922761646366103e-07,
+      "loss": 0.3389,
+      "step": 16300
+    },
+    {
+      "epoch": 0.9413869729589334,
+      "grad_norm": 4.810430576924547,
+      "learning_rate": 5.865034924666628e-07,
+      "loss": 0.3273,
+      "step": 16310
+    },
+    {
+      "epoch": 0.9419641568785894,
+      "grad_norm": 16.667420292327467,
+      "learning_rate": 5.807308202967154e-07,
+      "loss": 0.3333,
+      "step": 16320
+    },
+    {
+      "epoch": 0.9425413407982454,
+      "grad_norm": 4.476842181947663,
+      "learning_rate": 5.749581481267679e-07,
+      "loss": 0.3319,
+      "step": 16330
+    },
+    {
+      "epoch": 0.9431185247179014,
+      "grad_norm": 2.753605309820116,
+      "learning_rate": 5.691854759568204e-07,
+      "loss": 0.324,
+      "step": 16340
+    },
+    {
+      "epoch": 0.9436957086375574,
+      "grad_norm": 7.258998836239923,
+      "learning_rate": 5.63412803786873e-07,
+      "loss": 0.3317,
+      "step": 16350
+    },
+    {
+      "epoch": 0.9442728925572134,
+      "grad_norm": 3.3810984203362513,
+      "learning_rate": 5.576401316169255e-07,
+      "loss": 0.3308,
+      "step": 16360
+    },
+    {
+      "epoch": 0.9448500764768694,
+      "grad_norm": 10.448881769543355,
+      "learning_rate": 5.518674594469781e-07,
+      "loss": 0.3389,
+      "step": 16370
+    },
+    {
+      "epoch": 0.9454272603965254,
+      "grad_norm": 7.485384669846898,
+      "learning_rate": 5.460947872770306e-07,
+      "loss": 0.3225,
+      "step": 16380
+    },
+    {
+      "epoch": 0.9460044443161814,
+      "grad_norm": 7.048976390316521,
+      "learning_rate": 5.403221151070831e-07,
+      "loss": 0.3316,
+      "step": 16390
+    },
+    {
+      "epoch": 0.9465816282358374,
+      "grad_norm": 5.343786757617583,
+      "learning_rate": 5.345494429371357e-07,
+      "loss": 0.324,
+      "step": 16400
+    },
+    {
+      "epoch": 0.9471588121554934,
+      "grad_norm": 8.606885533079547,
+      "learning_rate": 5.287767707671882e-07,
+      "loss": 0.339,
+      "step": 16410
+    },
+    {
+      "epoch": 0.9477359960751494,
+      "grad_norm": 5.1476348800105205,
+      "learning_rate": 5.230040985972408e-07,
+      "loss": 0.3316,
+      "step": 16420
+    },
+    {
+      "epoch": 0.9483131799948054,
+      "grad_norm": 3.9518622190419386,
+      "learning_rate": 5.172314264272933e-07,
+      "loss": 0.3257,
+      "step": 16430
+    },
+    {
+      "epoch": 0.9488903639144614,
+      "grad_norm": 4.759115152912856,
+      "learning_rate": 5.114587542573457e-07,
+      "loss": 0.3252,
+      "step": 16440
+    },
+    {
+      "epoch": 0.9494675478341174,
+      "grad_norm": 2.3577377378728936,
+      "learning_rate": 5.056860820873983e-07,
+      "loss": 0.3354,
+      "step": 16450
+    },
+    {
+      "epoch": 0.9500447317537734,
+      "grad_norm": 8.233491658698778,
+      "learning_rate": 4.999134099174508e-07,
+      "loss": 0.3269,
+      "step": 16460
+    },
+    {
+      "epoch": 0.9506219156734294,
+      "grad_norm": 5.830593517325124,
+      "learning_rate": 4.941407377475033e-07,
+      "loss": 0.3303,
+      "step": 16470
+    },
+    {
+      "epoch": 0.9511990995930854,
+      "grad_norm": 4.5976658225857205,
+      "learning_rate": 4.883680655775559e-07,
+      "loss": 0.3164,
+      "step": 16480
+    },
+    {
+      "epoch": 0.9517762835127414,
+      "grad_norm": 2.683143832655395,
+      "learning_rate": 4.825953934076084e-07,
+      "loss": 0.3389,
+      "step": 16490
+    },
+    {
+      "epoch": 0.9523534674323973,
+      "grad_norm": 6.256568384332184,
+      "learning_rate": 4.7682272123766096e-07,
+      "loss": 0.3379,
+      "step": 16500
+    },
+    {
+      "epoch": 0.9529306513520533,
