diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12054 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 7.999534233814625,
+  "eval_steps": 500,
+  "global_step": 8584,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004657661853749418,
+      "grad_norm": 5.635344410943957,
+      "learning_rate": 2.910360884749709e-07,
+      "loss": 1.3166,
+      "step": 5
+    },
+    {
+      "epoch": 0.009315323707498836,
+      "grad_norm": 5.5648851943522954,
+      "learning_rate": 5.820721769499418e-07,
+      "loss": 1.302,
+      "step": 10
+    },
+    {
+      "epoch": 0.013972985561248253,
+      "grad_norm": 5.3123150312900895,
+      "learning_rate": 8.731082654249127e-07,
+      "loss": 1.2882,
+      "step": 15
+    },
+    {
+      "epoch": 0.018630647414997672,
+      "grad_norm": 4.237789240817698,
+      "learning_rate": 1.1641443538998836e-06,
+      "loss": 1.2574,
+      "step": 20
+    },
+    {
+      "epoch": 0.02328830926874709,
+      "grad_norm": 2.4660698848230447,
+      "learning_rate": 1.4551804423748545e-06,
+      "loss": 1.2017,
+      "step": 25
+    },
+    {
+      "epoch": 0.027945971122496506,
+      "grad_norm": 2.056462779510106,
+      "learning_rate": 1.7462165308498253e-06,
+      "loss": 1.1262,
+      "step": 30
+    },
+    {
+      "epoch": 0.032603632976245925,
+      "grad_norm": 1.4327847209433977,
+      "learning_rate": 2.037252619324796e-06,
+      "loss": 1.078,
+      "step": 35
+    },
+    {
+      "epoch": 0.037261294829995344,
+      "grad_norm": 1.3173231724943057,
+      "learning_rate": 2.3282887077997673e-06,
+      "loss": 1.0082,
+      "step": 40
+    },
+    {
+      "epoch": 0.04191895668374476,
+      "grad_norm": 0.8876525406538337,
+      "learning_rate": 2.6193247962747383e-06,
+      "loss": 0.9316,
+      "step": 45
+    },
+    {
+      "epoch": 0.04657661853749418,
+      "grad_norm": 0.7198347171125989,
+      "learning_rate": 2.910360884749709e-06,
+      "loss": 0.8925,
+      "step": 50
+    },
+    {
+      "epoch": 0.05123428039124359,
+      "grad_norm": 0.7051557927465876,
+      "learning_rate": 3.2013969732246805e-06,
+      "loss": 0.8458,
+      "step": 55
+    },
+    {
+      "epoch": 0.05589194224499301,
+      "grad_norm": 0.5078638639252016,
+      "learning_rate": 3.4924330616996507e-06,
+      "loss": 0.8224,
+      "step": 60
+    },
+    {
+      "epoch": 0.06054960409874243,
+      "grad_norm": 0.5201115702404338,
+      "learning_rate": 3.7834691501746217e-06,
+      "loss": 0.7945,
+      "step": 65
+    },
+    {
+      "epoch": 0.06520726595249185,
+      "grad_norm": 0.4026038703421235,
+      "learning_rate": 4.074505238649592e-06,
+      "loss": 0.7554,
+      "step": 70
+    },
+    {
+      "epoch": 0.06986492780624126,
+      "grad_norm": 0.35147895058479645,
+      "learning_rate": 4.3655413271245635e-06,
+      "loss": 0.7448,
+      "step": 75
+    },
+    {
+      "epoch": 0.07452258965999069,
+      "grad_norm": 0.3792473243721313,
+      "learning_rate": 4.6565774155995345e-06,
+      "loss": 0.7502,
+      "step": 80
+    },
+    {
+      "epoch": 0.0791802515137401,
+      "grad_norm": 0.33218038449745835,
+      "learning_rate": 4.947613504074506e-06,
+      "loss": 0.7138,
+      "step": 85
+    },
+    {
+      "epoch": 0.08383791336748952,
+      "grad_norm": 0.375352695478229,
+      "learning_rate": 5.238649592549477e-06,
+      "loss": 0.7143,
+      "step": 90
+    },
+    {
+      "epoch": 0.08849557522123894,
+      "grad_norm": 0.3853692458917314,
+      "learning_rate": 5.529685681024447e-06,
+      "loss": 0.7169,
+      "step": 95
+    },
+    {
+      "epoch": 0.09315323707498836,
+      "grad_norm": 0.33731584102558443,
+      "learning_rate": 5.820721769499418e-06,
+      "loss": 0.6978,
+      "step": 100
+    },
+    {
+      "epoch": 0.09781089892873777,
+      "grad_norm": 0.33360385089561806,
+      "learning_rate": 6.111757857974389e-06,
+      "loss": 0.6988,
+      "step": 105
+    },
+    {
+      "epoch": 0.10246856078248719,
+      "grad_norm": 0.3322529338881453,
+      "learning_rate": 6.402793946449361e-06,
+      "loss": 0.6786,
+      "step": 110
+    },
+    {
+      "epoch": 0.10712622263623661,
+      "grad_norm": 0.33345730335216767,
+      "learning_rate": 6.693830034924331e-06,
+      "loss": 0.6742,
+      "step": 115
+    },
+    {
+      "epoch": 0.11178388448998602,
+      "grad_norm": 0.3480904112152002,
+      "learning_rate": 6.984866123399301e-06,
+      "loss": 0.6843,
+      "step": 120
+    },
+    {
+      "epoch": 0.11644154634373545,
+      "grad_norm": 0.35538266924148043,
+      "learning_rate": 7.275902211874273e-06,
+      "loss": 0.6553,
+      "step": 125
+    },
+    {
+      "epoch": 0.12109920819748486,
+      "grad_norm": 0.36029823100280123,
+      "learning_rate": 7.5669383003492435e-06,
+      "loss": 0.6565,
+      "step": 130
+    },
+    {
+      "epoch": 0.1257568700512343,
+      "grad_norm": 0.3882208414109668,
+      "learning_rate": 7.857974388824214e-06,
+      "loss": 0.6634,
+      "step": 135
+    },
+    {
+      "epoch": 0.1304145319049837,
+      "grad_norm": 0.37180571772425014,
+      "learning_rate": 8.149010477299185e-06,
+      "loss": 0.6519,
+      "step": 140
+    },
+    {
+      "epoch": 0.1350721937587331,
+      "grad_norm": 0.3776764291908277,
+      "learning_rate": 8.440046565774158e-06,
+      "loss": 0.663,
+      "step": 145
+    },
+    {
+      "epoch": 0.13972985561248252,
+      "grad_norm": 0.37473220243738176,
+      "learning_rate": 8.731082654249127e-06,
+      "loss": 0.6526,
+      "step": 150
+    },
+    {
+      "epoch": 0.14438751746623196,
+      "grad_norm": 0.4110351084409799,
+      "learning_rate": 9.022118742724098e-06,
+      "loss": 0.6283,
+      "step": 155
+    },
+    {
+      "epoch": 0.14904517931998137,
+      "grad_norm": 0.4199064132815328,
+      "learning_rate": 9.313154831199069e-06,
+      "loss": 0.6344,
+      "step": 160
+    },
+    {
+      "epoch": 0.1537028411737308,
+      "grad_norm": 0.36807908599975947,
+      "learning_rate": 9.60419091967404e-06,
+      "loss": 0.6343,
+      "step": 165
+    },
+    {
+      "epoch": 0.1583605030274802,
+      "grad_norm": 0.3840459636079555,
+      "learning_rate": 9.895227008149011e-06,
+      "loss": 0.6303,
+      "step": 170
+    },
+    {
+      "epoch": 0.1630181648812296,
+      "grad_norm": 0.4201103965332282,
+      "learning_rate": 1.0186263096623982e-05,
+      "loss": 0.6418,
+      "step": 175
+    },
+    {
+      "epoch": 0.16767582673497905,
+      "grad_norm": 0.4114690700055479,
+      "learning_rate": 1.0477299185098953e-05,
+      "loss": 0.6335,
+      "step": 180
+    },
+    {
+      "epoch": 0.17233348858872846,
+      "grad_norm": 0.4136408854558498,
+      "learning_rate": 1.0768335273573923e-05,
+      "loss": 0.6299,
+      "step": 185
+    },
+    {
+      "epoch": 0.17699115044247787,
+      "grad_norm": 0.42259582362571485,
+      "learning_rate": 1.1059371362048894e-05,
+      "loss": 0.6224,
+      "step": 190
+    },
+    {
+      "epoch": 0.18164881229622729,
+      "grad_norm": 0.4423031089666256,
+      "learning_rate": 1.1350407450523866e-05,
+      "loss": 0.6274,
+      "step": 195
+    },
+    {
+      "epoch": 0.18630647414997673,
+      "grad_norm": 0.403271923196446,
+      "learning_rate": 1.1641443538998836e-05,
+      "loss": 0.594,
+      "step": 200
+    },
+    {
+      "epoch": 0.19096413600372614,
+      "grad_norm": 0.436282244225205,
+      "learning_rate": 1.1932479627473807e-05,
+      "loss": 0.6187,
+      "step": 205
+    },
+    {
+      "epoch": 0.19562179785747555,
+      "grad_norm": 0.39792760577510866,
+      "learning_rate": 1.2223515715948778e-05,
+      "loss": 0.6059,
+      "step": 210
+    },
+    {
+      "epoch": 0.20027945971122496,
+      "grad_norm": 0.39812469275850443,
+      "learning_rate": 1.2514551804423749e-05,
+      "loss": 0.6181,
+      "step": 215
+    },
+    {
+      "epoch": 0.20493712156497437,
+      "grad_norm": 0.39376887358317875,
+      "learning_rate": 1.2805587892898722e-05,
+      "loss": 0.6108,
+      "step": 220
+    },
+    {
+      "epoch": 0.2095947834187238,
+      "grad_norm": 0.464652404653219,
+      "learning_rate": 1.309662398137369e-05,
+      "loss": 0.614,
+      "step": 225
+    },
+    {
+      "epoch": 0.21425244527247322,
+      "grad_norm": 0.4023604229121632,
+      "learning_rate": 1.3387660069848662e-05,
+      "loss": 0.6108,
+      "step": 230
+    },
+    {
+      "epoch": 0.21891010712622264,
+      "grad_norm": 0.4360445909537957,
+      "learning_rate": 1.3678696158323633e-05,
+      "loss": 0.6049,
+      "step": 235
+    },
+    {
+      "epoch": 0.22356776897997205,
+      "grad_norm": 0.4373905365842894,
+      "learning_rate": 1.3969732246798603e-05,
+      "loss": 0.603,
+      "step": 240
+    },
+    {
+      "epoch": 0.22822543083372146,
+      "grad_norm": 0.39179031021650224,
+      "learning_rate": 1.4260768335273575e-05,
+      "loss": 0.59,
+      "step": 245
+    },
+    {
+      "epoch": 0.2328830926874709,
+      "grad_norm": 0.4196746075080395,
+      "learning_rate": 1.4551804423748547e-05,
+      "loss": 0.5865,
+      "step": 250
+    },
+    {
+      "epoch": 0.2375407545412203,
+      "grad_norm": 0.4303813452229446,
+      "learning_rate": 1.4842840512223516e-05,
+      "loss": 0.5984,
+      "step": 255
+    },
+    {
+      "epoch": 0.24219841639496972,
+      "grad_norm": 0.47523169063383375,
+      "learning_rate": 1.5133876600698487e-05,
+      "loss": 0.6036,
+      "step": 260
+    },
+    {
+      "epoch": 0.24685607824871914,
+      "grad_norm": 0.4335910841568507,
+      "learning_rate": 1.5424912689173458e-05,
+      "loss": 0.5947,
+      "step": 265
+    },
+    {
+      "epoch": 0.2515137401024686,
+      "grad_norm": 0.41341980426138514,
+      "learning_rate": 1.5715948777648427e-05,
+      "loss": 0.5813,
+      "step": 270
+    },
+    {
+      "epoch": 0.25617140195621796,
+      "grad_norm": 0.5160767565953956,
+      "learning_rate": 1.60069848661234e-05,
+      "loss": 0.6011,
+      "step": 275
+    },
+    {
+      "epoch": 0.2608290638099674,
+      "grad_norm": 0.5529678248494052,
+      "learning_rate": 1.629802095459837e-05,
+      "loss": 0.5876,
+      "step": 280
+    },
+    {
+      "epoch": 0.26548672566371684,
+      "grad_norm": 0.47547593485478673,
+      "learning_rate": 1.6589057043073342e-05,
+      "loss": 0.584,
+      "step": 285
+    },
+    {
+      "epoch": 0.2701443875174662,
+      "grad_norm": 0.45222675935849044,
+      "learning_rate": 1.6880093131548315e-05,
+      "loss": 0.5823,
+      "step": 290
+    },
+    {
+      "epoch": 0.27480204937121566,
+      "grad_norm": 0.4510297094925413,
+      "learning_rate": 1.717112922002328e-05,
+      "loss": 0.5742,
+      "step": 295
+    },
+    {
+      "epoch": 0.27945971122496505,
+      "grad_norm": 0.5521894478552463,
+      "learning_rate": 1.7462165308498254e-05,
+      "loss": 0.5866,
+      "step": 300
+    },
+    {
+      "epoch": 0.2841173730787145,
+      "grad_norm": 0.44020499832083604,
+      "learning_rate": 1.7753201396973227e-05,
+      "loss": 0.5941,
+      "step": 305
+    },
+    {
+      "epoch": 0.2887750349324639,
+      "grad_norm": 0.4401227459843441,
+      "learning_rate": 1.8044237485448196e-05,
+      "loss": 0.5884,
+      "step": 310
+    },
+    {
+      "epoch": 0.2934326967862133,
+      "grad_norm": 0.42288303099486596,
+      "learning_rate": 1.833527357392317e-05,
+      "loss": 0.5751,
+      "step": 315
+    },
+    {
+      "epoch": 0.29809035863996275,
+      "grad_norm": 0.44248419559906405,
+      "learning_rate": 1.8626309662398138e-05,
+      "loss": 0.5843,
+      "step": 320
+    },
+    {
+      "epoch": 0.30274802049371213,
+      "grad_norm": 0.4192435308637777,
+      "learning_rate": 1.8917345750873107e-05,
+      "loss": 0.5868,
+      "step": 325
+    },
+    {
+      "epoch": 0.3074056823474616,
+      "grad_norm": 0.4499232353603837,
+      "learning_rate": 1.920838183934808e-05,
+      "loss": 0.5847,
+      "step": 330
+    },
+    {
+      "epoch": 0.312063344201211,
+      "grad_norm": 0.4256491052871394,
+      "learning_rate": 1.9499417927823053e-05,
+      "loss": 0.5645,
+      "step": 335
+    },
+    {
+      "epoch": 0.3167210060549604,
+      "grad_norm": 0.4862344891338101,
+      "learning_rate": 1.9790454016298022e-05,
+      "loss": 0.5799,
+      "step": 340
+    },
+    {
+      "epoch": 0.32137866790870984,
+      "grad_norm": 0.44409885524054266,
+      "learning_rate": 2.0081490104772992e-05,
+      "loss": 0.5676,
+      "step": 345
+    },
+    {
+      "epoch": 0.3260363297624592,
+      "grad_norm": 0.5273940404660536,
+      "learning_rate": 2.0372526193247964e-05,
+      "loss": 0.5705,
+      "step": 350
+    },
+    {
+      "epoch": 0.33069399161620866,
+      "grad_norm": 0.4506295177797624,
+      "learning_rate": 2.0663562281722934e-05,
+      "loss": 0.5663,
+      "step": 355
+    },
+    {
+      "epoch": 0.3353516534699581,
+      "grad_norm": 0.4429860512388572,
+      "learning_rate": 2.0954598370197907e-05,
+      "loss": 0.5777,
+      "step": 360
+    },
+    {
+      "epoch": 0.3400093153237075,
+      "grad_norm": 0.47551309129603747,
+      "learning_rate": 2.124563445867288e-05,
+      "loss": 0.5771,
+      "step": 365
+    },
+    {
+      "epoch": 0.3446669771774569,
+      "grad_norm": 0.5109560446100108,
+      "learning_rate": 2.1536670547147845e-05,
+      "loss": 0.588,
+      "step": 370
+    },
+    {
+      "epoch": 0.3493246390312063,
+      "grad_norm": 0.5013967031612557,
+      "learning_rate": 2.1827706635622818e-05,
+      "loss": 0.5805,
+      "step": 375
+    },
+    {
+      "epoch": 0.35398230088495575,
+      "grad_norm": 0.5029584087507263,
+      "learning_rate": 2.2118742724097787e-05,
+      "loss": 0.5655,
+      "step": 380
+    },
+    {
+      "epoch": 0.3586399627387052,
+      "grad_norm": 0.392267985885871,
+      "learning_rate": 2.240977881257276e-05,
+      "loss": 0.5588,
+      "step": 385
+    },
+    {
+      "epoch": 0.36329762459245457,
+      "grad_norm": 0.49346124052819235,
+      "learning_rate": 2.2700814901047733e-05,
+      "loss": 0.5654,
+      "step": 390
+    },
+    {
+      "epoch": 0.367955286446204,
+      "grad_norm": 0.5440269465150551,
+      "learning_rate": 2.2991850989522702e-05,
+      "loss": 0.5576,
+      "step": 395
+    },
+    {
+      "epoch": 0.37261294829995345,
+      "grad_norm": 0.47389647089010667,
+      "learning_rate": 2.3282887077997672e-05,
+      "loss": 0.5544,
+      "step": 400
+    },
+    {
+      "epoch": 0.37727061015370283,
+      "grad_norm": 0.5060217377835531,
+      "learning_rate": 2.3573923166472644e-05,
+      "loss": 0.5581,
+      "step": 405
+    },
+    {
+      "epoch": 0.3819282720074523,
+      "grad_norm": 0.4280491572297177,
+      "learning_rate": 2.3864959254947614e-05,
+      "loss": 0.5648,
+      "step": 410
+    },
+    {
+      "epoch": 0.38658593386120166,
+      "grad_norm": 0.4546099384450869,
+      "learning_rate": 2.4155995343422587e-05,
+      "loss": 0.5648,
+      "step": 415
+    },
+    {
+      "epoch": 0.3912435957149511,
+      "grad_norm": 0.44032956083221475,
+      "learning_rate": 2.4447031431897556e-05,
+      "loss": 0.5523,
+      "step": 420
+    },
+    {
+      "epoch": 0.39590125756870054,
+      "grad_norm": 0.4418506911683883,
+      "learning_rate": 2.4738067520372525e-05,
+      "loss": 0.5711,
+      "step": 425
+    },
+    {
+      "epoch": 0.4005589194224499,
+      "grad_norm": 0.521278363776306,
+      "learning_rate": 2.5029103608847498e-05,
+      "loss": 0.5643,
+      "step": 430
+    },
+    {
+      "epoch": 0.40521658127619936,
+      "grad_norm": 0.532595160062641,
+      "learning_rate": 2.532013969732247e-05,
+      "loss": 0.5665,
+      "step": 435
+    },
+    {
+      "epoch": 0.40987424312994875,
+      "grad_norm": 0.41035785696363714,
+      "learning_rate": 2.5611175785797444e-05,
+      "loss": 0.5776,
+      "step": 440
+    },
+    {
+      "epoch": 0.4145319049836982,
+      "grad_norm": 0.4116095623614095,
+      "learning_rate": 2.590221187427241e-05,
+      "loss": 0.5502,
+      "step": 445
+    },
+    {
+      "epoch": 0.4191895668374476,
+      "grad_norm": 0.5062712817242586,
+      "learning_rate": 2.619324796274738e-05,
+      "loss": 0.5552,
+      "step": 450
+    },
+    {
+      "epoch": 0.423847228691197,
+      "grad_norm": 0.5033640845621997,
+      "learning_rate": 2.6484284051222352e-05,
+      "loss": 0.5762,
+      "step": 455
+    },
+    {
+      "epoch": 0.42850489054494645,
+      "grad_norm": 0.5035654454052618,
+      "learning_rate": 2.6775320139697325e-05,
+      "loss": 0.5546,
+      "step": 460
+    },
+    {
+      "epoch": 0.43316255239869583,
+      "grad_norm": 0.4336291085369849,
+      "learning_rate": 2.7066356228172297e-05,
+      "loss": 0.5641,
+      "step": 465
+    },
+    {
+      "epoch": 0.43782021425244527,
+      "grad_norm": 0.4699748065337065,
+      "learning_rate": 2.7357392316647267e-05,
+      "loss": 0.5551,
+      "step": 470
+    },
+    {
+      "epoch": 0.4424778761061947,
+      "grad_norm": 0.5259808485283921,
+      "learning_rate": 2.7648428405122233e-05,
+      "loss": 0.5655,
+      "step": 475
+    },
+    {
+      "epoch": 0.4471355379599441,
+      "grad_norm": 0.4170440552225455,
+      "learning_rate": 2.7939464493597205e-05,
+      "loss": 0.5495,
+      "step": 480
+    },
+    {
+      "epoch": 0.45179319981369354,
+      "grad_norm": 0.4402168017776964,
+      "learning_rate": 2.8230500582072178e-05,
+      "loss": 0.5522,
+      "step": 485
+    },
+    {
+      "epoch": 0.4564508616674429,
+      "grad_norm": 0.4569052068490754,
+      "learning_rate": 2.852153667054715e-05,
+      "loss": 0.5565,
+      "step": 490
+    },
+    {
+      "epoch": 0.46110852352119236,
+      "grad_norm": 0.4978898390597469,
+      "learning_rate": 2.881257275902212e-05,
+      "loss": 0.5466,
+      "step": 495
+    },
+    {
+      "epoch": 0.4657661853749418,
+      "grad_norm": 0.4229733715419873,
+      "learning_rate": 2.9103608847497093e-05,
+      "loss": 0.5554,
+      "step": 500
+    },
+    {
+      "epoch": 0.4704238472286912,
+      "grad_norm": 0.43529337252388356,
+      "learning_rate": 2.939464493597206e-05,
+      "loss": 0.5482,
+      "step": 505
+    },
+    {
+      "epoch": 0.4750815090824406,
+      "grad_norm": 0.5160534559430204,
+      "learning_rate": 2.9685681024447032e-05,
+      "loss": 0.5444,
+      "step": 510
+    },
+    {
+      "epoch": 0.47973917093619,
+      "grad_norm": 0.443484452067975,
+      "learning_rate": 2.9976717112922005e-05,
+      "loss": 0.5572,
+      "step": 515
+    },
+    {
+      "epoch": 0.48439683278993945,
+      "grad_norm": 0.5944864965501068,
+      "learning_rate": 3.0267753201396974e-05,
+      "loss": 0.5524,
+      "step": 520
+    },
+    {
+      "epoch": 0.4890544946436889,
+      "grad_norm": 0.5071449662346628,
+      "learning_rate": 3.055878928987195e-05,
+      "loss": 0.5638,
+      "step": 525
+    },
+    {
+      "epoch": 0.49371215649743827,
+      "grad_norm": 0.4939766509456816,
+      "learning_rate": 3.0849825378346916e-05,
+      "loss": 0.5466,
+      "step": 530
+    },
+    {
+      "epoch": 0.4983698183511877,
+      "grad_norm": 0.454732349457497,
+      "learning_rate": 3.1140861466821885e-05,
+      "loss": 0.5533,
+      "step": 535
+    },
+    {
+      "epoch": 0.5030274802049371,
+      "grad_norm": 0.47844922111423144,
+      "learning_rate": 3.1431897555296855e-05,
+      "loss": 0.5575,
+      "step": 540
+    },
+    {
+      "epoch": 0.5076851420586865,
+      "grad_norm": 0.48361551844718786,
+      "learning_rate": 3.172293364377183e-05,
+      "loss": 0.5452,
+      "step": 545
+    },
+    {
+      "epoch": 0.5123428039124359,
+      "grad_norm": 0.5181559294500859,
+      "learning_rate": 3.20139697322468e-05,
+      "loss": 0.5618,
+      "step": 550
+    },
+    {
+      "epoch": 0.5170004657661854,
+      "grad_norm": 0.4899936584071396,
+      "learning_rate": 3.2305005820721776e-05,
+      "loss": 0.5481,
+      "step": 555
+    },
+    {
+      "epoch": 0.5216581276199348,
+      "grad_norm": 0.5506585569549096,
+      "learning_rate": 3.259604190919674e-05,
+      "loss": 0.5538,
+      "step": 560
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.6754186581156059,
+      "learning_rate": 3.288707799767171e-05,
+      "loss": 0.5657,
+      "step": 565
+    },
+    {
+      "epoch": 0.5309734513274337,
+      "grad_norm": 0.6458262406019906,
+      "learning_rate": 3.3178114086146685e-05,
+      "loss": 0.5804,
+      "step": 570
+    },
+    {
+      "epoch": 0.5356311131811831,
+      "grad_norm": 0.5677532060663508,
+      "learning_rate": 3.3469150174621654e-05,
+      "loss": 0.5509,
+      "step": 575
+    },
+    {
+      "epoch": 0.5402887750349324,
+      "grad_norm": 0.5140073417715375,
+      "learning_rate": 3.376018626309663e-05,
+      "loss": 0.5498,
+      "step": 580
+    },
+    {
+      "epoch": 0.5449464368886818,
+      "grad_norm": 0.47973716619213574,
+      "learning_rate": 3.40512223515716e-05,
+      "loss": 0.5437,
+      "step": 585
+    },
+    {
+      "epoch": 0.5496040987424313,
+      "grad_norm": 2.850114498375805,
+      "learning_rate": 3.434225844004656e-05,
+      "loss": 0.5621,
+      "step": 590
+    },
+    {
+      "epoch": 0.5542617605961807,
+      "grad_norm": 0.7132559839205367,
+      "learning_rate": 3.463329452852154e-05,
+      "loss": 0.6687,
+      "step": 595
+    },
+    {
+      "epoch": 0.5589194224499301,
+      "grad_norm": 0.6423899235862081,
+      "learning_rate": 3.492433061699651e-05,
+      "loss": 0.5595,
+      "step": 600
+    },
+    {
+      "epoch": 0.5635770843036796,
+      "grad_norm": 0.698995384460383,
+      "learning_rate": 3.5215366705471484e-05,
+      "loss": 0.5625,
+      "step": 605
+    },
+    {
+      "epoch": 0.568234746157429,
+      "grad_norm": 0.4903608237478288,
+      "learning_rate": 3.550640279394645e-05,
+      "loss": 0.5516,
+      "step": 610
+    },
+    {
+      "epoch": 0.5728924080111784,
+      "grad_norm": 4.379274402774887,
+      "learning_rate": 3.579743888242142e-05,
+      "loss": 0.5441,
+      "step": 615
+    },
+    {
+      "epoch": 0.5775500698649279,
+      "grad_norm": 0.512657850098662,
+      "learning_rate": 3.608847497089639e-05,
+      "loss": 0.5458,
+      "step": 620
+    },
+    {
+      "epoch": 0.5822077317186772,
+      "grad_norm": 0.43346421191572665,
+      "learning_rate": 3.637951105937136e-05,
+      "loss": 0.5569,
+      "step": 625
+    },
+    {
+      "epoch": 0.5868653935724266,
+      "grad_norm": 0.531831351876456,
+      "learning_rate": 3.667054714784634e-05,
+      "loss": 0.5499,
+      "step": 630
+    },
+    {
+      "epoch": 0.5915230554261761,
+      "grad_norm": 0.6622562456603708,
+      "learning_rate": 3.696158323632131e-05,
+      "loss": 0.5594,
+      "step": 635
+    },
+    {
+      "epoch": 0.5961807172799255,
+      "grad_norm": 0.44507715538737447,
+      "learning_rate": 3.7252619324796276e-05,
+      "loss": 0.5487,
+      "step": 640
+    },
+    {
+      "epoch": 0.6008383791336749,
+      "grad_norm": 5.752742943024781,
+      "learning_rate": 3.7543655413271246e-05,
+      "loss": 0.5868,
+      "step": 645
+    },
+    {
+      "epoch": 0.6054960409874243,
+      "grad_norm": 0.5592925415997116,
+      "learning_rate": 3.7834691501746215e-05,
+      "loss": 0.545,
+      "step": 650
+    },
+    {
+      "epoch": 0.6101537028411738,
+      "grad_norm": 0.3957341710601061,
+      "learning_rate": 3.812572759022119e-05,
+      "loss": 0.5493,
+      "step": 655
+    },
+    {
+      "epoch": 0.6148113646949231,
+      "grad_norm": 0.45043284107748643,
+      "learning_rate": 3.841676367869616e-05,
+      "loss": 0.5519,
+      "step": 660
+    },
+    {
+      "epoch": 0.6194690265486725,
+      "grad_norm": 0.4597678841607699,
+      "learning_rate": 3.870779976717113e-05,
+      "loss": 0.5451,
+      "step": 665
+    },
+    {
+      "epoch": 0.624126688402422,
+      "grad_norm": 0.4401269476751836,
+      "learning_rate": 3.8998835855646106e-05,
+      "loss": 0.5478,
+      "step": 670
+    },
+    {
+      "epoch": 0.6287843502561714,
+      "grad_norm": 0.4517164819813474,
+      "learning_rate": 3.928987194412107e-05,
+      "loss": 0.5377,
+      "step": 675
+    },
+    {
+      "epoch": 0.6334420121099208,
+      "grad_norm": 0.42856901369174305,
+      "learning_rate": 3.9580908032596045e-05,
+      "loss": 0.5419,
+      "step": 680
+    },
+    {
+      "epoch": 0.6380996739636703,
+      "grad_norm": 0.4778106030676651,
+      "learning_rate": 3.9871944121071014e-05,
+      "loss": 0.5557,
+      "step": 685
+    },
+    {
+      "epoch": 0.6427573358174197,
+      "grad_norm": 0.4810252200454141,
+      "learning_rate": 4.0162980209545983e-05,
+      "loss": 0.5424,
+      "step": 690
+    },
+    {
+      "epoch": 0.6474149976711691,
+      "grad_norm": 0.42113883038821803,
+      "learning_rate": 4.045401629802096e-05,
+      "loss": 0.5421,
+      "step": 695
+    },
+    {
+      "epoch": 0.6520726595249184,
+      "grad_norm": 0.48920856808484403,
+      "learning_rate": 4.074505238649593e-05,
+      "loss": 0.5391,
+      "step": 700
+    },
+    {
+      "epoch": 0.6567303213786679,
+      "grad_norm": 0.48208747773248684,
+      "learning_rate": 4.10360884749709e-05,
+      "loss": 0.5354,
+      "step": 705
+    },
+    {
+      "epoch": 0.6613879832324173,
+      "grad_norm": 0.4687178891717093,
+      "learning_rate": 4.132712456344587e-05,
+      "loss": 0.5485,
+      "step": 710
+    },
+    {
+      "epoch": 0.6660456450861667,
+      "grad_norm": 0.45466895427802073,
+      "learning_rate": 4.161816065192084e-05,
+      "loss": 0.5429,
+      "step": 715
+    },
+    {
+      "epoch": 0.6707033069399162,
+      "grad_norm": 0.40939740583021633,
+      "learning_rate": 4.190919674039581e-05,
+      "loss": 0.5311,
+      "step": 720
+    },
+    {
+      "epoch": 0.6753609687936656,
+      "grad_norm": 0.4026942565114344,
+      "learning_rate": 4.220023282887078e-05,
+      "loss": 0.5365,
+      "step": 725
+    },
+    {
+      "epoch": 0.680018630647415,
+      "grad_norm": 0.44458418781623743,
+      "learning_rate": 4.249126891734576e-05,
+      "loss": 0.5445,
+      "step": 730
+    },
+    {
+      "epoch": 0.6846762925011645,
+      "grad_norm": 0.45908128679549104,
+      "learning_rate": 4.278230500582072e-05,
+      "loss": 0.5318,
+      "step": 735
+    },
+    {
+      "epoch": 0.6893339543549138,
+      "grad_norm": 0.4336786017781467,
+      "learning_rate": 4.307334109429569e-05,
+      "loss": 0.5392,
+      "step": 740
+    },
+    {
+      "epoch": 0.6939916162086632,
+      "grad_norm": 0.6430149710986586,
+      "learning_rate": 4.336437718277067e-05,
+      "loss": 0.5396,
+      "step": 745
+    },
+    {
+      "epoch": 0.6986492780624126,
+      "grad_norm": 0.5060406772414919,
+      "learning_rate": 4.3655413271245636e-05,
+      "loss": 0.542,
+      "step": 750
+    },
+    {
+      "epoch": 0.7033069399161621,
+      "grad_norm": 0.43575745452775133,
+      "learning_rate": 4.394644935972061e-05,
+      "loss": 0.5287,
+      "step": 755
+    },
+    {
+      "epoch": 0.7079646017699115,
+      "grad_norm": 0.4509480021266233,
+      "learning_rate": 4.4237485448195575e-05,
+      "loss": 0.5312,
+      "step": 760
+    },
+    {
+      "epoch": 0.7126222636236609,
+      "grad_norm": 0.39247296762258177,
+      "learning_rate": 4.452852153667055e-05,
+      "loss": 0.5428,
+      "step": 765
+    },
+    {
+      "epoch": 0.7172799254774104,
+      "grad_norm": 0.43065576347919365,
+      "learning_rate": 4.481955762514552e-05,
+      "loss": 0.5408,
+      "step": 770
+    },
+    {
+      "epoch": 0.7219375873311598,
+      "grad_norm": 0.4968073164264031,
+      "learning_rate": 4.511059371362049e-05,
+      "loss": 0.5277,
+      "step": 775
+    },
+    {
+      "epoch": 0.7265952491849091,
+      "grad_norm": 0.4752250835263456,
+      "learning_rate": 4.5401629802095466e-05,
+      "loss": 0.5461,
+      "step": 780
+    },
+    {
+      "epoch": 0.7312529110386586,
+      "grad_norm": 0.3917023142794256,
+      "learning_rate": 4.5692665890570435e-05,
+      "loss": 0.539,
+      "step": 785
+    },
+    {
+      "epoch": 0.735910572892408,
+      "grad_norm": 0.42438394966515364,
+      "learning_rate": 4.5983701979045405e-05,
+      "loss": 0.5303,
+      "step": 790
+    },
+    {
+      "epoch": 0.7405682347461574,
+      "grad_norm": 0.39706636448503807,
+      "learning_rate": 4.6274738067520374e-05,
+      "loss": 0.5371,
+      "step": 795
+    },
+    {
+      "epoch": 0.7452258965999069,
+      "grad_norm": 0.4034247813868715,
+      "learning_rate": 4.6565774155995343e-05,
+      "loss": 0.5334,
+      "step": 800
+    },
+    {
+      "epoch": 0.7498835584536563,
+      "grad_norm": 0.38931923519454176,
+      "learning_rate": 4.685681024447032e-05,
+      "loss": 0.5494,
+      "step": 805
+    },
+    {
+      "epoch": 0.7545412203074057,
+      "grad_norm": 0.4126988813279989,
+      "learning_rate": 4.714784633294529e-05,
+      "loss": 0.5329,
+      "step": 810
+    },
+    {
+      "epoch": 0.759198882161155,
+      "grad_norm": 0.41847891492257616,
+      "learning_rate": 4.743888242142026e-05,
+      "loss": 0.5368,
+      "step": 815
+    },
+    {
+      "epoch": 0.7638565440149045,
+      "grad_norm": 0.37832989246330034,
+      "learning_rate": 4.772991850989523e-05,
+      "loss": 0.5368,
+      "step": 820
+    },
+    {
+      "epoch": 0.7685142058686539,
+      "grad_norm": 0.4264254827761458,
+      "learning_rate": 4.80209545983702e-05,
+      "loss": 0.5222,
+      "step": 825
+    },
+    {
+      "epoch": 0.7731718677224033,
+      "grad_norm": 0.4215105843173847,
+      "learning_rate": 4.831199068684517e-05,
+      "loss": 0.5321,
+      "step": 830
+    },
+    {
+      "epoch": 0.7778295295761528,
+      "grad_norm": 0.40779532756054626,
+      "learning_rate": 4.860302677532014e-05,
+      "loss": 0.5386,
+      "step": 835
+    },
+    {
+      "epoch": 0.7824871914299022,
+      "grad_norm": 0.43656255019856294,
+      "learning_rate": 4.889406286379511e-05,
+      "loss": 0.5229,
+      "step": 840
+    },
+    {
+      "epoch": 0.7871448532836516,
+      "grad_norm": 0.46094480601401433,
+      "learning_rate": 4.918509895227008e-05,
+      "loss": 0.5301,
+      "step": 845
+    },
+    {
+      "epoch": 0.7918025151374011,
+      "grad_norm": 0.4382225357931327,
+      "learning_rate": 4.947613504074505e-05,
+      "loss": 0.5324,
+      "step": 850
+    },
+    {
+      "epoch": 0.7964601769911505,
+      "grad_norm": 0.37445428395677366,
+      "learning_rate": 4.976717112922003e-05,
+      "loss": 0.5215,
+      "step": 855
+    },
+    {
+      "epoch": 0.8011178388448998,
+      "grad_norm": 0.4172807893879857,
+      "learning_rate": 4.9993527508090615e-05,
+      "loss": 0.5264,
+      "step": 860
+    },
+    {
+      "epoch": 0.8057755006986492,
+      "grad_norm": 0.42493387898618923,
+      "learning_rate": 4.996116504854369e-05,
+      "loss": 0.5289,
+      "step": 865
+    },
+    {
+      "epoch": 0.8104331625523987,
+      "grad_norm": 0.4793713302760421,
+      "learning_rate": 4.992880258899677e-05,
+      "loss": 0.522,
+      "step": 870
+    },
+    {
+      "epoch": 0.8150908244061481,
+      "grad_norm": 0.42581510639579395,
+      "learning_rate": 4.989644012944984e-05,
+      "loss": 0.5353,
+      "step": 875
+    },
+    {
+      "epoch": 0.8197484862598975,
+      "grad_norm": 0.39346573629830267,
+      "learning_rate": 4.9864077669902914e-05,
+      "loss": 0.5143,
+      "step": 880
+    },
+    {
+      "epoch": 0.824406148113647,
+      "grad_norm": 0.46479332953603936,
+      "learning_rate": 4.983171521035599e-05,
+      "loss": 0.5343,
+      "step": 885
+    },
+    {
+      "epoch": 0.8290638099673964,
+      "grad_norm": 0.38234965919360414,
+      "learning_rate": 4.979935275080906e-05,
+      "loss": 0.5323,
+      "step": 890
+    },
+    {
+      "epoch": 0.8337214718211458,
+      "grad_norm": 0.38452753250908844,
+      "learning_rate": 4.976699029126214e-05,
+      "loss": 0.5323,
+      "step": 895
+    },
+    {
+      "epoch": 0.8383791336748952,
+      "grad_norm": 0.3558793576496668,
+      "learning_rate": 4.9734627831715214e-05,
+      "loss": 0.5258,
+      "step": 900
+    },
+    {
+      "epoch": 0.8430367955286446,
+      "grad_norm": 0.40705429734477666,
+      "learning_rate": 4.970226537216829e-05,
+      "loss": 0.5293,
+      "step": 905
+    },
+    {
+      "epoch": 0.847694457382394,
+      "grad_norm": 0.3681084398657872,
+      "learning_rate": 4.966990291262136e-05,
+      "loss": 0.5377,
+      "step": 910
+    },
+    {
+      "epoch": 0.8523521192361434,
+      "grad_norm": 0.4451189523688905,
+      "learning_rate": 4.963754045307444e-05,
+      "loss": 0.526,
+      "step": 915
+    },
+    {
+      "epoch": 0.8570097810898929,
+      "grad_norm": 0.4098553913425108,
+      "learning_rate": 4.9605177993527513e-05,
+      "loss": 0.5358,
+      "step": 920
+    },
+    {
+      "epoch": 0.8616674429436423,
+      "grad_norm": 0.4210630886877289,
+      "learning_rate": 4.957281553398058e-05,
+      "loss": 0.5285,
+      "step": 925
+    },
+    {
+      "epoch": 0.8663251047973917,
+      "grad_norm": 0.5682054260649863,
+      "learning_rate": 4.954045307443366e-05,
+      "loss": 0.531,
+      "step": 930
+    },
+    {
+      "epoch": 0.8709827666511412,
+      "grad_norm": 0.4003926309716626,
+      "learning_rate": 4.950809061488673e-05,
+      "loss": 0.516,
+      "step": 935
+    },
+    {
+      "epoch": 0.8756404285048905,
+      "grad_norm": 0.36583126445264214,
+      "learning_rate": 4.9475728155339806e-05,
+      "loss": 0.5147,
+      "step": 940
+    },
+    {
+      "epoch": 0.8802980903586399,
+      "grad_norm": 0.34710755584775155,
+      "learning_rate": 4.944336569579288e-05,
+      "loss": 0.521,
+      "step": 945
+    },
+    {
+      "epoch": 0.8849557522123894,
+      "grad_norm": 0.33969459451512213,
+      "learning_rate": 4.941100323624596e-05,
+      "loss": 0.508,
+      "step": 950
+    },
+    {
+      "epoch": 0.8896134140661388,
+      "grad_norm": 0.3667209891990349,
+      "learning_rate": 4.937864077669903e-05,
+      "loss": 0.5247,
+      "step": 955
+    },
+    {
+      "epoch": 0.8942710759198882,
+      "grad_norm": 0.3628895818072265,
+      "learning_rate": 4.9346278317152106e-05,
+      "loss": 0.5301,
+      "step": 960
+    },
+    {
+      "epoch": 0.8989287377736377,
+      "grad_norm": 0.3743557642518737,
+      "learning_rate": 4.931391585760518e-05,
+      "loss": 0.5236,
+      "step": 965
+    },
+    {
+      "epoch": 0.9035863996273871,
+      "grad_norm": 0.3571762409530922,
+      "learning_rate": 4.928155339805826e-05,
+      "loss": 0.527,
+      "step": 970
+    },
+    {
+      "epoch": 0.9082440614811365,
+      "grad_norm": 0.43902683870985737,
+      "learning_rate": 4.924919093851133e-05,
+      "loss": 0.5212,
+      "step": 975
+    },
+    {
+      "epoch": 0.9129017233348858,
+      "grad_norm": 0.39354640530264057,
+      "learning_rate": 4.9216828478964405e-05,
+      "loss": 0.5244,
+      "step": 980
+    },
+    {
+      "epoch": 0.9175593851886353,
+      "grad_norm": 0.3496362068420844,
+      "learning_rate": 4.918446601941748e-05,
+      "loss": 0.5295,
+      "step": 985
+    },
+    {
+      "epoch": 0.9222170470423847,
+      "grad_norm": 0.4353930009984086,
+      "learning_rate": 4.915210355987055e-05,
+      "loss": 0.4976,
+      "step": 990
+    },
+    {
+      "epoch": 0.9268747088961341,
+      "grad_norm": 0.3566344310918772,
+      "learning_rate": 4.911974110032363e-05,
+      "loss": 0.5152,
+      "step": 995
+    },
+    {
+      "epoch": 0.9315323707498836,
+      "grad_norm": 0.38805577658270657,
+      "learning_rate": 4.90873786407767e-05,
+      "loss": 0.519,
+      "step": 1000
+    },
+    {
+      "epoch": 0.936190032603633,
+      "grad_norm": 0.3883213881506984,
+      "learning_rate": 4.9055016181229774e-05,
+      "loss": 0.5304,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9408476944573824,
+      "grad_norm": 0.3729784939764993,
+      "learning_rate": 4.902265372168285e-05,
+      "loss": 0.5162,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9455053563111319,
+      "grad_norm": 0.39356886816717457,
+      "learning_rate": 4.899029126213592e-05,
+      "loss": 0.5194,
+      "step": 1015
+    },
+    {
+      "epoch": 0.9501630181648812,
+      "grad_norm": 0.37144096012401695,
+      "learning_rate": 4.8957928802589e-05,
+      "loss": 0.5292,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9548206800186306,
+      "grad_norm": 0.3960035947632593,
+      "learning_rate": 4.8925566343042074e-05,
+      "loss": 0.5193,
+      "step": 1025
+    },
+    {
+      "epoch": 0.95947834187238,
+      "grad_norm": 0.3625631338673291,
+      "learning_rate": 4.889320388349515e-05,
+      "loss": 0.5099,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9641360037261295,
+      "grad_norm": 0.37924346646168744,
+      "learning_rate": 4.886084142394822e-05,
+      "loss": 0.5146,
+      "step": 1035
+    },
+    {
+      "epoch": 0.9687936655798789,
+      "grad_norm": 0.35148813999705747,
+      "learning_rate": 4.88284789644013e-05,
+      "loss": 0.5117,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9734513274336283,
+      "grad_norm": 0.33290102128014737,
+      "learning_rate": 4.879611650485437e-05,
+      "loss": 0.514,
+      "step": 1045
+    },
+    {
+      "epoch": 0.9781089892873778,
+      "grad_norm": 0.36029983492128725,
+      "learning_rate": 4.876375404530745e-05,
+      "loss": 0.5128,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9827666511411272,
+      "grad_norm": 0.39114010829978446,
+      "learning_rate": 4.873139158576052e-05,
+      "loss": 0.5155,
+      "step": 1055
+    },