+      "grad_norm": 5.947037852710701,
+      "learning_rate": 4.710500490677135e-07,
+      "loss": 0.3336,
+      "step": 16510
+    },
+    {
+      "epoch": 0.9535078352717093,
+      "grad_norm": 3.9206339656766183,
+      "learning_rate": 4.65277376897766e-07,
+      "loss": 0.3416,
+      "step": 16520
+    },
+    {
+      "epoch": 0.9540850191913653,
+      "grad_norm": 5.305934878449426,
+      "learning_rate": 4.5950470472781854e-07,
+      "loss": 0.343,
+      "step": 16530
+    },
+    {
+      "epoch": 0.9546622031110213,
+      "grad_norm": 4.905648955862364,
+      "learning_rate": 4.53732032557871e-07,
+      "loss": 0.3326,
+      "step": 16540
+    },
+    {
+      "epoch": 0.9552393870306773,
+      "grad_norm": 6.934144679851784,
+      "learning_rate": 4.479593603879236e-07,
+      "loss": 0.3315,
+      "step": 16550
+    },
+    {
+      "epoch": 0.9558165709503333,
+      "grad_norm": 6.121333752853476,
+      "learning_rate": 4.4218668821797617e-07,
+      "loss": 0.3337,
+      "step": 16560
+    },
+    {
+      "epoch": 0.9563937548699893,
+      "grad_norm": 4.161869077945622,
+      "learning_rate": 4.3641401604802866e-07,
+      "loss": 0.354,
+      "step": 16570
+    },
+    {
+      "epoch": 0.9569709387896453,
+      "grad_norm": 4.792938959925312,
+      "learning_rate": 4.306413438780812e-07,
+      "loss": 0.3385,
+      "step": 16580
+    },
+    {
+      "epoch": 0.9575481227093013,
+      "grad_norm": 13.85786954380734,
+      "learning_rate": 4.248686717081337e-07,
+      "loss": 0.3206,
+      "step": 16590
+    },
+    {
+      "epoch": 0.9581253066289573,
+      "grad_norm": 21.263443082950594,
+      "learning_rate": 4.1909599953818624e-07,
+      "loss": 0.3325,
+      "step": 16600
+    },
+    {
+      "epoch": 0.9587024905486133,
+      "grad_norm": 3.660403999109124,
+      "learning_rate": 4.1332332736823884e-07,
+      "loss": 0.3227,
+      "step": 16610
+    },
+    {
+      "epoch": 0.9592796744682693,
+      "grad_norm": 3.9235176913649994,
+      "learning_rate": 4.075506551982913e-07,
+      "loss": 0.3283,
+      "step": 16620
+    },
+    {
+      "epoch": 0.9598568583879253,
+      "grad_norm": 5.6449372673837965,
+      "learning_rate": 4.0177798302834387e-07,
+      "loss": 0.3427,
+      "step": 16630
+    },
+    {
+      "epoch": 0.9604340423075813,
+      "grad_norm": 5.248416354277083,
+      "learning_rate": 3.9600531085839636e-07,
+      "loss": 0.3288,
+      "step": 16640
+    },
+    {
+      "epoch": 0.9610112262272373,
+      "grad_norm": 8.246345220378487,
+      "learning_rate": 3.902326386884489e-07,
+      "loss": 0.3354,
+      "step": 16650
+    },
+    {
+      "epoch": 0.9615884101468933,
+      "grad_norm": 13.274950590494653,
+      "learning_rate": 3.8445996651850145e-07,
+      "loss": 0.3312,
+      "step": 16660
+    },
+    {
+      "epoch": 0.9621655940665493,
+      "grad_norm": 15.537361667999631,
+      "learning_rate": 3.7868729434855394e-07,
+      "loss": 0.3237,
+      "step": 16670
+    },
+    {
+      "epoch": 0.9627427779862053,
+      "grad_norm": 17.30099668558429,
+      "learning_rate": 3.7291462217860654e-07,
+      "loss": 0.3447,
+      "step": 16680
+    },
+    {
+      "epoch": 0.9633199619058613,
+      "grad_norm": 6.00682814280853,
+      "learning_rate": 3.671419500086591e-07,
+      "loss": 0.3285,
+      "step": 16690
+    },
+    {
+      "epoch": 0.9638971458255173,
+      "grad_norm": 4.337349097771177,
+      "learning_rate": 3.613692778387116e-07,
+      "loss": 0.3406,
+      "step": 16700
+    },
+    {
+      "epoch": 0.9644743297451733,
+      "grad_norm": 5.112346889090425,
+      "learning_rate": 3.555966056687641e-07,
+      "loss": 0.3245,
+      "step": 16710
+    },
+    {
+      "epoch": 0.9650515136648293,
+      "grad_norm": 3.333315383479396,
+      "learning_rate": 3.498239334988166e-07,
+      "loss": 0.323,
+      "step": 16720
+    },
+    {
+      "epoch": 0.9656286975844853,
+      "grad_norm": 6.363838641104665,
+      "learning_rate": 3.4405126132886915e-07,
+      "loss": 0.3335,
+      "step": 16730
+    },
+    {
+      "epoch": 0.9662058815041413,
+      "grad_norm": 4.4727853159969095,
+      "learning_rate": 3.3827858915892164e-07,
+      "loss": 0.3195,
+      "step": 16740
+    },
+    {
+      "epoch": 0.9667830654237973,
+      "grad_norm": 2.9245486184268525,
+      "learning_rate": 3.3250591698897424e-07,
+      "loss": 0.334,
+      "step": 16750
+    },
+    {
+      "epoch": 0.9673602493434533,
+      "grad_norm": 8.292114221205217,
+      "learning_rate": 3.267332448190268e-07,
+      "loss": 0.3417,
+      "step": 16760
+    },
+    {
+      "epoch": 0.9679374332631093,
+      "grad_norm": 4.467404105277962,
+      "learning_rate": 3.209605726490793e-07,
+      "loss": 0.3273,
+      "step": 16770
+    },
+    {
+      "epoch": 0.9685146171827653,
+      "grad_norm": 3.555136063724782,
+      "learning_rate": 3.151879004791318e-07,
+      "loss": 0.3319,
+      "step": 16780
+    },
+    {
+      "epoch": 0.9690918011024213,
+      "grad_norm": 3.5864439964386206,
+      "learning_rate": 3.0941522830918436e-07,
+      "loss": 0.3151,
+      "step": 16790
+    },
+    {
+      "epoch": 0.9696689850220773,
+      "grad_norm": 4.286014953806982,
+      "learning_rate": 3.036425561392369e-07,
+      "loss": 0.3348,
+      "step": 16800
+    },
+    {
+      "epoch": 0.9702461689417333,
+      "grad_norm": 4.937397445129751,
+      "learning_rate": 2.978698839692894e-07,
+      "loss": 0.3447,
+      "step": 16810
+    },
+    {
+      "epoch": 0.9708233528613893,
+      "grad_norm": 4.053983936820117,
+      "learning_rate": 2.9209721179934194e-07,
+      "loss": 0.3234,
+      "step": 16820
+    },
+    {
+      "epoch": 0.9714005367810453,
+      "grad_norm": 4.668613443745286,
+      "learning_rate": 2.863245396293945e-07,
+      "loss": 0.3318,
+      "step": 16830
+    },
+    {
+      "epoch": 0.9719777207007013,
+      "grad_norm": 4.726308017445137,
+      "learning_rate": 2.80551867459447e-07,
+      "loss": 0.3384,
+      "step": 16840
+    },
+    {
+      "epoch": 0.9725549046203573,
+      "grad_norm": 5.20234640635383,
+      "learning_rate": 2.747791952894995e-07,
+      "loss": 0.3414,
+      "step": 16850
+    },
+    {
+      "epoch": 0.9731320885400133,
+      "grad_norm": 5.471268202808402,
+      "learning_rate": 2.6900652311955207e-07,
+      "loss": 0.3158,
+      "step": 16860
+    },
+    {
+      "epoch": 0.9737092724596693,
+      "grad_norm": 4.041775287210815,
+      "learning_rate": 2.632338509496046e-07,
+      "loss": 0.3353,
+      "step": 16870
+    },
+    {
+      "epoch": 0.9742864563793253,
+      "grad_norm": 6.9340722075810515,
+      "learning_rate": 2.574611787796571e-07,
+      "loss": 0.3292,
+      "step": 16880
+    },
+    {
+      "epoch": 0.9748636402989813,
+      "grad_norm": 5.462231359128078,
+      "learning_rate": 2.5168850660970965e-07,
+      "loss": 0.3382,
+      "step": 16890
+    },
+    {
+      "epoch": 0.9754408242186373,
+      "grad_norm": 5.016835747194534,
+      "learning_rate": 2.459158344397622e-07,
+      "loss": 0.3264,
+      "step": 16900
+    },
+    {
+      "epoch": 0.9760180081382933,
+      "grad_norm": 6.59783102359862,
+      "learning_rate": 2.4014316226981474e-07,
+      "loss": 0.3303,
+      "step": 16910
+    },
+    {
+      "epoch": 0.9765951920579493,
+      "grad_norm": 11.129353025607179,
+      "learning_rate": 2.3437049009986723e-07,