+    {
+      "epoch": 0.9874243129948765,
+      "grad_norm": 0.39656745624556927,
+      "learning_rate": 4.8699029126213596e-05,
+      "loss": 0.5187,
+      "step": 1060
+    },
+    {
+      "epoch": 0.992081974848626,
+      "grad_norm": 0.4294193162930166,
+      "learning_rate": 4.866666666666667e-05,
+      "loss": 0.5163,
+      "step": 1065
+    },
+    {
+      "epoch": 0.9967396367023754,
+      "grad_norm": 0.3982263305944793,
+      "learning_rate": 4.863430420711974e-05,
+      "loss": 0.5061,
+      "step": 1070
+    },
+    {
+      "epoch": 1.00093153237075,
+      "grad_norm": 0.5928038015870168,
+      "learning_rate": 4.860194174757281e-05,
+      "loss": 0.5253,
+      "step": 1075
+    },
+    {
+      "epoch": 1.0055891942244992,
+      "grad_norm": 0.45617357290049604,
+      "learning_rate": 4.856957928802589e-05,
+      "loss": 0.4908,
+      "step": 1080
+    },
+    {
+      "epoch": 1.0102468560782487,
+      "grad_norm": 0.39682435507090447,
+      "learning_rate": 4.8537216828478965e-05,
+      "loss": 0.4762,
+      "step": 1085
+    },
+    {
+      "epoch": 1.0149045179319982,
+      "grad_norm": 0.3918327414704858,
+      "learning_rate": 4.850485436893204e-05,
+      "loss": 0.4958,
+      "step": 1090
+    },
+    {
+      "epoch": 1.0195621797857475,
+      "grad_norm": 0.3836278471157271,
+      "learning_rate": 4.847249190938511e-05,
+      "loss": 0.4921,
+      "step": 1095
+    },
+    {
+      "epoch": 1.024219841639497,
+      "grad_norm": 0.3563021906682752,
+      "learning_rate": 4.844012944983819e-05,
+      "loss": 0.4944,
+      "step": 1100
+    },
+    {
+      "epoch": 1.0288775034932465,
+      "grad_norm": 0.35054296656861605,
+      "learning_rate": 4.8407766990291265e-05,
+      "loss": 0.4916,
+      "step": 1105
+    },
+    {
+      "epoch": 1.0335351653469957,
+      "grad_norm": 0.4068508503935885,
+      "learning_rate": 4.837540453074434e-05,
+      "loss": 0.4887,
+      "step": 1110
+    },
+    {
+      "epoch": 1.0381928272007452,
+      "grad_norm": 0.38370359471299786,
+      "learning_rate": 4.834304207119741e-05,
+      "loss": 0.4801,
+      "step": 1115
+    },
+    {
+      "epoch": 1.0428504890544947,
+      "grad_norm": 0.434600802675546,
+      "learning_rate": 4.831067961165049e-05,
+      "loss": 0.4904,
+      "step": 1120
+    },
+    {
+      "epoch": 1.047508150908244,
+      "grad_norm": 0.34054843848982447,
+      "learning_rate": 4.8278317152103564e-05,
+      "loss": 0.4967,
+      "step": 1125
+    },
+    {
+      "epoch": 1.0521658127619935,
+      "grad_norm": 0.3811863603092705,
+      "learning_rate": 4.824595469255664e-05,
+      "loss": 0.4838,
+      "step": 1130
+    },
+    {
+      "epoch": 1.056823474615743,
+      "grad_norm": 0.36035159110419396,
+      "learning_rate": 4.821359223300971e-05,
+      "loss": 0.4906,
+      "step": 1135
+    },
+    {
+      "epoch": 1.0614811364694923,
+      "grad_norm": 0.36681536116027774,
+      "learning_rate": 4.818122977346279e-05,
+      "loss": 0.4906,
+      "step": 1140
+    },
+    {
+      "epoch": 1.0661387983232418,
+      "grad_norm": 0.38080197685374506,
+      "learning_rate": 4.814886731391586e-05,
+      "loss": 0.4898,
+      "step": 1145
+    },
+    {
+      "epoch": 1.0707964601769913,
+      "grad_norm": 0.3583265769522901,
+      "learning_rate": 4.8116504854368934e-05,
+      "loss": 0.4886,
+      "step": 1150
+    },
+    {
+      "epoch": 1.0754541220307405,
+      "grad_norm": 0.4361753476463697,
+      "learning_rate": 4.808414239482201e-05,
+      "loss": 0.4902,
+      "step": 1155
+    },
+    {
+      "epoch": 1.08011178388449,
+      "grad_norm": 0.4099438655688987,
+      "learning_rate": 4.805177993527508e-05,
+      "loss": 0.4985,
+      "step": 1160
+    },
+    {
+      "epoch": 1.0847694457382393,
+      "grad_norm": 0.3625829222877311,
+      "learning_rate": 4.801941747572816e-05,
+      "loss": 0.4962,
+      "step": 1165
+    },
+    {
+      "epoch": 1.0894271075919888,
+      "grad_norm": 0.35932190699829,
+      "learning_rate": 4.798705501618123e-05,
+      "loss": 0.4922,
+      "step": 1170
+    },
+    {
+      "epoch": 1.0940847694457383,
+      "grad_norm": 0.38128309499882657,
+      "learning_rate": 4.795469255663431e-05,
+      "loss": 0.4944,
+      "step": 1175
+    },
+    {
+      "epoch": 1.0987424312994876,
+      "grad_norm": 0.3991505478965423,
+      "learning_rate": 4.792233009708738e-05,
+      "loss": 0.4856,
+      "step": 1180
+    },
+    {
+      "epoch": 1.103400093153237,
+      "grad_norm": 0.37327909814375987,
+      "learning_rate": 4.7889967637540456e-05,
+      "loss": 0.4835,
+      "step": 1185
+    },
+    {
+      "epoch": 1.1080577550069866,
+      "grad_norm": 0.33891412197653836,
+      "learning_rate": 4.785760517799353e-05,
+      "loss": 0.4658,
+      "step": 1190
+    },
+    {
+      "epoch": 1.1127154168607358,
+      "grad_norm": 0.3474094654759008,
+      "learning_rate": 4.78252427184466e-05,
+      "loss": 0.4768,
+      "step": 1195
+    },
+    {
+      "epoch": 1.1173730787144853,
+      "grad_norm": 0.36244721849020517,
+      "learning_rate": 4.779288025889968e-05,
+      "loss": 0.4836,
+      "step": 1200
+    },
+    {
+      "epoch": 1.1220307405682348,
+      "grad_norm": 0.4033010179698364,
+      "learning_rate": 4.7760517799352756e-05,
+      "loss": 0.497,
+      "step": 1205
+    },
+    {
+      "epoch": 1.126688402421984,
+      "grad_norm": 0.37809180198013964,
+      "learning_rate": 4.772815533980583e-05,
+      "loss": 0.4916,
+      "step": 1210
+    },
+    {
+      "epoch": 1.1313460642757336,
+      "grad_norm": 0.36933723525949036,
+      "learning_rate": 4.76957928802589e-05,
+      "loss": 0.4858,
+      "step": 1215
+    },
+    {
+      "epoch": 1.136003726129483,
+      "grad_norm": 0.36142771353078024,
+      "learning_rate": 4.766343042071197e-05,
+      "loss": 0.4938,
+      "step": 1220
+    },
+    {
+      "epoch": 1.1406613879832324,
+      "grad_norm": 0.39343722932343206,
+      "learning_rate": 4.763106796116505e-05,
+      "loss": 0.4846,
+      "step": 1225
+    },
+    {
+      "epoch": 1.1453190498369819,
+      "grad_norm": 0.38536600679449245,
+      "learning_rate": 4.7598705501618125e-05,
+      "loss": 0.4826,
+      "step": 1230
+    },
+    {
+      "epoch": 1.1499767116907313,
+      "grad_norm": 0.3208835229721257,
+      "learning_rate": 4.75663430420712e-05,
+      "loss": 0.4875,
+      "step": 1235
+    },
+    {
+      "epoch": 1.1546343735444806,
+      "grad_norm": 0.3045647033657388,
+      "learning_rate": 4.753398058252427e-05,
+      "loss": 0.4965,
+      "step": 1240
+    },
+    {
+      "epoch": 1.1592920353982301,
+      "grad_norm": 0.3330554785430036,
+      "learning_rate": 4.750161812297735e-05,
+      "loss": 0.4906,
+      "step": 1245
+    },
+    {
+      "epoch": 1.1639496972519794,
+      "grad_norm": 0.3751461447400107,
+      "learning_rate": 4.7469255663430424e-05,
+      "loss": 0.5014,
+      "step": 1250
+    },
+    {
+      "epoch": 1.1686073591057289,
+      "grad_norm": 0.34075871643973377,
+      "learning_rate": 4.74368932038835e-05,
+      "loss": 0.4878,
+      "step": 1255
+    },
+    {
+      "epoch": 1.1732650209594784,
+      "grad_norm": 0.3513756614443728,
+      "learning_rate": 4.740453074433657e-05,
+      "loss": 0.4967,
+      "step": 1260
+    },
+    {
+      "epoch": 1.1779226828132279,
+      "grad_norm": 0.3364889150024333,
+      "learning_rate": 4.737216828478965e-05,
+      "loss": 0.4827,
+      "step": 1265
+    },
+    {
+      "epoch": 1.1825803446669771,
+      "grad_norm": 0.32836418576161563,
+      "learning_rate": 4.7339805825242724e-05,
+      "loss": 0.475,
+      "step": 1270
+    },
+    {
+      "epoch": 1.1872380065207266,
+      "grad_norm": 0.34996683855701755,
+      "learning_rate": 4.73074433656958e-05,
+      "loss": 0.4812,
+      "step": 1275
+    },
+    {
+      "epoch": 1.191895668374476,
+      "grad_norm": 0.4025013401969635,
+      "learning_rate": 4.727508090614887e-05,
+      "loss": 0.4918,
+      "step": 1280
+    },
+    {
+      "epoch": 1.1965533302282254,
+      "grad_norm": 0.3401225034506895,
+      "learning_rate": 4.724271844660194e-05,
+      "loss": 0.4926,
+      "step": 1285
+    },
+    {
+      "epoch": 1.201210992081975,
+      "grad_norm": 0.31003980724430014,
+      "learning_rate": 4.7210355987055017e-05,
+      "loss": 0.4791,
+      "step": 1290
+    },
+    {
+      "epoch": 1.2058686539357242,
+      "grad_norm": 0.35347481534080316,
+      "learning_rate": 4.717799352750809e-05,
+      "loss": 0.485,
+      "step": 1295
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.35989687260611175,
+      "learning_rate": 4.714563106796116e-05,
+      "loss": 0.4804,
+      "step": 1300
+    },
+    {
+      "epoch": 1.2151839776432232,
+      "grad_norm": 0.35019515606373103,
+      "learning_rate": 4.711326860841424e-05,
+      "loss": 0.4915,
+      "step": 1305
+    },
+    {
+      "epoch": 1.2198416394969724,
+      "grad_norm": 0.3586166735739774,
+      "learning_rate": 4.7080906148867316e-05,
+      "loss": 0.4943,
+      "step": 1310
+    },
+    {
+      "epoch": 1.224499301350722,
+      "grad_norm": 0.30463953467892224,
+      "learning_rate": 4.704854368932039e-05,
+      "loss": 0.4836,
+      "step": 1315
+    },
+    {
+      "epoch": 1.2291569632044714,
+      "grad_norm": 0.3451363142622922,
+      "learning_rate": 4.701618122977346e-05,
+      "loss": 0.4955,
+      "step": 1320
+    },
+    {
+      "epoch": 1.2338146250582207,
+      "grad_norm": 0.34909714884580134,
+      "learning_rate": 4.698381877022654e-05,
+      "loss": 0.4803,
+      "step": 1325
+    },
+    {
+      "epoch": 1.2384722869119702,
+      "grad_norm": 0.36522864866273713,
+      "learning_rate": 4.6951456310679615e-05,
+      "loss": 0.4923,
+      "step": 1330
+    },
+    {
+      "epoch": 1.2431299487657197,
+      "grad_norm": 0.39221862966053167,
+      "learning_rate": 4.691909385113269e-05,
+      "loss": 0.4857,
+      "step": 1335
+    },
+    {
+      "epoch": 1.247787610619469,
+      "grad_norm": 0.39668592667455693,
+      "learning_rate": 4.688673139158576e-05,
+      "loss": 0.4836,
+      "step": 1340
+    },
+    {
+      "epoch": 1.2524452724732185,
+      "grad_norm": 0.3189900074415766,
+      "learning_rate": 4.685436893203884e-05,
+      "loss": 0.4733,
+      "step": 1345
+    },
+    {
+      "epoch": 1.257102934326968,
+      "grad_norm": 0.4752552185352753,
+      "learning_rate": 4.6822006472491915e-05,
+      "loss": 0.4837,
+      "step": 1350
+    },
+    {
+      "epoch": 1.2617605961807172,
+      "grad_norm": 0.34317843753923166,
+      "learning_rate": 4.678964401294499e-05,
+      "loss": 0.4762,
+      "step": 1355
+    },
+    {
+      "epoch": 1.2664182580344667,
+      "grad_norm": 0.331883711021671,
+      "learning_rate": 4.675728155339806e-05,
+      "loss": 0.4694,
+      "step": 1360
+    },
+    {
+      "epoch": 1.271075919888216,
+      "grad_norm": 0.31579302865236475,
+      "learning_rate": 4.672491909385113e-05,
+      "loss": 0.4874,
+      "step": 1365
+    },
+    {
+      "epoch": 1.2757335817419655,
+      "grad_norm": 0.33947923252160384,
+      "learning_rate": 4.669255663430421e-05,
+      "loss": 0.4833,
+      "step": 1370
+    },
+    {
+      "epoch": 1.280391243595715,
+      "grad_norm": 0.3561249490177733,
+      "learning_rate": 4.6660194174757284e-05,
+      "loss": 0.4831,
+      "step": 1375
+    },
+    {
+      "epoch": 1.2850489054494645,
+      "grad_norm": 0.3220762092182873,
+      "learning_rate": 4.6627831715210354e-05,
+      "loss": 0.4827,
+      "step": 1380
+    },
+    {
+      "epoch": 1.2897065673032138,
+      "grad_norm": 0.3251770474209212,
+      "learning_rate": 4.659546925566343e-05,
+      "loss": 0.484,
+      "step": 1385
+    },
+    {
+      "epoch": 1.2943642291569633,
+      "grad_norm": 0.36465163745930795,
+      "learning_rate": 4.656310679611651e-05,
+      "loss": 0.4922,
+      "step": 1390
+    },
+    {
+      "epoch": 1.2990218910107125,
+      "grad_norm": 0.32665155125372775,
+      "learning_rate": 4.6530744336569584e-05,
+      "loss": 0.476,
+      "step": 1395
+    },
+    {
+      "epoch": 1.303679552864462,
+      "grad_norm": 0.33996752375205314,
+      "learning_rate": 4.6498381877022653e-05,
+      "loss": 0.4903,
+      "step": 1400
+    },
+    {
+      "epoch": 1.3083372147182115,
+      "grad_norm": 0.31472307161627344,
+      "learning_rate": 4.646601941747573e-05,
+      "loss": 0.4793,
+      "step": 1405
+    },
+    {
+      "epoch": 1.312994876571961,
+      "grad_norm": 0.3373270351504413,
+      "learning_rate": 4.6433656957928807e-05,
+      "loss": 0.4762,
+      "step": 1410
+    },
+    {
+      "epoch": 1.3176525384257103,
+      "grad_norm": 0.33123321376978476,
+      "learning_rate": 4.640129449838188e-05,
+      "loss": 0.4967,
+      "step": 1415
+    },
+    {
+      "epoch": 1.3223102002794598,
+      "grad_norm": 0.31722496467412775,
+      "learning_rate": 4.636893203883495e-05,
+      "loss": 0.4839,
+      "step": 1420
+    },
+    {
+      "epoch": 1.326967862133209,
+      "grad_norm": 0.34349051465183933,
+      "learning_rate": 4.633656957928803e-05,
+      "loss": 0.4782,
+      "step": 1425
+    },
+    {
+      "epoch": 1.3316255239869585,
+      "grad_norm": 0.3587305409881593,
+      "learning_rate": 4.63042071197411e-05,
+      "loss": 0.4771,
+      "step": 1430
+    },
+    {
+      "epoch": 1.336283185840708,
+      "grad_norm": 0.3445600613600092,
+      "learning_rate": 4.6271844660194176e-05,
+      "loss": 0.482,
+      "step": 1435
+    },
+    {
+      "epoch": 1.3409408476944573,
+      "grad_norm": 0.3153330747521536,
+      "learning_rate": 4.623948220064725e-05,
+      "loss": 0.485,
+      "step": 1440
+    },
+    {
+      "epoch": 1.3455985095482068,
+      "grad_norm": 0.33287271332785706,
+      "learning_rate": 4.620711974110032e-05,
+      "loss": 0.4933,
+      "step": 1445
+    },
+    {
+      "epoch": 1.350256171401956,
+      "grad_norm": 0.34406012611925313,
+      "learning_rate": 4.61747572815534e-05,
+      "loss": 0.481,
+      "step": 1450
+    },
+    {
+      "epoch": 1.3549138332557056,
+      "grad_norm": 0.3575650526851001,
+      "learning_rate": 4.6142394822006475e-05,
+      "loss": 0.4914,
+      "step": 1455
+    },
+    {
+      "epoch": 1.359571495109455,
+      "grad_norm": 0.3197678788993809,
+      "learning_rate": 4.611003236245955e-05,
+      "loss": 0.4716,
+      "step": 1460
+    },
+    {
+      "epoch": 1.3642291569632046,
+      "grad_norm": 0.35048652815717235,
+      "learning_rate": 4.607766990291262e-05,
+      "loss": 0.4786,
+      "step": 1465
+    },
+    {
+      "epoch": 1.3688868188169538,
+      "grad_norm": 0.3299342177568497,
+      "learning_rate": 4.60453074433657e-05,
+      "loss": 0.4831,
+      "step": 1470
+    },
+    {
+      "epoch": 1.3735444806707033,
+      "grad_norm": 0.35020584888995054,
+      "learning_rate": 4.6012944983818775e-05,
+      "loss": 0.4751,
+      "step": 1475
+    },
+    {
+      "epoch": 1.3782021425244526,
+      "grad_norm": 0.34466146202234915,
+      "learning_rate": 4.5980582524271845e-05,
+      "loss": 0.4754,
+      "step": 1480
+    },
+    {
+      "epoch": 1.382859804378202,
+      "grad_norm": 0.3326105229693496,
+      "learning_rate": 4.594822006472492e-05,
+      "loss": 0.4814,
+      "step": 1485
+    },
+    {
+      "epoch": 1.3875174662319516,
+      "grad_norm": 0.3863251103968962,
+      "learning_rate": 4.5915857605178e-05,
+      "loss": 0.4856,
+      "step": 1490
+    },
+    {
+      "epoch": 1.392175128085701,
+      "grad_norm": 0.39869131402062374,
+      "learning_rate": 4.5883495145631074e-05,
+      "loss": 0.4872,
+      "step": 1495
+    },
+    {
+      "epoch": 1.3968327899394504,
+      "grad_norm": 0.3236670357940102,
+      "learning_rate": 4.5851132686084144e-05,
+      "loss": 0.4744,
+      "step": 1500
+    },
+    {
+      "epoch": 1.4014904517931999,
+      "grad_norm": 0.33671866535084927,
+      "learning_rate": 4.5818770226537214e-05,
+      "loss": 0.4782,
+      "step": 1505
+    },
+    {
+      "epoch": 1.4061481136469491,
+      "grad_norm": 0.3964201377988781,
+      "learning_rate": 4.578640776699029e-05,
+      "loss": 0.4782,
+      "step": 1510
+    },
+    {
+      "epoch": 1.4108057755006986,
+      "grad_norm": 0.3160900921197823,
+      "learning_rate": 4.575404530744337e-05,
+      "loss": 0.4895,
+      "step": 1515
+    },
+    {
+      "epoch": 1.4154634373544481,
+      "grad_norm": 0.32777499194109583,
+      "learning_rate": 4.5721682847896444e-05,
+      "loss": 0.4787,
+      "step": 1520
+    },
+    {
+      "epoch": 1.4201210992081974,
+      "grad_norm": 0.3903507215961828,
+      "learning_rate": 4.568932038834951e-05,
+      "loss": 0.4778,
+      "step": 1525
+    },
+    {
+      "epoch": 1.424778761061947,
+      "grad_norm": 0.35785839178718776,
+      "learning_rate": 4.565695792880259e-05,
+      "loss": 0.4712,
+      "step": 1530
+    },
+    {
+      "epoch": 1.4294364229156964,
+      "grad_norm": 0.3117173347583603,
+      "learning_rate": 4.5624595469255666e-05,
+      "loss": 0.4855,
+      "step": 1535
+    },
+    {
+      "epoch": 1.4340940847694457,
+      "grad_norm": 0.3068803654580654,
+      "learning_rate": 4.559223300970874e-05,
+      "loss": 0.4755,
+      "step": 1540
+    },
+    {
+      "epoch": 1.4387517466231952,
+      "grad_norm": 0.3429070075799174,
+      "learning_rate": 4.555987055016181e-05,
+      "loss": 0.4783,
+      "step": 1545
+    },
+    {
+      "epoch": 1.4434094084769447,
+      "grad_norm": 0.34774860374178423,
+      "learning_rate": 4.552750809061489e-05,
+      "loss": 0.4721,
+      "step": 1550
+    },
+    {
+      "epoch": 1.448067070330694,
+      "grad_norm": 0.3133469650282993,
+      "learning_rate": 4.5495145631067966e-05,
+      "loss": 0.468,
+      "step": 1555
+    },
+    {
+      "epoch": 1.4527247321844434,
+      "grad_norm": 0.32300725004307296,
+      "learning_rate": 4.546278317152104e-05,
+      "loss": 0.4754,
+      "step": 1560
+    },
+    {
+      "epoch": 1.4573823940381927,
+      "grad_norm": 0.3654645603530482,
+      "learning_rate": 4.543042071197411e-05,
+      "loss": 0.4856,
+      "step": 1565
+    },
+    {
+      "epoch": 1.4620400558919422,
+      "grad_norm": 0.3144382775868269,
+      "learning_rate": 4.539805825242719e-05,
+      "loss": 0.5011,
+      "step": 1570
+    },
+    {
+      "epoch": 1.4666977177456917,
+      "grad_norm": 0.32738808814400805,
+      "learning_rate": 4.536569579288026e-05,
+      "loss": 0.4796,
+      "step": 1575
+    },
+    {
+      "epoch": 1.4713553795994412,
+      "grad_norm": 0.33065024289653616,
+      "learning_rate": 4.5333333333333335e-05,
+      "loss": 0.467,
+      "step": 1580
+    },
+    {
+      "epoch": 1.4760130414531905,
+      "grad_norm": 0.31408876095328725,
+      "learning_rate": 4.5300970873786405e-05,
+      "loss": 0.4766,
+      "step": 1585
+    },
+    {
+      "epoch": 1.48067070330694,
+      "grad_norm": 0.33132227281018756,
+      "learning_rate": 4.526860841423948e-05,
+      "loss": 0.4774,
+      "step": 1590
+    },
+    {
+      "epoch": 1.4853283651606892,
+      "grad_norm": 0.3267028107171943,
+      "learning_rate": 4.523624595469256e-05,
+      "loss": 0.4721,
+      "step": 1595
+    },
+    {
+      "epoch": 1.4899860270144387,
+      "grad_norm": 0.34183969397437214,
+      "learning_rate": 4.5203883495145635e-05,
+      "loss": 0.4808,
+      "step": 1600
+    },
+    {
+      "epoch": 1.4946436888681882,
+      "grad_norm": 0.3341079510588814,
+      "learning_rate": 4.5171521035598705e-05,
+      "loss": 0.4712,
+      "step": 1605
+    },
+    {
+      "epoch": 1.4993013507219377,
+      "grad_norm": 0.3637929050334216,
+      "learning_rate": 4.513915857605178e-05,
+      "loss": 0.4769,
+      "step": 1610
+    },
+    {
+      "epoch": 1.503959012575687,
+      "grad_norm": 0.3614116465559769,
+      "learning_rate": 4.510679611650486e-05,
+      "loss": 0.4793,
+      "step": 1615
+    },
+    {
+      "epoch": 1.5086166744294365,
+      "grad_norm": 0.3282234536811554,
+      "learning_rate": 4.5074433656957934e-05,
+      "loss": 0.4756,
+      "step": 1620
+    },
+    {
+      "epoch": 1.5132743362831858,
+      "grad_norm": 0.3174096727244024,
+      "learning_rate": 4.5042071197411004e-05,
+      "loss": 0.4821,
+      "step": 1625
+    },
+    {
+      "epoch": 1.5179319981369352,
+      "grad_norm": 0.33054476333996435,
+      "learning_rate": 4.500970873786408e-05,
+      "loss": 0.4787,
+      "step": 1630
+    },
+    {
+      "epoch": 1.5225896599906847,
+      "grad_norm": 0.3143046040017771,
+      "learning_rate": 4.497734627831716e-05,
+      "loss": 0.4903,
+      "step": 1635
+    },
+    {
+      "epoch": 1.5272473218444342,
+      "grad_norm": 0.36640286783015863,
+      "learning_rate": 4.4944983818770234e-05,
+      "loss": 0.4833,
+      "step": 1640
+    },
+    {
+      "epoch": 1.5319049836981835,
+      "grad_norm": 0.3012687202528479,
+      "learning_rate": 4.4912621359223303e-05,
+      "loss": 0.4893,
+      "step": 1645
+    },
+    {
+      "epoch": 1.5365626455519328,
+      "grad_norm": 0.37938595625196225,
+      "learning_rate": 4.488025889967637e-05,
+      "loss": 0.4711,
+      "step": 1650
+    },
+    {
+      "epoch": 1.5412203074056823,
+      "grad_norm": 0.33746086958512767,
+      "learning_rate": 4.484789644012945e-05,
+      "loss": 0.469,
+      "step": 1655
+    },
+    {
+      "epoch": 1.5458779692594318,
+      "grad_norm": 0.3195215446211632,
+      "learning_rate": 4.4815533980582526e-05,
+      "loss": 0.4669,
+      "step": 1660
+    },
+    {
+      "epoch": 1.5505356311131813,
+      "grad_norm": 0.31017867155169293,
+      "learning_rate": 4.4783171521035596e-05,
+      "loss": 0.4813,
+      "step": 1665
+    },
+    {
+      "epoch": 1.5551932929669308,
+      "grad_norm": 0.35887769881626985,
+      "learning_rate": 4.475080906148867e-05,
+      "loss": 0.4752,
+      "step": 1670
+    },
+    {
+      "epoch": 1.55985095482068,
+      "grad_norm": 0.31528761907295816,
+      "learning_rate": 4.471844660194175e-05,
+      "loss": 0.4722,
+      "step": 1675
+    },
+    {
+      "epoch": 1.5645086166744293,
+      "grad_norm": 0.29713048688592464,
+      "learning_rate": 4.4686084142394826e-05,
+      "loss": 0.4816,
+      "step": 1680
+    },
+    {
+      "epoch": 1.5691662785281788,
+      "grad_norm": 0.34501437275635716,
+      "learning_rate": 4.4653721682847896e-05,
+      "loss": 0.4711,
+      "step": 1685
+    },
+    {
+      "epoch": 1.5738239403819283,
+      "grad_norm": 0.3193644835414442,
+      "learning_rate": 4.462135922330097e-05,
+      "loss": 0.4772,
+      "step": 1690
+    },
+    {
+      "epoch": 1.5784816022356778,
+      "grad_norm": 0.3525186050167732,
+      "learning_rate": 4.458899676375405e-05,
+      "loss": 0.4752,
+      "step": 1695
+    },
+    {
+      "epoch": 1.583139264089427,
+      "grad_norm": 0.3390490212094124,
+      "learning_rate": 4.4556634304207125e-05,
+      "loss": 0.4801,
+      "step": 1700
+    },
+    {
+      "epoch": 1.5877969259431766,
+      "grad_norm": 0.27199321463712367,
+      "learning_rate": 4.4524271844660195e-05,
+      "loss": 0.4732,
+      "step": 1705
+    },
+    {
+      "epoch": 1.5924545877969258,
+      "grad_norm": 0.33136111443589444,
+      "learning_rate": 4.449190938511327e-05,
+      "loss": 0.4794,
+      "step": 1710
+    },
+    {
+      "epoch": 1.5971122496506753,
+      "grad_norm": 0.33788165112244023,
+      "learning_rate": 4.445954692556635e-05,
+      "loss": 0.4869,
+      "step": 1715
+    },
+    {
+      "epoch": 1.6017699115044248,
+      "grad_norm": 0.2771764238057386,
+      "learning_rate": 4.442718446601942e-05,
+      "loss": 0.4701,
+      "step": 1720
+    },
+    {
+      "epoch": 1.6064275733581743,
+      "grad_norm": 0.3085685435375194,
+      "learning_rate": 4.4394822006472495e-05,
+      "loss": 0.4722,
+      "step": 1725
+    },
+    {
+      "epoch": 1.6110852352119236,
+      "grad_norm": 0.3213685919603364,
+      "learning_rate": 4.4362459546925564e-05,
+      "loss": 0.4854,
+      "step": 1730
+    },
+    {
+      "epoch": 1.6157428970656729,
+      "grad_norm": 0.3449226743664926,
+      "learning_rate": 4.433009708737864e-05,
+      "loss": 0.4831,
+      "step": 1735
+    },
+    {
+      "epoch": 1.6204005589194224,
+      "grad_norm": 0.2819191908463559,
+      "learning_rate": 4.429773462783172e-05,
+      "loss": 0.4626,
+      "step": 1740
+    },
+    {
+      "epoch": 1.6250582207731719,
+      "grad_norm": 0.2803367610061562,
+      "learning_rate": 4.4265372168284794e-05,
+      "loss": 0.4983,
+      "step": 1745
+    },
+    {
+      "epoch": 1.6297158826269214,
+      "grad_norm": 0.3032620461460716,
+      "learning_rate": 4.4233009708737864e-05,
+      "loss": 0.4688,
+      "step": 1750
+    },
+    {
+      "epoch": 1.6343735444806708,
+      "grad_norm": 0.31143778436372727,
+      "learning_rate": 4.420064724919094e-05,
+      "loss": 0.4878,
+      "step": 1755
+    },
+    {
+      "epoch": 1.6390312063344201,
+      "grad_norm": 0.29059732424768564,
+      "learning_rate": 4.416828478964402e-05,
+      "loss": 0.4662,
+      "step": 1760
+    },
+    {
+      "epoch": 1.6436888681881694,
+      "grad_norm": 0.30902263729331736,
+      "learning_rate": 4.4135922330097094e-05,
+      "loss": 0.4735,
+      "step": 1765
+    },
+    {
+      "epoch": 1.648346530041919,
+      "grad_norm": 0.3067477355060161,
+      "learning_rate": 4.410355987055016e-05,
+      "loss": 0.4659,
+      "step": 1770
+    },
+    {
+      "epoch": 1.6530041918956684,
+      "grad_norm": 0.30766225324156005,
+      "learning_rate": 4.407119741100324e-05,
+      "loss": 0.4839,
+      "step": 1775
+    },
+    {
+      "epoch": 1.6576618537494179,
+      "grad_norm": 0.3060356213831821,
+      "learning_rate": 4.4038834951456316e-05,
+      "loss": 0.4683,
+      "step": 1780
+    },
+    {
+      "epoch": 1.6623195156031674,
+      "grad_norm": 0.27993399458596613,
+      "learning_rate": 4.4006472491909386e-05,
+      "loss": 0.4787,
+      "step": 1785
+    },
+    {
+      "epoch": 1.6669771774569166,
+      "grad_norm": 0.2878538280302826,
+      "learning_rate": 4.397411003236246e-05,
+      "loss": 0.469,
+      "step": 1790
+    },
+    {
+      "epoch": 1.671634839310666,
+      "grad_norm": 0.3190507344099696,
+      "learning_rate": 4.394174757281553e-05,
+      "loss": 0.467,
+      "step": 1795
+    },
+    {
+      "epoch": 1.6762925011644154,
+      "grad_norm": 0.2740825398174676,
+      "learning_rate": 4.390938511326861e-05,
+      "loss": 0.4663,
+      "step": 1800
+    },
+    {
+      "epoch": 1.680950163018165,
+      "grad_norm": 0.3029113523627245,
+      "learning_rate": 4.3877022653721686e-05,
+      "loss": 0.473,
+      "step": 1805
+    },
+    {
+      "epoch": 1.6856078248719144,
+      "grad_norm": 0.30275582143151997,
+      "learning_rate": 4.3844660194174756e-05,
+      "loss": 0.4714,
+      "step": 1810
+    },
+    {
+      "epoch": 1.6902654867256637,
+      "grad_norm": 0.29018400666960237,
+      "learning_rate": 4.381229773462783e-05,
+      "loss": 0.4731,
+      "step": 1815
+    },
+    {
+      "epoch": 1.6949231485794132,
+      "grad_norm": 0.31944630955568515,
+      "learning_rate": 4.377993527508091e-05,
+      "loss": 0.4569,
+      "step": 1820
+    },
+    {
+      "epoch": 1.6995808104331624,
+      "grad_norm": 0.33676422137045936,
+      "learning_rate": 4.3747572815533985e-05,
+      "loss": 0.4633,
+      "step": 1825
+    },
+    {
+      "epoch": 1.704238472286912,
+      "grad_norm": 0.3396901223033333,
+      "learning_rate": 4.3715210355987055e-05,
+      "loss": 0.4643,
+      "step": 1830
+    },
+    {
+      "epoch": 1.7088961341406614,
+      "grad_norm": 0.3366830543419861,
+      "learning_rate": 4.368284789644013e-05,
+      "loss": 0.474,
+      "step": 1835
+    },
+    {
+      "epoch": 1.713553795994411,
+      "grad_norm": 0.3398454823387078,
+      "learning_rate": 4.365048543689321e-05,
+      "loss": 0.4716,
+      "step": 1840
+    },
+    {
+      "epoch": 1.7182114578481602,
+      "grad_norm": 0.3093101686513045,
+      "learning_rate": 4.3618122977346285e-05,
+      "loss": 0.4734,
+      "step": 1845
+    },
+    {
+      "epoch": 1.7228691197019095,
+      "grad_norm": 0.28865990425667754,
+      "learning_rate": 4.3585760517799354e-05,
+      "loss": 0.4843,
+      "step": 1850
+    },
+    {
+      "epoch": 1.727526781555659,
+      "grad_norm": 0.30938960750962463,
+      "learning_rate": 4.355339805825243e-05,
+      "loss": 0.4608,
+      "step": 1855
+    },
+    {
+      "epoch": 1.7321844434094085,
+      "grad_norm": 0.303570178840196,
+      "learning_rate": 4.352103559870551e-05,
+      "loss": 0.4938,
+      "step": 1860
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.320398560744746,
+      "learning_rate": 4.348867313915858e-05,
+      "loss": 0.4687,
+      "step": 1865
+    },
+    {
+      "epoch": 1.7414997671169075,
+      "grad_norm": 0.28934725217636076,
+      "learning_rate": 4.345631067961165e-05,
+      "loss": 0.4767,
+      "step": 1870
+    },
+    {
+      "epoch": 1.7461574289706567,
+      "grad_norm": 0.3082083156804532,
+      "learning_rate": 4.3423948220064724e-05,
+      "loss": 0.4917,
+      "step": 1875
+    },
+    {
+      "epoch": 1.750815090824406,
+      "grad_norm": 0.33073903630019985,
+      "learning_rate": 4.33915857605178e-05,
+      "loss": 0.48,
+      "step": 1880
+    },
+    {
+      "epoch": 1.7554727526781555,
+      "grad_norm": 0.3795705747571329,
+      "learning_rate": 4.335922330097088e-05,
+      "loss": 0.4818,
+      "step": 1885
+    },
+    {
+      "epoch": 1.760130414531905,
+      "grad_norm": 0.33196205439252985,
+      "learning_rate": 4.332686084142395e-05,
+      "loss": 0.4833,
+      "step": 1890
+    },
+    {
+      "epoch": 1.7647880763856545,
+      "grad_norm": 0.3755261858357244,
+      "learning_rate": 4.329449838187702e-05,
+      "loss": 0.4882,
+      "step": 1895
+    },
+    {
+      "epoch": 1.7694457382394038,
+      "grad_norm": 0.3036362546197639,
+      "learning_rate": 4.32621359223301e-05,
+      "loss": 0.4758,
+      "step": 1900
+    },
+    {
+      "epoch": 1.7741034000931533,
+      "grad_norm": 0.30502050664737707,
+      "learning_rate": 4.3229773462783176e-05,
+      "loss": 0.475,
+      "step": 1905
+    },
+    {
+      "epoch": 1.7787610619469025,
+      "grad_norm": 0.3157357177209518,
+      "learning_rate": 4.3197411003236246e-05,
+      "loss": 0.466,
+      "step": 1910
+    },
+    {
+      "epoch": 1.783418723800652,
+      "grad_norm": 0.3040131152683334,
+      "learning_rate": 4.316504854368932e-05,
+      "loss": 0.4577,
+      "step": 1915
+    },
+    {
+      "epoch": 1.7880763856544015,
+      "grad_norm": 0.3187185944812107,
+      "learning_rate": 4.31326860841424e-05,
+      "loss": 0.4669,
+      "step": 1920
+    },
+    {
+      "epoch": 1.792734047508151,
+      "grad_norm": 0.33931947982120053,
+      "learning_rate": 4.3100323624595476e-05,
+      "loss": 0.4735,
+      "step": 1925
+    },
+    {
+      "epoch": 1.7973917093619003,
+      "grad_norm": 0.32734347976657,
+      "learning_rate": 4.3067961165048546e-05,
+      "loss": 0.4672,
+      "step": 1930
+    },
+    {
+      "epoch": 1.8020493712156498,
+      "grad_norm": 0.4692029460469582,
+      "learning_rate": 4.3035598705501615e-05,
+      "loss": 0.4613,
+      "step": 1935
+    },
+    {
+      "epoch": 1.806707033069399,
+      "grad_norm": 0.32483802088569746,
+      "learning_rate": 4.300323624595469e-05,
+      "loss": 0.4636,
+      "step": 1940
+    },
+    {
+      "epoch": 1.8113646949231486,
+      "grad_norm": 0.302012305602097,
+      "learning_rate": 4.297087378640777e-05,
+      "loss": 0.4772,
+      "step": 1945
+    },
+    {
+      "epoch": 1.816022356776898,
+      "grad_norm": 0.31280787882848426,
+      "learning_rate": 4.2938511326860845e-05,
+      "loss": 0.4667,
+      "step": 1950
+    },
+    {
+      "epoch": 1.8206800186306475,
+      "grad_norm": 0.2833157297650169,
+      "learning_rate": 4.2906148867313915e-05,
+      "loss": 0.4685,
+      "step": 1955
+    },
+    {
+      "epoch": 1.8253376804843968,
+      "grad_norm": 0.3057640134824448,
+      "learning_rate": 4.287378640776699e-05,
+      "loss": 0.4713,
+      "step": 1960
+    },
+    {
+      "epoch": 1.829995342338146,
+      "grad_norm": 0.28708015146494686,
+      "learning_rate": 4.284142394822007e-05,
+      "loss": 0.4739,
+      "step": 1965
+    },
+    {
+      "epoch": 1.8346530041918956,
+      "grad_norm": 0.30123275415917106,
+      "learning_rate": 4.280906148867314e-05,
+      "loss": 0.4695,
+      "step": 1970
+    },
+    {
+      "epoch": 1.839310666045645,
+      "grad_norm": 0.3099654672262839,
+      "learning_rate": 4.2776699029126214e-05,
+      "loss": 0.4749,
+      "step": 1975
+    },
+    {
+      "epoch": 1.8439683278993946,
+      "grad_norm": 0.3152945452256832,
+      "learning_rate": 4.274433656957929e-05,
+      "loss": 0.4671,
+      "step": 1980
+    },
+    {
+      "epoch": 1.848625989753144,
+      "grad_norm": 0.31539137515997484,
+      "learning_rate": 4.271197411003237e-05,
+      "loss": 0.4843,
+      "step": 1985
+    },
+    {
+      "epoch": 1.8532836516068933,
+      "grad_norm": 0.33743023774825115,
+      "learning_rate": 4.267961165048544e-05,
+      "loss": 0.4742,
+      "step": 1990
+    },
+    {
+      "epoch": 1.8579413134606426,
+      "grad_norm": 0.278496500840638,
+      "learning_rate": 4.2647249190938514e-05,
+      "loss": 0.4571,
+      "step": 1995
+    },
+    {
+      "epoch": 1.8625989753143921,
+      "grad_norm": 0.2998795993570895,
+      "learning_rate": 4.261488673139159e-05,
+      "loss": 0.471,
+      "step": 2000
+    },
+    {
+      "epoch": 1.8672566371681416,
+      "grad_norm": 0.30700849405318176,
+      "learning_rate": 4.258252427184467e-05,
+      "loss": 0.4817,
+      "step": 2005
+    },
+    {
+      "epoch": 1.871914299021891,
+      "grad_norm": 0.27629264600348125,
+      "learning_rate": 4.255016181229774e-05,
+      "loss": 0.4632,
+      "step": 2010
+    },
+    {
+      "epoch": 1.8765719608756404,
+      "grad_norm": 0.28518434221399247,
+      "learning_rate": 4.2517799352750807e-05,
+      "loss": 0.4827,
+      "step": 2015
+    },
+    {
+      "epoch": 1.8812296227293899,
+      "grad_norm": 0.3021920950459163,
+      "learning_rate": 4.248543689320388e-05,
+      "loss": 0.4642,
+      "step": 2020
+    },
+    {
+      "epoch": 1.8858872845831391,
+      "grad_norm": 0.28974303173801413,
+      "learning_rate": 4.245307443365696e-05,
+      "loss": 0.4707,
+      "step": 2025
+    },
+    {
+      "epoch": 1.8905449464368886,
+      "grad_norm": 0.2944523033092523,
+      "learning_rate": 4.2420711974110036e-05,
+      "loss": 0.4664,
+      "step": 2030
+    },
+    {
+      "epoch": 1.8952026082906381,
+      "grad_norm": 1.2596267443116471,
+      "learning_rate": 4.2388349514563106e-05,
+      "loss": 0.463,
+      "step": 2035
+    },
+    {
+      "epoch": 1.8998602701443876,
+      "grad_norm": 0.3251364871994849,
+      "learning_rate": 4.235598705501618e-05,
+      "loss": 0.4769,
+      "step": 2040
+    },
+    {
+      "epoch": 1.904517931998137,
+      "grad_norm": 0.28708893882914116,
+      "learning_rate": 4.232362459546926e-05,
+      "loss": 0.4813,
+      "step": 2045
+    },
+    {
+      "epoch": 1.9091755938518864,
+      "grad_norm": 0.300276772974375,
+      "learning_rate": 4.2291262135922336e-05,
+      "loss": 0.4686,
+      "step": 2050
+    },
+    {
+      "epoch": 1.9138332557056357,
+      "grad_norm": 0.27252504469429434,
+      "learning_rate": 4.2258899676375405e-05,
+      "loss": 0.4681,
+      "step": 2055
+    },
+    {
+      "epoch": 1.9184909175593852,
+      "grad_norm": 0.2770826992578108,
+      "learning_rate": 4.222653721682848e-05,
+      "loss": 0.4632,
+      "step": 2060
+    },
+    {
+      "epoch": 1.9231485794131347,
+      "grad_norm": 0.2866673869336603,
+      "learning_rate": 4.219417475728156e-05,
+      "loss": 0.4678,
+      "step": 2065
+    },
+    {
+      "epoch": 1.9278062412668842,
+      "grad_norm": 0.3148912097597905,
+      "learning_rate": 4.2161812297734635e-05,
+      "loss": 0.4707,
+      "step": 2070
+    },
+    {
+      "epoch": 1.9324639031206334,
+      "grad_norm": 0.2856113904672565,
+      "learning_rate": 4.2129449838187705e-05,
+      "loss": 0.4682,
+      "step": 2075
+    },
+    {
+      "epoch": 1.9371215649743827,
+      "grad_norm": 0.28939502907971115,
+      "learning_rate": 4.2097087378640775e-05,
+      "loss": 0.4603,
+      "step": 2080
+    },
+    {
+      "epoch": 1.9417792268281322,
+      "grad_norm": 0.2833963329099756,
+      "learning_rate": 4.206472491909385e-05,
+      "loss": 0.4753,
+      "step": 2085
+    },
+    {
+      "epoch": 1.9464368886818817,
+      "grad_norm": 0.2792282266609358,
+      "learning_rate": 4.203236245954693e-05,
+      "loss": 0.4526,
+      "step": 2090
+    },
+    {
+      "epoch": 1.9510945505356312,
+      "grad_norm": 0.27854417338326687,
+      "learning_rate": 4.2e-05,
+      "loss": 0.4653,
+      "step": 2095
+    },
+    {
+      "epoch": 1.9557522123893807,
+      "grad_norm": 0.31580009741056303,
+      "learning_rate": 4.1967637540453074e-05,
+      "loss": 0.4634,
+      "step": 2100
+    },
+    {
+      "epoch": 1.96040987424313,
+      "grad_norm": 0.2919023246129842,
+      "learning_rate": 4.193527508090615e-05,
+      "loss": 0.4739,
+      "step": 2105
+    },
+    {
+      "epoch": 1.9650675360968792,
+      "grad_norm": 0.30597531862356964,
+      "learning_rate": 4.190291262135923e-05,
+      "loss": 0.4558,
+      "step": 2110
+    },
+    {
+      "epoch": 1.9697251979506287,
+      "grad_norm": 0.29143179368567523,
+      "learning_rate": 4.18705501618123e-05,
+      "loss": 0.4691,
+      "step": 2115
+    },
+    {
+      "epoch": 1.9743828598043782,
+      "grad_norm": 0.3179819476003419,
+      "learning_rate": 4.1838187702265374e-05,
+      "loss": 0.4723,
+      "step": 2120
+    },
+    {
+      "epoch": 1.9790405216581277,
+      "grad_norm": 0.31806911713554653,
+      "learning_rate": 4.180582524271845e-05,
+      "loss": 0.4618,
+      "step": 2125
+    },
+    {
+      "epoch": 1.983698183511877,
+      "grad_norm": 0.2831599208032219,
+      "learning_rate": 4.177346278317153e-05,
+      "loss": 0.465,
+      "step": 2130
+    },
+    {
+      "epoch": 1.9883558453656265,
+      "grad_norm": 0.30209993683958286,
+      "learning_rate": 4.17411003236246e-05,
+      "loss": 0.4723,
+      "step": 2135
+    },
+    {
+      "epoch": 1.9930135072193758,
+      "grad_norm": 0.2789050959016751,
+      "learning_rate": 4.170873786407767e-05,
+      "loss": 0.4709,
+      "step": 2140
+    },
+    {
+      "epoch": 1.9976711690731253,
+      "grad_norm": 0.3214563107858177,
+      "learning_rate": 4.167637540453075e-05,
+      "loss": 0.4705,
+      "step": 2145
+    },
+    {
+      "epoch": 2.0018630647415,
+      "grad_norm": 0.34324871696573384,
+      "learning_rate": 4.1644012944983826e-05,
+      "loss": 0.4626,
+      "step": 2150
+    },
+    {
+      "epoch": 2.0065207265952494,
+      "grad_norm": 0.33538344509233275,
+      "learning_rate": 4.161165048543689e-05,
+      "loss": 0.4303,
+      "step": 2155
+    },
+    {
+      "epoch": 2.0111783884489984,
+      "grad_norm": 0.3454953096884954,
+      "learning_rate": 4.1579288025889966e-05,
+      "loss": 0.4225,
+      "step": 2160
+    },
+    {
+      "epoch": 2.015836050302748,
+      "grad_norm": 0.32907002129649626,
+      "learning_rate": 4.154692556634304e-05,
+      "loss": 0.429,
+      "step": 2165
+    },
+    {
+      "epoch": 2.0204937121564974,
+      "grad_norm": 0.30918096504052195,
+      "learning_rate": 4.151456310679612e-05,
+      "loss": 0.4152,
+      "step": 2170
+    },
+    {
+      "epoch": 2.025151374010247,
+      "grad_norm": 0.29841535792838386,
+      "learning_rate": 4.148220064724919e-05,
+      "loss": 0.4362,
+      "step": 2175
+    },
+    {
+      "epoch": 2.0298090358639964,
+      "grad_norm": 0.31922764691509886,
+      "learning_rate": 4.1449838187702265e-05,
+      "loss": 0.4226,
+      "step": 2180
+    },
+    {
+      "epoch": 2.034466697717746,
+      "grad_norm": 0.33711068057941024,
+      "learning_rate": 4.141747572815534e-05,
+      "loss": 0.4322,
+      "step": 2185
+    },
+    {
+      "epoch": 2.039124359571495,
+      "grad_norm": 0.3615982081697286,
+      "learning_rate": 4.138511326860842e-05,
+      "loss": 0.429,
+      "step": 2190
+    },
+    {
+      "epoch": 2.0437820214252445,
+      "grad_norm": 0.2894804248909392,
+      "learning_rate": 4.135275080906149e-05,
+      "loss": 0.4268,
+      "step": 2195
+    },
+    {
+      "epoch": 2.048439683278994,
+      "grad_norm": 0.3094882269270303,
+      "learning_rate": 4.1320388349514565e-05,
+      "loss": 0.4262,
+      "step": 2200
+    },
+    {
+      "epoch": 2.0530973451327434,
+      "grad_norm": 0.3228500916846504,
+      "learning_rate": 4.128802588996764e-05,
+      "loss": 0.4323,
+      "step": 2205
+    },
+    {
+      "epoch": 2.057755006986493,
+      "grad_norm": 0.32065541360901384,
+      "learning_rate": 4.125566343042072e-05,
+      "loss": 0.4179,
+      "step": 2210
+    },
+    {
+      "epoch": 2.062412668840242,
+      "grad_norm": 0.37823517394569717,
+      "learning_rate": 4.122330097087379e-05,
+      "loss": 0.4251,
+      "step": 2215
+    },
+    {
+      "epoch": 2.0670703306939915,
+      "grad_norm": 0.29478585632572873,
+      "learning_rate": 4.1190938511326864e-05,
+      "loss": 0.4291,
+      "step": 2220
+    },
+    {
+      "epoch": 2.071727992547741,
+      "grad_norm": 0.2991406315394289,
+      "learning_rate": 4.1158576051779934e-05,
+      "loss": 0.4259,
+      "step": 2225
+    },
+    {
+      "epoch": 2.0763856544014905,
+      "grad_norm": 0.28207541248218293,
+      "learning_rate": 4.112621359223301e-05,
+      "loss": 0.4272,
+      "step": 2230
+    },
+    {
+      "epoch": 2.08104331625524,
+      "grad_norm": 0.3039635186974918,
+      "learning_rate": 4.109385113268609e-05,
+      "loss": 0.421,
+      "step": 2235
+    },
+    {
+      "epoch": 2.0857009781089895,
+      "grad_norm": 0.29727358441688373,
+      "learning_rate": 4.106148867313916e-05,
+      "loss": 0.4321,
+      "step": 2240
+    },
+    {
+      "epoch": 2.0903586399627385,
+      "grad_norm": 0.2999892387740095,
+      "learning_rate": 4.1029126213592234e-05,
+      "loss": 0.4319,
+      "step": 2245
+    },
+    {
+      "epoch": 2.095016301816488,
+      "grad_norm": 0.3098725249210994,
+      "learning_rate": 4.099676375404531e-05,
+      "loss": 0.424,
+      "step": 2250
+    },
+    {
+      "epoch": 2.0996739636702375,
+      "grad_norm": 0.33377558074464014,
+      "learning_rate": 4.096440129449839e-05,
+      "loss": 0.436,
+      "step": 2255
+    },
+    {
+      "epoch": 2.104331625523987,
+      "grad_norm": 0.3364161596111139,
+      "learning_rate": 4.0932038834951457e-05,
+      "loss": 0.4232,
+      "step": 2260
+    },
+    {
+      "epoch": 2.1089892873777365,
+      "grad_norm": 0.3131140562229925,
+      "learning_rate": 4.089967637540453e-05,
+      "loss": 0.4381,
+      "step": 2265
+    },
+    {
+      "epoch": 2.113646949231486,
+      "grad_norm": 0.32109915934435745,
+      "learning_rate": 4.086731391585761e-05,
+      "loss": 0.4265,
+      "step": 2270
+    },
+    {
+      "epoch": 2.118304611085235,
+      "grad_norm": 0.29268954120346086,
+      "learning_rate": 4.083495145631068e-05,
+      "loss": 0.4279,
+      "step": 2275
+    },
+    {
+      "epoch": 2.1229622729389845,
+      "grad_norm": 0.3249364007907747,
+      "learning_rate": 4.0802588996763756e-05,
+      "loss": 0.428,
+      "step": 2280
+    },
+    {
+      "epoch": 2.127619934792734,
+      "grad_norm": 0.31460519638919976,
+      "learning_rate": 4.077022653721683e-05,
+      "loss": 0.4329,
+      "step": 2285
+    },
+    {
+      "epoch": 2.1322775966464835,
+      "grad_norm": 0.32332636152884364,
+      "learning_rate": 4.073786407766991e-05,
+      "loss": 0.4301,
+      "step": 2290
+    },
+    {
+      "epoch": 2.136935258500233,
+      "grad_norm": 0.28748649337095494,
+      "learning_rate": 4.070550161812298e-05,
+      "loss": 0.4387,
+      "step": 2295
+    },
+    {
+      "epoch": 2.1415929203539825,
+      "grad_norm": 0.29107742775138384,
+      "learning_rate": 4.067313915857605e-05,
+      "loss": 0.4387,
+      "step": 2300
+    },
+    {
+      "epoch": 2.1462505822077316,
+      "grad_norm": 0.2844782439437948,
+      "learning_rate": 4.0640776699029125e-05,
+      "loss": 0.4295,
+      "step": 2305
+    },
+    {
+      "epoch": 2.150908244061481,
+      "grad_norm": 0.27362329963575716,
+      "learning_rate": 4.06084142394822e-05,
+      "loss": 0.4137,
+      "step": 2310
+    },
+    {
+      "epoch": 2.1555659059152306,
+      "grad_norm": 0.28922030486408484,
+      "learning_rate": 4.057605177993528e-05,
+      "loss": 0.4211,
+      "step": 2315
+    },
+    {
+      "epoch": 2.16022356776898,
+      "grad_norm": 0.2783657206027948,
+      "learning_rate": 4.054368932038835e-05,
+      "loss": 0.4248,
+      "step": 2320
+    },
+    {
+      "epoch": 2.1648812296227296,
+      "grad_norm": 0.30625672965034256,
+      "learning_rate": 4.0511326860841425e-05,
+      "loss": 0.4344,
+      "step": 2325
+    },
+    {
+      "epoch": 2.1695388914764786,
+      "grad_norm": 0.29237291366395135,
+      "learning_rate": 4.04789644012945e-05,
+      "loss": 0.4367,
+      "step": 2330
+    },
+    {
+      "epoch": 2.174196553330228,
+      "grad_norm": 0.30061144676071067,
+      "learning_rate": 4.044660194174758e-05,
+      "loss": 0.4193,
+      "step": 2335
+    },
+    {
+      "epoch": 2.1788542151839776,
+      "grad_norm": 0.3034112135905207,
+      "learning_rate": 4.041423948220065e-05,
+      "loss": 0.433,
+      "step": 2340
+    },
+    {
+      "epoch": 2.183511877037727,
+      "grad_norm": 0.2784621691367184,
+      "learning_rate": 4.0381877022653724e-05,
+      "loss": 0.4318,
+      "step": 2345
+    },
+    {
+      "epoch": 2.1881695388914766,
+      "grad_norm": 0.28920312172259255,
+      "learning_rate": 4.03495145631068e-05,
+      "loss": 0.4303,
+      "step": 2350
+    },
+    {
+      "epoch": 2.192827200745226,
+      "grad_norm": 0.2884542372811835,
+      "learning_rate": 4.031715210355988e-05,
+      "loss": 0.4237,
+      "step": 2355
+    },
+    {
+      "epoch": 2.197484862598975,
+      "grad_norm": 0.29695407946896535,
+      "learning_rate": 4.028478964401295e-05,
+      "loss": 0.4339,
+      "step": 2360
+    },
+    {
+      "epoch": 2.2021425244527246,
+      "grad_norm": 0.30202143737915377,
+      "learning_rate": 4.0252427184466024e-05,
+      "loss": 0.4301,
+      "step": 2365
+    },
+    {
+      "epoch": 2.206800186306474,
+      "grad_norm": 0.2935803126492083,
+      "learning_rate": 4.0220064724919093e-05,
+      "loss": 0.4353,
+      "step": 2370
+    },
+    {
+      "epoch": 2.2114578481602236,
+      "grad_norm": 0.2911223567144398,
+      "learning_rate": 4.018770226537217e-05,
+      "loss": 0.436,
+      "step": 2375
+    },
+    {
+      "epoch": 2.216115510013973,
+      "grad_norm": 0.31911706724850875,
+      "learning_rate": 4.015533980582524e-05,
+      "loss": 0.4265,
+      "step": 2380
+    },
+    {
+      "epoch": 2.2207731718677226,
+      "grad_norm": 0.28529775833448395,
+      "learning_rate": 4.0122977346278316e-05,
+      "loss": 0.4409,
+      "step": 2385
+    },
+    {
+      "epoch": 2.2254308337214717,
+      "grad_norm": 0.25695809085703514,
+      "learning_rate": 4.009061488673139e-05,
+      "loss": 0.4304,
+      "step": 2390
+    },
+    {
+      "epoch": 2.230088495575221,
+      "grad_norm": 0.29657749009103374,
+      "learning_rate": 4.005825242718447e-05,
+      "loss": 0.4296,
+      "step": 2395
+    },
+    {
+      "epoch": 2.2347461574289706,
+      "grad_norm": 0.29804875797050473,
+      "learning_rate": 4.002588996763754e-05,
+      "loss": 0.4209,
+      "step": 2400
+    },
+    {
+      "epoch": 2.23940381928272,
+      "grad_norm": 0.29220612134590873,
+      "learning_rate": 3.9993527508090616e-05,
+      "loss": 0.4311,
+      "step": 2405
+    },
+    {
+      "epoch": 2.2440614811364696,
+      "grad_norm": 0.2838098626320004,
+      "learning_rate": 3.996116504854369e-05,
+      "loss": 0.4299,
+      "step": 2410
+    },
+    {
+      "epoch": 2.248719142990219,
+      "grad_norm": 0.2717759340416869,
+      "learning_rate": 3.992880258899677e-05,
+      "loss": 0.4261,
+      "step": 2415
+    },
+    {
+      "epoch": 2.253376804843968,
+      "grad_norm": 0.2729701123231704,
+      "learning_rate": 3.989644012944984e-05,
+      "loss": 0.4138,
+      "step": 2420
+    },
+    {
+      "epoch": 2.2580344666977177,
+      "grad_norm": 0.2955858122750388,
+      "learning_rate": 3.9864077669902915e-05,
+      "loss": 0.4392,
+      "step": 2425
+    },
+    {
+      "epoch": 2.262692128551467,
+      "grad_norm": 0.2870127555384206,
+      "learning_rate": 3.983171521035599e-05,
+      "loss": 0.4217,
+      "step": 2430
+    },
+    {
+      "epoch": 2.2673497904052167,
+      "grad_norm": 0.29714494202478464,
+      "learning_rate": 3.979935275080907e-05,
+      "loss": 0.4307,
+      "step": 2435
+    },
+    {
+      "epoch": 2.272007452258966,
+      "grad_norm": 0.31323215155035516,
+      "learning_rate": 3.976699029126214e-05,
+      "loss": 0.4284,
+      "step": 2440
+    },
+    {
+      "epoch": 2.276665114112715,
+      "grad_norm": 0.30731519196804635,
+      "learning_rate": 3.973462783171521e-05,
+      "loss": 0.4269,
+      "step": 2445
+    },
+    {
+      "epoch": 2.2813227759664647,
+      "grad_norm": 0.2976558685973644,
+      "learning_rate": 3.9702265372168285e-05,
+      "loss": 0.4275,
+      "step": 2450
+    },
+    {
+      "epoch": 2.285980437820214,
+      "grad_norm": 0.31150191739658856,
+      "learning_rate": 3.966990291262136e-05,
+      "loss": 0.4232,
+      "step": 2455
+    },
+    {
+      "epoch": 2.2906380996739637,
+      "grad_norm": 0.3068690355886579,
+      "learning_rate": 3.963754045307443e-05,
+      "loss": 0.4194,
+      "step": 2460
+    },
+    {
+      "epoch": 2.295295761527713,
+      "grad_norm": 0.30564701433441577,
+      "learning_rate": 3.960517799352751e-05,
+      "loss": 0.4415,
+      "step": 2465
+    },
+    {
+      "epoch": 2.2999534233814627,
+      "grad_norm": 0.2832601518769638,
+      "learning_rate": 3.9572815533980584e-05,
+      "loss": 0.4325,
+      "step": 2470
+    },
+    {
+      "epoch": 2.3046110852352117,
+      "grad_norm": 0.26612149177883143,
+      "learning_rate": 3.954045307443366e-05,
+      "loss": 0.4348,
+      "step": 2475
+    },
+    {
+      "epoch": 2.3092687470889612,
+      "grad_norm": 0.3016054945452254,
+      "learning_rate": 3.950809061488673e-05,
+      "loss": 0.428,
+      "step": 2480
+    },
+    {
+      "epoch": 2.3139264089427107,
+      "grad_norm": 0.2934348638470389,
+      "learning_rate": 3.947572815533981e-05,
+      "loss": 0.44,
+      "step": 2485
+    },
+    {
+      "epoch": 2.3185840707964602,
+      "grad_norm": 0.2820481413068723,
+      "learning_rate": 3.9443365695792884e-05,
+      "loss": 0.4385,
+      "step": 2490
+    },
+    {
+      "epoch": 2.3232417326502097,
+      "grad_norm": 0.3118026197566741,
+      "learning_rate": 3.941100323624596e-05,
+      "loss": 0.4292,
+      "step": 2495
+    },
+    {
+      "epoch": 2.3278993945039588,
+      "grad_norm": 0.3056236723096069,
+      "learning_rate": 3.937864077669903e-05,
+      "loss": 0.4166,
+      "step": 2500
+    },
+    {
+      "epoch": 2.3325570563577083,
+      "grad_norm": 0.3024595378867525,
+      "learning_rate": 3.9346278317152106e-05,
+      "loss": 0.428,
+      "step": 2505
+    },
+    {
+      "epoch": 2.3372147182114578,
+      "grad_norm": 0.3016160148497247,
+      "learning_rate": 3.931391585760518e-05,
+      "loss": 0.4291,
+      "step": 2510
+    },
+    {
+      "epoch": 2.3418723800652073,
+      "grad_norm": 0.2766409436577404,
+      "learning_rate": 3.928155339805825e-05,
+      "loss": 0.4304,
+      "step": 2515
+    },
+    {
+      "epoch": 2.3465300419189568,
+      "grad_norm": 0.26733233943446133,
+      "learning_rate": 3.924919093851133e-05,
+      "loss": 0.4349,
+      "step": 2520
+    },
+    {
+      "epoch": 2.3511877037727063,
+      "grad_norm": 0.2690198456721221,
+      "learning_rate": 3.92168284789644e-05,
+      "loss": 0.4305,
+      "step": 2525
+    },
+    {
+      "epoch": 2.3558453656264557,
+      "grad_norm": 0.2945860386231397,
+      "learning_rate": 3.9184466019417476e-05,
+      "loss": 0.4313,
+      "step": 2530
+    },
+    {
+      "epoch": 2.360503027480205,
+      "grad_norm": 0.2887479801089616,
+      "learning_rate": 3.915210355987055e-05,
+      "loss": 0.4185,
+      "step": 2535
+    },
+    {
+      "epoch": 2.3651606893339543,
+      "grad_norm": 0.2657832811996791,
+      "learning_rate": 3.911974110032363e-05,
+      "loss": 0.4169,
+      "step": 2540
+    },
+    {
+      "epoch": 2.369818351187704,
+      "grad_norm": 0.28653108542984546,
+      "learning_rate": 3.90873786407767e-05,
+      "loss": 0.4324,
+      "step": 2545
+    },
+    {
+      "epoch": 2.3744760130414533,
+      "grad_norm": 0.28649792966496657,
+      "learning_rate": 3.9055016181229775e-05,
+      "loss": 0.4329,
+      "step": 2550
+    },
+    {
+      "epoch": 2.3791336748952028,
+      "grad_norm": 0.31285273419976656,
+      "learning_rate": 3.902265372168285e-05,
+      "loss": 0.4239,
+      "step": 2555
+    },
+    {
+      "epoch": 2.383791336748952,
+      "grad_norm": 0.29394295890338806,
+      "learning_rate": 3.899029126213593e-05,
+      "loss": 0.427,
+      "step": 2560
+    },
+    {
+      "epoch": 2.3884489986027013,
+      "grad_norm": 0.2751190301995722,
+      "learning_rate": 3.8957928802589e-05,
+      "loss": 0.4304,
+      "step": 2565
+    },
+    {
+      "epoch": 2.393106660456451,
+      "grad_norm": 0.28394635868081375,
+      "learning_rate": 3.8925566343042075e-05,
+      "loss": 0.4283,
+      "step": 2570
+    },
+    {
+      "epoch": 2.3977643223102003,
+      "grad_norm": 0.26088780818246793,
+      "learning_rate": 3.889320388349515e-05,
+      "loss": 0.4313,
+      "step": 2575
+    },
+    {
+      "epoch": 2.40242198416395,
+      "grad_norm": 0.27408334765565273,
+      "learning_rate": 3.886084142394822e-05,
+      "loss": 0.4211,
+      "step": 2580
+    },
+    {
+      "epoch": 2.4070796460176993,
+      "grad_norm": 0.30743853711271846,
+      "learning_rate": 3.882847896440129e-05,
+      "loss": 0.4228,
+      "step": 2585
+    },
+    {
+      "epoch": 2.4117373078714484,
+      "grad_norm": 0.28436879894690836,
+      "learning_rate": 3.879611650485437e-05,
+      "loss": 0.4173,
+      "step": 2590
+    },
+    {
+      "epoch": 2.416394969725198,
+      "grad_norm": 0.31784449370435436,
+      "learning_rate": 3.8763754045307444e-05,
+      "loss": 0.4366,
+      "step": 2595
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.2784369425125365,
+      "learning_rate": 3.873139158576052e-05,
+      "loss": 0.4325,
+      "step": 2600
+    },
+    {
+      "epoch": 2.425710293432697,
+      "grad_norm": 0.30370530331971196,
+      "learning_rate": 3.869902912621359e-05,
+      "loss": 0.4232,
+      "step": 2605
+    },
+    {
+      "epoch": 2.4303679552864463,
+      "grad_norm": 0.3064281445791429,
+      "learning_rate": 3.866666666666667e-05,
+      "loss": 0.419,
+      "step": 2610
+    },
+    {
+      "epoch": 2.4350256171401954,
+      "grad_norm": 0.33279912691579955,
+      "learning_rate": 3.8634304207119743e-05,
+      "loss": 0.4228,
+      "step": 2615
+    },
+    {
+      "epoch": 2.439683278993945,
+      "grad_norm": 0.30559746775800656,
+      "learning_rate": 3.860194174757282e-05,
+      "loss": 0.4279,
+      "step": 2620
+    },
+    {
+      "epoch": 2.4443409408476944,
+      "grad_norm": 0.2821489347743237,
+      "learning_rate": 3.856957928802589e-05,
+      "loss": 0.4273,
+      "step": 2625
+    },
+    {
+      "epoch": 2.448998602701444,
+      "grad_norm": 0.30707651894613214,
+      "learning_rate": 3.8537216828478966e-05,
+      "loss": 0.421,
+      "step": 2630
+    },
+    {
+      "epoch": 2.4536562645551934,
+      "grad_norm": 0.2961175138557341,
+      "learning_rate": 3.850485436893204e-05,
+      "loss": 0.433,
+      "step": 2635
+    },
+    {
+      "epoch": 2.458313926408943,
+      "grad_norm": 0.30038582825826193,
+      "learning_rate": 3.847249190938512e-05,
+      "loss": 0.4313,
+      "step": 2640
+    },
+    {
+      "epoch": 2.4629715882626924,
+      "grad_norm": 0.28346270038788274,
+      "learning_rate": 3.844012944983819e-05,
+      "loss": 0.4309,
+      "step": 2645
+    },
+    {
+      "epoch": 2.4676292501164414,
+      "grad_norm": 0.3057343243492832,
+      "learning_rate": 3.8407766990291266e-05,
+      "loss": 0.4228,
+      "step": 2650
+    },
+    {
+      "epoch": 2.472286911970191,
+      "grad_norm": 0.2838210234509721,
+      "learning_rate": 3.837540453074434e-05,
+      "loss": 0.4312,
+      "step": 2655
+    },
+    {
+      "epoch": 2.4769445738239404,
+      "grad_norm": 0.27759231451015043,
+      "learning_rate": 3.834304207119741e-05,
+      "loss": 0.4239,
+      "step": 2660
+    },
+    {
+      "epoch": 2.48160223567769,
+      "grad_norm": 0.30874594722339815,
+      "learning_rate": 3.831067961165048e-05,
+      "loss": 0.4239,
+      "step": 2665
+    },
+    {
+      "epoch": 2.4862598975314394,
+      "grad_norm": 0.28049362889434204,
+      "learning_rate": 3.827831715210356e-05,
+      "loss": 0.4313,
+      "step": 2670
+    },
+    {
+      "epoch": 2.4909175593851884,
+      "grad_norm": 0.3088921111059484,
+      "learning_rate": 3.8245954692556635e-05,
+      "loss": 0.4216,
+      "step": 2675
+    },
+    {
+      "epoch": 2.495575221238938,
+      "grad_norm": 0.3149293091667282,
+      "learning_rate": 3.821359223300971e-05,
+      "loss": 0.4235,
+      "step": 2680
+    },
+    {
+      "epoch": 2.5002328830926874,
+      "grad_norm": 0.28238147922335444,
+      "learning_rate": 3.818122977346278e-05,
+      "loss": 0.4271,
+      "step": 2685
+    },
+    {
+      "epoch": 2.504890544946437,
+      "grad_norm": 0.2956984739353203,
+      "learning_rate": 3.814886731391586e-05,
+      "loss": 0.4144,
+      "step": 2690
+    },
+    {
+      "epoch": 2.5095482068001864,
+      "grad_norm": 0.3037219452671658,
+      "learning_rate": 3.8116504854368935e-05,
+      "loss": 0.4256,
+      "step": 2695
+    },
+    {
+      "epoch": 2.514205868653936,
+      "grad_norm": 0.2808798231051215,
+      "learning_rate": 3.808414239482201e-05,
+      "loss": 0.4207,
+      "step": 2700
+    },
+    {
+      "epoch": 2.5188635305076854,
+      "grad_norm": 0.3439874362565095,
+      "learning_rate": 3.805177993527508e-05,
+      "loss": 0.4328,
+      "step": 2705
+    },
+    {
+      "epoch": 2.5235211923614345,
+      "grad_norm": 0.28499473571752737,
+      "learning_rate": 3.801941747572816e-05,
+      "loss": 0.4245,
+      "step": 2710
+    },
+    {
+      "epoch": 2.528178854215184,
+      "grad_norm": 0.2792589165257693,
+      "learning_rate": 3.7987055016181234e-05,
+      "loss": 0.4339,
+      "step": 2715
+    },
+    {
+      "epoch": 2.5328365160689335,
+      "grad_norm": 0.2768222961452844,
+      "learning_rate": 3.795469255663431e-05,
+      "loss": 0.4436,
+      "step": 2720
+    },
+    {
+      "epoch": 2.537494177922683,
+      "grad_norm": 0.2866344604425688,
+      "learning_rate": 3.792233009708738e-05,
+      "loss": 0.4263,
+      "step": 2725
+    },
+    {
+      "epoch": 2.542151839776432,
+      "grad_norm": 0.3126347316719504,
+      "learning_rate": 3.788996763754045e-05,
+      "loss": 0.4362,
+      "step": 2730
+    },
+    {
+      "epoch": 2.5468095016301815,
+      "grad_norm": 0.28054583912793213,
+      "learning_rate": 3.785760517799353e-05,
+      "loss": 0.4252,
+      "step": 2735
+    },
+    {
+      "epoch": 2.551467163483931,
+      "grad_norm": 0.3105960286486963,
+      "learning_rate": 3.78252427184466e-05,
+      "loss": 0.4235,
+      "step": 2740
+    },
+    {
+      "epoch": 2.5561248253376805,
+      "grad_norm": 0.2824370855630529,
+      "learning_rate": 3.779288025889968e-05,
+      "loss": 0.4224,
+      "step": 2745
+    },
+    {
+      "epoch": 2.56078248719143,
+      "grad_norm": 0.2694991578625555,
+      "learning_rate": 3.776051779935275e-05,
+      "loss": 0.4226,
+      "step": 2750
+    },
+    {
+      "epoch": 2.5654401490451795,
+      "grad_norm": 0.30479670220041877,
+      "learning_rate": 3.7728155339805826e-05,
+      "loss": 0.4176,
+      "step": 2755
+    },
+    {
+      "epoch": 2.570097810898929,
+      "grad_norm": 0.2838119776784065,
+      "learning_rate": 3.76957928802589e-05,
+      "loss": 0.431,
+      "step": 2760
+    },
+    {
+      "epoch": 2.574755472752678,
+      "grad_norm": 0.2902762884026888,
+      "learning_rate": 3.766343042071197e-05,
+      "loss": 0.422,
+      "step": 2765
+    },
+    {
+      "epoch": 2.5794131346064275,
+      "grad_norm": 0.2819819286550231,
+      "learning_rate": 3.763106796116505e-05,
+      "loss": 0.4247,
+      "step": 2770
+    },
+    {
+      "epoch": 2.584070796460177,
+      "grad_norm": 0.30369792665580253,
+      "learning_rate": 3.7598705501618126e-05,
+      "loss": 0.4225,
+      "step": 2775
+    },
+    {
+      "epoch": 2.5887284583139265,
+      "grad_norm": 0.29196706684716095,
+      "learning_rate": 3.75663430420712e-05,
+      "loss": 0.4299,
+      "step": 2780
+    },
+    {
+      "epoch": 2.5933861201676756,
+      "grad_norm": 0.2788017675818027,
+      "learning_rate": 3.753398058252427e-05,
+      "loss": 0.4346,
+      "step": 2785
+    },
+    {
+      "epoch": 2.598043782021425,
+      "grad_norm": 0.2916211716964687,
+      "learning_rate": 3.750161812297735e-05,
+      "loss": 0.435,
+      "step": 2790
+    },
+    {
+      "epoch": 2.6027014438751745,
+      "grad_norm": 0.278499551317208,
+      "learning_rate": 3.7469255663430425e-05,
+      "loss": 0.4244,
+      "step": 2795
+    },
+    {
+      "epoch": 2.607359105728924,
+      "grad_norm": 0.2711130362986734,
+      "learning_rate": 3.74368932038835e-05,
+      "loss": 0.4253,
+      "step": 2800
+    },
+    {
+      "epoch": 2.6120167675826735,
+      "grad_norm": 0.27338890570112695,
+      "learning_rate": 3.740453074433657e-05,
+      "loss": 0.4214,
+      "step": 2805
+    },
+    {
+      "epoch": 2.616674429436423,
+      "grad_norm": 0.2894722582711034,
+      "learning_rate": 3.737216828478964e-05,
+      "loss": 0.4288,
+      "step": 2810
+    },
+    {
+      "epoch": 2.6213320912901725,
+      "grad_norm": 0.24846741147246218,
+      "learning_rate": 3.733980582524272e-05,
+      "loss": 0.422,
+      "step": 2815
+    },
+    {
+      "epoch": 2.625989753143922,
+      "grad_norm": 0.2875890641223257,
+      "learning_rate": 3.7307443365695794e-05,
+      "loss": 0.4336,
+      "step": 2820
+    },
+    {
+      "epoch": 2.630647414997671,
+      "grad_norm": 0.2799915068598867,
+      "learning_rate": 3.727508090614887e-05,
+      "loss": 0.4331,
+      "step": 2825
+    },
+    {
+      "epoch": 2.6353050768514206,
+      "grad_norm": 0.3201306836255748,
+      "learning_rate": 3.724271844660194e-05,
+      "loss": 0.4227,
+      "step": 2830
+    },
+    {
+      "epoch": 2.63996273870517,
+      "grad_norm": 0.28169259531778595,
+      "learning_rate": 3.721035598705502e-05,
+      "loss": 0.4338,
+      "step": 2835
+    },
+    {
+      "epoch": 2.6446204005589196,
+      "grad_norm": 0.2587167611906602,
+      "learning_rate": 3.7177993527508094e-05,
+      "loss": 0.4366,
+      "step": 2840
+    },
+    {
+      "epoch": 2.6492780624126686,
+      "grad_norm": 0.27229247764700637,
+      "learning_rate": 3.714563106796117e-05,
+      "loss": 0.4313,
+      "step": 2845
+    },
+    {
+      "epoch": 2.653935724266418,
+      "grad_norm": 0.28467939462359715,
+      "learning_rate": 3.711326860841424e-05,
+      "loss": 0.4325,
+      "step": 2850
+    },
+    {
+      "epoch": 2.6585933861201676,
+      "grad_norm": 0.26758982806172527,
+      "learning_rate": 3.708090614886732e-05,
+      "loss": 0.426,
+      "step": 2855
+    },
+    {
+      "epoch": 2.663251047973917,
+      "grad_norm": 0.2743846459839584,
+      "learning_rate": 3.7048543689320393e-05,
+      "loss": 0.4322,
+      "step": 2860
+    },
+    {
+      "epoch": 2.6679087098276666,
+      "grad_norm": 0.284576576366453,
+      "learning_rate": 3.701618122977347e-05,
+      "loss": 0.4311,
+      "step": 2865
+    },
+    {
+      "epoch": 2.672566371681416,
+      "grad_norm": 0.2810259475558676,
+      "learning_rate": 3.698381877022654e-05,
+      "loss": 0.4226,
+      "step": 2870
+    },
+    {
+      "epoch": 2.6772240335351656,
+      "grad_norm": 0.2704861868580724,
+      "learning_rate": 3.695145631067961e-05,
+      "loss": 0.428,
+      "step": 2875
+    },
+    {
+      "epoch": 2.6818816953889146,
+      "grad_norm": 0.26025855572697787,
+      "learning_rate": 3.6919093851132686e-05,
+      "loss": 0.4198,
+      "step": 2880
+    },
+    {
+      "epoch": 2.686539357242664,
+      "grad_norm": 0.2783715023932399,
+      "learning_rate": 3.688673139158576e-05,
+      "loss": 0.4372,
+      "step": 2885
+    },
+    {
+      "epoch": 2.6911970190964136,
+      "grad_norm": 0.2519929327579714,
+      "learning_rate": 3.685436893203883e-05,
+      "loss": 0.4234,
+      "step": 2890
+    },
+    {
+      "epoch": 2.695854680950163,
+      "grad_norm": 0.31313597074446853,
+      "learning_rate": 3.682200647249191e-05,
+      "loss": 0.4312,
+      "step": 2895
+    },
+    {
+      "epoch": 2.700512342803912,
+      "grad_norm": 0.29198060410841264,
+      "learning_rate": 3.6789644012944986e-05,
+      "loss": 0.4256,
+      "step": 2900
+    },
+    {
+      "epoch": 2.7051700046576617,
+      "grad_norm": 0.2834423205145289,
+      "learning_rate": 3.675728155339806e-05,
+      "loss": 0.42,
+      "step": 2905
+    },
+    {
+      "epoch": 2.709827666511411,
+      "grad_norm": 0.2685259132702088,
+      "learning_rate": 3.672491909385113e-05,
+      "loss": 0.4233,
+      "step": 2910
+    },
+    {
+      "epoch": 2.7144853283651607,
+      "grad_norm": 0.2852612186401845,
+      "learning_rate": 3.669255663430421e-05,
+      "loss": 0.433,
+      "step": 2915
+    },
+    {
+      "epoch": 2.71914299021891,
+      "grad_norm": 0.2927922523300444,
+      "learning_rate": 3.6660194174757285e-05,
+      "loss": 0.4241,
+      "step": 2920
+    },
+    {
+      "epoch": 2.7238006520726596,
+      "grad_norm": 0.29404001370657046,
+      "learning_rate": 3.662783171521036e-05,
+      "loss": 0.4369,
+      "step": 2925
+    },
+    {
+      "epoch": 2.728458313926409,
+      "grad_norm": 0.2625429798487282,
+      "learning_rate": 3.659546925566343e-05,
+      "loss": 0.4356,
+      "step": 2930
+    },
+    {
+      "epoch": 2.7331159757801586,
+      "grad_norm": 0.26283427925119807,
+      "learning_rate": 3.656310679611651e-05,
+      "loss": 0.4241,
+      "step": 2935
+    },
+    {
+      "epoch": 2.7377736376339077,
+      "grad_norm": 0.2704873739007011,
+      "learning_rate": 3.6530744336569585e-05,
+      "loss": 0.424,
+      "step": 2940
+    },
+    {
+      "epoch": 2.742431299487657,
+      "grad_norm": 0.2656977810974544,
+      "learning_rate": 3.649838187702266e-05,
+      "loss": 0.4335,
+      "step": 2945
+    },
+    {
+      "epoch": 2.7470889613414067,
+      "grad_norm": 0.3103876140772922,
+      "learning_rate": 3.6466019417475724e-05,
+      "loss": 0.4253,
+      "step": 2950
+    },
+    {
+      "epoch": 2.751746623195156,
+      "grad_norm": 0.2829566677570478,
+      "learning_rate": 3.64336569579288e-05,
+      "loss": 0.4215,
+      "step": 2955
+    },
+    {
+      "epoch": 2.7564042850489052,
+      "grad_norm": 0.2841068708096097,
+      "learning_rate": 3.640129449838188e-05,
+      "loss": 0.4276,
+      "step": 2960
+    },
+    {
+      "epoch": 2.7610619469026547,
+      "grad_norm": 0.26485793653651674,
+      "learning_rate": 3.6368932038834954e-05,
+      "loss": 0.4263,
+      "step": 2965
+    },
+    {
+      "epoch": 2.765719608756404,
+      "grad_norm": 0.28162603988775287,
+      "learning_rate": 3.6336569579288024e-05,
+      "loss": 0.4283,
+      "step": 2970
+    },
+    {
+      "epoch": 2.7703772706101537,
+      "grad_norm": 0.2822946670919207,
+      "learning_rate": 3.63042071197411e-05,
+      "loss": 0.4207,
+      "step": 2975
+    },
+    {
+      "epoch": 2.775034932463903,
+      "grad_norm": 0.29086642862827244,
+      "learning_rate": 3.627184466019418e-05,
+      "loss": 0.4217,
+      "step": 2980
+    },
+    {
+      "epoch": 2.7796925943176527,
+      "grad_norm": 0.3044051581811522,
+      "learning_rate": 3.623948220064725e-05,
+      "loss": 0.432,
+      "step": 2985
+    },
+    {
+      "epoch": 2.784350256171402,
+      "grad_norm": 0.28576952931281374,
+      "learning_rate": 3.620711974110032e-05,
+      "loss": 0.4248,
+      "step": 2990
+    },
+    {
+      "epoch": 2.7890079180251512,
+      "grad_norm": 0.28252026904913935,
+      "learning_rate": 3.61747572815534e-05,
+      "loss": 0.4224,
+      "step": 2995
+    },
+    {
+      "epoch": 2.7936655798789007,
+      "grad_norm": 0.2863225210504502,
+      "learning_rate": 3.6142394822006476e-05,
+      "loss": 0.4342,
+      "step": 3000
+    },
+    {
+      "epoch": 2.7983232417326502,
+      "grad_norm": 0.25690143326129644,
+      "learning_rate": 3.611003236245955e-05,
+      "loss": 0.4201,
+      "step": 3005
+    },
+    {
+      "epoch": 2.8029809035863997,
+      "grad_norm": 0.2609803133859486,
+      "learning_rate": 3.607766990291262e-05,
+      "loss": 0.4354,
+      "step": 3010
+    },
+    {
+      "epoch": 2.807638565440149,
+      "grad_norm": 0.2763251649760853,
+      "learning_rate": 3.60453074433657e-05,
+      "loss": 0.4123,
+      "step": 3015
+    },
+    {
+      "epoch": 2.8122962272938983,
+      "grad_norm": 0.2552305839808659,
+      "learning_rate": 3.601294498381877e-05,
+      "loss": 0.4304,
+      "step": 3020
+    },
+    {
+      "epoch": 2.8169538891476478,
+      "grad_norm": 0.2615547694390696,
+      "learning_rate": 3.5980582524271845e-05,
+      "loss": 0.4268,
+      "step": 3025
+    },
+    {
+      "epoch": 2.8216115510013973,
+      "grad_norm": 0.2797157438024088,
+      "learning_rate": 3.594822006472492e-05,
+      "loss": 0.4281,
+      "step": 3030
+    },
+    {
+      "epoch": 2.8262692128551468,
+      "grad_norm": 0.27294965373088204,
+      "learning_rate": 3.591585760517799e-05,
+      "loss": 0.4292,
+      "step": 3035
+    },
+    {
+      "epoch": 2.8309268747088963,
+      "grad_norm": 0.3033447910940389,
+      "learning_rate": 3.588349514563107e-05,
+      "loss": 0.439,
+      "step": 3040
+    },
+    {
+      "epoch": 2.8355845365626458,
+      "grad_norm": 0.27535370122112035,
+      "learning_rate": 3.5851132686084145e-05,
+      "loss": 0.4254,
+      "step": 3045
+    },
+    {
+      "epoch": 2.840242198416395,
+      "grad_norm": 0.2872687513105721,
+      "learning_rate": 3.581877022653722e-05,
+      "loss": 0.4302,
+      "step": 3050
+    },
+    {
+      "epoch": 2.8448998602701443,
+      "grad_norm": 0.282487449246132,
+      "learning_rate": 3.578640776699029e-05,
+      "loss": 0.4315,
+      "step": 3055
+    },
+    {
+      "epoch": 2.849557522123894,
+      "grad_norm": 0.27478843406031495,
+      "learning_rate": 3.575404530744337e-05,
+      "loss": 0.4335,
+      "step": 3060
+    },
+    {
+      "epoch": 2.8542151839776433,
+      "grad_norm": 0.26905353589793823,
+      "learning_rate": 3.5721682847896444e-05,
+      "loss": 0.4275,
+      "step": 3065
+    },
+    {
+      "epoch": 2.858872845831393,
+      "grad_norm": 0.26383496267923806,
+      "learning_rate": 3.5689320388349514e-05,
+      "loss": 0.4179,
+      "step": 3070
+    },
+    {
+      "epoch": 2.863530507685142,
+      "grad_norm": 0.28002246896820876,
+      "learning_rate": 3.565695792880259e-05,
+      "loss": 0.4289,
+      "step": 3075
+    },
+    {
+      "epoch": 2.8681881695388913,
+      "grad_norm": 0.27936359297466795,
+      "learning_rate": 3.562459546925567e-05,
+      "loss": 0.4174,
+      "step": 3080
+    },
+    {
+      "epoch": 2.872845831392641,
+      "grad_norm": 0.2781403896702408,
+      "learning_rate": 3.5592233009708744e-05,
+      "loss": 0.4292,
+      "step": 3085
+    },
+    {
+      "epoch": 2.8775034932463903,
+      "grad_norm": 0.26905538248570515,
+      "learning_rate": 3.5559870550161814e-05,
+      "loss": 0.4332,
+      "step": 3090
+    },
+    {
+      "epoch": 2.88216115510014,
+      "grad_norm": 0.28429896396917936,
+      "learning_rate": 3.5527508090614884e-05,
+      "loss": 0.4196,
+      "step": 3095
+    },
+    {
+      "epoch": 2.8868188169538893,
+      "grad_norm": 0.2841084121770271,
+      "learning_rate": 3.549514563106796e-05,
+      "loss": 0.4273,
+      "step": 3100
+    },
+    {
+      "epoch": 2.891476478807639,
+      "grad_norm": 0.2670969391614412,
+      "learning_rate": 3.546278317152104e-05,
+      "loss": 0.4333,
+      "step": 3105
+    },
+    {
+      "epoch": 2.896134140661388,
+      "grad_norm": 0.2781049518759708,
+      "learning_rate": 3.543042071197411e-05,
+      "loss": 0.4213,
+      "step": 3110
+    },
+    {
+      "epoch": 2.9007918025151374,
+      "grad_norm": 0.2907894011753527,
+      "learning_rate": 3.539805825242718e-05,
+      "loss": 0.417,
+      "step": 3115
+    },
+    {
+      "epoch": 2.905449464368887,
+      "grad_norm": 0.2775449067053298,
+      "learning_rate": 3.536569579288026e-05,
+      "loss": 0.4256,
+      "step": 3120
+    },
+    {
+      "epoch": 2.9101071262226363,
+      "grad_norm": 0.26484076532516715,
+      "learning_rate": 3.5333333333333336e-05,
+      "loss": 0.4311,
+      "step": 3125
+    },
+    {
+      "epoch": 2.9147647880763854,
+      "grad_norm": 0.24962832810099225,
+      "learning_rate": 3.530097087378641e-05,
+      "loss": 0.414,
+      "step": 3130
+    },
+    {
+      "epoch": 2.919422449930135,
+      "grad_norm": 0.2728547334939587,
+      "learning_rate": 3.526860841423948e-05,
+      "loss": 0.4198,
+      "step": 3135
+    },
+    {
+      "epoch": 2.9240801117838844,
+      "grad_norm": 0.26480567082735235,
+      "learning_rate": 3.523624595469256e-05,
+      "loss": 0.4324,
+      "step": 3140
+    },
+    {
+      "epoch": 2.928737773637634,
+      "grad_norm": 0.277450964142194,
+      "learning_rate": 3.5203883495145636e-05,
+      "loss": 0.4347,
+      "step": 3145
+    },
+    {
+      "epoch": 2.9333954354913834,
+      "grad_norm": 0.30278907294138124,
+      "learning_rate": 3.517152103559871e-05,
+      "loss": 0.4323,
+      "step": 3150
+    },
+    {
+      "epoch": 2.938053097345133,
+      "grad_norm": 0.32235080879387107,
+      "learning_rate": 3.513915857605178e-05,
+      "loss": 0.4167,
+      "step": 3155
+    },
+    {
+      "epoch": 2.9427107591988824,
+      "grad_norm": 0.29246494531502704,
+      "learning_rate": 3.510679611650486e-05,
+      "loss": 0.4289,
+      "step": 3160
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.28676116742888197,
+      "learning_rate": 3.507443365695793e-05,
+      "loss": 0.4298,
+      "step": 3165
+    },
+    {
+      "epoch": 2.952026082906381,
+      "grad_norm": 0.2642095649143648,
+      "learning_rate": 3.5042071197411005e-05,
+      "loss": 0.4262,
+      "step": 3170
+    },
+    {
+      "epoch": 2.9566837447601304,
+      "grad_norm": 0.2699739868562258,
+      "learning_rate": 3.5009708737864075e-05,