+      "loss": 0.3196,
+      "step": 16920
+    },
+    {
+      "epoch": 0.9771723759776053,
+      "grad_norm": 5.828386789897742,
+      "learning_rate": 2.285978179299198e-07,
+      "loss": 0.3319,
+      "step": 16930
+    },
+    {
+      "epoch": 0.9777495598972613,
+      "grad_norm": 2.710691717608737,
+      "learning_rate": 2.2282514575997232e-07,
+      "loss": 0.3319,
+      "step": 16940
+    },
+    {
+      "epoch": 0.9783267438169173,
+      "grad_norm": 5.1520597373996715,
+      "learning_rate": 2.1705247359002483e-07,
+      "loss": 0.3314,
+      "step": 16950
+    },
+    {
+      "epoch": 0.9789039277365733,
+      "grad_norm": 3.185806720570308,
+      "learning_rate": 2.1127980142007738e-07,
+      "loss": 0.3296,
+      "step": 16960
+    },
+    {
+      "epoch": 0.9794811116562293,
+      "grad_norm": 6.515634970555692,
+      "learning_rate": 2.055071292501299e-07,
+      "loss": 0.3312,
+      "step": 16970
+    },
+    {
+      "epoch": 0.9800582955758853,
+      "grad_norm": 10.485461655446002,
+      "learning_rate": 1.997344570801824e-07,
+      "loss": 0.3422,
+      "step": 16980
+    },
+    {
+      "epoch": 0.9806354794955413,
+      "grad_norm": 3.8847690688300727,
+      "learning_rate": 1.9396178491023498e-07,
+      "loss": 0.3283,
+      "step": 16990
+    },
+    {
+      "epoch": 0.9812126634151973,
+      "grad_norm": 9.994920110996672,
+      "learning_rate": 1.881891127402875e-07,
+      "loss": 0.3327,
+      "step": 17000
+    },
+    {
+      "epoch": 0.9817898473348533,
+      "grad_norm": 5.025058096183087,
+      "learning_rate": 1.8241644057034002e-07,
+      "loss": 0.3207,
+      "step": 17010
+    },
+    {
+      "epoch": 0.9823670312545093,
+      "grad_norm": 11.136774459380124,
+      "learning_rate": 1.7664376840039256e-07,
+      "loss": 0.3233,
+      "step": 17020
+    },
+    {
+      "epoch": 0.9829442151741652,
+      "grad_norm": 2.906968928719714,
+      "learning_rate": 1.708710962304451e-07,
+      "loss": 0.335,
+      "step": 17030
+    },
+    {
+      "epoch": 0.9835213990938212,
+      "grad_norm": 47.22379036340986,
+      "learning_rate": 1.6509842406049762e-07,
+      "loss": 0.326,
+      "step": 17040
+    },
+    {
+      "epoch": 0.9840985830134772,
+      "grad_norm": 3.5068334227447537,
+      "learning_rate": 1.5932575189055014e-07,
+      "loss": 0.3308,
+      "step": 17050
+    },
+    {
+      "epoch": 0.9846757669331332,
+      "grad_norm": 6.058207244664307,
+      "learning_rate": 1.5355307972060266e-07,
+      "loss": 0.336,
+      "step": 17060
+    },
+    {
+      "epoch": 0.9852529508527892,
+      "grad_norm": 14.19694786309551,
+      "learning_rate": 1.477804075506552e-07,
+      "loss": 0.3482,
+      "step": 17070
+    },
+    {
+      "epoch": 0.9858301347724452,
+      "grad_norm": 4.351680537742745,
+      "learning_rate": 1.4200773538070775e-07,
+      "loss": 0.3533,
+      "step": 17080
+    },
+    {
+      "epoch": 0.9864073186921012,
+      "grad_norm": 11.029248449585278,
+      "learning_rate": 1.3623506321076027e-07,
+      "loss": 0.3196,
+      "step": 17090
+    },
+    {
+      "epoch": 0.9869845026117572,
+      "grad_norm": 5.081238129188481,
+      "learning_rate": 1.304623910408128e-07,
+      "loss": 0.333,
+      "step": 17100
+    },
+    {
+      "epoch": 0.9875616865314132,
+      "grad_norm": 23.011375672208313,
+      "learning_rate": 1.2468971887086533e-07,
+      "loss": 0.3353,
+      "step": 17110
+    },
+    {
+      "epoch": 0.9881388704510692,
+      "grad_norm": 3.615170936112003,
+      "learning_rate": 1.1891704670091786e-07,
+      "loss": 0.3293,
+      "step": 17120
+    },
+    {
+      "epoch": 0.9887160543707252,