+      "loss": 0.4313,
+      "step": 3175
+    },
+    {
+      "epoch": 2.96134140661388,
+      "grad_norm": 0.2660895898372268,
+      "learning_rate": 3.497734627831715e-05,
+      "loss": 0.4225,
+      "step": 3180
+    },
+    {
+      "epoch": 2.9659990684676294,
+      "grad_norm": 0.263658290507509,
+      "learning_rate": 3.494498381877023e-05,
+      "loss": 0.4277,
+      "step": 3185
+    },
+    {
+      "epoch": 2.9706567303213784,
+      "grad_norm": 0.2737982059810963,
+      "learning_rate": 3.4912621359223304e-05,
+      "loss": 0.4166,
+      "step": 3190
+    },
+    {
+      "epoch": 2.975314392175128,
+      "grad_norm": 0.2639166192385112,
+      "learning_rate": 3.4880258899676374e-05,
+      "loss": 0.4337,
+      "step": 3195
+    },
+    {
+      "epoch": 2.9799720540288774,
+      "grad_norm": 0.25426316889153205,
+      "learning_rate": 3.484789644012945e-05,
+      "loss": 0.4306,
+      "step": 3200
+    },
+    {
+      "epoch": 2.984629715882627,
+      "grad_norm": 0.30651398901580795,
+      "learning_rate": 3.481553398058253e-05,
+      "loss": 0.4377,
+      "step": 3205
+    },
+    {
+      "epoch": 2.9892873777363764,
+      "grad_norm": 0.2578203086881995,
+      "learning_rate": 3.4783171521035604e-05,
+      "loss": 0.4102,
+      "step": 3210
+    },
+    {
+      "epoch": 2.993945039590126,
+      "grad_norm": 0.25777786368813266,
+      "learning_rate": 3.4750809061488674e-05,
+      "loss": 0.4109,
+      "step": 3215
+    },
+    {
+      "epoch": 2.9986027014438754,
+      "grad_norm": 0.2943705743126533,
+      "learning_rate": 3.471844660194175e-05,
+      "loss": 0.4308,
+      "step": 3220
+    },
+    {
+      "epoch": 3.0027945971122496,
+      "grad_norm": 0.3189228181467205,
+      "learning_rate": 3.468608414239483e-05,
+      "loss": 0.3913,
+      "step": 3225
+    },
+    {
+      "epoch": 3.007452258965999,
+      "grad_norm": 0.32071113650327926,
+      "learning_rate": 3.46537216828479e-05,
+      "loss": 0.3748,
+      "step": 3230
+    },
+    {
+      "epoch": 3.0121099208197486,
+      "grad_norm": 0.296343837965841,
+      "learning_rate": 3.462135922330097e-05,
+      "loss": 0.3798,
+      "step": 3235
+    },
+    {
+      "epoch": 3.016767582673498,
+      "grad_norm": 0.2691270837455306,
+      "learning_rate": 3.458899676375404e-05,
+      "loss": 0.3692,
+      "step": 3240
+    },
+    {
+      "epoch": 3.021425244527247,
+      "grad_norm": 0.2881057731000792,
+      "learning_rate": 3.455663430420712e-05,
+      "loss": 0.3795,
+      "step": 3245
+    },
+    {
+      "epoch": 3.0260829063809966,
+      "grad_norm": 0.28863008786419747,
+      "learning_rate": 3.4524271844660196e-05,
+      "loss": 0.3805,
+      "step": 3250
+    },
+    {
+      "epoch": 3.030740568234746,
+      "grad_norm": 0.27177752021649265,
+      "learning_rate": 3.4491909385113266e-05,
+      "loss": 0.3819,
+      "step": 3255
+    },
+    {
+      "epoch": 3.0353982300884956,
+      "grad_norm": 0.2679652027957279,
+      "learning_rate": 3.445954692556634e-05,
+      "loss": 0.3828,
+      "step": 3260
+    },
+    {
+      "epoch": 3.040055891942245,
+      "grad_norm": 0.2868363828990796,
+      "learning_rate": 3.442718446601942e-05,
+      "loss": 0.3853,
+      "step": 3265
+    },
+    {
+      "epoch": 3.0447135537959946,
+      "grad_norm": 0.2890105371189473,
+      "learning_rate": 3.4394822006472495e-05,
+      "loss": 0.3732,
+      "step": 3270
+    },
+    {
+      "epoch": 3.0493712156497437,
+      "grad_norm": 0.27203805412745935,
+      "learning_rate": 3.4362459546925565e-05,
+      "loss": 0.3788,
+      "step": 3275
+    },
+    {
+      "epoch": 3.054028877503493,
+      "grad_norm": 0.2602642985237527,
+      "learning_rate": 3.433009708737864e-05,
+      "loss": 0.373,
+      "step": 3280
+    },
+    {
+      "epoch": 3.0586865393572427,
+      "grad_norm": 0.27094020224990417,
+      "learning_rate": 3.429773462783172e-05,
+      "loss": 0.3918,
+      "step": 3285
+    },
+    {
+      "epoch": 3.063344201210992,
+      "grad_norm": 0.26783933249771047,
+      "learning_rate": 3.4265372168284795e-05,
+      "loss": 0.3777,
+      "step": 3290
+    },
+    {
+      "epoch": 3.0680018630647417,
+      "grad_norm": 0.2756467871454317,
+      "learning_rate": 3.4233009708737865e-05,
+      "loss": 0.3823,
+      "step": 3295
+    },
+    {
+      "epoch": 3.0726595249184907,
+      "grad_norm": 0.31533387411110464,
+      "learning_rate": 3.420064724919094e-05,
+      "loss": 0.3748,
+      "step": 3300
+    },
+    {
+      "epoch": 3.07731718677224,
+      "grad_norm": 0.3032337273006253,
+      "learning_rate": 3.416828478964402e-05,
+      "loss": 0.3722,
+      "step": 3305
+    },
+    {
+      "epoch": 3.0819748486259897,
+      "grad_norm": 0.29797556470822706,
+      "learning_rate": 3.413592233009709e-05,
+      "loss": 0.3774,
+      "step": 3310
+    },
+    {
+      "epoch": 3.086632510479739,
+      "grad_norm": 0.2907185676087362,
+      "learning_rate": 3.4103559870550164e-05,
+      "loss": 0.3744,
+      "step": 3315
+    },
+    {
+      "epoch": 3.0912901723334887,
+      "grad_norm": 0.2797688048331569,
+      "learning_rate": 3.4071197411003234e-05,
+      "loss": 0.368,
+      "step": 3320
+    },
+    {
+      "epoch": 3.095947834187238,
+      "grad_norm": 0.280107318388062,
+      "learning_rate": 3.403883495145631e-05,
+      "loss": 0.3854,
+      "step": 3325
+    },
+    {
+      "epoch": 3.1006054960409872,
+      "grad_norm": 0.31397208430684337,
+      "learning_rate": 3.400647249190939e-05,
+      "loss": 0.3802,
+      "step": 3330
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.310995000296094,
+      "learning_rate": 3.3974110032362464e-05,
+      "loss": 0.3753,
+      "step": 3335
+    },
+    {
+      "epoch": 3.109920819748486,
+      "grad_norm": 0.28716773816954755,
+      "learning_rate": 3.3941747572815533e-05,
+      "loss": 0.3733,
+      "step": 3340
+    },
+    {
+      "epoch": 3.1145784816022357,
+      "grad_norm": 0.2959267581584993,
+      "learning_rate": 3.390938511326861e-05,
+      "loss": 0.3795,
+      "step": 3345
+    },
+    {
+      "epoch": 3.119236143455985,
+      "grad_norm": 0.27780045387843744,
+      "learning_rate": 3.3877022653721687e-05,
+      "loss": 0.3937,
+      "step": 3350
+    },
+    {
+      "epoch": 3.1238938053097347,
+      "grad_norm": 0.29258888948818523,
+      "learning_rate": 3.384466019417476e-05,
+      "loss": 0.3799,
+      "step": 3355
+    },
+    {
+      "epoch": 3.1285514671634838,
+      "grad_norm": 0.286240193461029,
+      "learning_rate": 3.381229773462783e-05,
+      "loss": 0.3798,
+      "step": 3360
+    },
+    {
+      "epoch": 3.1332091290172333,
+      "grad_norm": 0.29269590854130606,
+      "learning_rate": 3.377993527508091e-05,
+      "loss": 0.3873,
+      "step": 3365
+    },
+    {
+      "epoch": 3.1378667908709827,
+      "grad_norm": 0.2675117207246662,
+      "learning_rate": 3.3747572815533986e-05,
+      "loss": 0.3814,
+      "step": 3370
+    },
+    {
+      "epoch": 3.1425244527247322,
+      "grad_norm": 0.2655466234635772,
+      "learning_rate": 3.3715210355987056e-05,
+      "loss": 0.3814,
+      "step": 3375
+    },
+    {
+      "epoch": 3.1471821145784817,
+      "grad_norm": 0.2993931745308231,
+      "learning_rate": 3.3682847896440126e-05,
+      "loss": 0.3806,
+      "step": 3380
+    },
+    {
+      "epoch": 3.1518397764322312,
+      "grad_norm": 0.2737577998173692,
+      "learning_rate": 3.36504854368932e-05,
+      "loss": 0.3784,
+      "step": 3385
+    },
+    {
+      "epoch": 3.1564974382859803,
+      "grad_norm": 0.2852074345495303,
+      "learning_rate": 3.361812297734628e-05,
+      "loss": 0.3837,
+      "step": 3390
+    },
+    {
+      "epoch": 3.16115510013973,
+      "grad_norm": 0.27338540600457273,
+      "learning_rate": 3.3585760517799355e-05,
+      "loss": 0.3747,
+      "step": 3395
+    },
+    {
+      "epoch": 3.1658127619934793,
+      "grad_norm": 0.2768239874769818,
+      "learning_rate": 3.3553398058252425e-05,
+      "loss": 0.3839,
+      "step": 3400
+    },
+    {
+      "epoch": 3.1704704238472288,
+      "grad_norm": 0.25522005482265103,
+      "learning_rate": 3.35210355987055e-05,
+      "loss": 0.387,
+      "step": 3405
+    },
+    {
+      "epoch": 3.1751280857009783,
+      "grad_norm": 0.27748340384064507,
+      "learning_rate": 3.348867313915858e-05,
+      "loss": 0.3919,
+      "step": 3410
+    },
+    {
+      "epoch": 3.1797857475547273,
+      "grad_norm": 0.2730075292402206,
+      "learning_rate": 3.3456310679611655e-05,
+      "loss": 0.3769,
+      "step": 3415
+    },
+    {
+      "epoch": 3.184443409408477,
+      "grad_norm": 0.33155555731665076,
+      "learning_rate": 3.3423948220064725e-05,
+      "loss": 0.3917,
+      "step": 3420
+    },
+    {
+      "epoch": 3.1891010712622263,
+      "grad_norm": 0.3224328620915219,
+      "learning_rate": 3.33915857605178e-05,
+      "loss": 0.3815,
+      "step": 3425
+    },
+    {
+      "epoch": 3.193758733115976,
+      "grad_norm": 0.32625083415496575,
+      "learning_rate": 3.335922330097088e-05,
+      "loss": 0.3921,
+      "step": 3430
+    },
+    {
+      "epoch": 3.1984163949697253,
+      "grad_norm": 0.287751067070024,
+      "learning_rate": 3.3326860841423954e-05,
+      "loss": 0.3894,
+      "step": 3435
+    },
+    {
+      "epoch": 3.203074056823475,
+      "grad_norm": 0.3025426180601656,
+      "learning_rate": 3.3294498381877024e-05,
+      "loss": 0.3875,
+      "step": 3440
+    },
+    {
+      "epoch": 3.207731718677224,
+      "grad_norm": 0.2820344404437468,
+      "learning_rate": 3.32621359223301e-05,
+      "loss": 0.3748,
+      "step": 3445
+    },
+    {
+      "epoch": 3.2123893805309733,
+      "grad_norm": 0.27489434938772717,
+      "learning_rate": 3.322977346278318e-05,
+      "loss": 0.3837,
+      "step": 3450
+    },
+    {
+      "epoch": 3.217047042384723,
+      "grad_norm": 0.2697069074678676,
+      "learning_rate": 3.319741100323625e-05,
+      "loss": 0.3869,
+      "step": 3455
+    },
+    {
+      "epoch": 3.2217047042384723,
+      "grad_norm": 0.2711007301779973,
+      "learning_rate": 3.316504854368932e-05,
+      "loss": 0.3854,
+      "step": 3460
+    },
+    {
+      "epoch": 3.226362366092222,
+      "grad_norm": 0.30944819540145513,
+      "learning_rate": 3.313268608414239e-05,
+      "loss": 0.3802,
+      "step": 3465
+    },
+    {
+      "epoch": 3.2310200279459713,
+      "grad_norm": 0.29374337119158195,
+      "learning_rate": 3.310032362459547e-05,
+      "loss": 0.3875,
+      "step": 3470
+    },
+    {
+      "epoch": 3.2356776897997204,
+      "grad_norm": 0.26681487298679607,
+      "learning_rate": 3.3067961165048546e-05,
+      "loss": 0.38,
+      "step": 3475
+    },
+    {
+      "epoch": 3.24033535165347,
+      "grad_norm": 0.2630307816642084,
+      "learning_rate": 3.3035598705501616e-05,
+      "loss": 0.3854,
+      "step": 3480
+    },
+    {
+      "epoch": 3.2449930135072194,
+      "grad_norm": 0.28504157556514736,
+      "learning_rate": 3.300323624595469e-05,
+      "loss": 0.3939,
+      "step": 3485
+    },
+    {
+      "epoch": 3.249650675360969,
+      "grad_norm": 0.29762962713937785,
+      "learning_rate": 3.297087378640777e-05,
+      "loss": 0.3979,
+      "step": 3490
+    },
+    {
+      "epoch": 3.2543083372147183,
+      "grad_norm": 0.2967087683421554,
+      "learning_rate": 3.2938511326860846e-05,
+      "loss": 0.3958,
+      "step": 3495
+    },
+    {
+      "epoch": 3.258965999068468,
+      "grad_norm": 0.30491724136449017,
+      "learning_rate": 3.2906148867313916e-05,
+      "loss": 0.3944,
+      "step": 3500
+    },
+    {
+      "epoch": 3.263623660922217,
+      "grad_norm": 0.29395653686814566,
+      "learning_rate": 3.287378640776699e-05,
+      "loss": 0.3969,
+      "step": 3505
+    },
+    {
+      "epoch": 3.2682813227759664,
+      "grad_norm": 0.30303390983399636,
+      "learning_rate": 3.284142394822007e-05,
+      "loss": 0.3967,
+      "step": 3510
+    },
+    {
+      "epoch": 3.272938984629716,
+      "grad_norm": 0.27927982759731645,
+      "learning_rate": 3.2809061488673145e-05,
+      "loss": 0.3891,
+      "step": 3515
+    },
+    {
+      "epoch": 3.2775966464834654,
+      "grad_norm": 0.2678751340501679,
+      "learning_rate": 3.2776699029126215e-05,
+      "loss": 0.3874,
+      "step": 3520
+    },
+    {
+      "epoch": 3.282254308337215,
+      "grad_norm": 0.2900994579568571,
+      "learning_rate": 3.2744336569579285e-05,
+      "loss": 0.4021,
+      "step": 3525
+    },
+    {
+      "epoch": 3.286911970190964,
+      "grad_norm": 0.27271565228674416,
+      "learning_rate": 3.271197411003236e-05,
+      "loss": 0.3786,
+      "step": 3530
+    },
+    {
+      "epoch": 3.2915696320447134,
+      "grad_norm": 0.28652170030917257,
+      "learning_rate": 3.267961165048544e-05,
+      "loss": 0.3764,
+      "step": 3535
+    },
+    {
+      "epoch": 3.296227293898463,
+      "grad_norm": 0.2942415884741508,
+      "learning_rate": 3.2647249190938515e-05,
+      "loss": 0.3804,
+      "step": 3540
+    },
+    {
+      "epoch": 3.3008849557522124,
+      "grad_norm": 0.28215749117316613,
+      "learning_rate": 3.2614886731391584e-05,
+      "loss": 0.3899,
+      "step": 3545
+    },
+    {
+      "epoch": 3.305542617605962,
+      "grad_norm": 0.28326309598504706,
+      "learning_rate": 3.258252427184466e-05,
+      "loss": 0.3839,
+      "step": 3550
+    },
+    {
+      "epoch": 3.3102002794597114,
+      "grad_norm": 0.28041612783208264,
+      "learning_rate": 3.255016181229774e-05,
+      "loss": 0.3751,
+      "step": 3555
+    },
+    {
+      "epoch": 3.3148579413134605,
+      "grad_norm": 0.28751641561182656,
+      "learning_rate": 3.251779935275081e-05,
+      "loss": 0.3845,
+      "step": 3560
+    },
+    {
+      "epoch": 3.31951560316721,
+      "grad_norm": 0.2916078642342049,
+      "learning_rate": 3.2485436893203884e-05,
+      "loss": 0.3858,
+      "step": 3565
+    },
+    {
+      "epoch": 3.3241732650209594,
+      "grad_norm": 0.2746505539943781,
+      "learning_rate": 3.245307443365696e-05,
+      "loss": 0.3816,
+      "step": 3570
+    },
+    {
+      "epoch": 3.328830926874709,
+      "grad_norm": 0.266878501429854,
+      "learning_rate": 3.242071197411004e-05,
+      "loss": 0.3923,
+      "step": 3575
+    },
+    {
+      "epoch": 3.3334885887284584,
+      "grad_norm": 0.29417958059476884,
+      "learning_rate": 3.238834951456311e-05,
+      "loss": 0.3874,
+      "step": 3580
+    },
+    {
+      "epoch": 3.3381462505822075,
+      "grad_norm": 0.30459617589281895,
+      "learning_rate": 3.2355987055016183e-05,
+      "loss": 0.3897,
+      "step": 3585
+    },
+    {
+      "epoch": 3.342803912435957,
+      "grad_norm": 0.2619547631377024,
+      "learning_rate": 3.232362459546926e-05,
+      "loss": 0.382,
+      "step": 3590
+    },
+    {
+      "epoch": 3.3474615742897065,
+      "grad_norm": 0.2709779198156142,
+      "learning_rate": 3.2291262135922337e-05,
+      "loss": 0.3814,
+      "step": 3595
+    },
+    {
+      "epoch": 3.352119236143456,
+      "grad_norm": 0.29945087804261317,
+      "learning_rate": 3.2258899676375406e-05,
+      "loss": 0.3822,
+      "step": 3600
+    },
+    {
+      "epoch": 3.3567768979972055,
+      "grad_norm": 0.293632639764198,
+      "learning_rate": 3.2226537216828476e-05,
+      "loss": 0.3748,
+      "step": 3605
+    },
+    {
+      "epoch": 3.361434559850955,
+      "grad_norm": 0.2860156609777422,
+      "learning_rate": 3.219417475728155e-05,
+      "loss": 0.386,
+      "step": 3610
+    },
+    {
+      "epoch": 3.3660922217047045,
+      "grad_norm": 0.2911454668206605,
+      "learning_rate": 3.216181229773463e-05,
+      "loss": 0.3826,
+      "step": 3615
+    },
+    {
+      "epoch": 3.3707498835584535,
+      "grad_norm": 0.30221815752117726,
+      "learning_rate": 3.2129449838187706e-05,
+      "loss": 0.3822,
+      "step": 3620
+    },
+    {
+      "epoch": 3.375407545412203,
+      "grad_norm": 0.2794516738068307,
+      "learning_rate": 3.2097087378640776e-05,
+      "loss": 0.3919,
+      "step": 3625
+    },
+    {
+      "epoch": 3.3800652072659525,
+      "grad_norm": 0.3011158384207955,
+      "learning_rate": 3.206472491909385e-05,
+      "loss": 0.3847,
+      "step": 3630
+    },
+    {
+      "epoch": 3.384722869119702,
+      "grad_norm": 0.2739006993400138,
+      "learning_rate": 3.203236245954693e-05,
+      "loss": 0.3949,
+      "step": 3635
+    },
+    {
+      "epoch": 3.3893805309734515,
+      "grad_norm": 0.25395578448798356,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.3723,
+      "step": 3640
+    },
+    {
+      "epoch": 3.3940381928272005,
+      "grad_norm": 0.29246657684261074,
+      "learning_rate": 3.1967637540453075e-05,
+      "loss": 0.3889,
+      "step": 3645
+    },
+    {
+      "epoch": 3.39869585468095,
+      "grad_norm": 0.2853839589454004,
+      "learning_rate": 3.193527508090615e-05,
+      "loss": 0.3795,
+      "step": 3650
+    },
+    {
+      "epoch": 3.4033535165346995,
+      "grad_norm": 0.27538025167907765,
+      "learning_rate": 3.190291262135923e-05,
+      "loss": 0.3964,
+      "step": 3655
+    },
+    {
+      "epoch": 3.408011178388449,
+      "grad_norm": 0.2793994213733302,
+      "learning_rate": 3.1870550161812305e-05,
+      "loss": 0.3813,
+      "step": 3660
+    },
+    {
+      "epoch": 3.4126688402421985,
+      "grad_norm": 0.27325487831809636,
+      "learning_rate": 3.1838187702265375e-05,
+      "loss": 0.3817,
+      "step": 3665
+    },
+    {
+      "epoch": 3.417326502095948,
+      "grad_norm": 0.27544264466801227,
+      "learning_rate": 3.1805825242718444e-05,
+      "loss": 0.3959,
+      "step": 3670
+    },
+    {
+      "epoch": 3.421984163949697,
+      "grad_norm": 0.29269746788784207,
+      "learning_rate": 3.177346278317152e-05,
+      "loss": 0.3827,
+      "step": 3675
+    },
+    {
+      "epoch": 3.4266418258034466,
+      "grad_norm": 0.2716112401157802,
+      "learning_rate": 3.17411003236246e-05,
+      "loss": 0.3757,
+      "step": 3680
+    },
+    {
+      "epoch": 3.431299487657196,
+      "grad_norm": 0.2722022576763241,
+      "learning_rate": 3.170873786407767e-05,
+      "loss": 0.3903,
+      "step": 3685
+    },
+    {
+      "epoch": 3.4359571495109456,
+      "grad_norm": 0.27203797852537565,
+      "learning_rate": 3.1676375404530744e-05,
+      "loss": 0.3815,
+      "step": 3690
+    },
+    {
+      "epoch": 3.440614811364695,
+      "grad_norm": 0.2977608399442475,
+      "learning_rate": 3.164401294498382e-05,
+      "loss": 0.384,
+      "step": 3695
+    },
+    {
+      "epoch": 3.445272473218444,
+      "grad_norm": 0.2744028942156667,
+      "learning_rate": 3.16116504854369e-05,
+      "loss": 0.3928,
+      "step": 3700
+    },
+    {
+      "epoch": 3.4499301350721936,
+      "grad_norm": 0.2530992346697563,
+      "learning_rate": 3.157928802588997e-05,
+      "loss": 0.3945,
+      "step": 3705
+    },
+    {
+      "epoch": 3.454587796925943,
+      "grad_norm": 0.2668517004773288,
+      "learning_rate": 3.154692556634304e-05,
+      "loss": 0.39,
+      "step": 3710
+    },
+    {
+      "epoch": 3.4592454587796926,
+      "grad_norm": 0.2850822875751711,
+      "learning_rate": 3.151456310679612e-05,
+      "loss": 0.3803,
+      "step": 3715
+    },
+    {
+      "epoch": 3.463903120633442,
+      "grad_norm": 0.27417386624832685,
+      "learning_rate": 3.1482200647249196e-05,
+      "loss": 0.3831,
+      "step": 3720
+    },
+    {
+      "epoch": 3.4685607824871916,
+      "grad_norm": 0.27493865538797385,
+      "learning_rate": 3.1449838187702266e-05,
+      "loss": 0.3931,
+      "step": 3725
+    },
+    {
+      "epoch": 3.473218444340941,
+      "grad_norm": 0.28177535307624046,
+      "learning_rate": 3.141747572815534e-05,
+      "loss": 0.3879,
+      "step": 3730
+    },
+    {
+      "epoch": 3.47787610619469,
+      "grad_norm": 0.2829667377578548,
+      "learning_rate": 3.138511326860842e-05,
+      "loss": 0.3923,
+      "step": 3735
+    },
+    {
+      "epoch": 3.4825337680484396,
+      "grad_norm": 0.2674923275403164,
+      "learning_rate": 3.135275080906149e-05,
+      "loss": 0.3828,
+      "step": 3740
+    },
+    {
+      "epoch": 3.487191429902189,
+      "grad_norm": 0.27103683627144687,
+      "learning_rate": 3.132038834951456e-05,
+      "loss": 0.3879,
+      "step": 3745
+    },
+    {
+      "epoch": 3.4918490917559386,
+      "grad_norm": 0.28765532743751443,
+      "learning_rate": 3.1288025889967636e-05,
+      "loss": 0.392,
+      "step": 3750
+    },
+    {
+      "epoch": 3.496506753609688,
+      "grad_norm": 0.25654156325072325,
+      "learning_rate": 3.125566343042071e-05,
+      "loss": 0.3876,
+      "step": 3755
+    },
+    {
+      "epoch": 3.501164415463437,
+      "grad_norm": 0.26945747468313935,
+      "learning_rate": 3.122330097087379e-05,
+      "loss": 0.3858,
+      "step": 3760
+    },
+    {
+      "epoch": 3.5058220773171866,
+      "grad_norm": 0.2820454488625458,
+      "learning_rate": 3.119093851132686e-05,
+      "loss": 0.3787,
+      "step": 3765
+    },
+    {
+      "epoch": 3.510479739170936,
+      "grad_norm": 0.29939288089971683,
+      "learning_rate": 3.1158576051779935e-05,
+      "loss": 0.3925,
+      "step": 3770
+    },
+    {
+      "epoch": 3.5151374010246856,
+      "grad_norm": 0.2794718494657326,
+      "learning_rate": 3.112621359223301e-05,
+      "loss": 0.3813,
+      "step": 3775
+    },
+    {
+      "epoch": 3.519795062878435,
+      "grad_norm": 0.2552939514511198,
+      "learning_rate": 3.109385113268609e-05,
+      "loss": 0.3704,
+      "step": 3780
+    },
+    {
+      "epoch": 3.5244527247321846,
+      "grad_norm": 0.26867450601995413,
+      "learning_rate": 3.106148867313916e-05,
+      "loss": 0.3902,
+      "step": 3785
+    },
+    {
+      "epoch": 3.529110386585934,
+      "grad_norm": 0.26301103443758594,
+      "learning_rate": 3.1029126213592234e-05,
+      "loss": 0.3847,
+      "step": 3790
+    },
+    {
+      "epoch": 3.533768048439683,
+      "grad_norm": 0.26969144672133993,
+      "learning_rate": 3.099676375404531e-05,
+      "loss": 0.3814,
+      "step": 3795
+    },
+    {
+      "epoch": 3.5384257102934327,
+      "grad_norm": 0.25876446057457136,
+      "learning_rate": 3.096440129449839e-05,
+      "loss": 0.3933,
+      "step": 3800
+    },
+    {
+      "epoch": 3.543083372147182,
+      "grad_norm": 0.26528598864564545,
+      "learning_rate": 3.093203883495146e-05,
+      "loss": 0.3999,
+      "step": 3805
+    },
+    {
+      "epoch": 3.5477410340009317,
+      "grad_norm": 0.25395748015440367,
+      "learning_rate": 3.0899676375404534e-05,
+      "loss": 0.3922,
+      "step": 3810
+    },
+    {
+      "epoch": 3.5523986958546807,
+      "grad_norm": 0.2539446461729014,
+      "learning_rate": 3.0867313915857604e-05,
+      "loss": 0.3901,
+      "step": 3815
+    },
+    {
+      "epoch": 3.55705635770843,
+      "grad_norm": 0.25587630611112755,
+      "learning_rate": 3.083495145631068e-05,
+      "loss": 0.3809,
+      "step": 3820
+    },
+    {
+      "epoch": 3.5617140195621797,
+      "grad_norm": 0.2646247097511274,
+      "learning_rate": 3.080258899676376e-05,
+      "loss": 0.3952,
+      "step": 3825
+    },
+    {
+      "epoch": 3.566371681415929,
+      "grad_norm": 0.2657954284334948,
+      "learning_rate": 3.077022653721683e-05,
+      "loss": 0.3889,
+      "step": 3830
+    },
+    {
+      "epoch": 3.5710293432696787,
+      "grad_norm": 0.2536065507085372,
+      "learning_rate": 3.07378640776699e-05,
+      "loss": 0.3922,
+      "step": 3835
+    },
+    {
+      "epoch": 3.575687005123428,
+      "grad_norm": 0.28212563334640933,
+      "learning_rate": 3.070550161812298e-05,
+      "loss": 0.3853,
+      "step": 3840
+    },
+    {
+      "epoch": 3.5803446669771777,
+      "grad_norm": 0.2891735614020263,
+      "learning_rate": 3.0673139158576056e-05,
+      "loss": 0.3806,
+      "step": 3845
+    },
+    {
+      "epoch": 3.5850023288309267,
+      "grad_norm": 0.27718664019958394,
+      "learning_rate": 3.0640776699029126e-05,
+      "loss": 0.3865,
+      "step": 3850
+    },
+    {
+      "epoch": 3.5896599906846762,
+      "grad_norm": 0.2603433525061592,
+      "learning_rate": 3.06084142394822e-05,
+      "loss": 0.3957,
+      "step": 3855
+    },
+    {
+      "epoch": 3.5943176525384257,
+      "grad_norm": 0.2875738160319884,
+      "learning_rate": 3.057605177993528e-05,
+      "loss": 0.3889,
+      "step": 3860
+    },
+    {
+      "epoch": 3.598975314392175,
+      "grad_norm": 0.29818723262331726,
+      "learning_rate": 3.054368932038835e-05,
+      "loss": 0.3936,
+      "step": 3865
+    },
+    {
+      "epoch": 3.6036329762459243,
+      "grad_norm": 0.25405865808063205,
+      "learning_rate": 3.0511326860841426e-05,
+      "loss": 0.389,
+      "step": 3870
+    },
+    {
+      "epoch": 3.6082906380996738,
+      "grad_norm": 0.26055927208792357,
+      "learning_rate": 3.0478964401294502e-05,
+      "loss": 0.3889,
+      "step": 3875
+    },
+    {
+      "epoch": 3.6129482999534233,
+      "grad_norm": 0.2700398384860422,
+      "learning_rate": 3.0446601941747575e-05,
+      "loss": 0.3914,
+      "step": 3880
+    },
+    {
+      "epoch": 3.6176059618071728,
+      "grad_norm": 0.270324050559537,
+      "learning_rate": 3.0414239482200645e-05,
+      "loss": 0.3862,
+      "step": 3885
+    },
+    {
+      "epoch": 3.6222636236609222,
+      "grad_norm": 0.2601990175502833,
+      "learning_rate": 3.0381877022653722e-05,
+      "loss": 0.3861,
+      "step": 3890
+    },
+    {
+      "epoch": 3.6269212855146717,
+      "grad_norm": 0.26161758347757474,
+      "learning_rate": 3.0349514563106795e-05,
+      "loss": 0.3892,
+      "step": 3895
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.2489041481801808,
+      "learning_rate": 3.031715210355987e-05,
+      "loss": 0.3762,
+      "step": 3900
+    },
+    {
+      "epoch": 3.6362366092221707,
+      "grad_norm": 0.26774224401033997,
+      "learning_rate": 3.0284789644012945e-05,
+      "loss": 0.3959,
+      "step": 3905
+    },
+    {
+      "epoch": 3.64089427107592,
+      "grad_norm": 0.26793933237265394,
+      "learning_rate": 3.025242718446602e-05,
+      "loss": 0.3864,
+      "step": 3910
+    },
+    {
+      "epoch": 3.6455519329296693,
+      "grad_norm": 0.2585379407540758,
+      "learning_rate": 3.0220064724919094e-05,
+      "loss": 0.3905,
+      "step": 3915
+    },
+    {
+      "epoch": 3.6502095947834188,
+      "grad_norm": 0.26491393957203,
+      "learning_rate": 3.018770226537217e-05,
+      "loss": 0.3812,
+      "step": 3920
+    },
+    {
+      "epoch": 3.6548672566371683,
+      "grad_norm": 0.271400161850077,
+      "learning_rate": 3.0155339805825244e-05,
+      "loss": 0.3689,
+      "step": 3925
+    },
+    {
+      "epoch": 3.6595249184909173,
+      "grad_norm": 0.25988470519759704,
+      "learning_rate": 3.012297734627832e-05,
+      "loss": 0.3873,
+      "step": 3930
+    },
+    {
+      "epoch": 3.664182580344667,
+      "grad_norm": 0.2687443667800867,
+      "learning_rate": 3.0090614886731394e-05,
+      "loss": 0.3852,
+      "step": 3935
+    },
+    {
+      "epoch": 3.6688402421984163,
+      "grad_norm": 0.24634214735412785,
+      "learning_rate": 3.005825242718447e-05,
+      "loss": 0.3884,
+      "step": 3940
+    },
+    {
+      "epoch": 3.673497904052166,
+      "grad_norm": 0.2798971204953813,
+      "learning_rate": 3.0025889967637544e-05,
+      "loss": 0.3843,
+      "step": 3945
+    },
+    {
+      "epoch": 3.6781555659059153,
+      "grad_norm": 0.2606601001038783,
+      "learning_rate": 2.999352750809062e-05,
+      "loss": 0.3773,
+      "step": 3950
+    },
+    {
+      "epoch": 3.682813227759665,
+      "grad_norm": 0.25113689388450894,
+      "learning_rate": 2.9961165048543693e-05,
+      "loss": 0.3911,
+      "step": 3955
+    },
+    {
+      "epoch": 3.6874708896134143,
+      "grad_norm": 0.31146062762708887,
+      "learning_rate": 2.9928802588996763e-05,
+      "loss": 0.3932,
+      "step": 3960
+    },
+    {
+      "epoch": 3.6921285514671633,
+      "grad_norm": 0.2667929766530086,
+      "learning_rate": 2.9896440129449836e-05,
+      "loss": 0.3862,
+      "step": 3965
+    },
+    {
+      "epoch": 3.696786213320913,
+      "grad_norm": 0.2659300669598008,
+      "learning_rate": 2.9864077669902913e-05,
+      "loss": 0.3951,
+      "step": 3970
+    },
+    {
+      "epoch": 3.7014438751746623,
+      "grad_norm": 0.2505557400077424,
+      "learning_rate": 2.9831715210355986e-05,
+      "loss": 0.3762,
+      "step": 3975
+    },
+    {
+      "epoch": 3.706101537028412,
+      "grad_norm": 0.26179678001104023,
+      "learning_rate": 2.9799352750809063e-05,
+      "loss": 0.3819,
+      "step": 3980
+    },
+    {
+      "epoch": 3.710759198882161,
+      "grad_norm": 0.2655073543872324,
+      "learning_rate": 2.9766990291262136e-05,
+      "loss": 0.4004,
+      "step": 3985
+    },
+    {
+      "epoch": 3.7154168607359104,
+      "grad_norm": 0.27257211675153703,
+      "learning_rate": 2.9734627831715212e-05,
+      "loss": 0.3887,
+      "step": 3990
+    },
+    {
+      "epoch": 3.72007452258966,
+      "grad_norm": 0.2530649928660773,
+      "learning_rate": 2.9702265372168285e-05,
+      "loss": 0.3826,
+      "step": 3995
+    },
+    {
+      "epoch": 3.7247321844434094,
+      "grad_norm": 0.25498988538656236,
+      "learning_rate": 2.9669902912621362e-05,
+      "loss": 0.3865,
+      "step": 4000
+    },
+    {
+      "epoch": 3.729389846297159,
+      "grad_norm": 0.27534519248019096,
+      "learning_rate": 2.9637540453074435e-05,
+      "loss": 0.3861,
+      "step": 4005
+    },
+    {
+      "epoch": 3.7340475081509084,
+      "grad_norm": 0.30351479908172263,
+      "learning_rate": 2.9605177993527512e-05,
+      "loss": 0.3936,
+      "step": 4010
+    },
+    {
+      "epoch": 3.738705170004658,
+      "grad_norm": 0.26297033246002766,
+      "learning_rate": 2.9572815533980585e-05,
+      "loss": 0.3794,
+      "step": 4015
+    },
+    {
+      "epoch": 3.7433628318584073,
+      "grad_norm": 0.2570828812364618,
+      "learning_rate": 2.954045307443366e-05,
+      "loss": 0.3935,
+      "step": 4020
+    },
+    {
+      "epoch": 3.7480204937121564,
+      "grad_norm": 0.26413532586678856,
+      "learning_rate": 2.9508090614886735e-05,
+      "loss": 0.3902,
+      "step": 4025
+    },
+    {
+      "epoch": 3.752678155565906,
+      "grad_norm": 0.2469097728757831,
+      "learning_rate": 2.9475728155339804e-05,
+      "loss": 0.3871,
+      "step": 4030
+    },
+    {
+      "epoch": 3.7573358174196554,
+      "grad_norm": 0.25017081158294974,
+      "learning_rate": 2.944336569579288e-05,
+      "loss": 0.3844,
+      "step": 4035
+    },
+    {
+      "epoch": 3.761993479273405,
+      "grad_norm": 0.2618028697384336,
+      "learning_rate": 2.9411003236245954e-05,
+      "loss": 0.3816,
+      "step": 4040
+    },
+    {
+      "epoch": 3.766651141127154,
+      "grad_norm": 0.2589025120578359,
+      "learning_rate": 2.9378640776699027e-05,
+      "loss": 0.3987,
+      "step": 4045
+    },
+    {
+      "epoch": 3.7713088029809034,
+      "grad_norm": 0.25799806537651776,
+      "learning_rate": 2.9346278317152104e-05,
+      "loss": 0.3922,
+      "step": 4050
+    },
+    {
+      "epoch": 3.775966464834653,
+      "grad_norm": 0.2663926898446476,
+      "learning_rate": 2.9313915857605177e-05,
+      "loss": 0.3962,
+      "step": 4055
+    },
+    {
+      "epoch": 3.7806241266884024,
+      "grad_norm": 0.2603178754321631,
+      "learning_rate": 2.9281553398058254e-05,
+      "loss": 0.3857,
+      "step": 4060
+    },
+    {
+      "epoch": 3.785281788542152,
+      "grad_norm": 0.26731069796356594,
+      "learning_rate": 2.9249190938511327e-05,
+      "loss": 0.381,
+      "step": 4065
+    },
+    {
+      "epoch": 3.7899394503959014,
+      "grad_norm": 0.26696106303225803,
+      "learning_rate": 2.9216828478964403e-05,
+      "loss": 0.3862,
+      "step": 4070
+    },
+    {
+      "epoch": 3.794597112249651,
+      "grad_norm": 0.2815662554612125,
+      "learning_rate": 2.9184466019417477e-05,
+      "loss": 0.3826,
+      "step": 4075
+    },
+    {
+      "epoch": 3.7992547741034,
+      "grad_norm": 0.2827948401073868,
+      "learning_rate": 2.9152103559870553e-05,
+      "loss": 0.3891,
+      "step": 4080
+    },
+    {
+      "epoch": 3.8039124359571495,
+      "grad_norm": 0.2581943852613212,
+      "learning_rate": 2.9119741100323626e-05,
+      "loss": 0.3883,
+      "step": 4085
+    },
+    {
+      "epoch": 3.808570097810899,
+      "grad_norm": 0.2392873592198476,
+      "learning_rate": 2.9087378640776703e-05,
+      "loss": 0.3863,
+      "step": 4090
+    },
+    {
+      "epoch": 3.8132277596646484,
+      "grad_norm": 0.2584092344259934,
+      "learning_rate": 2.9055016181229776e-05,
+      "loss": 0.3811,
+      "step": 4095
+    },
+    {
+      "epoch": 3.8178854215183975,
+      "grad_norm": 0.27433228887528766,
+      "learning_rate": 2.9022653721682853e-05,
+      "loss": 0.3832,
+      "step": 4100
+    },
+    {
+      "epoch": 3.822543083372147,
+      "grad_norm": 0.2555586480404662,
+      "learning_rate": 2.8990291262135922e-05,
+      "loss": 0.3823,
+      "step": 4105
+    },
+    {
+      "epoch": 3.8272007452258965,
+      "grad_norm": 0.2562741954663834,
+      "learning_rate": 2.8957928802588996e-05,
+      "loss": 0.385,
+      "step": 4110
+    },
+    {
+      "epoch": 3.831858407079646,
+      "grad_norm": 0.2747688007417003,
+      "learning_rate": 2.8925566343042072e-05,
+      "loss": 0.3929,
+      "step": 4115
+    },
+    {
+      "epoch": 3.8365160689333955,
+      "grad_norm": 0.2751272213075399,
+      "learning_rate": 2.8893203883495145e-05,
+      "loss": 0.3802,
+      "step": 4120
+    },
+    {
+      "epoch": 3.841173730787145,
+      "grad_norm": 0.2866223815197106,
+      "learning_rate": 2.8860841423948222e-05,
+      "loss": 0.3918,
+      "step": 4125
+    },
+    {
+      "epoch": 3.8458313926408945,
+      "grad_norm": 0.28003334091875887,
+      "learning_rate": 2.8828478964401295e-05,
+      "loss": 0.3947,
+      "step": 4130
+    },
+    {
+      "epoch": 3.850489054494644,
+      "grad_norm": 0.27526826453198927,
+      "learning_rate": 2.879611650485437e-05,
+      "loss": 0.3922,
+      "step": 4135
+    },
+    {
+      "epoch": 3.855146716348393,
+      "grad_norm": 0.2693495968482176,
+      "learning_rate": 2.8763754045307445e-05,
+      "loss": 0.3895,
+      "step": 4140
+    },
+    {
+      "epoch": 3.8598043782021425,
+      "grad_norm": 0.26923804782011435,
+      "learning_rate": 2.873139158576052e-05,
+      "loss": 0.3921,
+      "step": 4145
+    },
+    {
+      "epoch": 3.864462040055892,
+      "grad_norm": 0.2569074358143511,
+      "learning_rate": 2.8699029126213595e-05,
+      "loss": 0.3805,
+      "step": 4150
+    },
+    {
+      "epoch": 3.8691197019096415,
+      "grad_norm": 0.26120815562925404,
+      "learning_rate": 2.8666666666666668e-05,
+      "loss": 0.3841,
+      "step": 4155
+    },
+    {
+      "epoch": 3.8737773637633905,
+      "grad_norm": 0.2656997451883316,
+      "learning_rate": 2.8634304207119744e-05,
+      "loss": 0.4023,
+      "step": 4160
+    },
+    {
+      "epoch": 3.87843502561714,
+      "grad_norm": 0.27576343279594656,
+      "learning_rate": 2.8601941747572818e-05,
+      "loss": 0.3942,
+      "step": 4165
+    },
+    {
+      "epoch": 3.8830926874708895,
+      "grad_norm": 0.2708624364712976,
+      "learning_rate": 2.8569579288025894e-05,
+      "loss": 0.3849,
+      "step": 4170
+    },
+    {
+      "epoch": 3.887750349324639,
+      "grad_norm": 0.2692406892963796,
+      "learning_rate": 2.8537216828478964e-05,
+      "loss": 0.393,
+      "step": 4175
+    },
+    {
+      "epoch": 3.8924080111783885,
+      "grad_norm": 0.25240465246199195,
+      "learning_rate": 2.8504854368932037e-05,
+      "loss": 0.3879,
+      "step": 4180
+    },
+    {
+      "epoch": 3.897065673032138,
+      "grad_norm": 0.2612808657935316,
+      "learning_rate": 2.8472491909385114e-05,
+      "loss": 0.3895,
+      "step": 4185
+    },
+    {
+      "epoch": 3.9017233348858875,
+      "grad_norm": 0.24624516475912692,
+      "learning_rate": 2.8440129449838187e-05,
+      "loss": 0.388,
+      "step": 4190
+    },
+    {
+      "epoch": 3.9063809967396366,
+      "grad_norm": 0.25821877656832976,
+      "learning_rate": 2.8407766990291263e-05,
+      "loss": 0.3802,
+      "step": 4195
+    },
+    {
+      "epoch": 3.911038658593386,
+      "grad_norm": 0.2706771888411189,
+      "learning_rate": 2.8375404530744337e-05,
+      "loss": 0.3835,
+      "step": 4200
+    },
+    {
+      "epoch": 3.9156963204471356,
+      "grad_norm": 0.26183511758896194,
+      "learning_rate": 2.8343042071197413e-05,
+      "loss": 0.3885,
+      "step": 4205
+    },
+    {
+      "epoch": 3.920353982300885,
+      "grad_norm": 0.25796419196414927,
+      "learning_rate": 2.8310679611650486e-05,
+      "loss": 0.3837,
+      "step": 4210
+    },
+    {
+      "epoch": 3.925011644154634,
+      "grad_norm": 0.25902566823988277,
+      "learning_rate": 2.8278317152103563e-05,
+      "loss": 0.3817,
+      "step": 4215
+    },
+    {
+      "epoch": 3.9296693060083836,
+      "grad_norm": 0.2616577356628845,
+      "learning_rate": 2.8245954692556636e-05,
+      "loss": 0.3853,
+      "step": 4220
+    },
+    {
+      "epoch": 3.934326967862133,
+      "grad_norm": 0.27134376397369114,
+      "learning_rate": 2.8213592233009713e-05,
+      "loss": 0.383,
+      "step": 4225
+    },
+    {
+      "epoch": 3.9389846297158826,
+      "grad_norm": 0.26252907445849977,
+      "learning_rate": 2.8181229773462786e-05,
+      "loss": 0.3864,
+      "step": 4230
+    },
+    {
+      "epoch": 3.943642291569632,
+      "grad_norm": 0.2727230567833132,
+      "learning_rate": 2.8148867313915862e-05,
+      "loss": 0.3848,
+      "step": 4235
+    },
+    {
+      "epoch": 3.9482999534233816,
+      "grad_norm": 0.2523302047492875,
+      "learning_rate": 2.8116504854368935e-05,
+      "loss": 0.3816,
+      "step": 4240
+    },
+    {
+      "epoch": 3.952957615277131,
+      "grad_norm": 0.2528613621966959,
+      "learning_rate": 2.8084142394822012e-05,
+      "loss": 0.3819,
+      "step": 4245
+    },
+    {
+      "epoch": 3.9576152771308806,
+      "grad_norm": 0.2550015777299334,
+      "learning_rate": 2.805177993527508e-05,
+      "loss": 0.3868,
+      "step": 4250
+    },
+    {
+      "epoch": 3.9622729389846296,
+      "grad_norm": 0.2606325297186323,
+      "learning_rate": 2.8019417475728155e-05,
+      "loss": 0.3871,
+      "step": 4255
+    },
+    {
+      "epoch": 3.966930600838379,
+      "grad_norm": 0.2641775467222413,
+      "learning_rate": 2.7987055016181228e-05,
+      "loss": 0.3938,
+      "step": 4260
+    },
+    {
+      "epoch": 3.9715882626921286,
+      "grad_norm": 0.26248135354387697,
+      "learning_rate": 2.7954692556634305e-05,
+      "loss": 0.3933,
+      "step": 4265
+    },
+    {
+      "epoch": 3.976245924545878,
+      "grad_norm": 0.2489884177697842,
+      "learning_rate": 2.7922330097087378e-05,
+      "loss": 0.3879,
+      "step": 4270
+    },
+    {
+      "epoch": 3.980903586399627,
+      "grad_norm": 0.26477408983487055,
+      "learning_rate": 2.7889967637540454e-05,
+      "loss": 0.3948,
+      "step": 4275
+    },
+    {
+      "epoch": 3.9855612482533767,
+      "grad_norm": 0.2996748671144824,
+      "learning_rate": 2.7857605177993528e-05,
+      "loss": 0.392,
+      "step": 4280
+    },
+    {
+      "epoch": 3.990218910107126,
+      "grad_norm": 0.2577741645300303,
+      "learning_rate": 2.7825242718446604e-05,
+      "loss": 0.3825,
+      "step": 4285
+    },
+    {
+      "epoch": 3.9948765719608756,
+      "grad_norm": 0.2604730744906666,
+      "learning_rate": 2.7792880258899677e-05,
+      "loss": 0.389,
+      "step": 4290
+    },
+    {
+      "epoch": 3.999534233814625,
+      "grad_norm": 0.2554837869509496,
+      "learning_rate": 2.7760517799352754e-05,
+      "loss": 0.39,
+      "step": 4295
+    },
+    {
+      "epoch": 4.003726129483,
+      "grad_norm": 0.30922681334743707,
+      "learning_rate": 2.7728155339805827e-05,
+      "loss": 0.3427,
+      "step": 4300
+    },
+    {
+      "epoch": 4.008383791336749,
+      "grad_norm": 0.301227570163748,
+      "learning_rate": 2.7695792880258904e-05,
+      "loss": 0.3331,
+      "step": 4305
+    },
+    {
+      "epoch": 4.013041453190499,
+      "grad_norm": 0.2991715513747825,
+      "learning_rate": 2.7663430420711977e-05,
+      "loss": 0.3375,
+      "step": 4310
+    },
+    {
+      "epoch": 4.017699115044247,
+      "grad_norm": 0.28312685959454287,
+      "learning_rate": 2.7631067961165053e-05,
+      "loss": 0.3322,
+      "step": 4315
+    },
+    {
+      "epoch": 4.022356776897997,
+      "grad_norm": 0.29405199530826526,
+      "learning_rate": 2.7598705501618123e-05,
+      "loss": 0.3345,
+      "step": 4320
+    },
+    {
+      "epoch": 4.027014438751746,
+      "grad_norm": 0.29429720520135993,
+      "learning_rate": 2.7566343042071196e-05,
+      "loss": 0.3259,
+      "step": 4325
+    },
+    {
+      "epoch": 4.031672100605496,
+      "grad_norm": 0.2637447275165894,
+      "learning_rate": 2.7533980582524273e-05,
+      "loss": 0.3359,
+      "step": 4330
+    },
+    {
+      "epoch": 4.036329762459245,
+      "grad_norm": 0.27150321164414054,
+      "learning_rate": 2.7501618122977346e-05,
+      "loss": 0.3387,
+      "step": 4335
+    },
+    {
+      "epoch": 4.040987424312995,
+      "grad_norm": 0.2884600945776732,
+      "learning_rate": 2.7469255663430423e-05,
+      "loss": 0.337,
+      "step": 4340
+    },
+    {
+      "epoch": 4.045645086166744,
+      "grad_norm": 0.27462833663202424,
+      "learning_rate": 2.7436893203883496e-05,
+      "loss": 0.3332,
+      "step": 4345
+    },
+    {
+      "epoch": 4.050302748020494,
+      "grad_norm": 0.2770260272624138,
+      "learning_rate": 2.740453074433657e-05,
+      "loss": 0.34,
+      "step": 4350
+    },
+    {
+      "epoch": 4.054960409874243,
+      "grad_norm": 0.2719819262086612,
+      "learning_rate": 2.7372168284789646e-05,
+      "loss": 0.3448,
+      "step": 4355
+    },
+    {
+      "epoch": 4.059618071727993,
+      "grad_norm": 0.2937422211184152,
+      "learning_rate": 2.733980582524272e-05,
+      "loss": 0.3453,
+      "step": 4360
+    },
+    {
+      "epoch": 4.064275733581742,
+      "grad_norm": 0.2883040671392013,
+      "learning_rate": 2.7307443365695795e-05,
+      "loss": 0.3419,
+      "step": 4365
+    },
+    {
+      "epoch": 4.068933395435492,
+      "grad_norm": 0.28028997899065444,
+      "learning_rate": 2.727508090614887e-05,
+      "loss": 0.3416,
+      "step": 4370
+    },
+    {
+      "epoch": 4.07359105728924,
+      "grad_norm": 0.2747243694429081,
+      "learning_rate": 2.7242718446601945e-05,
+      "loss": 0.3367,
+      "step": 4375
+    },
+    {
+      "epoch": 4.07824871914299,
+      "grad_norm": 0.28036218802618706,
+      "learning_rate": 2.7210355987055018e-05,
+      "loss": 0.336,
+      "step": 4380
+    },
+    {
+      "epoch": 4.082906380996739,
+      "grad_norm": 0.288073842340157,
+      "learning_rate": 2.7177993527508095e-05,
+      "loss": 0.335,
+      "step": 4385
+    },
+    {
+      "epoch": 4.087564042850489,
+      "grad_norm": 0.2840079345646035,
+      "learning_rate": 2.7145631067961165e-05,
+      "loss": 0.3428,
+      "step": 4390
+    },
+    {
+      "epoch": 4.092221704704238,
+      "grad_norm": 0.274581937492412,
+      "learning_rate": 2.7113268608414238e-05,
+      "loss": 0.3394,
+      "step": 4395
+    },
+    {
+      "epoch": 4.096879366557988,
+      "grad_norm": 0.2772673030668756,
+      "learning_rate": 2.7080906148867314e-05,
+      "loss": 0.3394,
+      "step": 4400
+    },
+    {
+      "epoch": 4.101537028411737,
+      "grad_norm": 0.27849278411701783,
+      "learning_rate": 2.7048543689320388e-05,
+      "loss": 0.3386,
+      "step": 4405
+    },
+    {
+      "epoch": 4.106194690265487,
+      "grad_norm": 0.2796044183413613,
+      "learning_rate": 2.7016181229773464e-05,
+      "loss": 0.3415,
+      "step": 4410
+    },
+    {
+      "epoch": 4.110852352119236,
+      "grad_norm": 0.275451109703783,
+      "learning_rate": 2.6983818770226537e-05,
+      "loss": 0.3387,
+      "step": 4415
+    },
+    {
+      "epoch": 4.115510013972986,
+      "grad_norm": 0.2961787378625091,
+      "learning_rate": 2.6951456310679614e-05,
+      "loss": 0.3472,
+      "step": 4420
+    },
+    {
+      "epoch": 4.120167675826735,
+      "grad_norm": 0.2834404893264586,
+      "learning_rate": 2.6919093851132687e-05,
+      "loss": 0.3419,
+      "step": 4425
+    },
+    {
+      "epoch": 4.124825337680484,
+      "grad_norm": 0.2770535443435555,
+      "learning_rate": 2.6886731391585764e-05,
+      "loss": 0.3459,
+      "step": 4430
+    },
+    {
+      "epoch": 4.1294829995342335,
+      "grad_norm": 0.276070086645229,
+      "learning_rate": 2.6854368932038837e-05,
+      "loss": 0.3323,
+      "step": 4435
+    },
+    {
+      "epoch": 4.134140661387983,
+      "grad_norm": 0.26597945758759256,
+      "learning_rate": 2.6822006472491913e-05,
+      "loss": 0.3415,
+      "step": 4440
+    },
+    {
+      "epoch": 4.1387983232417325,
+      "grad_norm": 0.28650783158027116,
+      "learning_rate": 2.6789644012944986e-05,
+      "loss": 0.3422,
+      "step": 4445
+    },
+    {
+      "epoch": 4.143455985095482,
+      "grad_norm": 0.2731057495828754,
+      "learning_rate": 2.6757281553398063e-05,
+      "loss": 0.3429,
+      "step": 4450
+    },
+    {
+      "epoch": 4.1481136469492315,
+      "grad_norm": 0.2785744835419355,
+      "learning_rate": 2.6724919093851136e-05,
+      "loss": 0.3412,
+      "step": 4455
+    },
+    {
+      "epoch": 4.152771308802981,
+      "grad_norm": 0.27582090334567066,
+      "learning_rate": 2.669255663430421e-05,
+      "loss": 0.3403,
+      "step": 4460
+    },
+    {
+      "epoch": 4.1574289706567304,
+      "grad_norm": 0.2761543634938907,
+      "learning_rate": 2.666019417475728e-05,
+      "loss": 0.35,
+      "step": 4465
+    },
+    {
+      "epoch": 4.16208663251048,
+      "grad_norm": 0.2720827997393211,
+      "learning_rate": 2.6627831715210356e-05,
+      "loss": 0.3482,
+      "step": 4470
+    },
+    {
+      "epoch": 4.166744294364229,
+      "grad_norm": 0.2887979203991473,
+      "learning_rate": 2.659546925566343e-05,
+      "loss": 0.3406,
+      "step": 4475
+    },
+    {
+      "epoch": 4.171401956217979,
+      "grad_norm": 0.2827047506321513,
+      "learning_rate": 2.6563106796116505e-05,
+      "loss": 0.3402,
+      "step": 4480
+    },
+    {
+      "epoch": 4.1760596180717275,
+      "grad_norm": 0.2772822053172354,
+      "learning_rate": 2.653074433656958e-05,
+      "loss": 0.3299,
+      "step": 4485
+    },
+    {
+      "epoch": 4.180717279925477,
+      "grad_norm": 0.29272264893784194,
+      "learning_rate": 2.6498381877022655e-05,
+      "loss": 0.3354,
+      "step": 4490
+    },
+    {
+      "epoch": 4.1853749417792265,
+      "grad_norm": 0.29375422060955253,
+      "learning_rate": 2.646601941747573e-05,
+      "loss": 0.3453,
+      "step": 4495
+    },
+    {
+      "epoch": 4.190032603632976,
+      "grad_norm": 0.29150751325996316,
+      "learning_rate": 2.6433656957928805e-05,
+      "loss": 0.3484,
+      "step": 4500
+    },
+    {
+      "epoch": 4.1946902654867255,
+      "grad_norm": 0.2817529454463103,
+      "learning_rate": 2.6401294498381878e-05,
+      "loss": 0.3371,
+      "step": 4505
+    },
+    {
+      "epoch": 4.199347927340475,
+      "grad_norm": 0.28966074414240217,
+      "learning_rate": 2.6368932038834955e-05,
+      "loss": 0.3451,
+      "step": 4510
+    },
+    {
+      "epoch": 4.2040055891942245,
+      "grad_norm": 0.27391509686820326,
+      "learning_rate": 2.6336569579288028e-05,
+      "loss": 0.3416,
+      "step": 4515
+    },
+    {
+      "epoch": 4.208663251047974,
+      "grad_norm": 0.29148215450693565,
+      "learning_rate": 2.6304207119741104e-05,
+      "loss": 0.3414,
+      "step": 4520
+    },
+    {
+      "epoch": 4.2133209129017235,
+      "grad_norm": 0.2689398936859596,
+      "learning_rate": 2.6271844660194178e-05,
+      "loss": 0.3385,
+      "step": 4525
+    },
+    {
+      "epoch": 4.217978574755473,
+      "grad_norm": 0.296520115240737,
+      "learning_rate": 2.6239482200647254e-05,
+      "loss": 0.34,
+      "step": 4530
+    },
+    {
+      "epoch": 4.2226362366092225,
+      "grad_norm": 0.2819120898954846,
+      "learning_rate": 2.620711974110032e-05,
+      "loss": 0.3413,
+      "step": 4535
+    },
+    {
+      "epoch": 4.227293898462972,
+      "grad_norm": 0.3030595014424858,
+      "learning_rate": 2.6174757281553397e-05,
+      "loss": 0.3446,
+      "step": 4540
+    },
+    {
+      "epoch": 4.231951560316721,
+      "grad_norm": 0.28596944705353855,
+      "learning_rate": 2.614239482200647e-05,
+      "loss": 0.3447,
+      "step": 4545
+    },
+    {
+      "epoch": 4.23660922217047,
+      "grad_norm": 0.2820542054589958,
+      "learning_rate": 2.6110032362459547e-05,
+      "loss": 0.3424,
+      "step": 4550
+    },
+    {
+      "epoch": 4.24126688402422,
+      "grad_norm": 0.3143861183376282,
+      "learning_rate": 2.607766990291262e-05,
+      "loss": 0.3464,
+      "step": 4555
+    },
+    {
+      "epoch": 4.245924545877969,
+      "grad_norm": 0.28997006041292267,
+      "learning_rate": 2.6045307443365697e-05,
+      "loss": 0.3503,
+      "step": 4560
+    },
+    {
+      "epoch": 4.250582207731719,
+      "grad_norm": 0.29817368907759895,
+      "learning_rate": 2.601294498381877e-05,
+      "loss": 0.3433,
+      "step": 4565
+    },
+    {
+      "epoch": 4.255239869585468,
+      "grad_norm": 0.28644175397133737,
+      "learning_rate": 2.5980582524271846e-05,
+      "loss": 0.3504,
+      "step": 4570
+    },
+    {
+      "epoch": 4.259897531439218,
+      "grad_norm": 0.3080648006556646,
+      "learning_rate": 2.594822006472492e-05,
+      "loss": 0.35,
+      "step": 4575
+    },
+    {
+      "epoch": 4.264555193292967,
+      "grad_norm": 0.28181312998638813,
+      "learning_rate": 2.5915857605177996e-05,
+      "loss": 0.347,
+      "step": 4580
+    },
+    {
+      "epoch": 4.269212855146717,
+      "grad_norm": 0.26855612943955576,
+      "learning_rate": 2.588349514563107e-05,
+      "loss": 0.3456,
+      "step": 4585
+    },
+    {
+      "epoch": 4.273870517000466,
+      "grad_norm": 0.2813470462397564,
+      "learning_rate": 2.5851132686084146e-05,
+      "loss": 0.3424,
+      "step": 4590
+    },
+    {
+      "epoch": 4.2785281788542155,
+      "grad_norm": 0.27916875991725926,
+      "learning_rate": 2.581877022653722e-05,
+      "loss": 0.3466,
+      "step": 4595
+    },
+    {
+      "epoch": 4.283185840707965,
+      "grad_norm": 0.2698085291333574,
+      "learning_rate": 2.5786407766990296e-05,
+      "loss": 0.3474,
+      "step": 4600
+    },
+    {
+      "epoch": 4.287843502561714,
+      "grad_norm": 0.265844277649668,
+      "learning_rate": 2.575404530744337e-05,
+      "loss": 0.3524,
+      "step": 4605
+    },
+    {
+      "epoch": 4.292501164415463,
+      "grad_norm": 0.2697893276218071,
+      "learning_rate": 2.572168284789644e-05,
+      "loss": 0.3432,
+      "step": 4610
+    },
+    {
+      "epoch": 4.297158826269213,
+      "grad_norm": 0.3027818789291518,
+      "learning_rate": 2.5689320388349515e-05,
+      "loss": 0.3435,
+      "step": 4615
+    },
+    {
+      "epoch": 4.301816488122962,
+      "grad_norm": 0.3097305617204279,
+      "learning_rate": 2.5656957928802588e-05,
+      "loss": 0.3409,
+      "step": 4620
+    },
+    {
+      "epoch": 4.306474149976712,
+      "grad_norm": 0.28377648125088384,
+      "learning_rate": 2.5624595469255665e-05,
+      "loss": 0.3456,
+      "step": 4625
+    },
+    {
+      "epoch": 4.311131811830461,
+      "grad_norm": 0.30292279319148513,
+      "learning_rate": 2.5592233009708738e-05,
+      "loss": 0.3485,
+      "step": 4630
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.2783600029192179,
+      "learning_rate": 2.5559870550161815e-05,
+      "loss": 0.3439,
+      "step": 4635
+    },
+    {
+      "epoch": 4.32044713553796,
+      "grad_norm": 0.25988299657809694,
+      "learning_rate": 2.5527508090614888e-05,
+      "loss": 0.3428,
+      "step": 4640
+    },
+    {
+      "epoch": 4.32510479739171,
+      "grad_norm": 0.281776274609003,
+      "learning_rate": 2.549514563106796e-05,
+      "loss": 0.3423,
+      "step": 4645
+    },
+    {
+      "epoch": 4.329762459245459,
+      "grad_norm": 0.2848401214493112,
+      "learning_rate": 2.5462783171521038e-05,
+      "loss": 0.343,
+      "step": 4650
+    },
+    {
+      "epoch": 4.334420121099209,
+      "grad_norm": 0.26808676440877494,
+      "learning_rate": 2.543042071197411e-05,
+      "loss": 0.3423,
+      "step": 4655
+    },
+    {
+      "epoch": 4.339077782952957,
+      "grad_norm": 0.26625329707544965,
+      "learning_rate": 2.5398058252427187e-05,
+      "loss": 0.3468,
+      "step": 4660
+    },
+    {
+      "epoch": 4.343735444806707,
+      "grad_norm": 0.2760378023774553,
+      "learning_rate": 2.536569579288026e-05,
+      "loss": 0.3526,
+      "step": 4665
+    },
+    {
+      "epoch": 4.348393106660456,
+      "grad_norm": 0.27430108921888785,
+      "learning_rate": 2.5333333333333337e-05,
+      "loss": 0.3481,
+      "step": 4670
+    },
+    {
+      "epoch": 4.353050768514206,
+      "grad_norm": 0.2890731689439146,
+      "learning_rate": 2.530097087378641e-05,
+      "loss": 0.3384,
+      "step": 4675
+    },
+    {
+      "epoch": 4.357708430367955,
+      "grad_norm": 0.2850035961825501,
+      "learning_rate": 2.526860841423948e-05,
+      "loss": 0.3511,
+      "step": 4680
+    },
+    {
+      "epoch": 4.362366092221705,
+      "grad_norm": 0.28371174995581067,
+      "learning_rate": 2.5236245954692557e-05,
+      "loss": 0.341,
+      "step": 4685
+    },
+    {
+      "epoch": 4.367023754075454,
+      "grad_norm": 0.2798334297146016,
+      "learning_rate": 2.520388349514563e-05,
+      "loss": 0.3457,
+      "step": 4690
+    },
+    {
+      "epoch": 4.371681415929204,
+      "grad_norm": 0.2792780079485981,
+      "learning_rate": 2.5171521035598706e-05,
+      "loss": 0.3421,
+      "step": 4695
+    },
+    {
+      "epoch": 4.376339077782953,
+      "grad_norm": 0.2789377083550942,
+      "learning_rate": 2.513915857605178e-05,
+      "loss": 0.3526,
+      "step": 4700
+    },
+    {
+      "epoch": 4.380996739636703,
+      "grad_norm": 0.27147206127013335,
+      "learning_rate": 2.5106796116504856e-05,
+      "loss": 0.3441,
+      "step": 4705
+    },
+    {
+      "epoch": 4.385654401490452,
+      "grad_norm": 0.2902924692605414,
+      "learning_rate": 2.507443365695793e-05,
+      "loss": 0.3529,
+      "step": 4710
+    },
+    {
+      "epoch": 4.390312063344201,
+      "grad_norm": 0.27929557612231476,
+      "learning_rate": 2.5042071197411006e-05,
+      "loss": 0.3378,
+      "step": 4715
+    },
+    {
+      "epoch": 4.39496972519795,
+      "grad_norm": 0.2701750459913034,
+      "learning_rate": 2.500970873786408e-05,
+      "loss": 0.3412,
+      "step": 4720
+    },
+    {
+      "epoch": 4.3996273870517,
+      "grad_norm": 0.2799269464041573,
+      "learning_rate": 2.4977346278317155e-05,
+      "loss": 0.3375,
+      "step": 4725
+    },
+    {
+      "epoch": 4.404285048905449,
+      "grad_norm": 0.28078679287279756,
+      "learning_rate": 2.494498381877023e-05,
+      "loss": 0.3526,
+      "step": 4730
+    },
+    {
+      "epoch": 4.408942710759199,
+      "grad_norm": 0.28802952806823595,
+      "learning_rate": 2.4912621359223302e-05,
+      "loss": 0.3514,
+      "step": 4735
+    },
+    {
+      "epoch": 4.413600372612948,
+      "grad_norm": 0.28721454128911067,
+      "learning_rate": 2.4880258899676375e-05,
+      "loss": 0.3351,
+      "step": 4740
+    },
+    {
+      "epoch": 4.418258034466698,
+      "grad_norm": 0.2906670016825979,
+      "learning_rate": 2.484789644012945e-05,
+      "loss": 0.347,
+      "step": 4745
+    },
+    {
+      "epoch": 4.422915696320447,
+      "grad_norm": 0.2780805909530155,
+      "learning_rate": 2.4815533980582525e-05,
+      "loss": 0.3412,
+      "step": 4750
+    },
+    {
+      "epoch": 4.427573358174197,
+      "grad_norm": 0.3024089657369089,
+      "learning_rate": 2.47831715210356e-05,
+      "loss": 0.3547,
+      "step": 4755
+    },
+    {
+      "epoch": 4.432231020027946,
+      "grad_norm": 0.2807347536015423,
+      "learning_rate": 2.4750809061488674e-05,
+      "loss": 0.3461,
+      "step": 4760
+    },
+    {
+      "epoch": 4.436888681881696,
+      "grad_norm": 0.27572550054354583,
+      "learning_rate": 2.471844660194175e-05,
+      "loss": 0.3462,
+      "step": 4765
+    },
+    {
+      "epoch": 4.441546343735445,
+      "grad_norm": 0.2845970279662639,
+      "learning_rate": 2.468608414239482e-05,
+      "loss": 0.3372,
+      "step": 4770
+    },
+    {
+      "epoch": 4.446204005589194,
+      "grad_norm": 0.2798998146127669,
+      "learning_rate": 2.4653721682847897e-05,
+      "loss": 0.3453,
+      "step": 4775
+    },
+    {
+      "epoch": 4.450861667442943,
+      "grad_norm": 0.2951093743961065,
+      "learning_rate": 2.462135922330097e-05,
+      "loss": 0.3476,
+      "step": 4780
+    },
+    {
+      "epoch": 4.455519329296693,
+      "grad_norm": 0.2587475261460553,
+      "learning_rate": 2.4588996763754047e-05,
+      "loss": 0.3525,
+      "step": 4785
+    },
+    {
+      "epoch": 4.460176991150442,
+      "grad_norm": 0.2772934314044936,
+      "learning_rate": 2.455663430420712e-05,
+      "loss": 0.3441,
+      "step": 4790
+    },
+    {
+      "epoch": 4.464834653004192,
+      "grad_norm": 0.28261470030353486,
+      "learning_rate": 2.4524271844660197e-05,
+      "loss": 0.3473,
+      "step": 4795
+    },
+    {
+      "epoch": 4.469492314857941,
+      "grad_norm": 0.26471592322331733,
+      "learning_rate": 2.449190938511327e-05,
+      "loss": 0.3505,
+      "step": 4800
+    },
+    {
+      "epoch": 4.474149976711691,
+      "grad_norm": 0.28724363547599,
+      "learning_rate": 2.4459546925566343e-05,
+      "loss": 0.3487,
+      "step": 4805
+    },
+    {
+      "epoch": 4.47880763856544,
+      "grad_norm": 0.2959670533347613,
+      "learning_rate": 2.4427184466019416e-05,
+      "loss": 0.3451,
+      "step": 4810
+    },
+    {
+      "epoch": 4.48346530041919,
+      "grad_norm": 0.27338912373615454,
+      "learning_rate": 2.4394822006472493e-05,
+      "loss": 0.3406,
+      "step": 4815
+    },
+    {
+      "epoch": 4.488122962272939,
+      "grad_norm": 0.27552301008710245,
+      "learning_rate": 2.4362459546925566e-05,
+      "loss": 0.352,
+      "step": 4820
+    },
+    {
+      "epoch": 4.492780624126689,
+      "grad_norm": 0.5518125177948608,
+      "learning_rate": 2.4330097087378643e-05,
+      "loss": 0.3506,
+      "step": 4825
+    },
+    {
+      "epoch": 4.497438285980438,
+      "grad_norm": 0.27618297840980366,
+      "learning_rate": 2.4297734627831716e-05,
+      "loss": 0.3528,
+      "step": 4830
+    },
+    {
+      "epoch": 4.502095947834187,
+      "grad_norm": 0.2725652864541843,
+      "learning_rate": 2.4265372168284792e-05,
+      "loss": 0.3477,
+      "step": 4835
+    },
+    {
+      "epoch": 4.506753609687936,
+      "grad_norm": 0.2688666248203594,
+      "learning_rate": 2.4233009708737866e-05,
+      "loss": 0.3492,
+      "step": 4840
+    },
+    {
+      "epoch": 4.511411271541686,
+      "grad_norm": 0.2675264289953943,
+      "learning_rate": 2.420064724919094e-05,
+      "loss": 0.3463,
+      "step": 4845
+    },
+    {
+      "epoch": 4.516068933395435,
+      "grad_norm": 0.28151463623190387,
+      "learning_rate": 2.4168284789644012e-05,
+      "loss": 0.3398,
+      "step": 4850
+    },
+    {
+      "epoch": 4.520726595249185,
+      "grad_norm": 0.2566560814830423,
+      "learning_rate": 2.413592233009709e-05,
+      "loss": 0.3482,
+      "step": 4855
+    },
+    {
+      "epoch": 4.525384257102934,
+      "grad_norm": 0.2898263552733961,
+      "learning_rate": 2.4103559870550162e-05,
+      "loss": 0.3418,
+      "step": 4860
+    },
+    {
+      "epoch": 4.530041918956684,
+      "grad_norm": 0.2588298954480233,
+      "learning_rate": 2.4071197411003238e-05,
+      "loss": 0.347,
+      "step": 4865
+    },
+    {
+      "epoch": 4.534699580810433,
+      "grad_norm": 0.28811816872383217,
+      "learning_rate": 2.403883495145631e-05,
+      "loss": 0.3489,
+      "step": 4870
+    },
+    {
+      "epoch": 4.539357242664183,
+      "grad_norm": 0.27020662162198467,
+      "learning_rate": 2.4006472491909388e-05,
+      "loss": 0.3498,
+      "step": 4875
+    },
+    {
+      "epoch": 4.544014904517932,
+      "grad_norm": 0.2736851785899911,
+      "learning_rate": 2.397411003236246e-05,
+      "loss": 0.3505,
+      "step": 4880
+    },
+    {
+      "epoch": 4.548672566371682,
+      "grad_norm": 0.270114303022711,
+      "learning_rate": 2.3941747572815534e-05,
+      "loss": 0.3485,
+      "step": 4885
+    },
+    {
+      "epoch": 4.55333022822543,
+      "grad_norm": 0.274061760735231,
+      "learning_rate": 2.3909385113268608e-05,
+      "loss": 0.3454,
+      "step": 4890
+    },
+    {
+      "epoch": 4.55798789007918,
+      "grad_norm": 0.27592510610491877,
+      "learning_rate": 2.3877022653721684e-05,
+      "loss": 0.3435,
+      "step": 4895
+    },
+    {
+      "epoch": 4.562645551932929,
+      "grad_norm": 0.26265886365244917,
+      "learning_rate": 2.3844660194174757e-05,
+      "loss": 0.3381,
+      "step": 4900
+    },
+    {
+      "epoch": 4.567303213786679,
+      "grad_norm": 0.2675318112084943,
+      "learning_rate": 2.3812297734627834e-05,
+      "loss": 0.3479,
+      "step": 4905
+    },
+    {
+      "epoch": 4.571960875640428,
+      "grad_norm": 0.2756756004486905,
+      "learning_rate": 2.3779935275080907e-05,
+      "loss": 0.3454,
+      "step": 4910
+    },
+    {
+      "epoch": 4.576618537494178,
+      "grad_norm": 0.2744517804738508,
+      "learning_rate": 2.374757281553398e-05,
+      "loss": 0.3422,
+      "step": 4915
+    },
+    {
+      "epoch": 4.581276199347927,
+      "grad_norm": 0.2956500056572687,
+      "learning_rate": 2.3715210355987057e-05,
+      "loss": 0.3431,
+      "step": 4920
+    },
+    {
+      "epoch": 4.585933861201677,
+      "grad_norm": 0.2790678584215101,
+      "learning_rate": 2.368284789644013e-05,
+      "loss": 0.3496,
+      "step": 4925
+    },
+    {
+      "epoch": 4.590591523055426,
+      "grad_norm": 0.2765275041694904,
+      "learning_rate": 2.3650485436893206e-05,
+      "loss": 0.3423,
+      "step": 4930
+    },
+    {
+      "epoch": 4.595249184909176,
+      "grad_norm": 0.24998145054628612,
+      "learning_rate": 2.361812297734628e-05,
+      "loss": 0.3504,
+      "step": 4935
+    },
+    {
+      "epoch": 4.599906846762925,
+      "grad_norm": 0.2687141625914439,
+      "learning_rate": 2.3585760517799356e-05,
+      "loss": 0.3413,
+      "step": 4940
+    },
+    {
+      "epoch": 4.604564508616674,
+      "grad_norm": 0.25987178416925466,
+      "learning_rate": 2.355339805825243e-05,
+      "loss": 0.3569,
+      "step": 4945
+    },
+    {
+      "epoch": 4.6092221704704235,
+      "grad_norm": 0.2784445234026133,
+      "learning_rate": 2.3521035598705503e-05,
+      "loss": 0.3437,
+      "step": 4950
+    },
+    {
+      "epoch": 4.613879832324173,
+      "grad_norm": 0.2639561735354289,
+      "learning_rate": 2.3488673139158576e-05,
+      "loss": 0.3507,
+      "step": 4955
+    },
+    {
+      "epoch": 4.6185374941779225,
+      "grad_norm": 0.26122970993507066,
+      "learning_rate": 2.3456310679611652e-05,
+      "loss": 0.3426,
+      "step": 4960
+    },
+    {
+      "epoch": 4.623195156031672,
+      "grad_norm": 0.2692150538527661,
+      "learning_rate": 2.3423948220064725e-05,
+      "loss": 0.3546,
+      "step": 4965
+    },
+    {
+      "epoch": 4.6278528178854215,
+      "grad_norm": 0.2733463787678732,
+      "learning_rate": 2.3391585760517802e-05,
+      "loss": 0.3525,
+      "step": 4970
+    },
+    {
+      "epoch": 4.632510479739171,
+      "grad_norm": 0.27159880615014076,
+      "learning_rate": 2.3359223300970875e-05,
+      "loss": 0.3391,
+      "step": 4975
+    },
+    {
+      "epoch": 4.6371681415929205,
+      "grad_norm": 0.2727071919572588,
+      "learning_rate": 2.3326860841423952e-05,
+      "loss": 0.3407,
+      "step": 4980
+    },
+    {
+      "epoch": 4.64182580344667,
+      "grad_norm": 0.2700456284616228,
+      "learning_rate": 2.3294498381877025e-05,
+      "loss": 0.355,
+      "step": 4985
+    },
+    {
+      "epoch": 4.6464834653004194,
+      "grad_norm": 0.2676835694946083,
+      "learning_rate": 2.3262135922330098e-05,
+      "loss": 0.3413,
+      "step": 4990
+    },
+    {
+      "epoch": 4.651141127154169,
+      "grad_norm": 0.27720368166900317,
+      "learning_rate": 2.322977346278317e-05,
+      "loss": 0.3489,
+      "step": 4995
+    },
+    {
+      "epoch": 4.6557987890079175,
+      "grad_norm": 0.28112682498727826,
+      "learning_rate": 2.3197411003236248e-05,
+      "loss": 0.3498,
+      "step": 5000
+    },
+    {
+      "epoch": 4.660456450861668,
+      "grad_norm": 0.27696629138280643,
+      "learning_rate": 2.316504854368932e-05,
+      "loss": 0.3482,
+      "step": 5005
+    },
+    {
+      "epoch": 4.6651141127154165,
+      "grad_norm": 0.26509317898455204,
+      "learning_rate": 2.3132686084142398e-05,
+      "loss": 0.3526,
+      "step": 5010
+    },
+    {
+      "epoch": 4.669771774569166,
+      "grad_norm": 0.2661649794027691,
+      "learning_rate": 2.310032362459547e-05,
+      "loss": 0.3471,
+      "step": 5015
+    },
+    {
+      "epoch": 4.6744294364229155,
+      "grad_norm": 0.2742085688046022,
+      "learning_rate": 2.3067961165048547e-05,
+      "loss": 0.348,
+      "step": 5020
+    },
+    {
+      "epoch": 4.679087098276665,
+      "grad_norm": 0.2631833779800896,
+      "learning_rate": 2.3035598705501617e-05,
+      "loss": 0.3491,
+      "step": 5025
+    },
+    {
+      "epoch": 4.6837447601304145,
+      "grad_norm": 0.27825788482078895,
+      "learning_rate": 2.3003236245954694e-05,
+      "loss": 0.3453,
+      "step": 5030
+    },
+    {
+      "epoch": 4.688402421984164,
+      "grad_norm": 0.28050281118191694,
+      "learning_rate": 2.2970873786407767e-05,
+      "loss": 0.3462,
+      "step": 5035
+    },
+    {
+      "epoch": 4.6930600838379135,
+      "grad_norm": 0.2577816537435826,
+      "learning_rate": 2.2938511326860843e-05,
+      "loss": 0.3447,
+      "step": 5040
+    },
+    {
+      "epoch": 4.697717745691663,
+      "grad_norm": 0.2682148115169009,
+      "learning_rate": 2.2906148867313917e-05,
+      "loss": 0.3409,
+      "step": 5045
+    },
+    {
+      "epoch": 4.7023754075454125,
+      "grad_norm": 0.2716317439276806,
+      "learning_rate": 2.2873786407766993e-05,
+      "loss": 0.3431,
+      "step": 5050
+    },
+    {
+      "epoch": 4.707033069399162,
+      "grad_norm": 0.26484905556040395,
+      "learning_rate": 2.2841423948220066e-05,
+      "loss": 0.344,
+      "step": 5055
+    },
+    {
+      "epoch": 4.7116907312529115,
+      "grad_norm": 0.2582513351019725,
+      "learning_rate": 2.280906148867314e-05,
+      "loss": 0.3438,
+      "step": 5060
+    },
+    {
+      "epoch": 4.71634839310666,
+      "grad_norm": 0.27263840516335347,
+      "learning_rate": 2.2776699029126213e-05,
+      "loss": 0.3405,
+      "step": 5065
+    },
+    {
+      "epoch": 4.72100605496041,
+      "grad_norm": 0.2636007854508017,
+      "learning_rate": 2.274433656957929e-05,
+      "loss": 0.3444,
+      "step": 5070
+    },
+    {
+      "epoch": 4.725663716814159,
+      "grad_norm": 0.28450549420320026,
+      "learning_rate": 2.2711974110032362e-05,
+      "loss": 0.3495,
+      "step": 5075
+    },
+    {
+      "epoch": 4.730321378667909,
+      "grad_norm": 0.29049335289255,
+      "learning_rate": 2.267961165048544e-05,
+      "loss": 0.3501,
+      "step": 5080
+    },
+    {
+      "epoch": 4.734979040521658,
+      "grad_norm": 0.2617782030129591,
+      "learning_rate": 2.2647249190938512e-05,
+      "loss": 0.3412,
+      "step": 5085
+    },
+    {
+      "epoch": 4.739636702375408,
+      "grad_norm": 0.26619791820681776,
+      "learning_rate": 2.261488673139159e-05,
+      "loss": 0.3465,
+      "step": 5090
+    },
+    {
+      "epoch": 4.744294364229157,
+      "grad_norm": 0.258803310255694,
+      "learning_rate": 2.258252427184466e-05,
+      "loss": 0.3401,
+      "step": 5095
+    },
+    {
+      "epoch": 4.748952026082907,
+      "grad_norm": 0.26462008992087993,
+      "learning_rate": 2.2550161812297735e-05,
+      "loss": 0.3485,
+      "step": 5100
+    },
+    {
+      "epoch": 4.753609687936656,
+      "grad_norm": 0.28270089001402976,
+      "learning_rate": 2.2517799352750808e-05,
+      "loss": 0.3501,
+      "step": 5105
+    },
+    {
+      "epoch": 4.7582673497904056,
+      "grad_norm": 0.2768800118594251,
+      "learning_rate": 2.2485436893203885e-05,
+      "loss": 0.3483,
+      "step": 5110
+    },
+    {
+      "epoch": 4.762925011644155,
+      "grad_norm": 0.2827237369746407,
+      "learning_rate": 2.2453074433656958e-05,
+      "loss": 0.3455,
+      "step": 5115
+    },
+    {
+      "epoch": 4.767582673497904,
+      "grad_norm": 0.2586423898934621,
+      "learning_rate": 2.2420711974110035e-05,
+      "loss": 0.3404,
+      "step": 5120
+    },
+    {
+      "epoch": 4.772240335351653,
+      "grad_norm": 0.2695792090511496,
+      "learning_rate": 2.2388349514563108e-05,
+      "loss": 0.3469,
+      "step": 5125
+    },
+    {
+      "epoch": 4.776897997205403,
+      "grad_norm": 0.2728769912631749,
+      "learning_rate": 2.235598705501618e-05,
+      "loss": 0.3497,
+      "step": 5130
+    },
+    {
+      "epoch": 4.781555659059152,
+      "grad_norm": 0.26886469462732937,
+      "learning_rate": 2.2323624595469254e-05,
+      "loss": 0.3584,
+      "step": 5135
+    },
+    {
+      "epoch": 4.786213320912902,
+      "grad_norm": 0.2569094746861235,
+      "learning_rate": 2.229126213592233e-05,
+      "loss": 0.341,
+      "step": 5140
+    },
+    {
+      "epoch": 4.790870982766651,
+      "grad_norm": 0.27323842192783715,
+      "learning_rate": 2.2258899676375404e-05,
+      "loss": 0.3473,
+      "step": 5145
+    },
+    {
+      "epoch": 4.795528644620401,
+      "grad_norm": 0.2779329948321792,
+      "learning_rate": 2.222653721682848e-05,
+      "loss": 0.3341,
+      "step": 5150
+    },
+    {
+      "epoch": 4.80018630647415,
+      "grad_norm": 0.27433246496954844,
+      "learning_rate": 2.2194174757281554e-05,
+      "loss": 0.3397,
+      "step": 5155
+    },
+    {
+      "epoch": 4.8048439683279,
+      "grad_norm": 0.2751671836627462,
+      "learning_rate": 2.216181229773463e-05,
+      "loss": 0.34,
+      "step": 5160
+    },
+    {
+      "epoch": 4.809501630181649,
+      "grad_norm": 0.2581618089575914,
+      "learning_rate": 2.2129449838187703e-05,
+      "loss": 0.351,
+      "step": 5165
+    },
+    {
+      "epoch": 4.814159292035399,
+      "grad_norm": 0.28479693685585583,
+      "learning_rate": 2.2097087378640777e-05,
+      "loss": 0.3525,
+      "step": 5170
+    },
+    {
+      "epoch": 4.818816953889147,
+      "grad_norm": 0.30270734662749793,
+      "learning_rate": 2.2064724919093853e-05,
+      "loss": 0.3553,
+      "step": 5175
+    },
+    {
+      "epoch": 4.823474615742897,
+      "grad_norm": 0.2592937709832869,
+      "learning_rate": 2.2032362459546926e-05,
+      "loss": 0.3429,
+      "step": 5180
+    },
+    {
+      "epoch": 4.828132277596646,
+      "grad_norm": 0.263554513112591,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.3427,
+      "step": 5185
+    },
+    {
+      "epoch": 4.832789939450396,
+      "grad_norm": 0.27555447618027107,
+      "learning_rate": 2.1967637540453076e-05,
+      "loss": 0.3568,
+      "step": 5190
+    },
+    {
+      "epoch": 4.837447601304145,
+      "grad_norm": 0.27340251424360784,
+      "learning_rate": 2.193527508090615e-05,
+      "loss": 0.346,
+      "step": 5195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.25997046231793325,
+      "learning_rate": 2.1902912621359226e-05,
+      "loss": 0.3418,
+      "step": 5200
+    },
+    {
+      "epoch": 4.846762925011644,
+      "grad_norm": 0.2610912590644731,
+      "learning_rate": 2.18705501618123e-05,
+      "loss": 0.3515,
+      "step": 5205
+    },
+    {
+      "epoch": 4.851420586865394,
+      "grad_norm": 0.27044344081098914,
+      "learning_rate": 2.1838187702265372e-05,
+      "loss": 0.356,
+      "step": 5210
+    },
+    {
+      "epoch": 4.856078248719143,
+      "grad_norm": 0.2823486349622019,
+      "learning_rate": 2.180582524271845e-05,
+      "loss": 0.3483,
+      "step": 5215
+    },
+    {
+      "epoch": 4.860735910572893,
+      "grad_norm": 0.26661920969071545,
+      "learning_rate": 2.1773462783171522e-05,
+      "loss": 0.3495,
+      "step": 5220
+    },
+    {
+      "epoch": 4.865393572426642,
+      "grad_norm": 0.2601414397083011,
+      "learning_rate": 2.17411003236246e-05,
+      "loss": 0.3409,
+      "step": 5225
+    },
+    {
+      "epoch": 4.870051234280391,
+      "grad_norm": 0.25772456138046534,
+      "learning_rate": 2.170873786407767e-05,
+      "loss": 0.3476,
+      "step": 5230
+    },
+    {
+      "epoch": 4.874708896134141,
+      "grad_norm": 0.2801575006987134,
+      "learning_rate": 2.1676375404530748e-05,
+      "loss": 0.3535,
+      "step": 5235
+    },
+    {
+      "epoch": 4.87936655798789,
+      "grad_norm": 0.2713153646610175,
+      "learning_rate": 2.1644012944983818e-05,
+      "loss": 0.3455,
+      "step": 5240
+    },
+    {
+      "epoch": 4.884024219841639,
+      "grad_norm": 0.2756939128532273,
+      "learning_rate": 2.1611650485436894e-05,
+      "loss": 0.3581,
+      "step": 5245
+    },
+    {
+      "epoch": 4.888681881695389,
+      "grad_norm": 0.27593933048729546,
+      "learning_rate": 2.1579288025889968e-05,
+      "loss": 0.3559,
+      "step": 5250
+    },
+    {
+      "epoch": 4.893339543549138,
+      "grad_norm": 0.28570776066307496,
+      "learning_rate": 2.1546925566343044e-05,
+      "loss": 0.3478,
+      "step": 5255
+    },
+    {
+      "epoch": 4.897997205402888,
+      "grad_norm": 0.27757994536033176,
+      "learning_rate": 2.1514563106796117e-05,
+      "loss": 0.3413,
+      "step": 5260
+    },
+    {
+      "epoch": 4.902654867256637,
+      "grad_norm": 0.2656187787344544,
+      "learning_rate": 2.1482200647249194e-05,
+      "loss": 0.3592,
+      "step": 5265
+    },
+    {
+      "epoch": 4.907312529110387,
+      "grad_norm": 0.2658878897171109,
+      "learning_rate": 2.1449838187702267e-05,
+      "loss": 0.3506,
+      "step": 5270
+    },
+    {
+      "epoch": 4.911970190964136,
+      "grad_norm": 0.2689864272819296,
+      "learning_rate": 2.141747572815534e-05,
+      "loss": 0.3482,
+      "step": 5275
+    },
+    {
+      "epoch": 4.916627852817886,
+      "grad_norm": 0.25500517516010174,
+      "learning_rate": 2.1385113268608413e-05,
+      "loss": 0.3499,
+      "step": 5280
+    },
+    {
+      "epoch": 4.921285514671635,
+      "grad_norm": 0.26771225102965857,
+      "learning_rate": 2.135275080906149e-05,
+      "loss": 0.3453,
+      "step": 5285
+    },
+    {
+      "epoch": 4.925943176525385,
+      "grad_norm": 0.27129980615123656,
+      "learning_rate": 2.1320388349514563e-05,
+      "loss": 0.3539,
+      "step": 5290
+    },
+    {
+      "epoch": 4.930600838379133,
+      "grad_norm": 0.2696112163961816,