+      "grad_norm": 6.0241909835288645,
+      "learning_rate": 1.131443745309704e-07,
+      "loss": 0.3305,
+      "step": 17130
+    },
+    {
+      "epoch": 0.9892932382903812,
+      "grad_norm": 6.4242440293309,
+      "learning_rate": 1.0737170236102292e-07,
+      "loss": 0.3229,
+      "step": 17140
+    },
+    {
+      "epoch": 0.9898704222100372,
+      "grad_norm": 4.8207248692315465,
+      "learning_rate": 1.0159903019107546e-07,
+      "loss": 0.3317,
+      "step": 17150
+    },
+    {
+      "epoch": 0.9904476061296932,
+      "grad_norm": 4.674342685671797,
+      "learning_rate": 9.5826358021128e-08,
+      "loss": 0.328,
+      "step": 17160
+    },
+    {
+      "epoch": 0.9910247900493492,
+      "grad_norm": 5.524320830604144,
+      "learning_rate": 9.005368585118051e-08,
+      "loss": 0.3297,
+      "step": 17170
+    },
+    {
+      "epoch": 0.9916019739690052,
+      "grad_norm": 6.1310872624369175,
+      "learning_rate": 8.428101368123306e-08,
+      "loss": 0.3342,
+      "step": 17180
+    },
+    {
+      "epoch": 0.9921791578886612,
+      "grad_norm": 4.736837397124582,
+      "learning_rate": 7.850834151128557e-08,
+      "loss": 0.3261,
+      "step": 17190
+    },
+    {
+      "epoch": 0.9927563418083172,
+      "grad_norm": 3.3135028507038498,
+      "learning_rate": 7.273566934133812e-08,
+      "loss": 0.3125,
+      "step": 17200
+    },
+    {
+      "epoch": 0.9933335257279732,
+      "grad_norm": 4.80914101916905,
+      "learning_rate": 6.696299717139064e-08,
+      "loss": 0.3233,
+      "step": 17210
+    },
+    {
+      "epoch": 0.9939107096476292,
+      "grad_norm": 4.178676432109751,
+      "learning_rate": 6.119032500144317e-08,
+      "loss": 0.316,
+      "step": 17220
+    },
+    {
+      "epoch": 0.9944878935672852,
+      "grad_norm": 4.648278510632473,
+      "learning_rate": 5.5417652831495705e-08,
+      "loss": 0.3247,
+      "step": 17230
+    },
+    {
+      "epoch": 0.9950650774869412,
+      "grad_norm": 6.4221153929916515,
+      "learning_rate": 4.9644980661548236e-08,
+      "loss": 0.3252,
+      "step": 17240
+    },
+    {
+      "epoch": 0.9956422614065972,
+      "grad_norm": 4.57910460292981,
+      "learning_rate": 4.3872308491600766e-08,
+      "loss": 0.346,
+      "step": 17250
+    },
+    {
+      "epoch": 0.9962194453262532,
+      "grad_norm": 6.013404774431674,
+      "learning_rate": 3.80996363216533e-08,
+      "loss": 0.3286,
+      "step": 17260
+    },
+    {
+      "epoch": 0.9967966292459092,
+      "grad_norm": 4.660240371593313,
+      "learning_rate": 3.232696415170583e-08,
+      "loss": 0.3238,
+      "step": 17270
+    },
+    {
+      "epoch": 0.9973738131655652,
+      "grad_norm": 3.0170050218995734,
+      "learning_rate": 2.6554291981758356e-08,
+      "loss": 0.3399,
+      "step": 17280
+    },
+    {
+      "epoch": 0.9979509970852212,
+      "grad_norm": 3.235155684245671,
+      "learning_rate": 2.078161981181089e-08,
+      "loss": 0.343,
+      "step": 17290
+    },
+    {
+      "epoch": 0.9985281810048772,
+      "grad_norm": 3.668276198786728,
+      "learning_rate": 1.500894764186342e-08,
+      "loss": 0.3299,
+      "step": 17300
+    },
+    {
+      "epoch": 0.9991053649245332,
+      "grad_norm": 5.359004024139474,
+      "learning_rate": 9.23627547191595e-09,
+      "loss": 0.3202,
+      "step": 17310
+    },
+    {
+      "epoch": 0.9996825488441892,
+      "grad_norm": 5.70686001442703,
+      "learning_rate": 3.4636033019684815e-09,
+      "loss": 0.3161,
+      "step": 17320
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 17325,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9935909437571072.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}