+      "learning_rate": 2.128802588996764e-05,
+      "loss": 0.3435,
+      "step": 5295
+    },
+    {
+      "epoch": 4.935258500232883,
+      "grad_norm": 0.2538508151162818,
+      "learning_rate": 2.1255663430420713e-05,
+      "loss": 0.348,
+      "step": 5300
+    },
+    {
+      "epoch": 4.939916162086632,
+      "grad_norm": 0.2758886557259427,
+      "learning_rate": 2.122330097087379e-05,
+      "loss": 0.3517,
+      "step": 5305
+    },
+    {
+      "epoch": 4.944573823940382,
+      "grad_norm": 0.27270125908117077,
+      "learning_rate": 2.119093851132686e-05,
+      "loss": 0.3467,
+      "step": 5310
+    },
+    {
+      "epoch": 4.949231485794131,
+      "grad_norm": 0.26641381930592944,
+      "learning_rate": 2.1158576051779936e-05,
+      "loss": 0.3395,
+      "step": 5315
+    },
+    {
+      "epoch": 4.953889147647881,
+      "grad_norm": 0.2671391952368209,
+      "learning_rate": 2.112621359223301e-05,
+      "loss": 0.348,
+      "step": 5320
+    },
+    {
+      "epoch": 4.95854680950163,
+      "grad_norm": 0.27534994898249343,
+      "learning_rate": 2.1093851132686086e-05,
+      "loss": 0.3505,
+      "step": 5325
+    },
+    {
+      "epoch": 4.96320447135538,
+      "grad_norm": 0.26284212092487064,
+      "learning_rate": 2.106148867313916e-05,
+      "loss": 0.3497,
+      "step": 5330
+    },
+    {
+      "epoch": 4.967862133209129,
+      "grad_norm": 0.2528206388631256,
+      "learning_rate": 2.1029126213592235e-05,
+      "loss": 0.3379,
+      "step": 5335
+    },
+    {
+      "epoch": 4.972519795062879,
+      "grad_norm": 0.25393533764045556,
+      "learning_rate": 2.099676375404531e-05,
+      "loss": 0.3485,
+      "step": 5340
+    },
+    {
+      "epoch": 4.977177456916628,
+      "grad_norm": 0.26312186320913344,
+      "learning_rate": 2.0964401294498385e-05,
+      "loss": 0.3434,
+      "step": 5345
+    },
+    {
+      "epoch": 4.981835118770377,
+      "grad_norm": 0.2675083313597985,
+      "learning_rate": 2.0932038834951455e-05,
+      "loss": 0.3388,
+      "step": 5350
+    },
+    {
+      "epoch": 4.986492780624126,
+      "grad_norm": 0.2656481802559307,
+      "learning_rate": 2.089967637540453e-05,
+      "loss": 0.3492,
+      "step": 5355
+    },
+    {
+      "epoch": 4.991150442477876,
+      "grad_norm": 0.2715437365012533,
+      "learning_rate": 2.0867313915857605e-05,
+      "loss": 0.3347,
+      "step": 5360
+    },
+    {
+      "epoch": 4.995808104331625,
+      "grad_norm": 0.27608544405178936,
+      "learning_rate": 2.083495145631068e-05,
+      "loss": 0.3497,
+      "step": 5365
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.26313815985857575,
+      "learning_rate": 2.0802588996763754e-05,
+      "loss": 0.3378,
+      "step": 5370
+    },
+    {
+      "epoch": 5.0046576618537495,
+      "grad_norm": 0.35923261940748996,
+      "learning_rate": 2.077022653721683e-05,
+      "loss": 0.3066,
+      "step": 5375
+    },
+    {
+      "epoch": 5.009315323707499,
+      "grad_norm": 0.32640043938876945,
+      "learning_rate": 2.0737864077669904e-05,
+      "loss": 0.3022,
+      "step": 5380
+    },
+    {
+      "epoch": 5.0139729855612485,
+      "grad_norm": 0.32506467004826806,
+      "learning_rate": 2.0705501618122977e-05,
+      "loss": 0.299,
+      "step": 5385
+    },
+    {
+      "epoch": 5.018630647414998,
+      "grad_norm": 0.3047003724926295,
+      "learning_rate": 2.067313915857605e-05,
+      "loss": 0.3079,
+      "step": 5390
+    },
+    {
+      "epoch": 5.0232883092687475,
+      "grad_norm": 0.2890942126979149,
+      "learning_rate": 2.0640776699029127e-05,
+      "loss": 0.2949,
+      "step": 5395
+    },
+    {
+      "epoch": 5.027945971122496,
+      "grad_norm": 0.28379140694538696,
+      "learning_rate": 2.06084142394822e-05,
+      "loss": 0.2933,
+      "step": 5400
+    },
+    {
+      "epoch": 5.032603632976246,
+      "grad_norm": 0.29498361686431485,
+      "learning_rate": 2.0576051779935277e-05,
+      "loss": 0.2944,
+      "step": 5405
+    },
+    {
+      "epoch": 5.037261294829995,
+      "grad_norm": 0.2857039981729688,
+      "learning_rate": 2.054368932038835e-05,
+      "loss": 0.296,
+      "step": 5410
+    },
+    {
+      "epoch": 5.041918956683745,
+      "grad_norm": 0.3213640227256369,
+      "learning_rate": 2.0511326860841426e-05,
+      "loss": 0.2968,
+      "step": 5415
+    },
+    {
+      "epoch": 5.046576618537494,
+      "grad_norm": 0.2935458031818663,
+      "learning_rate": 2.04789644012945e-05,
+      "loss": 0.292,
+      "step": 5420
+    },
+    {
+      "epoch": 5.051234280391244,
+      "grad_norm": 0.278771134528813,
+      "learning_rate": 2.0446601941747573e-05,
+      "loss": 0.2963,
+      "step": 5425
+    },
+    {
+      "epoch": 5.055891942244993,
+      "grad_norm": 0.28885954698934563,
+      "learning_rate": 2.041423948220065e-05,
+      "loss": 0.3027,
+      "step": 5430
+    },
+    {
+      "epoch": 5.0605496040987425,
+      "grad_norm": 0.2985162846167875,
+      "learning_rate": 2.0381877022653723e-05,
+      "loss": 0.2855,
+      "step": 5435
+    },
+    {
+      "epoch": 5.065207265952492,
+      "grad_norm": 0.2779933359640956,
+      "learning_rate": 2.0349514563106796e-05,
+      "loss": 0.2899,
+      "step": 5440
+    },
+    {
+      "epoch": 5.0698649278062415,
+      "grad_norm": 0.2794291563396326,
+      "learning_rate": 2.0317152103559872e-05,
+      "loss": 0.2948,
+      "step": 5445
+    },
+    {
+      "epoch": 5.074522589659991,
+      "grad_norm": 0.2782833031018338,
+      "learning_rate": 2.0284789644012945e-05,
+      "loss": 0.3055,
+      "step": 5450
+    },
+    {
+      "epoch": 5.0791802515137405,
+      "grad_norm": 0.3084450159275583,
+      "learning_rate": 2.025242718446602e-05,
+      "loss": 0.3048,
+      "step": 5455
+    },
+    {
+      "epoch": 5.083837913367489,
+      "grad_norm": 0.28616079154167284,
+      "learning_rate": 2.0220064724919095e-05,
+      "loss": 0.2873,
+      "step": 5460
+    },
+    {
+      "epoch": 5.088495575221239,
+      "grad_norm": 0.2933049647365567,
+      "learning_rate": 2.018770226537217e-05,
+      "loss": 0.2995,
+      "step": 5465
+    },
+    {
+      "epoch": 5.093153237074988,
+      "grad_norm": 0.2985619669338097,
+      "learning_rate": 2.0155339805825245e-05,
+      "loss": 0.303,
+      "step": 5470
+    },
+    {
+      "epoch": 5.097810898928738,
+      "grad_norm": 0.287924432538155,
+      "learning_rate": 2.0122977346278318e-05,
+      "loss": 0.304,
+      "step": 5475
+    },
+    {
+      "epoch": 5.102468560782487,
+      "grad_norm": 0.2822730331503238,
+      "learning_rate": 2.0090614886731395e-05,
+      "loss": 0.3096,
+      "step": 5480
+    },
+    {
+      "epoch": 5.107126222636237,
+      "grad_norm": 0.2881739389332346,
+      "learning_rate": 2.0058252427184468e-05,
+      "loss": 0.2994,
+      "step": 5485
+    },
+    {
+      "epoch": 5.111783884489986,
+      "grad_norm": 0.27816726175124123,
+      "learning_rate": 2.0025889967637544e-05,
+      "loss": 0.2958,
+      "step": 5490
+    },
+    {
+      "epoch": 5.116441546343736,
+      "grad_norm": 0.3030122639921386,
+      "learning_rate": 1.9993527508090614e-05,
+      "loss": 0.3091,
+      "step": 5495
+    },
+    {
+      "epoch": 5.121099208197485,
+      "grad_norm": 0.2966032003252686,
+      "learning_rate": 1.996116504854369e-05,
+      "loss": 0.3026,
+      "step": 5500
+    },
+    {
+      "epoch": 5.125756870051235,
+      "grad_norm": 0.28524766975263094,
+      "learning_rate": 1.9928802588996764e-05,
+      "loss": 0.301,
+      "step": 5505
+    },
+    {
+      "epoch": 5.130414531904984,
+      "grad_norm": 0.28364105344562807,
+      "learning_rate": 1.989644012944984e-05,
+      "loss": 0.3103,
+      "step": 5510
+    },
+    {
+      "epoch": 5.135072193758733,
+      "grad_norm": 0.28685104135946954,
+      "learning_rate": 1.9864077669902914e-05,
+      "loss": 0.2932,
+      "step": 5515
+    },
+    {
+      "epoch": 5.139729855612482,
+      "grad_norm": 0.28023219847238934,
+      "learning_rate": 1.983171521035599e-05,
+      "loss": 0.3046,
+      "step": 5520
+    },
+    {
+      "epoch": 5.144387517466232,
+      "grad_norm": 0.2983844446151215,
+      "learning_rate": 1.9799352750809063e-05,
+      "loss": 0.3006,
+      "step": 5525
+    },
+    {
+      "epoch": 5.149045179319981,
+      "grad_norm": 0.2888586797330622,
+      "learning_rate": 1.9766990291262137e-05,
+      "loss": 0.2979,
+      "step": 5530
+    },
+    {
+      "epoch": 5.153702841173731,
+      "grad_norm": 0.2991770317047864,
+      "learning_rate": 1.973462783171521e-05,
+      "loss": 0.3026,
+      "step": 5535
+    },
+    {
+      "epoch": 5.15836050302748,
+      "grad_norm": 0.29147195431068673,
+      "learning_rate": 1.9702265372168286e-05,
+      "loss": 0.2988,
+      "step": 5540
+    },
+    {
+      "epoch": 5.16301816488123,
+      "grad_norm": 0.29769143133613796,
+      "learning_rate": 1.966990291262136e-05,
+      "loss": 0.3001,
+      "step": 5545
+    },
+    {
+      "epoch": 5.167675826734979,
+      "grad_norm": 0.28832867719266875,
+      "learning_rate": 1.9637540453074436e-05,
+      "loss": 0.3064,
+      "step": 5550
+    },
+    {
+      "epoch": 5.172333488588729,
+      "grad_norm": 0.2969967378852645,
+      "learning_rate": 1.960517799352751e-05,
+      "loss": 0.3038,
+      "step": 5555
+    },
+    {
+      "epoch": 5.176991150442478,
+      "grad_norm": 0.2887367922059911,
+      "learning_rate": 1.9572815533980586e-05,
+      "loss": 0.3145,
+      "step": 5560
+    },
+    {
+      "epoch": 5.181648812296228,
+      "grad_norm": 0.2738537643409762,
+      "learning_rate": 1.9540453074433656e-05,
+      "loss": 0.2996,
+      "step": 5565
+    },
+    {
+      "epoch": 5.186306474149977,
+      "grad_norm": 0.2823866205348465,
+      "learning_rate": 1.9508090614886732e-05,
+      "loss": 0.3011,
+      "step": 5570
+    },
+    {
+      "epoch": 5.190964136003726,
+      "grad_norm": 0.26814278929933427,
+      "learning_rate": 1.9475728155339805e-05,
+      "loss": 0.303,
+      "step": 5575
+    },
+    {
+      "epoch": 5.195621797857475,
+      "grad_norm": 0.2878790608953289,
+      "learning_rate": 1.9443365695792882e-05,
+      "loss": 0.2982,
+      "step": 5580
+    },
+    {
+      "epoch": 5.200279459711225,
+      "grad_norm": 0.28478623464513214,
+      "learning_rate": 1.9411003236245955e-05,
+      "loss": 0.3002,
+      "step": 5585
+    },
+    {
+      "epoch": 5.204937121564974,
+      "grad_norm": 0.29267564895392156,
+      "learning_rate": 1.937864077669903e-05,
+      "loss": 0.3043,
+      "step": 5590
+    },
+    {
+      "epoch": 5.209594783418724,
+      "grad_norm": 0.30782221674287835,
+      "learning_rate": 1.9346278317152105e-05,
+      "loss": 0.3035,
+      "step": 5595
+    },
+    {
+      "epoch": 5.214252445272473,
+      "grad_norm": 0.2862863533195408,
+      "learning_rate": 1.9313915857605178e-05,
+      "loss": 0.307,
+      "step": 5600
+    },
+    {
+      "epoch": 5.218910107126223,
+      "grad_norm": 0.2849550623168926,
+      "learning_rate": 1.928155339805825e-05,
+      "loss": 0.3069,
+      "step": 5605
+    },
+    {
+      "epoch": 5.223567768979972,
+      "grad_norm": 0.29920473766249195,
+      "learning_rate": 1.9249190938511328e-05,
+      "loss": 0.3099,
+      "step": 5610
+    },
+    {
+      "epoch": 5.228225430833722,
+      "grad_norm": 0.2899994453541254,
+      "learning_rate": 1.92168284789644e-05,
+      "loss": 0.2988,
+      "step": 5615
+    },
+    {
+      "epoch": 5.232883092687471,
+      "grad_norm": 0.2745311500475131,
+      "learning_rate": 1.9184466019417478e-05,
+      "loss": 0.3013,
+      "step": 5620
+    },
+    {
+      "epoch": 5.237540754541221,
+      "grad_norm": 0.2905887239671658,
+      "learning_rate": 1.915210355987055e-05,
+      "loss": 0.2939,
+      "step": 5625
+    },
+    {
+      "epoch": 5.242198416394969,
+      "grad_norm": 0.2756889111772824,
+      "learning_rate": 1.9119741100323627e-05,
+      "loss": 0.2998,
+      "step": 5630
+    },
+    {
+      "epoch": 5.246856078248719,
+      "grad_norm": 0.29560085688678456,
+      "learning_rate": 1.9087378640776697e-05,
+      "loss": 0.3051,
+      "step": 5635
+    },
+    {
+      "epoch": 5.251513740102468,
+      "grad_norm": 0.28723648162439525,
+      "learning_rate": 1.9055016181229774e-05,
+      "loss": 0.2974,
+      "step": 5640
+    },
+    {
+      "epoch": 5.256171401956218,
+      "grad_norm": 0.275091484400794,
+      "learning_rate": 1.9022653721682847e-05,
+      "loss": 0.3014,
+      "step": 5645
+    },
+    {
+      "epoch": 5.260829063809967,
+      "grad_norm": 0.2924049738193532,
+      "learning_rate": 1.8990291262135923e-05,
+      "loss": 0.3022,
+      "step": 5650
+    },
+    {
+      "epoch": 5.265486725663717,
+      "grad_norm": 0.29754127995135604,
+      "learning_rate": 1.8957928802588997e-05,
+      "loss": 0.2999,
+      "step": 5655
+    },
+    {
+      "epoch": 5.270144387517466,
+      "grad_norm": 0.2798089481799948,
+      "learning_rate": 1.8925566343042073e-05,
+      "loss": 0.2991,
+      "step": 5660
+    },
+    {
+      "epoch": 5.274802049371216,
+      "grad_norm": 0.2749452867239249,
+      "learning_rate": 1.8893203883495146e-05,
+      "loss": 0.3046,
+      "step": 5665
+    },
+    {
+      "epoch": 5.279459711224965,
+      "grad_norm": 0.29331315085638715,
+      "learning_rate": 1.8860841423948223e-05,
+      "loss": 0.3,
+      "step": 5670
+    },
+    {
+      "epoch": 5.284117373078715,
+      "grad_norm": 0.2928312452903107,
+      "learning_rate": 1.8828478964401296e-05,
+      "loss": 0.2964,
+      "step": 5675
+    },
+    {
+      "epoch": 5.288775034932464,
+      "grad_norm": 0.3008544248483986,
+      "learning_rate": 1.879611650485437e-05,
+      "loss": 0.3098,
+      "step": 5680
+    },
+    {
+      "epoch": 5.293432696786214,
+      "grad_norm": 0.28553735790247403,
+      "learning_rate": 1.8763754045307442e-05,
+      "loss": 0.3047,
+      "step": 5685
+    },
+    {
+      "epoch": 5.298090358639962,
+      "grad_norm": 0.2887305220130655,
+      "learning_rate": 1.873139158576052e-05,
+      "loss": 0.3044,
+      "step": 5690
+    },
+    {
+      "epoch": 5.302748020493712,
+      "grad_norm": 0.27448891372378204,
+      "learning_rate": 1.8699029126213592e-05,
+      "loss": 0.3035,
+      "step": 5695
+    },
+    {
+      "epoch": 5.307405682347461,
+      "grad_norm": 0.28295005491441444,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 0.3005,
+      "step": 5700
+    },
+    {
+      "epoch": 5.312063344201211,
+      "grad_norm": 0.2725195115220492,
+      "learning_rate": 1.8634304207119742e-05,
+      "loss": 0.3001,
+      "step": 5705
+    },
+    {
+      "epoch": 5.31672100605496,
+      "grad_norm": 0.27847794549500177,
+      "learning_rate": 1.8601941747572815e-05,
+      "loss": 0.305,
+      "step": 5710
+    },
+    {
+      "epoch": 5.32137866790871,
+      "grad_norm": 0.2839018387739063,
+      "learning_rate": 1.856957928802589e-05,
+      "loss": 0.3049,
+      "step": 5715
+    },
+    {
+      "epoch": 5.326036329762459,
+      "grad_norm": 0.3021144133018839,
+      "learning_rate": 1.8537216828478965e-05,
+      "loss": 0.3037,
+      "step": 5720
+    },
+    {
+      "epoch": 5.330693991616209,
+      "grad_norm": 0.2972052723877232,
+      "learning_rate": 1.850485436893204e-05,
+      "loss": 0.3066,
+      "step": 5725
+    },
+    {
+      "epoch": 5.335351653469958,
+      "grad_norm": 0.2856206514179065,
+      "learning_rate": 1.8472491909385114e-05,
+      "loss": 0.3065,
+      "step": 5730
+    },
+    {
+      "epoch": 5.340009315323708,
+      "grad_norm": 0.28392937584839906,
+      "learning_rate": 1.844012944983819e-05,
+      "loss": 0.3058,
+      "step": 5735
+    },
+    {
+      "epoch": 5.344666977177457,
+      "grad_norm": 0.2842067899622778,
+      "learning_rate": 1.8407766990291264e-05,
+      "loss": 0.3063,
+      "step": 5740
+    },
+    {
+      "epoch": 5.349324639031206,
+      "grad_norm": 0.2824119276233377,
+      "learning_rate": 1.8375404530744337e-05,
+      "loss": 0.3024,
+      "step": 5745
+    },
+    {
+      "epoch": 5.353982300884955,
+      "grad_norm": 0.2930347933934265,
+      "learning_rate": 1.834304207119741e-05,
+      "loss": 0.3005,
+      "step": 5750
+    },
+    {
+      "epoch": 5.358639962738705,
+      "grad_norm": 0.29347353585391656,
+      "learning_rate": 1.8310679611650487e-05,
+      "loss": 0.3017,
+      "step": 5755
+    },
+    {
+      "epoch": 5.363297624592454,
+      "grad_norm": 0.2926875315209124,
+      "learning_rate": 1.827831715210356e-05,
+      "loss": 0.3028,
+      "step": 5760
+    },
+    {
+      "epoch": 5.367955286446204,
+      "grad_norm": 0.2901551394692628,
+      "learning_rate": 1.8245954692556637e-05,
+      "loss": 0.3014,
+      "step": 5765
+    },
+    {
+      "epoch": 5.372612948299953,
+      "grad_norm": 0.28209903392114477,
+      "learning_rate": 1.821359223300971e-05,
+      "loss": 0.3046,
+      "step": 5770
+    },
+    {
+      "epoch": 5.377270610153703,
+      "grad_norm": 0.2749568426111605,
+      "learning_rate": 1.8181229773462787e-05,
+      "loss": 0.3032,
+      "step": 5775
+    },
+    {
+      "epoch": 5.381928272007452,
+      "grad_norm": 0.3002171746802142,
+      "learning_rate": 1.8148867313915856e-05,
+      "loss": 0.3027,
+      "step": 5780
+    },
+    {
+      "epoch": 5.386585933861202,
+      "grad_norm": 0.3008333108433299,
+      "learning_rate": 1.8116504854368933e-05,
+      "loss": 0.3042,
+      "step": 5785
+    },
+    {
+      "epoch": 5.391243595714951,
+      "grad_norm": 0.2897571641211582,
+      "learning_rate": 1.8084142394822006e-05,
+      "loss": 0.3103,
+      "step": 5790
+    },
+    {
+      "epoch": 5.395901257568701,
+      "grad_norm": 0.2850028644112056,
+      "learning_rate": 1.8051779935275083e-05,
+      "loss": 0.3093,
+      "step": 5795
+    },
+    {
+      "epoch": 5.4005589194224495,
+      "grad_norm": 0.29898701097962505,
+      "learning_rate": 1.8019417475728156e-05,
+      "loss": 0.3027,
+      "step": 5800
+    },
+    {
+      "epoch": 5.405216581276199,
+      "grad_norm": 0.28638137759427745,
+      "learning_rate": 1.7987055016181232e-05,
+      "loss": 0.3187,
+      "step": 5805
+    },
+    {
+      "epoch": 5.4098742431299485,
+      "grad_norm": 0.2767947983890502,
+      "learning_rate": 1.7954692556634306e-05,
+      "loss": 0.301,
+      "step": 5810
+    },
+    {
+      "epoch": 5.414531904983698,
+      "grad_norm": 0.293008143502894,
+      "learning_rate": 1.7922330097087382e-05,
+      "loss": 0.3012,
+      "step": 5815
+    },
+    {
+      "epoch": 5.4191895668374475,
+      "grad_norm": 0.29107156446017896,
+      "learning_rate": 1.7889967637540452e-05,
+      "loss": 0.3018,
+      "step": 5820
+    },
+    {
+      "epoch": 5.423847228691197,
+      "grad_norm": 0.2836867406762743,
+      "learning_rate": 1.785760517799353e-05,
+      "loss": 0.3056,
+      "step": 5825
+    },
+    {
+      "epoch": 5.4285048905449464,
+      "grad_norm": 0.3141053505194596,
+      "learning_rate": 1.7825242718446602e-05,
+      "loss": 0.3012,
+      "step": 5830
+    },
+    {
+      "epoch": 5.433162552398696,
+      "grad_norm": 0.2880531038635791,
+      "learning_rate": 1.7792880258899678e-05,
+      "loss": 0.3032,
+      "step": 5835
+    },
+    {
+      "epoch": 5.437820214252445,
+      "grad_norm": 0.2833450738832913,
+      "learning_rate": 1.776051779935275e-05,
+      "loss": 0.3148,
+      "step": 5840
+    },
+    {
+      "epoch": 5.442477876106195,
+      "grad_norm": 0.2888814119260653,
+      "learning_rate": 1.7728155339805828e-05,
+      "loss": 0.301,
+      "step": 5845
+    },
+    {
+      "epoch": 5.447135537959944,
+      "grad_norm": 0.28688991478590337,
+      "learning_rate": 1.76957928802589e-05,
+      "loss": 0.309,
+      "step": 5850
+    },
+    {
+      "epoch": 5.451793199813694,
+      "grad_norm": 0.290900394259756,
+      "learning_rate": 1.7663430420711974e-05,
+      "loss": 0.3071,
+      "step": 5855
+    },
+    {
+      "epoch": 5.4564508616674425,
+      "grad_norm": 0.2760376265018021,
+      "learning_rate": 1.7631067961165048e-05,
+      "loss": 0.3064,
+      "step": 5860
+    },
+    {
+      "epoch": 5.461108523521192,
+      "grad_norm": 0.2961192697428574,
+      "learning_rate": 1.7598705501618124e-05,
+      "loss": 0.3147,
+      "step": 5865
+    },
+    {
+      "epoch": 5.4657661853749415,
+      "grad_norm": 0.30113261224880655,
+      "learning_rate": 1.7566343042071197e-05,
+      "loss": 0.305,
+      "step": 5870
+    },
+    {
+      "epoch": 5.470423847228691,
+      "grad_norm": 0.30236627699114277,
+      "learning_rate": 1.7533980582524274e-05,
+      "loss": 0.2918,
+      "step": 5875
+    },
+    {
+      "epoch": 5.4750815090824405,
+      "grad_norm": 0.28786989286640924,
+      "learning_rate": 1.7501618122977347e-05,
+      "loss": 0.3029,
+      "step": 5880
+    },
+    {
+      "epoch": 5.47973917093619,
+      "grad_norm": 0.285207802940617,
+      "learning_rate": 1.7469255663430424e-05,
+      "loss": 0.3045,
+      "step": 5885
+    },
+    {
+      "epoch": 5.4843968327899395,
+      "grad_norm": 0.2704341023495142,
+      "learning_rate": 1.7436893203883493e-05,
+      "loss": 0.3085,
+      "step": 5890
+    },
+    {
+      "epoch": 5.489054494643689,
+      "grad_norm": 0.2717516312446038,
+      "learning_rate": 1.740453074433657e-05,
+      "loss": 0.3099,
+      "step": 5895
+    },
+    {
+      "epoch": 5.4937121564974385,
+      "grad_norm": 0.2753605154349603,
+      "learning_rate": 1.7372168284789643e-05,
+      "loss": 0.3023,
+      "step": 5900
+    },
+    {
+      "epoch": 5.498369818351188,
+      "grad_norm": 0.29454768766505524,
+      "learning_rate": 1.733980582524272e-05,
+      "loss": 0.315,
+      "step": 5905
+    },
+    {
+      "epoch": 5.5030274802049375,
+      "grad_norm": 0.2780881770733559,
+      "learning_rate": 1.7307443365695793e-05,
+      "loss": 0.2996,
+      "step": 5910
+    },
+    {
+      "epoch": 5.507685142058687,
+      "grad_norm": 0.29014626764984414,
+      "learning_rate": 1.727508090614887e-05,
+      "loss": 0.3051,
+      "step": 5915
+    },
+    {
+      "epoch": 5.512342803912436,
+      "grad_norm": 0.2837262516203258,
+      "learning_rate": 1.7242718446601943e-05,
+      "loss": 0.3008,
+      "step": 5920
+    },
+    {
+      "epoch": 5.517000465766185,
+      "grad_norm": 0.264402190574167,
+      "learning_rate": 1.7210355987055016e-05,
+      "loss": 0.3074,
+      "step": 5925
+    },
+    {
+      "epoch": 5.521658127619935,
+      "grad_norm": 0.2738935632023862,
+      "learning_rate": 1.717799352750809e-05,
+      "loss": 0.303,
+      "step": 5930
+    },
+    {
+      "epoch": 5.526315789473684,
+      "grad_norm": 0.27687565798806185,
+      "learning_rate": 1.7145631067961165e-05,
+      "loss": 0.3028,
+      "step": 5935
+    },
+    {
+      "epoch": 5.530973451327434,
+      "grad_norm": 0.28930865751975293,
+      "learning_rate": 1.711326860841424e-05,
+      "loss": 0.3041,
+      "step": 5940
+    },
+    {
+      "epoch": 5.535631113181183,
+      "grad_norm": 0.28596312771077054,
+      "learning_rate": 1.7080906148867315e-05,
+      "loss": 0.3065,
+      "step": 5945
+    },
+    {
+      "epoch": 5.5402887750349326,
+      "grad_norm": 0.2820486218168636,
+      "learning_rate": 1.704854368932039e-05,
+      "loss": 0.3158,
+      "step": 5950
+    },
+    {
+      "epoch": 5.544946436888682,
+      "grad_norm": 0.29890310082210714,
+      "learning_rate": 1.7016181229773465e-05,
+      "loss": 0.2946,
+      "step": 5955
+    },
+    {
+      "epoch": 5.5496040987424315,
+      "grad_norm": 0.2786619683844314,
+      "learning_rate": 1.6983818770226538e-05,
+      "loss": 0.3048,
+      "step": 5960
+    },
+    {
+      "epoch": 5.554261760596181,
+      "grad_norm": 0.27659802253619753,
+      "learning_rate": 1.695145631067961e-05,
+      "loss": 0.3039,
+      "step": 5965
+    },
+    {
+      "epoch": 5.5589194224499305,
+      "grad_norm": 0.2829305638289971,
+      "learning_rate": 1.6919093851132688e-05,
+      "loss": 0.3018,
+      "step": 5970
+    },
+    {
+      "epoch": 5.563577084303679,
+      "grad_norm": 0.2706599135302209,
+      "learning_rate": 1.688673139158576e-05,
+      "loss": 0.3068,
+      "step": 5975
+    },
+    {
+      "epoch": 5.568234746157429,
+      "grad_norm": 0.28211649730594046,
+      "learning_rate": 1.6854368932038838e-05,
+      "loss": 0.3063,
+      "step": 5980
+    },
+    {
+      "epoch": 5.572892408011178,
+      "grad_norm": 0.2951191578671079,
+      "learning_rate": 1.682200647249191e-05,
+      "loss": 0.3057,
+      "step": 5985
+    },
+    {
+      "epoch": 5.577550069864928,
+      "grad_norm": 0.2748692573835518,
+      "learning_rate": 1.6789644012944984e-05,
+      "loss": 0.3026,
+      "step": 5990
+    },
+    {
+      "epoch": 5.582207731718677,
+      "grad_norm": 0.2939313528747144,
+      "learning_rate": 1.675728155339806e-05,
+      "loss": 0.3021,
+      "step": 5995
+    },
+    {
+      "epoch": 5.586865393572427,
+      "grad_norm": 0.3031816982790779,
+      "learning_rate": 1.6724919093851134e-05,
+      "loss": 0.3119,
+      "step": 6000
+    },
+    {
+      "epoch": 5.591523055426176,
+      "grad_norm": 0.2752378797516827,
+      "learning_rate": 1.6692556634304207e-05,
+      "loss": 0.3096,
+      "step": 6005
+    },
+    {
+      "epoch": 5.596180717279926,
+      "grad_norm": 0.2903316590367377,
+      "learning_rate": 1.6660194174757283e-05,
+      "loss": 0.3076,
+      "step": 6010
+    },
+    {
+      "epoch": 5.600838379133675,
+      "grad_norm": 0.28716777283864764,
+      "learning_rate": 1.6627831715210357e-05,
+      "loss": 0.3094,
+      "step": 6015
+    },
+    {
+      "epoch": 5.605496040987425,
+      "grad_norm": 0.2891692730334035,
+      "learning_rate": 1.6595469255663433e-05,
+      "loss": 0.3038,
+      "step": 6020
+    },
+    {
+      "epoch": 5.610153702841174,
+      "grad_norm": 0.2982118804535516,
+      "learning_rate": 1.6563106796116506e-05,
+      "loss": 0.2994,
+      "step": 6025
+    },
+    {
+      "epoch": 5.614811364694923,
+      "grad_norm": 0.29024791500115066,
+      "learning_rate": 1.6530744336569583e-05,
+      "loss": 0.3035,
+      "step": 6030
+    },
+    {
+      "epoch": 5.619469026548672,
+      "grad_norm": 0.2780711941813935,
+      "learning_rate": 1.6498381877022653e-05,
+      "loss": 0.3004,
+      "step": 6035
+    },
+    {
+      "epoch": 5.624126688402422,
+      "grad_norm": 0.27659165989238843,
+      "learning_rate": 1.646601941747573e-05,
+      "loss": 0.3019,
+      "step": 6040
+    },
+    {
+      "epoch": 5.628784350256171,
+      "grad_norm": 0.28984546199561667,
+      "learning_rate": 1.6433656957928802e-05,
+      "loss": 0.306,
+      "step": 6045
+    },
+    {
+      "epoch": 5.633442012109921,
+      "grad_norm": 0.29164650641720846,
+      "learning_rate": 1.640129449838188e-05,
+      "loss": 0.3135,
+      "step": 6050
+    },
+    {
+      "epoch": 5.63809967396367,
+      "grad_norm": 0.26750630815804854,
+      "learning_rate": 1.6368932038834952e-05,
+      "loss": 0.3031,
+      "step": 6055
+    },
+    {
+      "epoch": 5.64275733581742,
+      "grad_norm": 0.2946599458080474,
+      "learning_rate": 1.633656957928803e-05,
+      "loss": 0.3049,
+      "step": 6060
+    },
+    {
+      "epoch": 5.647414997671169,
+      "grad_norm": 0.28010853242725814,
+      "learning_rate": 1.6304207119741102e-05,
+      "loss": 0.303,
+      "step": 6065
+    },
+    {
+      "epoch": 5.652072659524919,
+      "grad_norm": 0.2742657873539159,
+      "learning_rate": 1.6271844660194175e-05,
+      "loss": 0.3044,
+      "step": 6070
+    },
+    {
+      "epoch": 5.656730321378668,
+      "grad_norm": 0.2753030283239919,
+      "learning_rate": 1.6239482200647248e-05,
+      "loss": 0.3076,
+      "step": 6075
+    },
+    {
+      "epoch": 5.661387983232418,
+      "grad_norm": 0.2771477193319902,
+      "learning_rate": 1.6207119741100325e-05,
+      "loss": 0.3076,
+      "step": 6080
+    },
+    {
+      "epoch": 5.666045645086166,
+      "grad_norm": 0.2873816399993061,
+      "learning_rate": 1.6174757281553398e-05,
+      "loss": 0.3114,
+      "step": 6085
+    },
+    {
+      "epoch": 5.670703306939917,
+      "grad_norm": 0.2715615426840634,
+      "learning_rate": 1.6142394822006475e-05,
+      "loss": 0.2998,
+      "step": 6090
+    },
+    {
+      "epoch": 5.675360968793665,
+      "grad_norm": 0.2725590859049447,
+      "learning_rate": 1.6110032362459548e-05,
+      "loss": 0.3003,
+      "step": 6095
+    },
+    {
+      "epoch": 5.680018630647415,
+      "grad_norm": 0.27208932930615526,
+      "learning_rate": 1.6077669902912624e-05,
+      "loss": 0.3055,
+      "step": 6100
+    },
+    {
+      "epoch": 5.684676292501164,
+      "grad_norm": 0.28401672646025355,
+      "learning_rate": 1.6045307443365694e-05,
+      "loss": 0.3101,
+      "step": 6105
+    },
+    {
+      "epoch": 5.689333954354914,
+      "grad_norm": 0.28584083494492163,
+      "learning_rate": 1.601294498381877e-05,
+      "loss": 0.3061,
+      "step": 6110
+    },
+    {
+      "epoch": 5.693991616208663,
+      "grad_norm": 0.2866753336676237,
+      "learning_rate": 1.5980582524271844e-05,
+      "loss": 0.3025,
+      "step": 6115
+    },
+    {
+      "epoch": 5.698649278062413,
+      "grad_norm": 0.2771789772174729,
+      "learning_rate": 1.594822006472492e-05,
+      "loss": 0.3093,
+      "step": 6120
+    },
+    {
+      "epoch": 5.703306939916162,
+      "grad_norm": 0.3001232150493061,
+      "learning_rate": 1.5915857605177994e-05,
+      "loss": 0.3084,
+      "step": 6125
+    },
+    {
+      "epoch": 5.707964601769912,
+      "grad_norm": 0.26727045280123496,
+      "learning_rate": 1.588349514563107e-05,
+      "loss": 0.3015,
+      "step": 6130
+    },
+    {
+      "epoch": 5.712622263623661,
+      "grad_norm": 0.2913315766767915,
+      "learning_rate": 1.5851132686084143e-05,
+      "loss": 0.3073,
+      "step": 6135
+    },
+    {
+      "epoch": 5.717279925477411,
+      "grad_norm": 0.2850823152288255,
+      "learning_rate": 1.581877022653722e-05,
+      "loss": 0.3084,
+      "step": 6140
+    },
+    {
+      "epoch": 5.72193758733116,
+      "grad_norm": 0.2864258330057128,
+      "learning_rate": 1.578640776699029e-05,
+      "loss": 0.3017,
+      "step": 6145
+    },
+    {
+      "epoch": 5.726595249184909,
+      "grad_norm": 0.28175280652917506,
+      "learning_rate": 1.5754045307443366e-05,
+      "loss": 0.3084,
+      "step": 6150
+    },
+    {
+      "epoch": 5.731252911038658,
+      "grad_norm": 0.2807399096732578,
+      "learning_rate": 1.572168284789644e-05,
+      "loss": 0.3133,
+      "step": 6155
+    },
+    {
+      "epoch": 5.735910572892408,
+      "grad_norm": 0.2724600629396077,
+      "learning_rate": 1.5689320388349516e-05,
+      "loss": 0.3107,
+      "step": 6160
+    },
+    {
+      "epoch": 5.740568234746157,
+      "grad_norm": 0.2700028287114075,
+      "learning_rate": 1.565695792880259e-05,
+      "loss": 0.312,
+      "step": 6165
+    },
+    {
+      "epoch": 5.745225896599907,
+      "grad_norm": 0.27130060678321727,
+      "learning_rate": 1.5624595469255666e-05,
+      "loss": 0.3012,
+      "step": 6170
+    },
+    {
+      "epoch": 5.749883558453656,
+      "grad_norm": 0.27951703064101047,
+      "learning_rate": 1.559223300970874e-05,
+      "loss": 0.3082,
+      "step": 6175
+    },
+    {
+      "epoch": 5.754541220307406,
+      "grad_norm": 0.27592930561792794,
+      "learning_rate": 1.5559870550161812e-05,
+      "loss": 0.3072,
+      "step": 6180
+    },
+    {
+      "epoch": 5.759198882161155,
+      "grad_norm": 0.3465026871622147,
+      "learning_rate": 1.5527508090614885e-05,
+      "loss": 0.3059,
+      "step": 6185
+    },
+    {
+      "epoch": 5.763856544014905,
+      "grad_norm": 0.2746005126989584,
+      "learning_rate": 1.5495145631067962e-05,
+      "loss": 0.3016,
+      "step": 6190
+    },
+    {
+      "epoch": 5.768514205868654,
+      "grad_norm": 0.2772282824990714,
+      "learning_rate": 1.5462783171521035e-05,
+      "loss": 0.312,
+      "step": 6195
+    },
+    {
+      "epoch": 5.773171867722404,
+      "grad_norm": 0.2743844708127403,
+      "learning_rate": 1.543042071197411e-05,
+      "loss": 0.309,
+      "step": 6200
+    },
+    {
+      "epoch": 5.777829529576152,
+      "grad_norm": 0.2877371452033368,
+      "learning_rate": 1.5398058252427185e-05,
+      "loss": 0.3133,
+      "step": 6205
+    },
+    {
+      "epoch": 5.782487191429902,
+      "grad_norm": 0.2622623411875907,
+      "learning_rate": 1.536569579288026e-05,
+      "loss": 0.3028,
+      "step": 6210
+    },
+    {
+      "epoch": 5.787144853283651,
+      "grad_norm": 0.2710226518065163,
+      "learning_rate": 1.5333333333333334e-05,
+      "loss": 0.3031,
+      "step": 6215
+    },
+    {
+      "epoch": 5.791802515137401,
+      "grad_norm": 0.27172885126665913,
+      "learning_rate": 1.5300970873786408e-05,
+      "loss": 0.3032,
+      "step": 6220
+    },
+    {
+      "epoch": 5.79646017699115,
+      "grad_norm": 0.26208199703980944,
+      "learning_rate": 1.526860841423948e-05,
+      "loss": 0.3084,
+      "step": 6225
+    },
+    {
+      "epoch": 5.8011178388449,
+      "grad_norm": 0.2721236700379943,
+      "learning_rate": 1.5236245954692557e-05,
+      "loss": 0.3149,
+      "step": 6230
+    },
+    {
+      "epoch": 5.805775500698649,
+      "grad_norm": 0.27870200535732587,
+      "learning_rate": 1.5203883495145632e-05,
+      "loss": 0.3111,
+      "step": 6235
+    },
+    {
+      "epoch": 5.810433162552399,
+      "grad_norm": 0.28270581614522816,
+      "learning_rate": 1.5171521035598707e-05,
+      "loss": 0.3077,
+      "step": 6240
+    },
+    {
+      "epoch": 5.815090824406148,
+      "grad_norm": 0.2735990234754489,
+      "learning_rate": 1.5139158576051782e-05,
+      "loss": 0.3062,
+      "step": 6245
+    },
+    {
+      "epoch": 5.819748486259898,
+      "grad_norm": 0.27679598567312363,
+      "learning_rate": 1.5106796116504853e-05,
+      "loss": 0.3079,
+      "step": 6250
+    },
+    {
+      "epoch": 5.824406148113647,
+      "grad_norm": 0.28298106354491503,
+      "learning_rate": 1.5074433656957928e-05,
+      "loss": 0.3153,
+      "step": 6255
+    },
+    {
+      "epoch": 5.829063809967396,
+      "grad_norm": 0.2846618481258775,
+      "learning_rate": 1.5042071197411003e-05,
+      "loss": 0.302,
+      "step": 6260
+    },
+    {
+      "epoch": 5.833721471821145,
+      "grad_norm": 0.2851511747212997,
+      "learning_rate": 1.5009708737864078e-05,
+      "loss": 0.3073,
+      "step": 6265
+    },
+    {
+      "epoch": 5.838379133674895,
+      "grad_norm": 0.28636420923696226,
+      "learning_rate": 1.4977346278317153e-05,
+      "loss": 0.3143,
+      "step": 6270
+    },
+    {
+      "epoch": 5.843036795528644,
+      "grad_norm": 0.2746846190129639,
+      "learning_rate": 1.4944983818770228e-05,
+      "loss": 0.3072,
+      "step": 6275
+    },
+    {
+      "epoch": 5.847694457382394,
+      "grad_norm": 0.28020487777908365,
+      "learning_rate": 1.4912621359223303e-05,
+      "loss": 0.3073,
+      "step": 6280
+    },
+    {
+      "epoch": 5.852352119236143,
+      "grad_norm": 0.2792427758363,
+      "learning_rate": 1.4880258899676374e-05,
+      "loss": 0.3087,
+      "step": 6285
+    },
+    {
+      "epoch": 5.857009781089893,
+      "grad_norm": 0.2782388073052526,
+      "learning_rate": 1.4847896440129449e-05,
+      "loss": 0.3032,
+      "step": 6290
+    },
+    {
+      "epoch": 5.861667442943642,
+      "grad_norm": 0.2579244666091734,
+      "learning_rate": 1.4815533980582524e-05,
+      "loss": 0.3012,
+      "step": 6295
+    },
+    {
+      "epoch": 5.866325104797392,
+      "grad_norm": 0.2656839646498183,
+      "learning_rate": 1.4783171521035599e-05,
+      "loss": 0.3042,
+      "step": 6300
+    },
+    {
+      "epoch": 5.870982766651141,
+      "grad_norm": 0.2808985029331758,
+      "learning_rate": 1.4750809061488674e-05,
+      "loss": 0.2979,
+      "step": 6305
+    },
+    {
+      "epoch": 5.875640428504891,
+      "grad_norm": 0.2806107094803382,
+      "learning_rate": 1.4718446601941749e-05,
+      "loss": 0.3125,
+      "step": 6310
+    },
+    {
+      "epoch": 5.8802980903586395,
+      "grad_norm": 0.2795338031099193,
+      "learning_rate": 1.4686084142394823e-05,
+      "loss": 0.2961,
+      "step": 6315
+    },
+    {
+      "epoch": 5.88495575221239,
+      "grad_norm": 0.2760213410981478,
+      "learning_rate": 1.4653721682847898e-05,
+      "loss": 0.3046,
+      "step": 6320
+    },
+    {
+      "epoch": 5.8896134140661385,
+      "grad_norm": 0.2932730180826483,
+      "learning_rate": 1.462135922330097e-05,
+      "loss": 0.3049,
+      "step": 6325
+    },
+    {
+      "epoch": 5.894271075919888,
+      "grad_norm": 0.28127802161670795,
+      "learning_rate": 1.4588996763754045e-05,
+      "loss": 0.3146,
+      "step": 6330
+    },
+    {
+      "epoch": 5.8989287377736375,
+      "grad_norm": 0.2792155247730604,
+      "learning_rate": 1.455663430420712e-05,
+      "loss": 0.3021,
+      "step": 6335
+    },
+    {
+      "epoch": 5.903586399627387,
+      "grad_norm": 0.27856019558478023,
+      "learning_rate": 1.4524271844660194e-05,
+      "loss": 0.3015,
+      "step": 6340
+    },
+    {
+      "epoch": 5.9082440614811365,
+      "grad_norm": 0.28321062046344186,
+      "learning_rate": 1.449190938511327e-05,
+      "loss": 0.2967,
+      "step": 6345
+    },
+    {
+      "epoch": 5.912901723334886,
+      "grad_norm": 0.27330655251511077,
+      "learning_rate": 1.4459546925566344e-05,
+      "loss": 0.3061,
+      "step": 6350
+    },
+    {
+      "epoch": 5.9175593851886354,
+      "grad_norm": 0.27751395380672944,
+      "learning_rate": 1.4427184466019419e-05,
+      "loss": 0.306,
+      "step": 6355
+    },
+    {
+      "epoch": 5.922217047042385,
+      "grad_norm": 0.27478548080495335,
+      "learning_rate": 1.4394822006472492e-05,
+      "loss": 0.2994,
+      "step": 6360
+    },
+    {
+      "epoch": 5.926874708896134,
+      "grad_norm": 0.2823478583161036,
+      "learning_rate": 1.4362459546925567e-05,
+      "loss": 0.3118,
+      "step": 6365
+    },
+    {
+      "epoch": 5.931532370749884,
+      "grad_norm": 0.28287174374847096,
+      "learning_rate": 1.4330097087378642e-05,
+      "loss": 0.299,
+      "step": 6370
+    },
+    {
+      "epoch": 5.936190032603633,
+      "grad_norm": 0.28444922456277477,
+      "learning_rate": 1.4297734627831717e-05,
+      "loss": 0.3157,
+      "step": 6375
+    },
+    {
+      "epoch": 5.940847694457382,
+      "grad_norm": 0.28803878164195124,
+      "learning_rate": 1.426537216828479e-05,
+      "loss": 0.3044,
+      "step": 6380
+    },
+    {
+      "epoch": 5.9455053563111315,
+      "grad_norm": 0.2891142857695731,
+      "learning_rate": 1.4233009708737865e-05,
+      "loss": 0.3087,
+      "step": 6385
+    },
+    {
+      "epoch": 5.950163018164881,
+      "grad_norm": 0.2867316511263794,
+      "learning_rate": 1.420064724919094e-05,
+      "loss": 0.3166,
+      "step": 6390
+    },
+    {
+      "epoch": 5.9548206800186305,
+      "grad_norm": 0.28433359030956057,
+      "learning_rate": 1.4168284789644013e-05,
+      "loss": 0.3113,
+      "step": 6395
+    },
+    {
+      "epoch": 5.95947834187238,
+      "grad_norm": 0.27053434327134906,
+      "learning_rate": 1.4135922330097088e-05,
+      "loss": 0.3088,
+      "step": 6400
+    },
+    {
+      "epoch": 5.9641360037261295,
+      "grad_norm": 0.2645493590042334,
+      "learning_rate": 1.4103559870550163e-05,
+      "loss": 0.3027,
+      "step": 6405
+    },
+    {
+      "epoch": 5.968793665579879,
+      "grad_norm": 0.27412814807463415,
+      "learning_rate": 1.4071197411003237e-05,
+      "loss": 0.3054,
+      "step": 6410
+    },
+    {
+      "epoch": 5.9734513274336285,
+      "grad_norm": 0.2714124141236671,
+      "learning_rate": 1.4038834951456312e-05,
+      "loss": 0.296,
+      "step": 6415
+    },
+    {
+      "epoch": 5.978108989287378,
+      "grad_norm": 0.27945140158886544,
+      "learning_rate": 1.4006472491909387e-05,
+      "loss": 0.299,
+      "step": 6420
+    },
+    {
+      "epoch": 5.9827666511411275,
+      "grad_norm": 0.2713442676698441,
+      "learning_rate": 1.3974110032362462e-05,
+      "loss": 0.3057,
+      "step": 6425
+    },
+    {
+      "epoch": 5.987424312994877,
+      "grad_norm": 0.2827835607802788,
+      "learning_rate": 1.3941747572815534e-05,
+      "loss": 0.3135,
+      "step": 6430
+    },
+    {
+      "epoch": 5.992081974848626,
+      "grad_norm": 0.2716682297183809,
+      "learning_rate": 1.3909385113268608e-05,
+      "loss": 0.3083,
+      "step": 6435
+    },
+    {
+      "epoch": 5.996739636702375,
+      "grad_norm": 0.27225571037293367,
+      "learning_rate": 1.3877022653721683e-05,
+      "loss": 0.3047,
+      "step": 6440
+    },
+    {
+      "epoch": 6.00093153237075,
+      "grad_norm": 0.35495868274581827,
+      "learning_rate": 1.3844660194174758e-05,
+      "loss": 0.2923,
+      "step": 6445
+    },
+    {
+      "epoch": 6.005589194224499,
+      "grad_norm": 0.3297878781680354,
+      "learning_rate": 1.3812297734627833e-05,
+      "loss": 0.2578,
+      "step": 6450
+    },
+    {
+      "epoch": 6.010246856078249,
+      "grad_norm": 0.3049393990291519,
+      "learning_rate": 1.3779935275080908e-05,
+      "loss": 0.2647,
+      "step": 6455
+    },
+    {
+      "epoch": 6.014904517931998,
+      "grad_norm": 0.3045770601314326,
+      "learning_rate": 1.3747572815533983e-05,
+      "loss": 0.2593,
+      "step": 6460
+    },
+    {
+      "epoch": 6.019562179785748,
+      "grad_norm": 0.29522401789966235,
+      "learning_rate": 1.3715210355987058e-05,
+      "loss": 0.2614,
+      "step": 6465
+    },
+    {
+      "epoch": 6.024219841639497,
+      "grad_norm": 0.2992223453383009,
+      "learning_rate": 1.3682847896440129e-05,
+      "loss": 0.2572,
+      "step": 6470
+    },
+    {
+      "epoch": 6.028877503493247,
+      "grad_norm": 0.3104405294123856,
+      "learning_rate": 1.3650485436893204e-05,
+      "loss": 0.2639,
+      "step": 6475
+    },
+    {
+      "epoch": 6.033535165346996,
+      "grad_norm": 0.29912511415470894,
+      "learning_rate": 1.3618122977346279e-05,
+      "loss": 0.2592,
+      "step": 6480
+    },
+    {
+      "epoch": 6.038192827200745,
+      "grad_norm": 0.3021541526670289,
+      "learning_rate": 1.3585760517799354e-05,
+      "loss": 0.2628,
+      "step": 6485
+    },
+    {
+      "epoch": 6.042850489054494,
+      "grad_norm": 0.3133637204965224,
+      "learning_rate": 1.3553398058252429e-05,
+      "loss": 0.2609,
+      "step": 6490
+    },
+    {
+      "epoch": 6.047508150908244,
+      "grad_norm": 0.31960446564212824,
+      "learning_rate": 1.3521035598705503e-05,
+      "loss": 0.2629,
+      "step": 6495
+    },
+    {
+      "epoch": 6.052165812761993,
+      "grad_norm": 0.30991779756370824,
+      "learning_rate": 1.3488673139158578e-05,
+      "loss": 0.2641,
+      "step": 6500
+    },
+    {
+      "epoch": 6.056823474615743,
+      "grad_norm": 0.3264487591304554,
+      "learning_rate": 1.345631067961165e-05,
+      "loss": 0.2642,
+      "step": 6505
+    },
+    {
+      "epoch": 6.061481136469492,
+      "grad_norm": 0.28858785783185686,
+      "learning_rate": 1.3423948220064725e-05,
+      "loss": 0.2617,
+      "step": 6510
+    },
+    {
+      "epoch": 6.066138798323242,
+      "grad_norm": 0.3105777896361813,
+      "learning_rate": 1.33915857605178e-05,
+      "loss": 0.2622,
+      "step": 6515
+    },
+    {
+      "epoch": 6.070796460176991,
+      "grad_norm": 0.29554505620954064,
+      "learning_rate": 1.3359223300970874e-05,
+      "loss": 0.2511,
+      "step": 6520
+    },
+    {
+      "epoch": 6.075454122030741,
+      "grad_norm": 0.2953799747572498,
+      "learning_rate": 1.332686084142395e-05,
+      "loss": 0.2624,
+      "step": 6525
+    },
+    {
+      "epoch": 6.08011178388449,
+      "grad_norm": 0.29749608780623976,
+      "learning_rate": 1.3294498381877024e-05,
+      "loss": 0.2654,
+      "step": 6530
+    },
+    {
+      "epoch": 6.08476944573824,
+      "grad_norm": 0.3045485523189274,
+      "learning_rate": 1.3262135922330099e-05,
+      "loss": 0.2635,
+      "step": 6535
+    },
+    {
+      "epoch": 6.089427107591989,
+      "grad_norm": 0.30079742188595193,
+      "learning_rate": 1.322977346278317e-05,
+      "loss": 0.2627,
+      "step": 6540
+    },
+    {
+      "epoch": 6.094084769445738,
+      "grad_norm": 0.2997340622747308,
+      "learning_rate": 1.3197411003236245e-05,
+      "loss": 0.2595,
+      "step": 6545
+    },
+    {
+      "epoch": 6.098742431299487,
+      "grad_norm": 0.31052606162390783,
+      "learning_rate": 1.316504854368932e-05,
+      "loss": 0.2644,
+      "step": 6550
+    },
+    {
+      "epoch": 6.103400093153237,
+      "grad_norm": 0.29891365265398806,
+      "learning_rate": 1.3132686084142395e-05,
+      "loss": 0.258,
+      "step": 6555
+    },
+    {
+      "epoch": 6.108057755006986,
+      "grad_norm": 0.2977288901620322,
+      "learning_rate": 1.310032362459547e-05,
+      "loss": 0.2687,
+      "step": 6560
+    },
+    {
+      "epoch": 6.112715416860736,
+      "grad_norm": 0.3290928356537451,
+      "learning_rate": 1.3067961165048545e-05,
+      "loss": 0.2646,
+      "step": 6565
+    },
+    {
+      "epoch": 6.117373078714485,
+      "grad_norm": 0.2888676419652737,
+      "learning_rate": 1.303559870550162e-05,
+      "loss": 0.26,
+      "step": 6570
+    },
+    {
+      "epoch": 6.122030740568235,
+      "grad_norm": 0.28503442473557816,
+      "learning_rate": 1.3003236245954691e-05,
+      "loss": 0.2599,
+      "step": 6575
+    },
+    {
+      "epoch": 6.126688402421984,
+      "grad_norm": 0.2905977574492153,
+      "learning_rate": 1.2970873786407766e-05,
+      "loss": 0.26,
+      "step": 6580
+    },
+    {
+      "epoch": 6.131346064275734,
+      "grad_norm": 0.3062366689782617,
+      "learning_rate": 1.2938511326860841e-05,
+      "loss": 0.2633,
+      "step": 6585
+    },
+    {
+      "epoch": 6.136003726129483,
+      "grad_norm": 0.29240949963563057,
+      "learning_rate": 1.2906148867313916e-05,
+      "loss": 0.2575,
+      "step": 6590
+    },
+    {
+      "epoch": 6.140661387983233,
+      "grad_norm": 0.3023468143758687,
+      "learning_rate": 1.287378640776699e-05,
+      "loss": 0.2635,
+      "step": 6595
+    },
+    {
+      "epoch": 6.145319049836981,
+      "grad_norm": 0.30685404285585544,
+      "learning_rate": 1.2841423948220066e-05,
+      "loss": 0.2671,
+      "step": 6600
+    },
+    {
+      "epoch": 6.149976711690731,
+      "grad_norm": 0.3013693735118242,
+      "learning_rate": 1.280906148867314e-05,
+      "loss": 0.268,
+      "step": 6605
+    },
+    {
+      "epoch": 6.15463437354448,
+      "grad_norm": 0.2924855300662006,
+      "learning_rate": 1.2776699029126214e-05,
+      "loss": 0.2627,
+      "step": 6610
+    },
+    {
+      "epoch": 6.15929203539823,
+      "grad_norm": 0.3079077929755618,
+      "learning_rate": 1.2744336569579288e-05,
+      "loss": 0.2642,
+      "step": 6615
+    },
+    {
+      "epoch": 6.163949697251979,
+      "grad_norm": 0.2996259816826425,
+      "learning_rate": 1.2711974110032363e-05,
+      "loss": 0.2619,
+      "step": 6620
+    },
+    {
+      "epoch": 6.168607359105729,
+      "grad_norm": 0.32401337024916305,
+      "learning_rate": 1.2679611650485437e-05,
+      "loss": 0.2671,
+      "step": 6625
+    },
+    {
+      "epoch": 6.173265020959478,
+      "grad_norm": 0.29220015779851977,
+      "learning_rate": 1.2647249190938511e-05,
+      "loss": 0.2608,
+      "step": 6630
+    },
+    {
+      "epoch": 6.177922682813228,
+      "grad_norm": 0.3098691399332797,
+      "learning_rate": 1.2614886731391586e-05,
+      "loss": 0.2702,
+      "step": 6635
+    },
+    {
+      "epoch": 6.182580344666977,
+      "grad_norm": 0.30089891579442063,
+      "learning_rate": 1.2582524271844661e-05,
+      "loss": 0.2631,
+      "step": 6640
+    },
+    {
+      "epoch": 6.187238006520727,
+      "grad_norm": 0.31490153801082493,
+      "learning_rate": 1.2550161812297736e-05,
+      "loss": 0.2666,
+      "step": 6645
+    },
+    {
+      "epoch": 6.191895668374476,
+      "grad_norm": 0.30465227236311726,
+      "learning_rate": 1.251779935275081e-05,
+      "loss": 0.2607,
+      "step": 6650
+    },
+    {
+      "epoch": 6.196553330228226,
+      "grad_norm": 0.293389503277096,
+      "learning_rate": 1.2485436893203884e-05,
+      "loss": 0.2574,
+      "step": 6655
+    },
+    {
+      "epoch": 6.2012109920819745,
+      "grad_norm": 0.29459904393256614,
+      "learning_rate": 1.2453074433656959e-05,
+      "loss": 0.2602,
+      "step": 6660
+    },
+    {
+      "epoch": 6.205868653935724,
+      "grad_norm": 0.2952166883304406,
+      "learning_rate": 1.2420711974110034e-05,
+      "loss": 0.2605,
+      "step": 6665
+    },
+    {
+      "epoch": 6.2105263157894735,
+      "grad_norm": 0.3011999710805045,
+      "learning_rate": 1.2388349514563109e-05,
+      "loss": 0.2601,
+      "step": 6670
+    },
+    {
+      "epoch": 6.215183977643223,
+      "grad_norm": 0.3041059442688845,
+      "learning_rate": 1.2355987055016182e-05,
+      "loss": 0.2621,
+      "step": 6675
+    },
+    {
+      "epoch": 6.219841639496972,
+      "grad_norm": 0.2994745867739228,
+      "learning_rate": 1.2323624595469257e-05,
+      "loss": 0.262,
+      "step": 6680
+    },
+    {
+      "epoch": 6.224499301350722,
+      "grad_norm": 0.3029753006698125,
+      "learning_rate": 1.2291262135922332e-05,
+      "loss": 0.2609,
+      "step": 6685
+    },
+    {
+      "epoch": 6.229156963204471,
+      "grad_norm": 0.2973074409724744,
+      "learning_rate": 1.2258899676375406e-05,
+      "loss": 0.2575,
+      "step": 6690
+    },
+    {
+      "epoch": 6.233814625058221,
+      "grad_norm": 0.2919733617641279,
+      "learning_rate": 1.222653721682848e-05,
+      "loss": 0.2627,
+      "step": 6695
+    },
+    {
+      "epoch": 6.23847228691197,
+      "grad_norm": 0.29864477756602675,
+      "learning_rate": 1.2194174757281554e-05,
+      "loss": 0.2638,
+      "step": 6700
+    },
+    {
+      "epoch": 6.24312994876572,
+      "grad_norm": 0.2986274093655223,
+      "learning_rate": 1.216181229773463e-05,
+      "loss": 0.2618,
+      "step": 6705
+    },
+    {
+      "epoch": 6.247787610619469,
+      "grad_norm": 0.28958751927297255,
+      "learning_rate": 1.2129449838187703e-05,
+      "loss": 0.2621,
+      "step": 6710
+    },
+    {
+      "epoch": 6.252445272473219,
+      "grad_norm": 0.3055074090931315,
+      "learning_rate": 1.2097087378640777e-05,
+      "loss": 0.2611,
+      "step": 6715
+    },
+    {
+      "epoch": 6.2571029343269675,
+      "grad_norm": 0.29902667150406925,
+      "learning_rate": 1.2064724919093852e-05,
+      "loss": 0.2654,
+      "step": 6720
+    },
+    {
+      "epoch": 6.261760596180717,
+      "grad_norm": 0.29407942984693936,
+      "learning_rate": 1.2032362459546927e-05,
+      "loss": 0.2632,
+      "step": 6725
+    },
+    {
+      "epoch": 6.2664182580344665,
+      "grad_norm": 0.30488411423530387,
+      "learning_rate": 1.2e-05,
+      "loss": 0.2643,
+      "step": 6730
+    },
+    {
+      "epoch": 6.271075919888216,
+      "grad_norm": 0.30663631113750467,
+      "learning_rate": 1.1967637540453075e-05,
+      "loss": 0.2715,
+      "step": 6735
+    },
+    {
+      "epoch": 6.2757335817419655,
+      "grad_norm": 0.2937405769189997,
+      "learning_rate": 1.193527508090615e-05,
+      "loss": 0.2639,
+      "step": 6740
+    },
+    {
+      "epoch": 6.280391243595715,
+      "grad_norm": 0.30589147682629475,
+      "learning_rate": 1.1902912621359223e-05,
+      "loss": 0.2645,
+      "step": 6745
+    },
+    {
+      "epoch": 6.2850489054494645,
+      "grad_norm": 0.297237299068367,
+      "learning_rate": 1.1870550161812298e-05,
+      "loss": 0.2636,
+      "step": 6750
+    },
+    {
+      "epoch": 6.289706567303214,
+      "grad_norm": 0.3004296527812095,
+      "learning_rate": 1.1838187702265373e-05,
+      "loss": 0.2634,
+      "step": 6755
+    },
+    {
+      "epoch": 6.2943642291569635,
+      "grad_norm": 0.3000745498565585,
+      "learning_rate": 1.1805825242718448e-05,
+      "loss": 0.2639,
+      "step": 6760
+    },
+    {
+      "epoch": 6.299021891010713,
+      "grad_norm": 0.2936673332032889,
+      "learning_rate": 1.1773462783171521e-05,
+      "loss": 0.2628,
+      "step": 6765
+    },
+    {
+      "epoch": 6.3036795528644625,
+      "grad_norm": 0.2956144303291978,
+      "learning_rate": 1.1741100323624596e-05,
+      "loss": 0.2623,
+      "step": 6770
+    },
+    {
+      "epoch": 6.308337214718211,
+      "grad_norm": 0.3037688284992403,
+      "learning_rate": 1.170873786407767e-05,
+      "loss": 0.2724,
+      "step": 6775
+    },
+    {
+      "epoch": 6.312994876571961,
+      "grad_norm": 0.2967836672965417,
+      "learning_rate": 1.1676375404530746e-05,
+      "loss": 0.267,
+      "step": 6780
+    },
+    {
+      "epoch": 6.31765253842571,
+      "grad_norm": 0.2938130888285839,
+      "learning_rate": 1.1644012944983819e-05,
+      "loss": 0.2594,
+      "step": 6785
+    },
+    {
+      "epoch": 6.32231020027946,
+      "grad_norm": 0.31339946077072456,
+      "learning_rate": 1.1611650485436894e-05,
+      "loss": 0.269,
+      "step": 6790
+    },
+    {
+      "epoch": 6.326967862133209,
+      "grad_norm": 0.31099128875742715,
+      "learning_rate": 1.1579288025889969e-05,
+      "loss": 0.2626,
+      "step": 6795
+    },
+    {
+      "epoch": 6.3316255239869585,
+      "grad_norm": 0.30323748956184154,
+      "learning_rate": 1.1546925566343042e-05,
+      "loss": 0.2633,
+      "step": 6800
+    },
+    {
+      "epoch": 6.336283185840708,
+      "grad_norm": 0.3081294010202703,
+      "learning_rate": 1.1514563106796117e-05,
+      "loss": 0.2689,
+      "step": 6805
+    },
+    {
+      "epoch": 6.3409408476944575,
+      "grad_norm": 0.30399877874074915,
+      "learning_rate": 1.1482200647249191e-05,
+      "loss": 0.2662,
+      "step": 6810
+    },
+    {
+      "epoch": 6.345598509548207,
+      "grad_norm": 0.29208443874938306,
+      "learning_rate": 1.1449838187702266e-05,
+      "loss": 0.2667,
+      "step": 6815
+    },
+    {
+      "epoch": 6.3502561714019565,
+      "grad_norm": 0.2977432429689944,
+      "learning_rate": 1.141747572815534e-05,
+      "loss": 0.2615,
+      "step": 6820
+    },
+    {
+      "epoch": 6.354913833255706,
+      "grad_norm": 0.3091236926197169,
+      "learning_rate": 1.1385113268608414e-05,
+      "loss": 0.2664,
+      "step": 6825
+    },
+    {
+      "epoch": 6.359571495109455,
+      "grad_norm": 0.29377145812895084,
+      "learning_rate": 1.135275080906149e-05,
+      "loss": 0.2722,
+      "step": 6830
+    },
+    {
+      "epoch": 6.364229156963204,
+      "grad_norm": 0.3060481901896544,
+      "learning_rate": 1.1320388349514564e-05,
+      "loss": 0.2676,
+      "step": 6835
+    },
+    {
+      "epoch": 6.368886818816954,
+      "grad_norm": 0.3044897451068506,
+      "learning_rate": 1.1288025889967637e-05,
+      "loss": 0.2655,
+      "step": 6840
+    },
+    {
+      "epoch": 6.373544480670703,
+      "grad_norm": 0.31723769869483387,
+      "learning_rate": 1.1255663430420712e-05,
+      "loss": 0.2739,
+      "step": 6845
+    },
+    {
+      "epoch": 6.378202142524453,
+      "grad_norm": 0.3003761183208017,
+      "learning_rate": 1.1223300970873787e-05,
+      "loss": 0.2676,
+      "step": 6850
+    },
+    {
+      "epoch": 6.382859804378202,
+      "grad_norm": 0.28548123147448823,
+      "learning_rate": 1.119093851132686e-05,
+      "loss": 0.2666,
+      "step": 6855
+    },
+    {
+      "epoch": 6.387517466231952,
+      "grad_norm": 0.30757258907864266,
+      "learning_rate": 1.1158576051779935e-05,
+      "loss": 0.2715,
+      "step": 6860
+    },
+    {
+      "epoch": 6.392175128085701,
+      "grad_norm": 0.2824807846329121,
+      "learning_rate": 1.112621359223301e-05,
+      "loss": 0.2641,
+      "step": 6865
+    },
+    {
+      "epoch": 6.396832789939451,
+      "grad_norm": 0.2987736875062339,
+      "learning_rate": 1.1093851132686085e-05,
+      "loss": 0.2648,
+      "step": 6870
+    },
+    {
+      "epoch": 6.4014904517932,
+      "grad_norm": 0.29263393638706314,
+      "learning_rate": 1.1061488673139158e-05,
+      "loss": 0.2571,
+      "step": 6875
+    },
+    {
+      "epoch": 6.40614811364695,
+      "grad_norm": 0.30484628646229844,
+      "learning_rate": 1.1029126213592233e-05,
+      "loss": 0.2621,
+      "step": 6880
+    },
+    {
+      "epoch": 6.410805775500698,
+      "grad_norm": 0.29353850152552546,
+      "learning_rate": 1.0996763754045308e-05,
+      "loss": 0.2656,
+      "step": 6885
+    },
+    {
+      "epoch": 6.415463437354448,
+      "grad_norm": 0.292195194398706,
+      "learning_rate": 1.0964401294498383e-05,
+      "loss": 0.2666,
+      "step": 6890
+    },
+    {
+      "epoch": 6.420121099208197,
+      "grad_norm": 0.29631019439162676,
+      "learning_rate": 1.0932038834951456e-05,
+      "loss": 0.264,
+      "step": 6895
+    },
+    {
+      "epoch": 6.424778761061947,
+      "grad_norm": 0.3018934528570828,
+      "learning_rate": 1.089967637540453e-05,
+      "loss": 0.2659,
+      "step": 6900
+    },
+    {
+      "epoch": 6.429436422915696,
+      "grad_norm": 0.2982611313693515,
+      "learning_rate": 1.0867313915857605e-05,
+      "loss": 0.2616,
+      "step": 6905
+    },
+    {
+      "epoch": 6.434094084769446,
+      "grad_norm": 0.31311872362980864,
+      "learning_rate": 1.083495145631068e-05,
+      "loss": 0.2723,
+      "step": 6910
+    },
+    {
+      "epoch": 6.438751746623195,
+      "grad_norm": 0.29715420694631617,
+      "learning_rate": 1.0802588996763755e-05,
+      "loss": 0.2686,
+      "step": 6915
+    },
+    {
+      "epoch": 6.443409408476945,
+      "grad_norm": 0.2973471857168713,
+      "learning_rate": 1.077022653721683e-05,
+      "loss": 0.2586,
+      "step": 6920
+    },
+    {
+      "epoch": 6.448067070330694,
+      "grad_norm": 0.29641440938775837,
+      "learning_rate": 1.0737864077669903e-05,
+      "loss": 0.265,
+      "step": 6925
+    },
+    {
+      "epoch": 6.452724732184444,
+      "grad_norm": 0.3233739194102652,
+      "learning_rate": 1.0705501618122978e-05,
+      "loss": 0.265,
+      "step": 6930
+    },
+    {
+      "epoch": 6.457382394038193,
+      "grad_norm": 0.29505157833257495,
+      "learning_rate": 1.0673139158576053e-05,
+      "loss": 0.2633,
+      "step": 6935
+    },
+    {
+      "epoch": 6.462040055891943,
+      "grad_norm": 0.2896116728878053,
+      "learning_rate": 1.0640776699029128e-05,
+      "loss": 0.262,
+      "step": 6940
+    },
+    {
+      "epoch": 6.466697717745692,
+      "grad_norm": 0.3063186852222229,
+      "learning_rate": 1.0608414239482201e-05,
+      "loss": 0.2666,
+      "step": 6945
+    },
+    {
+      "epoch": 6.471355379599441,
+      "grad_norm": 0.30838975092012333,
+      "learning_rate": 1.0576051779935276e-05,
+      "loss": 0.2756,
+      "step": 6950
+    },
+    {
+      "epoch": 6.47601304145319,
+      "grad_norm": 0.30807125812888864,
+      "learning_rate": 1.054368932038835e-05,
+      "loss": 0.2752,
+      "step": 6955
+    },
+    {
+      "epoch": 6.48067070330694,
+      "grad_norm": 0.306453526616044,
+      "learning_rate": 1.0511326860841426e-05,
+      "loss": 0.2583,
+      "step": 6960
+    },
+    {
+      "epoch": 6.485328365160689,
+      "grad_norm": 0.3015846241843987,
+      "learning_rate": 1.0478964401294499e-05,
+      "loss": 0.2645,
+      "step": 6965
+    },
+    {
+      "epoch": 6.489986027014439,
+      "grad_norm": 0.32405432934544387,
+      "learning_rate": 1.0446601941747574e-05,
+      "loss": 0.2663,
+      "step": 6970
+    },
+    {
+      "epoch": 6.494643688868188,
+      "grad_norm": 0.2887204349916844,
+      "learning_rate": 1.0414239482200649e-05,
+      "loss": 0.2681,
+      "step": 6975
+    },
+    {
+      "epoch": 6.499301350721938,
+      "grad_norm": 0.294867047881976,
+      "learning_rate": 1.0381877022653722e-05,
+      "loss": 0.2669,
+      "step": 6980
+    },
+    {
+      "epoch": 6.503959012575687,
+      "grad_norm": 0.2973771704685833,
+      "learning_rate": 1.0349514563106797e-05,
+      "loss": 0.2613,
+      "step": 6985
+    },
+    {
+      "epoch": 6.508616674429437,
+      "grad_norm": 0.29580255284489865,
+      "learning_rate": 1.0317152103559872e-05,
+      "loss": 0.2701,
+      "step": 6990
+    },
+    {
+      "epoch": 6.513274336283186,
+      "grad_norm": 0.28965926055244845,
+      "learning_rate": 1.0284789644012946e-05,
+      "loss": 0.2696,
+      "step": 6995
+    },
+    {
+      "epoch": 6.517931998136936,
+      "grad_norm": 0.2953950153666152,
+      "learning_rate": 1.025242718446602e-05,
+      "loss": 0.2726,
+      "step": 7000
+    },
+    {
+      "epoch": 6.522589659990684,
+      "grad_norm": 0.28930047453778795,
+      "learning_rate": 1.0220064724919094e-05,
+      "loss": 0.2654,
+      "step": 7005
+    },
+    {
+      "epoch": 6.527247321844434,
+      "grad_norm": 0.2972447245728253,
+      "learning_rate": 1.018770226537217e-05,
+      "loss": 0.2692,
+      "step": 7010
+    },
+    {
+      "epoch": 6.531904983698183,
+      "grad_norm": 0.3084742358881919,
+      "learning_rate": 1.0155339805825244e-05,
+      "loss": 0.2678,
+      "step": 7015
+    },
+    {
+      "epoch": 6.536562645551933,
+      "grad_norm": 0.29204430034154616,
+      "learning_rate": 1.0122977346278317e-05,
+      "loss": 0.2685,
+      "step": 7020
+    },
+    {
+      "epoch": 6.541220307405682,
+      "grad_norm": 0.29192158072861035,
+      "learning_rate": 1.0090614886731392e-05,
+      "loss": 0.265,
+      "step": 7025
+    },
+    {
+      "epoch": 6.545877969259432,
+      "grad_norm": 0.30029405364577455,
+      "learning_rate": 1.0058252427184467e-05,
+      "loss": 0.2717,
+      "step": 7030
+    },
+    {
+      "epoch": 6.550535631113181,
+      "grad_norm": 0.3015878432849847,
+      "learning_rate": 1.002588996763754e-05,
+      "loss": 0.2679,
+      "step": 7035
+    },
+    {
+      "epoch": 6.555193292966931,
+      "grad_norm": 0.2896341645771092,
+      "learning_rate": 9.993527508090615e-06,
+      "loss": 0.2603,
+      "step": 7040
+    },
+    {
+      "epoch": 6.55985095482068,
+      "grad_norm": 0.3056973985571281,
+      "learning_rate": 9.96116504854369e-06,
+      "loss": 0.2724,
+      "step": 7045
+    },
+    {
+      "epoch": 6.56450861667443,
+      "grad_norm": 0.2921427997727339,
+      "learning_rate": 9.928802588996765e-06,
+      "loss": 0.2629,
+      "step": 7050
+    },
+    {
+      "epoch": 6.569166278528179,
+      "grad_norm": 0.2884651901506424,
+      "learning_rate": 9.896440129449838e-06,
+      "loss": 0.2602,
+      "step": 7055
+    },
+    {
+      "epoch": 6.573823940381928,
+      "grad_norm": 0.30672936523261934,
+      "learning_rate": 9.864077669902913e-06,
+      "loss": 0.2682,
+      "step": 7060
+    },
+    {
+      "epoch": 6.578481602235677,
+      "grad_norm": 0.2994506278425497,
+      "learning_rate": 9.831715210355988e-06,
+      "loss": 0.269,
+      "step": 7065
+    },
+    {
+      "epoch": 6.583139264089427,
+      "grad_norm": 0.312676659203426,
+      "learning_rate": 9.799352750809061e-06,
+      "loss": 0.2658,
+      "step": 7070
+    },
+    {
+      "epoch": 6.587796925943176,
+      "grad_norm": 0.2861834121660456,
+      "learning_rate": 9.766990291262136e-06,
+      "loss": 0.2591,
+      "step": 7075
+    },
+    {
+      "epoch": 6.592454587796926,
+      "grad_norm": 0.3159432948596571,
+      "learning_rate": 9.73462783171521e-06,
+      "loss": 0.2665,
+      "step": 7080
+    },
+    {
+      "epoch": 6.597112249650675,
+      "grad_norm": 0.3106042391433404,
+      "learning_rate": 9.702265372168286e-06,
+      "loss": 0.2693,
+      "step": 7085
+    },
+    {
+      "epoch": 6.601769911504425,
+      "grad_norm": 0.2874804804924082,
+      "learning_rate": 9.669902912621359e-06,
+      "loss": 0.2648,
+      "step": 7090
+    },
+    {
+      "epoch": 6.606427573358174,
+      "grad_norm": 0.30044410174964736,
+      "learning_rate": 9.637540453074434e-06,
+      "loss": 0.2627,
+      "step": 7095
+    },
+    {
+      "epoch": 6.611085235211924,
+      "grad_norm": 0.29779385187282104,
+      "learning_rate": 9.605177993527508e-06,
+      "loss": 0.2646,
+      "step": 7100
+    },
+    {
+      "epoch": 6.615742897065673,
+      "grad_norm": 0.28021042242020305,
+      "learning_rate": 9.572815533980583e-06,
+      "loss": 0.2627,
+      "step": 7105
+    },
+    {
+      "epoch": 6.620400558919423,
+      "grad_norm": 0.2865437965680647,
+      "learning_rate": 9.540453074433657e-06,
+      "loss": 0.2644,
+      "step": 7110
+    },
+    {
+      "epoch": 6.625058220773171,
+      "grad_norm": 0.28847036183764996,
+      "learning_rate": 9.508090614886731e-06,
+      "loss": 0.2704,
+      "step": 7115
+    },
+    {
+      "epoch": 6.629715882626921,
+      "grad_norm": 0.29923828687211645,
+      "learning_rate": 9.475728155339806e-06,
+      "loss": 0.2669,
+      "step": 7120
+    },
+    {
+      "epoch": 6.63437354448067,
+      "grad_norm": 0.28995732729217844,
+      "learning_rate": 9.44336569579288e-06,
+      "loss": 0.2655,
+      "step": 7125
+    },
+    {
+      "epoch": 6.63903120633442,
+      "grad_norm": 0.29606046217236426,
+      "learning_rate": 9.411003236245954e-06,
+      "loss": 0.2673,
+      "step": 7130
+    },
+    {
+      "epoch": 6.643688868188169,
+      "grad_norm": 0.29502049733762253,
+      "learning_rate": 9.37864077669903e-06,
+      "loss": 0.2694,
+      "step": 7135
+    },
+    {
+      "epoch": 6.648346530041919,
+      "grad_norm": 0.30156900738632925,
+      "learning_rate": 9.346278317152104e-06,
+      "loss": 0.2643,
+      "step": 7140
+    },
+    {
+      "epoch": 6.653004191895668,
+      "grad_norm": 0.2865140989183644,
+      "learning_rate": 9.313915857605177e-06,
+      "loss": 0.2676,
+      "step": 7145
+    },
+    {
+      "epoch": 6.657661853749418,
+      "grad_norm": 0.2864300410942112,
+      "learning_rate": 9.281553398058252e-06,
+      "loss": 0.2567,
+      "step": 7150
+    },
+    {
+      "epoch": 6.662319515603167,
+      "grad_norm": 0.2963896335540961,
+      "learning_rate": 9.249190938511327e-06,
+      "loss": 0.2629,
+      "step": 7155
+    },
+    {
+      "epoch": 6.666977177456917,
+      "grad_norm": 0.2929631491412471,
+      "learning_rate": 9.216828478964402e-06,
+      "loss": 0.263,
+      "step": 7160
+    },
+    {
+      "epoch": 6.671634839310666,
+      "grad_norm": 0.2960008906124059,
+      "learning_rate": 9.184466019417477e-06,
+      "loss": 0.2665,
+      "step": 7165
+    },
+    {
+      "epoch": 6.676292501164415,
+      "grad_norm": 0.2957610946094985,
+      "learning_rate": 9.15210355987055e-06,
+      "loss": 0.2704,
+      "step": 7170
+    },
+    {
+      "epoch": 6.680950163018165,
+      "grad_norm": 0.28705408532948146,
+      "learning_rate": 9.119741100323625e-06,
+      "loss": 0.2594,
+      "step": 7175
+    },
+    {
+      "epoch": 6.685607824871914,
+      "grad_norm": 0.29626334225006756,
+      "learning_rate": 9.0873786407767e-06,
+      "loss": 0.2639,
+      "step": 7180
+    },
+    {
+      "epoch": 6.6902654867256635,
+      "grad_norm": 0.2826872422417214,
+      "learning_rate": 9.055016181229774e-06,
+      "loss": 0.2637,
+      "step": 7185
+    },
+    {
+      "epoch": 6.694923148579413,
+      "grad_norm": 0.29968315068623025,
+      "learning_rate": 9.02265372168285e-06,
+      "loss": 0.2678,
+      "step": 7190
+    },
+    {
+      "epoch": 6.6995808104331624,
+      "grad_norm": 0.29773778109848104,
+      "learning_rate": 8.990291262135924e-06,
+      "loss": 0.2712,
+      "step": 7195
+    },
+    {
+      "epoch": 6.704238472286912,
+      "grad_norm": 0.3120568855084784,
+      "learning_rate": 8.957928802588997e-06,
+      "loss": 0.2636,
+      "step": 7200
+    },
+    {
+      "epoch": 6.708896134140661,
+      "grad_norm": 0.2953788516265102,
+      "learning_rate": 8.925566343042072e-06,
+      "loss": 0.2626,
+      "step": 7205
+    },
+    {
+      "epoch": 6.713553795994411,
+      "grad_norm": 0.2946094057534323,
+      "learning_rate": 8.893203883495147e-06,
+      "loss": 0.2653,
+      "step": 7210
+    },
+    {
+      "epoch": 6.71821145784816,
+      "grad_norm": 0.28172021875277514,
+      "learning_rate": 8.86084142394822e-06,
+      "loss": 0.2586,
+      "step": 7215
+    },
+    {
+      "epoch": 6.72286911970191,
+      "grad_norm": 0.2837044630178436,
+      "learning_rate": 8.828478964401295e-06,
+      "loss": 0.2638,
+      "step": 7220
+    },
+    {
+      "epoch": 6.727526781555659,
+      "grad_norm": 0.2931079722970222,
+      "learning_rate": 8.79611650485437e-06,
+      "loss": 0.2702,
+      "step": 7225
+    },
+    {
+      "epoch": 6.732184443409409,
+      "grad_norm": 0.29377975665519773,
+      "learning_rate": 8.763754045307445e-06,
+      "loss": 0.2682,
+      "step": 7230
+    },
+    {
+      "epoch": 6.7368421052631575,
+      "grad_norm": 0.2957932794325564,
+      "learning_rate": 8.731391585760518e-06,
+      "loss": 0.2647,
+      "step": 7235
+    },
+    {
+      "epoch": 6.741499767116907,
+      "grad_norm": 0.28703326172482113,
+      "learning_rate": 8.699029126213593e-06,
+      "loss": 0.2646,
+      "step": 7240
+    },
+    {
+      "epoch": 6.7461574289706565,
+      "grad_norm": 0.29889415938281116,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.2692,
+      "step": 7245
+    },
+    {
+      "epoch": 6.750815090824406,
+      "grad_norm": 0.2924197702146725,
+      "learning_rate": 8.634304207119743e-06,
+      "loss": 0.2717,
+      "step": 7250
+    },
+    {
+      "epoch": 6.7554727526781555,
+      "grad_norm": 0.2950840807083608,
+      "learning_rate": 8.601941747572816e-06,
+      "loss": 0.2712,
+      "step": 7255
+    },
+    {
+      "epoch": 6.760130414531905,
+      "grad_norm": 0.2792814764684901,
+      "learning_rate": 8.56957928802589e-06,
+      "loss": 0.2657,
+      "step": 7260
+    },
+    {
+      "epoch": 6.7647880763856545,
+      "grad_norm": 0.30399814729641883,
+      "learning_rate": 8.537216828478966e-06,
+      "loss": 0.2697,
+      "step": 7265
+    },
+    {
+      "epoch": 6.769445738239404,
+      "grad_norm": 0.3003646990654973,
+      "learning_rate": 8.504854368932039e-06,
+      "loss": 0.272,
+      "step": 7270
+    },
+    {
+      "epoch": 6.7741034000931535,
+      "grad_norm": 0.29113757128782447,
+      "learning_rate": 8.472491909385114e-06,
+      "loss": 0.2714,
+      "step": 7275
+    },
+    {
+      "epoch": 6.778761061946903,
+      "grad_norm": 0.28732756484781796,
+      "learning_rate": 8.440129449838189e-06,
+      "loss": 0.2584,
+      "step": 7280
+    },
+    {
+      "epoch": 6.7834187238006525,
+      "grad_norm": 0.2864611200669503,
+      "learning_rate": 8.407766990291263e-06,
+      "loss": 0.2703,
+      "step": 7285
+    },
+    {
+      "epoch": 6.788076385654401,
+      "grad_norm": 0.2955687762426276,
+      "learning_rate": 8.375404530744337e-06,
+      "loss": 0.2599,
+      "step": 7290
+    },
+    {
+      "epoch": 6.792734047508151,
+      "grad_norm": 0.2951699494413524,
+      "learning_rate": 8.343042071197411e-06,
+      "loss": 0.2665,
+      "step": 7295
+    },
+    {
+      "epoch": 6.7973917093619,
+      "grad_norm": 0.28605632622518984,
+      "learning_rate": 8.310679611650486e-06,
+      "loss": 0.2574,
+      "step": 7300
+    },
+    {
+      "epoch": 6.80204937121565,
+      "grad_norm": 0.2962348449926179,
+      "learning_rate": 8.27831715210356e-06,
+      "loss": 0.2651,
+      "step": 7305
+    },
+    {
+      "epoch": 6.806707033069399,
+      "grad_norm": 0.29452805097334306,
+      "learning_rate": 8.245954692556634e-06,
+      "loss": 0.2733,
+      "step": 7310
+    },
+    {
+      "epoch": 6.8113646949231486,
+      "grad_norm": 0.3012791508232329,
+      "learning_rate": 8.21359223300971e-06,
+      "loss": 0.2683,
+      "step": 7315
+    },
+    {
+      "epoch": 6.816022356776898,
+      "grad_norm": 0.3128355013767348,
+      "learning_rate": 8.181229773462784e-06,
+      "loss": 0.2664,
+      "step": 7320
+    },
+    {
+      "epoch": 6.8206800186306475,
+      "grad_norm": 0.3024946632235964,
+      "learning_rate": 8.148867313915857e-06,
+      "loss": 0.2706,
+      "step": 7325
+    },
+    {
+      "epoch": 6.825337680484397,
+      "grad_norm": 0.30155698607280634,
+      "learning_rate": 8.116504854368932e-06,
+      "loss": 0.2649,
+      "step": 7330
+    },
+    {
+      "epoch": 6.8299953423381465,
+      "grad_norm": 0.283215740652698,
+      "learning_rate": 8.084142394822007e-06,
+      "loss": 0.2661,
+      "step": 7335
+    },
+    {
+      "epoch": 6.834653004191896,
+      "grad_norm": 0.2967537224876354,
+      "learning_rate": 8.051779935275082e-06,
+      "loss": 0.2702,
+      "step": 7340
+    },
+    {
+      "epoch": 6.839310666045645,
+      "grad_norm": 0.28931249674766857,
+      "learning_rate": 8.019417475728155e-06,
+      "loss": 0.2665,
+      "step": 7345
+    },
+    {
+      "epoch": 6.843968327899394,
+      "grad_norm": 0.2755925519692027,
+      "learning_rate": 7.98705501618123e-06,
+      "loss": 0.266,
+      "step": 7350
+    },
+    {
+      "epoch": 6.848625989753144,
+      "grad_norm": 0.3054564057994233,
+      "learning_rate": 7.954692556634305e-06,
+      "loss": 0.2689,
+      "step": 7355
+    },
+    {
+      "epoch": 6.853283651606893,
+      "grad_norm": 0.297366586058889,
+      "learning_rate": 7.922330097087378e-06,
+      "loss": 0.2609,
+      "step": 7360
+    },
+    {
+      "epoch": 6.857941313460643,
+      "grad_norm": 0.30458237552842815,
+      "learning_rate": 7.889967637540453e-06,
+      "loss": 0.2685,
+      "step": 7365
+    },
+    {
+      "epoch": 6.862598975314392,
+      "grad_norm": 0.2933080510933296,
+      "learning_rate": 7.857605177993528e-06,
+      "loss": 0.2691,
+      "step": 7370
+    },
+    {
+      "epoch": 6.867256637168142,
+      "grad_norm": 0.2885045348475768,
+      "learning_rate": 7.825242718446603e-06,
+      "loss": 0.262,
+      "step": 7375
+    },
+    {
+      "epoch": 6.871914299021891,
+      "grad_norm": 0.31863232410466397,
+      "learning_rate": 7.792880258899676e-06,
+      "loss": 0.2669,
+      "step": 7380
+    },
+    {
+      "epoch": 6.876571960875641,
+      "grad_norm": 0.28217303875134564,
+      "learning_rate": 7.76051779935275e-06,
+      "loss": 0.2687,
+      "step": 7385
+    },
+    {
+      "epoch": 6.88122962272939,
+      "grad_norm": 0.2692090947552198,
+      "learning_rate": 7.728155339805825e-06,
+      "loss": 0.2636,
+      "step": 7390
+    },
+    {
+      "epoch": 6.88588728458314,
+      "grad_norm": 0.30114752200919587,
+      "learning_rate": 7.695792880258899e-06,
+      "loss": 0.2687,
+      "step": 7395
+    },
+    {
+      "epoch": 6.890544946436888,
+      "grad_norm": 0.28975468868704063,
+      "learning_rate": 7.663430420711974e-06,
+      "loss": 0.2649,
+      "step": 7400
+    },
+    {
+      "epoch": 6.895202608290639,
+      "grad_norm": 0.2868613444787753,
+      "learning_rate": 7.631067961165048e-06,
+      "loss": 0.266,
+      "step": 7405
+    },
+    {
+      "epoch": 6.899860270144387,
+      "grad_norm": 0.29949397350159185,
+      "learning_rate": 7.598705501618124e-06,
+      "loss": 0.2738,
+      "step": 7410
+    },
+    {
+      "epoch": 6.904517931998137,
+      "grad_norm": 0.29353444896116215,
+      "learning_rate": 7.566343042071197e-06,
+      "loss": 0.2635,
+      "step": 7415
+    },
+    {
+      "epoch": 6.909175593851886,
+      "grad_norm": 0.2936811288174701,
+      "learning_rate": 7.533980582524272e-06,
+      "loss": 0.2629,
+      "step": 7420
+    },
+    {
+      "epoch": 6.913833255705636,
+      "grad_norm": 0.28723294135836375,
+      "learning_rate": 7.501618122977347e-06,
+      "loss": 0.2639,
+      "step": 7425
+    },
+    {
+      "epoch": 6.918490917559385,
+      "grad_norm": 0.28236578729616,
+      "learning_rate": 7.469255663430422e-06,
+      "loss": 0.2653,
+      "step": 7430
+    },
+    {
+      "epoch": 6.923148579413135,
+      "grad_norm": 0.28907354168684685,
+      "learning_rate": 7.436893203883495e-06,
+      "loss": 0.2688,
+      "step": 7435
+    },
+    {
+      "epoch": 6.927806241266884,
+      "grad_norm": 0.28564285514552595,
+      "learning_rate": 7.40453074433657e-06,
+      "loss": 0.2649,
+      "step": 7440
+    },
+    {
+      "epoch": 6.932463903120634,
+      "grad_norm": 0.2837275318483068,
+      "learning_rate": 7.372168284789645e-06,
+      "loss": 0.2637,
+      "step": 7445
+    },
+    {
+      "epoch": 6.937121564974383,
+      "grad_norm": 0.2938890945273722,
+      "learning_rate": 7.339805825242718e-06,
+      "loss": 0.2639,
+      "step": 7450
+    },
+    {
+      "epoch": 6.941779226828133,
+      "grad_norm": 0.2898250518862583,
+      "learning_rate": 7.307443365695793e-06,
+      "loss": 0.2676,
+      "step": 7455
+    },
+    {
+      "epoch": 6.946436888681882,
+      "grad_norm": 0.2840903723540283,
+      "learning_rate": 7.275080906148868e-06,
+      "loss": 0.2574,
+      "step": 7460
+    },
+    {
+      "epoch": 6.951094550535631,
+      "grad_norm": 0.29011946794308335,
+      "learning_rate": 7.242718446601943e-06,
+      "loss": 0.2641,
+      "step": 7465
+    },
+    {
+      "epoch": 6.95575221238938,
+      "grad_norm": 0.27449848997937937,
+      "learning_rate": 7.210355987055016e-06,
+      "loss": 0.2629,
+      "step": 7470
+    },
+    {
+      "epoch": 6.96040987424313,
+      "grad_norm": 0.284340271468117,
+      "learning_rate": 7.177993527508091e-06,
+      "loss": 0.2605,
+      "step": 7475
+    },
+    {
+      "epoch": 6.965067536096879,
+      "grad_norm": 0.2889739086826554,
+      "learning_rate": 7.1456310679611655e-06,
+      "loss": 0.2745,
+      "step": 7480
+    },
+    {
+      "epoch": 6.969725197950629,
+      "grad_norm": 0.28879334801466866,
+      "learning_rate": 7.1132686084142395e-06,
+      "loss": 0.2678,
+      "step": 7485
+    },
+    {
+      "epoch": 6.974382859804378,
+      "grad_norm": 0.28876981261647255,
+      "learning_rate": 7.0809061488673136e-06,
+      "loss": 0.2639,
+      "step": 7490
+    },
+    {
+      "epoch": 6.979040521658128,
+      "grad_norm": 0.2756354004750388,
+      "learning_rate": 7.0485436893203884e-06,
+      "loss": 0.2715,
+      "step": 7495
+    },
+    {
+      "epoch": 6.983698183511877,
+      "grad_norm": 0.2819723345873216,
+      "learning_rate": 7.016181229773463e-06,
+      "loss": 0.272,
+      "step": 7500
+    },
+    {
+      "epoch": 6.988355845365627,
+      "grad_norm": 0.2861866310764198,
+      "learning_rate": 6.983818770226537e-06,
+      "loss": 0.2666,
+      "step": 7505
+    },
+    {
+      "epoch": 6.993013507219376,
+      "grad_norm": 0.2958946961429871,
+      "learning_rate": 6.951456310679612e-06,
+      "loss": 0.2666,
+      "step": 7510
+    },
+    {
+      "epoch": 6.997671169073126,
+      "grad_norm": 0.2889127846949159,
+      "learning_rate": 6.919093851132687e-06,
+      "loss": 0.2719,
+      "step": 7515
+    },
+    {
+      "epoch": 7.008383791336749,
+      "grad_norm": 0.32096371180847455,
+      "learning_rate": 6.886731391585761e-06,
+      "loss": 0.2247,
+      "step": 7520
+    },
+    {
+      "epoch": 7.013041453190499,
+      "grad_norm": 0.3488405021055525,
+      "learning_rate": 6.854368932038835e-06,
+      "loss": 0.2307,
+      "step": 7525
+    },
+    {
+      "epoch": 7.017699115044247,
+      "grad_norm": 0.3327638817581441,
+      "learning_rate": 6.82200647249191e-06,
+      "loss": 0.233,
+      "step": 7530
+    },
+    {
+      "epoch": 7.022356776897997,
+      "grad_norm": 0.317729462852438,
+      "learning_rate": 6.789644012944985e-06,
+      "loss": 0.2325,
+      "step": 7535
+    },
+    {
+      "epoch": 7.027014438751746,
+      "grad_norm": 0.32110341569284023,
+      "learning_rate": 6.757281553398058e-06,
+      "loss": 0.2293,
+      "step": 7540
+    },
+    {
+      "epoch": 7.031672100605496,
+      "grad_norm": 0.304755330404921,
+      "learning_rate": 6.724919093851133e-06,
+      "loss": 0.2268,
+      "step": 7545
+    },
+    {
+      "epoch": 7.036329762459245,
+      "grad_norm": 0.3050925925720897,
+      "learning_rate": 6.692556634304208e-06,
+      "loss": 0.2291,
+      "step": 7550
+    },
+    {
+      "epoch": 7.040987424312995,
+      "grad_norm": 0.3031758245942809,
+      "learning_rate": 6.660194174757283e-06,
+      "loss": 0.2329,
+      "step": 7555
+    },
+    {
+      "epoch": 7.045645086166744,
+      "grad_norm": 0.3000571027142737,
+      "learning_rate": 6.627831715210356e-06,
+      "loss": 0.2377,
+      "step": 7560
+    },
+    {
+      "epoch": 7.050302748020494,
+      "grad_norm": 0.29397010498824716,
+      "learning_rate": 6.595469255663431e-06,
+      "loss": 0.2312,
+      "step": 7565
+    },
+    {
+      "epoch": 7.054960409874243,
+      "grad_norm": 0.3145370761461229,
+      "learning_rate": 6.5631067961165056e-06,
+      "loss": 0.2341,
+      "step": 7570
+    },
+    {
+      "epoch": 7.059618071727993,
+      "grad_norm": 0.3036224922667628,
+      "learning_rate": 6.53074433656958e-06,
+      "loss": 0.2268,
+      "step": 7575
+    },
+    {
+      "epoch": 7.064275733581742,
+      "grad_norm": 0.3059794559599267,
+      "learning_rate": 6.498381877022654e-06,
+      "loss": 0.2294,
+      "step": 7580
+    },
+    {
+      "epoch": 7.068933395435492,
+      "grad_norm": 0.31289234807793126,
+      "learning_rate": 6.4660194174757285e-06,
+      "loss": 0.2264,
+      "step": 7585
+    },
+    {
+      "epoch": 7.07359105728924,
+      "grad_norm": 0.29989915293078834,
+      "learning_rate": 6.433656957928803e-06,
+      "loss": 0.2354,
+      "step": 7590
+    },
+    {
+      "epoch": 7.07824871914299,
+      "grad_norm": 0.3075677478265152,
+      "learning_rate": 6.4012944983818765e-06,
+      "loss": 0.2369,
+      "step": 7595
+    },
+    {
+      "epoch": 7.082906380996739,
+      "grad_norm": 0.3048232778200601,
+      "learning_rate": 6.368932038834951e-06,
+      "loss": 0.2364,
+      "step": 7600
+    },
+    {
+      "epoch": 7.087564042850489,
+      "grad_norm": 0.3190828574703254,
+      "learning_rate": 6.336569579288026e-06,
+      "loss": 0.2296,
+      "step": 7605
+    },
+    {
+      "epoch": 7.092221704704238,
+      "grad_norm": 0.31184952505319163,
+      "learning_rate": 6.304207119741101e-06,
+      "loss": 0.2343,
+      "step": 7610
+    },
+    {
+      "epoch": 7.096879366557988,
+      "grad_norm": 0.30613800937594887,
+      "learning_rate": 6.271844660194174e-06,
+      "loss": 0.2303,
+      "step": 7615
+    },
+    {
+      "epoch": 7.101537028411737,
+      "grad_norm": 0.30498712611830453,
+      "learning_rate": 6.239482200647249e-06,
+      "loss": 0.2259,
+      "step": 7620
+    },
+    {
+      "epoch": 7.106194690265487,
+      "grad_norm": 0.3100049730029901,
+      "learning_rate": 6.207119741100323e-06,
+      "loss": 0.2251,
+      "step": 7625
+    },
+    {
+      "epoch": 7.110852352119236,
+      "grad_norm": 0.31023444487558577,
+      "learning_rate": 6.174757281553398e-06,
+      "loss": 0.2308,
+      "step": 7630
+    },
+    {
+      "epoch": 7.115510013972986,
+      "grad_norm": 0.29584613934244036,
+      "learning_rate": 6.142394822006473e-06,
+      "loss": 0.2304,
+      "step": 7635
+    },
+    {
+      "epoch": 7.120167675826735,
+      "grad_norm": 0.29448593108961485,
+      "learning_rate": 6.110032362459547e-06,
+      "loss": 0.2371,
+      "step": 7640
+    },
+    {
+      "epoch": 7.124825337680484,
+      "grad_norm": 0.29164372951556267,
+      "learning_rate": 6.077669902912622e-06,
+      "loss": 0.2342,
+      "step": 7645
+    },
+    {
+      "epoch": 7.1294829995342335,
+      "grad_norm": 0.29109888029207276,
+      "learning_rate": 6.045307443365697e-06,
+      "loss": 0.229,
+      "step": 7650
+    },
+    {
+      "epoch": 7.134140661387983,
+      "grad_norm": 0.30246909213948103,
+      "learning_rate": 6.012944983818771e-06,
+      "loss": 0.2303,
+      "step": 7655
+    },
+    {
+      "epoch": 7.1387983232417325,
+      "grad_norm": 0.30563015303434904,
+      "learning_rate": 5.980582524271846e-06,
+      "loss": 0.233,
+      "step": 7660
+    },
+    {
+      "epoch": 7.143455985095482,
+      "grad_norm": 0.2930342631619546,
+      "learning_rate": 5.94822006472492e-06,
+      "loss": 0.2298,
+      "step": 7665
+    },
+    {
+      "epoch": 7.1481136469492315,
+      "grad_norm": 0.2969918259211002,
+      "learning_rate": 5.915857605177994e-06,
+      "loss": 0.2348,
+      "step": 7670
+    },
+    {
+      "epoch": 7.152771308802981,
+      "grad_norm": 0.2938360434343845,
+      "learning_rate": 5.8834951456310685e-06,
+      "loss": 0.2251,
+      "step": 7675
+    },
+    {
+      "epoch": 7.1574289706567304,
+      "grad_norm": 0.30467062065402506,
+      "learning_rate": 5.8511326860841425e-06,
+      "loss": 0.2333,
+      "step": 7680
+    },
+    {
+      "epoch": 7.16208663251048,
+      "grad_norm": 0.3070298308597322,
+      "learning_rate": 5.818770226537217e-06,
+      "loss": 0.2317,
+      "step": 7685
+    },
+    {
+      "epoch": 7.166744294364229,
+      "grad_norm": 0.3199230292905365,
+      "learning_rate": 5.786407766990291e-06,
+      "loss": 0.234,
+      "step": 7690
+    },
+    {
+      "epoch": 7.171401956217979,
+      "grad_norm": 0.319527302015154,
+      "learning_rate": 5.754045307443366e-06,
+      "loss": 0.2344,
+      "step": 7695
+    },
+    {
+      "epoch": 7.1760596180717275,
+      "grad_norm": 0.2955541120587258,
+      "learning_rate": 5.72168284789644e-06,
+      "loss": 0.2272,
+      "step": 7700
+    },
+    {
+      "epoch": 7.180717279925477,
+      "grad_norm": 0.3013211631850419,
+      "learning_rate": 5.689320388349515e-06,
+      "loss": 0.2318,
+      "step": 7705
+    },
+    {
+      "epoch": 7.1853749417792265,
+      "grad_norm": 0.3003735465957072,
+      "learning_rate": 5.656957928802589e-06,
+      "loss": 0.2281,
+      "step": 7710
+    },
+    {
+      "epoch": 7.190032603632976,
+      "grad_norm": 0.29421673650976043,
+      "learning_rate": 5.624595469255663e-06,
+      "loss": 0.2268,
+      "step": 7715
+    },
+    {
+      "epoch": 7.1946902654867255,
+      "grad_norm": 0.3161682831115032,
+      "learning_rate": 5.592233009708738e-06,
+      "loss": 0.2282,
+      "step": 7720
+    },
+    {
+      "epoch": 7.199347927340475,
+      "grad_norm": 0.31838340427441747,
+      "learning_rate": 5.559870550161812e-06,
+      "loss": 0.233,
+      "step": 7725
+    },
+    {
+      "epoch": 7.2040055891942245,
+      "grad_norm": 0.3126739004760499,
+      "learning_rate": 5.527508090614887e-06,
+      "loss": 0.2335,
+      "step": 7730
+    },
+    {
+      "epoch": 7.208663251047974,
+      "grad_norm": 0.30809587611518674,
+      "learning_rate": 5.495145631067961e-06,
+      "loss": 0.2297,
+      "step": 7735
+    },
+    {
+      "epoch": 7.2133209129017235,
+      "grad_norm": 0.3178394441862703,
+      "learning_rate": 5.462783171521036e-06,
+      "loss": 0.2327,
+      "step": 7740
+    },
+    {
+      "epoch": 7.217978574755473,
+      "grad_norm": 0.29434533557144726,
+      "learning_rate": 5.43042071197411e-06,
+      "loss": 0.2333,
+      "step": 7745
+    },
+    {
+      "epoch": 7.2226362366092225,
+      "grad_norm": 0.30090154673070385,
+      "learning_rate": 5.398058252427185e-06,
+      "loss": 0.2333,
+      "step": 7750
+    },
+    {
+      "epoch": 7.227293898462972,
+      "grad_norm": 0.3002754882767805,
+      "learning_rate": 5.365695792880259e-06,
+      "loss": 0.2286,
+      "step": 7755
+    },
+    {
+      "epoch": 7.231951560316721,
+      "grad_norm": 0.29763020403980267,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.2305,
+      "step": 7760
+    },
+    {
+      "epoch": 7.23660922217047,
+      "grad_norm": 0.3044463012655862,
+      "learning_rate": 5.300970873786408e-06,
+      "loss": 0.2323,
+      "step": 7765
+    },
+    {
+      "epoch": 7.24126688402422,
+      "grad_norm": 0.33893341317409675,
+      "learning_rate": 5.2686084142394825e-06,
+      "loss": 0.23,
+      "step": 7770
+    },
+    {
+      "epoch": 7.245924545877969,
+      "grad_norm": 0.29788870867671846,
+      "learning_rate": 5.2362459546925566e-06,
+      "loss": 0.2271,
+      "step": 7775
+    },
+    {
+      "epoch": 7.250582207731719,
+      "grad_norm": 0.3001345810140399,
+      "learning_rate": 5.2038834951456314e-06,
+      "loss": 0.2356,
+      "step": 7780
+    },
+    {
+      "epoch": 7.255239869585468,
+      "grad_norm": 0.29624174126127084,
+      "learning_rate": 5.171521035598706e-06,
+      "loss": 0.2258,
+      "step": 7785
+    },
+    {
+      "epoch": 7.259897531439218,
+      "grad_norm": 0.2998103795335706,
+      "learning_rate": 5.13915857605178e-06,
+      "loss": 0.2261,
+      "step": 7790
+    },
+    {
+      "epoch": 7.264555193292967,
+      "grad_norm": 0.30170442123639113,
+      "learning_rate": 5.106796116504855e-06,
+      "loss": 0.2335,
+      "step": 7795
+    },
+    {
+      "epoch": 7.269212855146717,
+      "grad_norm": 0.3048136215688031,
+      "learning_rate": 5.074433656957929e-06,
+      "loss": 0.2291,
+      "step": 7800
+    },
+    {
+      "epoch": 7.273870517000466,
+      "grad_norm": 0.308336424611371,
+      "learning_rate": 5.042071197411004e-06,
+      "loss": 0.2351,
+      "step": 7805
+    },
+    {
+      "epoch": 7.2785281788542155,
+      "grad_norm": 0.2981761242984121,
+      "learning_rate": 5.009708737864078e-06,
+      "loss": 0.2296,
+      "step": 7810
+    },
+    {
+      "epoch": 7.283185840707965,
+      "grad_norm": 0.3125107994925421,
+      "learning_rate": 4.977346278317152e-06,
+      "loss": 0.2327,
+      "step": 7815
+    },
+    {
+      "epoch": 7.287843502561714,
+      "grad_norm": 0.29783242600642124,
+      "learning_rate": 4.944983818770227e-06,
+      "loss": 0.2339,
+      "step": 7820
+    },
+    {
+      "epoch": 7.292501164415463,
+      "grad_norm": 0.3070161169497212,
+      "learning_rate": 4.912621359223301e-06,
+      "loss": 0.224,
+      "step": 7825
+    },
+    {
+      "epoch": 7.297158826269213,
+      "grad_norm": 0.3066786806886577,
+      "learning_rate": 4.880258899676376e-06,
+      "loss": 0.2364,
+      "step": 7830
+    },
+    {
+      "epoch": 7.301816488122962,
+      "grad_norm": 0.3138937968754756,
+      "learning_rate": 4.84789644012945e-06,
+      "loss": 0.2366,
+      "step": 7835
+    },
+    {
+      "epoch": 7.306474149976712,
+      "grad_norm": 0.29513622504520076,
+      "learning_rate": 4.815533980582525e-06,
+      "loss": 0.2268,
+      "step": 7840
+    },
+    {
+      "epoch": 7.311131811830461,
+      "grad_norm": 0.3005379349455707,
+      "learning_rate": 4.783171521035599e-06,
+      "loss": 0.2315,
+      "step": 7845
+    },
+    {
+      "epoch": 7.315789473684211,
+      "grad_norm": 0.311541396432155,
+      "learning_rate": 4.750809061488674e-06,
+      "loss": 0.2353,
+      "step": 7850
+    },
+    {
+      "epoch": 7.32044713553796,
+      "grad_norm": 0.2946986117882064,
+      "learning_rate": 4.718446601941748e-06,
+      "loss": 0.228,
+      "step": 7855
+    },
+    {
+      "epoch": 7.32510479739171,
+      "grad_norm": 0.2863751901804945,
+      "learning_rate": 4.686084142394822e-06,
+      "loss": 0.2334,
+      "step": 7860
+    },
+    {
+      "epoch": 7.329762459245459,
+      "grad_norm": 0.2863240982722415,
+      "learning_rate": 4.653721682847897e-06,
+      "loss": 0.2276,
+      "step": 7865
+    },
+    {
+      "epoch": 7.334420121099209,
+      "grad_norm": 0.3045987014232288,
+      "learning_rate": 4.621359223300971e-06,
+      "loss": 0.233,
+      "step": 7870
+    },
+    {
+      "epoch": 7.339077782952957,
+      "grad_norm": 0.30961495259781646,
+      "learning_rate": 4.5889967637540455e-06,
+      "loss": 0.2325,
+      "step": 7875
+    },
+    {
+      "epoch": 7.343735444806707,
+      "grad_norm": 0.3004648632404717,
+      "learning_rate": 4.5566343042071195e-06,
+      "loss": 0.2365,
+      "step": 7880
+    },
+    {
+      "epoch": 7.348393106660456,
+      "grad_norm": 0.29199875078520454,
+      "learning_rate": 4.524271844660194e-06,
+      "loss": 0.2266,
+      "step": 7885
+    },
+    {
+      "epoch": 7.353050768514206,
+      "grad_norm": 0.30603013293302106,
+      "learning_rate": 4.491909385113268e-06,
+      "loss": 0.2333,
+      "step": 7890
+    },
+    {
+      "epoch": 7.357708430367955,
+      "grad_norm": 0.30955009017341273,
+      "learning_rate": 4.459546925566343e-06,
+      "loss": 0.2362,
+      "step": 7895
+    },
+    {
+      "epoch": 7.362366092221705,
+      "grad_norm": 0.29175598178699413,
+      "learning_rate": 4.427184466019417e-06,
+      "loss": 0.2356,
+      "step": 7900
+    },
+    {
+      "epoch": 7.367023754075454,
+      "grad_norm": 0.2924815398673306,
+      "learning_rate": 4.394822006472492e-06,
+      "loss": 0.2359,
+      "step": 7905
+    },
+    {
+      "epoch": 7.371681415929204,
+      "grad_norm": 0.3036714855580005,
+      "learning_rate": 4.362459546925567e-06,
+      "loss": 0.2278,
+      "step": 7910
+    },
+    {
+      "epoch": 7.376339077782953,
+      "grad_norm": 0.2957981028535776,
+      "learning_rate": 4.330097087378641e-06,
+      "loss": 0.2325,
+      "step": 7915
+    },
+    {
+      "epoch": 7.380996739636703,
+      "grad_norm": 0.29468279953281146,
+      "learning_rate": 4.297734627831716e-06,
+      "loss": 0.2331,
+      "step": 7920
+    },
+    {
+      "epoch": 7.385654401490452,
+      "grad_norm": 0.3007582685414487,
+      "learning_rate": 4.26537216828479e-06,
+      "loss": 0.2381,
+      "step": 7925
+    },
+    {
+      "epoch": 7.390312063344201,
+      "grad_norm": 0.299592184905258,
+      "learning_rate": 4.233009708737865e-06,
+      "loss": 0.2381,
+      "step": 7930
+    },
+    {
+      "epoch": 7.39496972519795,
+      "grad_norm": 0.30220768863760616,
+      "learning_rate": 4.200647249190939e-06,
+      "loss": 0.2362,
+      "step": 7935
+    },
+    {
+      "epoch": 7.3996273870517,
+      "grad_norm": 0.3125457994305197,
+      "learning_rate": 4.168284789644014e-06,
+      "loss": 0.233,
+      "step": 7940
+    },
+    {
+      "epoch": 7.404285048905449,
+      "grad_norm": 0.300485101651472,
+      "learning_rate": 4.135922330097088e-06,
+      "loss": 0.2258,
+      "step": 7945
+    },
+    {
+      "epoch": 7.408942710759199,
+      "grad_norm": 0.3002736163154933,
+      "learning_rate": 4.103559870550162e-06,
+      "loss": 0.2311,
+      "step": 7950
+    },
+    {
+      "epoch": 7.413600372612948,
+      "grad_norm": 0.2954509859677457,
+      "learning_rate": 4.071197411003237e-06,
+      "loss": 0.2389,
+      "step": 7955
+    },
+    {
+      "epoch": 7.418258034466698,
+      "grad_norm": 0.2968556612469226,
+      "learning_rate": 4.038834951456311e-06,
+      "loss": 0.2345,
+      "step": 7960
+    },
+    {
+      "epoch": 7.422915696320447,
+      "grad_norm": 0.3256328206048908,
+      "learning_rate": 4.0064724919093855e-06,
+      "loss": 0.2315,
+      "step": 7965
+    },
+    {
+      "epoch": 7.427573358174197,
+      "grad_norm": 0.30621774169996147,
+      "learning_rate": 3.9741100323624595e-06,
+      "loss": 0.2331,
+      "step": 7970
+    },
+    {
+      "epoch": 7.432231020027946,
+      "grad_norm": 0.2890683583664343,
+      "learning_rate": 3.941747572815534e-06,
+      "loss": 0.2346,
+      "step": 7975
+    },
+    {
+      "epoch": 7.436888681881696,
+      "grad_norm": 0.29389431766416,
+      "learning_rate": 3.9093851132686084e-06,
+      "loss": 0.2286,
+      "step": 7980
+    },
+    {
+      "epoch": 7.441546343735445,
+      "grad_norm": 0.2972520095660748,
+      "learning_rate": 3.877022653721683e-06,
+      "loss": 0.2355,
+      "step": 7985
+    },
+    {
+      "epoch": 7.446204005589194,
+      "grad_norm": 0.29555186694090657,
+      "learning_rate": 3.844660194174757e-06,
+      "loss": 0.2358,
+      "step": 7990
+    },
+    {
+      "epoch": 7.450861667442943,
+      "grad_norm": 0.3018674902048298,
+      "learning_rate": 3.8122977346278318e-06,
+      "loss": 0.2301,
+      "step": 7995
+    },
+    {
+      "epoch": 7.455519329296693,
+      "grad_norm": 0.2893659186658692,
+      "learning_rate": 3.7799352750809062e-06,
+      "loss": 0.2282,
+      "step": 8000
+    },
+    {
+      "epoch": 7.460176991150442,
+      "grad_norm": 0.29704410993243424,
+      "learning_rate": 3.7475728155339807e-06,
+      "loss": 0.2295,
+      "step": 8005
+    },
+    {
+      "epoch": 7.464834653004192,
+      "grad_norm": 0.299885140331387,
+      "learning_rate": 3.7152103559870555e-06,
+      "loss": 0.2309,
+      "step": 8010
+    },
+    {
+      "epoch": 7.469492314857941,
+      "grad_norm": 0.28340958647352826,
+      "learning_rate": 3.6828478964401296e-06,
+      "loss": 0.2257,
+      "step": 8015
+    },
+    {
+      "epoch": 7.474149976711691,
+      "grad_norm": 0.29171765309912573,
+      "learning_rate": 3.6504854368932044e-06,
+      "loss": 0.2292,
+      "step": 8020
+    },
+    {
+      "epoch": 7.47880763856544,
+      "grad_norm": 0.2949674377093234,
+      "learning_rate": 3.6181229773462785e-06,
+      "loss": 0.2402,
+      "step": 8025
+    },
+    {
+      "epoch": 7.48346530041919,
+      "grad_norm": 0.2959510821426171,
+      "learning_rate": 3.5857605177993533e-06,
+      "loss": 0.232,
+      "step": 8030
+    },
+    {
+      "epoch": 7.488122962272939,
+      "grad_norm": 0.30774189794162643,
+      "learning_rate": 3.5533980582524273e-06,
+      "loss": 0.2346,
+      "step": 8035
+    },
+    {
+      "epoch": 7.492780624126689,
+      "grad_norm": 0.2898875205533011,
+      "learning_rate": 3.5210355987055014e-06,
+      "loss": 0.2271,
+      "step": 8040
+    },
+    {
+      "epoch": 7.497438285980438,
+      "grad_norm": 0.2937295756525654,
+      "learning_rate": 3.4886731391585762e-06,
+      "loss": 0.2368,
+      "step": 8045
+    },
+    {
+      "epoch": 7.502095947834187,
+      "grad_norm": 0.29222859784505506,
+      "learning_rate": 3.4563106796116503e-06,
+      "loss": 0.2245,
+      "step": 8050
+    },
+    {
+      "epoch": 7.506753609687936,
+      "grad_norm": 0.3088248045967586,
+      "learning_rate": 3.423948220064725e-06,
+      "loss": 0.2356,
+      "step": 8055
+    },
+    {
+      "epoch": 7.511411271541686,
+      "grad_norm": 0.29795075855304565,
+      "learning_rate": 3.391585760517799e-06,
+      "loss": 0.234,
+      "step": 8060
+    },
+    {
+      "epoch": 7.516068933395435,
+      "grad_norm": 0.29499070200838384,
+      "learning_rate": 3.359223300970874e-06,
+      "loss": 0.2337,
+      "step": 8065
+    },
+    {
+      "epoch": 7.520726595249185,
+      "grad_norm": 0.29851001641469693,
+      "learning_rate": 3.3268608414239485e-06,
+      "loss": 0.2318,
+      "step": 8070
+    },
+    {
+      "epoch": 7.525384257102934,
+      "grad_norm": 0.31823647472711214,
+      "learning_rate": 3.294498381877023e-06,
+      "loss": 0.2379,
+      "step": 8075
+    },
+    {
+      "epoch": 7.530041918956684,
+      "grad_norm": 0.2949888501288617,
+      "learning_rate": 3.2621359223300974e-06,
+      "loss": 0.2382,
+      "step": 8080
+    },
+    {
+      "epoch": 7.534699580810433,
+      "grad_norm": 0.3079856554407537,
+      "learning_rate": 3.2297734627831714e-06,
+      "loss": 0.2328,
+      "step": 8085
+    },
+    {
+      "epoch": 7.539357242664183,
+      "grad_norm": 0.30063866398956485,
+      "learning_rate": 3.1974110032362463e-06,
+      "loss": 0.2296,
+      "step": 8090
+    },
+    {
+      "epoch": 7.544014904517932,
+      "grad_norm": 0.3079909250110497,
+      "learning_rate": 3.1650485436893203e-06,
+      "loss": 0.23,
+      "step": 8095
+    },
+    {
+      "epoch": 7.548672566371682,
+      "grad_norm": 0.2958436428090986,
+      "learning_rate": 3.132686084142395e-06,
+      "loss": 0.2333,
+      "step": 8100
+    },
+    {
+      "epoch": 7.55333022822543,
+      "grad_norm": 0.3053616138559035,
+      "learning_rate": 3.1003236245954696e-06,
+      "loss": 0.2317,
+      "step": 8105
+    },
+    {
+      "epoch": 7.55798789007918,
+      "grad_norm": 0.30532557136329946,
+      "learning_rate": 3.0679611650485436e-06,
+      "loss": 0.2326,
+      "step": 8110
+    },
+    {
+      "epoch": 7.562645551932929,
+      "grad_norm": 0.2979271334069836,
+      "learning_rate": 3.035598705501618e-06,
+      "loss": 0.2307,
+      "step": 8115
+    },
+    {
+      "epoch": 7.567303213786679,
+      "grad_norm": 0.2960491281556295,
+      "learning_rate": 3.0032362459546925e-06,
+      "loss": 0.2357,
+      "step": 8120
+    },
+    {
+      "epoch": 7.571960875640428,
+      "grad_norm": 0.28863064035013297,
+      "learning_rate": 2.970873786407767e-06,
+      "loss": 0.2252,
+      "step": 8125
+    },
+    {
+      "epoch": 7.576618537494178,
+      "grad_norm": 0.29019940510217623,
+      "learning_rate": 2.9385113268608414e-06,
+      "loss": 0.2306,
+      "step": 8130
+    },
+    {
+      "epoch": 7.581276199347927,
+      "grad_norm": 0.28983014326869905,
+      "learning_rate": 2.906148867313916e-06,
+      "loss": 0.232,
+      "step": 8135
+    },
+    {
+      "epoch": 7.585933861201677,
+      "grad_norm": 0.3016603416714441,
+      "learning_rate": 2.8737864077669903e-06,
+      "loss": 0.2397,
+      "step": 8140
+    },
+    {
+      "epoch": 7.590591523055426,
+      "grad_norm": 0.29625751997817373,
+      "learning_rate": 2.841423948220065e-06,
+      "loss": 0.2316,
+      "step": 8145
+    },
+    {
+      "epoch": 7.595249184909176,
+      "grad_norm": 0.29904315648276253,
+      "learning_rate": 2.8090614886731396e-06,
+      "loss": 0.2351,
+      "step": 8150
+    },
+    {
+      "epoch": 7.599906846762925,
+      "grad_norm": 0.29640256634448936,
+      "learning_rate": 2.7766990291262136e-06,
+      "loss": 0.2255,
+      "step": 8155
+    },
+    {
+      "epoch": 7.604564508616674,
+      "grad_norm": 0.3005793461785394,
+      "learning_rate": 2.744336569579288e-06,
+      "loss": 0.2341,
+      "step": 8160
+    },
+    {
+      "epoch": 7.6092221704704235,
+      "grad_norm": 0.28604597838088214,
+      "learning_rate": 2.7119741100323625e-06,
+      "loss": 0.2331,
+      "step": 8165
+    },
+    {
+      "epoch": 7.613879832324173,
+      "grad_norm": 0.30494875057396265,
+      "learning_rate": 2.679611650485437e-06,
+      "loss": 0.2302,
+      "step": 8170
+    },
+    {
+      "epoch": 7.6185374941779225,
+      "grad_norm": 0.29665569172774714,
+      "learning_rate": 2.6472491909385114e-06,
+      "loss": 0.2268,
+      "step": 8175
+    },
+    {
+      "epoch": 7.623195156031672,
+      "grad_norm": 0.2903936085950602,
+      "learning_rate": 2.614886731391586e-06,
+      "loss": 0.2305,
+      "step": 8180
+    },
+    {
+      "epoch": 7.6278528178854215,
+      "grad_norm": 0.2840449153885873,
+      "learning_rate": 2.5825242718446603e-06,
+      "loss": 0.2279,
+      "step": 8185
+    },
+    {
+      "epoch": 7.632510479739171,
+      "grad_norm": 0.29311363279163966,
+      "learning_rate": 2.5501618122977347e-06,
+      "loss": 0.232,
+      "step": 8190
+    },
+    {
+      "epoch": 7.6371681415929205,
+      "grad_norm": 0.3024133237930259,
+      "learning_rate": 2.517799352750809e-06,
+      "loss": 0.234,
+      "step": 8195
+    },
+    {
+      "epoch": 7.64182580344667,
+      "grad_norm": 0.29842457764289687,
+      "learning_rate": 2.4854368932038836e-06,
+      "loss": 0.2283,
+      "step": 8200
+    },
+    {
+      "epoch": 7.6464834653004194,
+      "grad_norm": 0.29404576936950766,
+      "learning_rate": 2.453074433656958e-06,
+      "loss": 0.2308,
+      "step": 8205
+    },
+    {
+      "epoch": 7.651141127154169,
+      "grad_norm": 0.30149197235212855,
+      "learning_rate": 2.4207119741100325e-06,
+      "loss": 0.2364,
+      "step": 8210
+    },
+    {
+      "epoch": 7.6557987890079175,
+      "grad_norm": 0.30022001517778296,
+      "learning_rate": 2.388349514563107e-06,
+      "loss": 0.2296,
+      "step": 8215
+    },
+    {
+      "epoch": 7.660456450861668,
+      "grad_norm": 0.2886569031218321,
+      "learning_rate": 2.3559870550161814e-06,
+      "loss": 0.2293,
+      "step": 8220
+    },
+    {
+      "epoch": 7.6651141127154165,
+      "grad_norm": 0.29162327843663,
+      "learning_rate": 2.323624595469256e-06,
+      "loss": 0.2314,
+      "step": 8225
+    },
+    {
+      "epoch": 7.669771774569166,
+      "grad_norm": 0.304313023599954,
+      "learning_rate": 2.2912621359223303e-06,
+      "loss": 0.2324,
+      "step": 8230
+    },
+    {
+      "epoch": 7.6744294364229155,
+      "grad_norm": 0.3155740911282473,
+      "learning_rate": 2.2588996763754048e-06,
+      "loss": 0.2368,
+      "step": 8235
+    },
+    {
+      "epoch": 7.679087098276665,
+      "grad_norm": 0.30076835045192435,
+      "learning_rate": 2.226537216828479e-06,
+      "loss": 0.2337,
+      "step": 8240
+    },
+    {
+      "epoch": 7.6837447601304145,
+      "grad_norm": 0.2967465249604805,
+      "learning_rate": 2.1941747572815537e-06,
+      "loss": 0.2291,
+      "step": 8245
+    },
+    {
+      "epoch": 7.688402421984164,
+      "grad_norm": 0.3040736391467545,
+      "learning_rate": 2.1618122977346277e-06,
+      "loss": 0.2303,
+      "step": 8250
+    },
+    {
+      "epoch": 7.6930600838379135,
+      "grad_norm": 0.30152847877289063,
+      "learning_rate": 2.129449838187702e-06,
+      "loss": 0.2305,
+      "step": 8255
+    },
+    {
+      "epoch": 7.697717745691663,
+      "grad_norm": 0.28538787401129073,
+      "learning_rate": 2.0970873786407766e-06,
+      "loss": 0.2291,
+      "step": 8260
+    },
+    {
+      "epoch": 7.7023754075454125,
+      "grad_norm": 0.2998642191190617,
+      "learning_rate": 2.064724919093851e-06,
+      "loss": 0.2383,
+      "step": 8265
+    },
+    {
+      "epoch": 7.707033069399162,
+      "grad_norm": 0.2917780958653254,
+      "learning_rate": 2.0323624595469255e-06,
+      "loss": 0.2332,
+      "step": 8270
+    },
+    {
+      "epoch": 7.7116907312529115,
+      "grad_norm": 0.298805259047907,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.2303,
+      "step": 8275
+    },
+    {
+      "epoch": 7.71634839310666,
+      "grad_norm": 0.2890465234614115,
+      "learning_rate": 1.9676375404530748e-06,
+      "loss": 0.2275,
+      "step": 8280
+    },
+    {
+      "epoch": 7.72100605496041,
+      "grad_norm": 0.29389457909621547,
+      "learning_rate": 1.9352750809061492e-06,
+      "loss": 0.2284,
+      "step": 8285
+    },
+    {
+      "epoch": 7.725663716814159,
+      "grad_norm": 0.2959644171394708,
+      "learning_rate": 1.9029126213592235e-06,
+      "loss": 0.2345,
+      "step": 8290
+    },
+    {
+      "epoch": 7.730321378667909,
+      "grad_norm": 0.29421810879944266,
+      "learning_rate": 1.8705501618122977e-06,
+      "loss": 0.2302,
+      "step": 8295
+    },
+    {
+      "epoch": 7.734979040521658,
+      "grad_norm": 0.2915200400387438,
+      "learning_rate": 1.8381877022653721e-06,
+      "loss": 0.2274,
+      "step": 8300
+    },
+    {
+      "epoch": 7.739636702375408,
+      "grad_norm": 0.30041840963474026,
+      "learning_rate": 1.8058252427184466e-06,
+      "loss": 0.2298,
+      "step": 8305
+    },
+    {
+      "epoch": 7.744294364229157,
+      "grad_norm": 0.2811313882467788,
+      "learning_rate": 1.773462783171521e-06,
+      "loss": 0.2266,
+      "step": 8310
+    },
+    {
+      "epoch": 7.748952026082907,
+      "grad_norm": 0.28925553908674034,
+      "learning_rate": 1.7411003236245955e-06,
+      "loss": 0.2319,
+      "step": 8315
+    },
+    {
+      "epoch": 7.753609687936656,
+      "grad_norm": 0.2906004808859835,
+      "learning_rate": 1.70873786407767e-06,
+      "loss": 0.2332,
+      "step": 8320
+    },
+    {
+      "epoch": 7.7582673497904056,
+      "grad_norm": 0.28951439849956767,
+      "learning_rate": 1.6763754045307446e-06,
+      "loss": 0.2324,
+      "step": 8325
+    },
+    {
+      "epoch": 7.762925011644155,
+      "grad_norm": 0.29773446769676193,
+      "learning_rate": 1.644012944983819e-06,
+      "loss": 0.2324,
+      "step": 8330
+    },
+    {
+      "epoch": 7.767582673497904,
+      "grad_norm": 0.3004545400619002,
+      "learning_rate": 1.6116504854368935e-06,
+      "loss": 0.2331,
+      "step": 8335
+    },
+    {
+      "epoch": 7.772240335351653,
+      "grad_norm": 0.2852230726291482,
+      "learning_rate": 1.5792880258899675e-06,
+      "loss": 0.2285,
+      "step": 8340
+    },
+    {
+      "epoch": 7.776897997205403,
+      "grad_norm": 0.3054806079257286,
+      "learning_rate": 1.5469255663430422e-06,
+      "loss": 0.2333,
+      "step": 8345
+    },
+    {
+      "epoch": 7.781555659059152,
+      "grad_norm": 0.2821067570432199,
+      "learning_rate": 1.5145631067961166e-06,
+      "loss": 0.231,
+      "step": 8350
+    },
+    {
+      "epoch": 7.786213320912902,
+      "grad_norm": 0.2734352593394025,
+      "learning_rate": 1.482200647249191e-06,
+      "loss": 0.2284,
+      "step": 8355
+    },
+    {
+      "epoch": 7.790870982766651,
+      "grad_norm": 0.2744502239977389,
+      "learning_rate": 1.4498381877022655e-06,
+      "loss": 0.2303,
+      "step": 8360
+    },
+    {
+      "epoch": 7.795528644620401,
+      "grad_norm": 0.2906796981483134,
+      "learning_rate": 1.41747572815534e-06,
+      "loss": 0.2329,
+      "step": 8365
+    },
+    {
+      "epoch": 7.80018630647415,
+      "grad_norm": 0.285778377684361,
+      "learning_rate": 1.3851132686084144e-06,
+      "loss": 0.2303,
+      "step": 8370
+    },
+    {
+      "epoch": 7.8048439683279,
+      "grad_norm": 0.2807787673889005,
+      "learning_rate": 1.3527508090614886e-06,
+      "loss": 0.2349,
+      "step": 8375
+    },
+    {
+      "epoch": 7.809501630181649,
+      "grad_norm": 0.30122596269049373,
+      "learning_rate": 1.320388349514563e-06,
+      "loss": 0.2365,
+      "step": 8380
+    },
+    {
+      "epoch": 7.814159292035399,
+      "grad_norm": 0.29920390405740077,
+      "learning_rate": 1.2880258899676375e-06,
+      "loss": 0.2324,
+      "step": 8385
+    },
+    {
+      "epoch": 7.818816953889147,
+      "grad_norm": 0.29268110303894623,
+      "learning_rate": 1.2556634304207122e-06,
+      "loss": 0.2306,
+      "step": 8390
+    },
+    {
+      "epoch": 7.823474615742897,
+      "grad_norm": 0.2891680650027946,
+      "learning_rate": 1.2233009708737866e-06,
+      "loss": 0.2308,
+      "step": 8395
+    },
+    {
+      "epoch": 7.828132277596646,
+      "grad_norm": 0.2965934137032846,
+      "learning_rate": 1.1909385113268608e-06,
+      "loss": 0.2376,
+      "step": 8400
+    },
+    {
+      "epoch": 7.832789939450396,
+      "grad_norm": 0.2932565825711091,
+      "learning_rate": 1.1585760517799353e-06,
+      "loss": 0.2337,
+      "step": 8405
+    },
+    {
+      "epoch": 7.837447601304145,
+      "grad_norm": 0.29773472915142823,
+      "learning_rate": 1.1262135922330097e-06,
+      "loss": 0.2293,
+      "step": 8410
+    },
+    {
+      "epoch": 7.842105263157895,
+      "grad_norm": 0.2934365641617045,
+      "learning_rate": 1.0938511326860842e-06,
+      "loss": 0.2339,
+      "step": 8415
+    },
+    {
+      "epoch": 7.846762925011644,
+      "grad_norm": 0.2960715244509101,
+      "learning_rate": 1.0614886731391586e-06,
+      "loss": 0.2331,
+      "step": 8420
+    },
+    {
+      "epoch": 7.851420586865394,
+      "grad_norm": 0.2954502286031332,
+      "learning_rate": 1.029126213592233e-06,
+      "loss": 0.2254,
+      "step": 8425
+    },
+    {
+      "epoch": 7.856078248719143,
+      "grad_norm": 0.3018314394803379,
+      "learning_rate": 9.967637540453075e-07,
+      "loss": 0.2329,
+      "step": 8430
+    },
+    {
+      "epoch": 7.860735910572893,
+      "grad_norm": 0.2880968499417839,
+      "learning_rate": 9.64401294498382e-07,
+      "loss": 0.2357,
+      "step": 8435
+    },
+    {
+      "epoch": 7.865393572426642,
+      "grad_norm": 0.294434898554631,
+      "learning_rate": 9.320388349514564e-07,
+      "loss": 0.2312,
+      "step": 8440
+    },
+    {
+      "epoch": 7.870051234280391,
+      "grad_norm": 0.2932235274867024,
+      "learning_rate": 8.996763754045308e-07,
+      "loss": 0.2342,
+      "step": 8445
+    },
+    {
+      "epoch": 7.874708896134141,
+      "grad_norm": 0.2884539693871812,
+      "learning_rate": 8.673139158576052e-07,
+      "loss": 0.2335,
+      "step": 8450
+    },
+    {
+      "epoch": 7.87936655798789,
+      "grad_norm": 0.2936378535586453,
+      "learning_rate": 8.349514563106797e-07,
+      "loss": 0.2341,
+      "step": 8455
+    },
+    {
+      "epoch": 7.884024219841639,
+      "grad_norm": 0.289241138931296,
+      "learning_rate": 8.025889967637541e-07,
+      "loss": 0.2259,
+      "step": 8460
+    },
+    {
+      "epoch": 7.888681881695389,
+      "grad_norm": 0.29803961484355557,
+      "learning_rate": 7.702265372168285e-07,
+      "loss": 0.2371,
+      "step": 8465
+    },
+    {
+      "epoch": 7.893339543549138,
+      "grad_norm": 0.28305757547326993,
+      "learning_rate": 7.37864077669903e-07,
+      "loss": 0.2333,
+      "step": 8470
+    },
+    {
+      "epoch": 7.897997205402888,
+      "grad_norm": 0.2891859316380593,
+      "learning_rate": 7.055016181229773e-07,
+      "loss": 0.2311,
+      "step": 8475
+    },
+    {
+      "epoch": 7.902654867256637,
+      "grad_norm": 0.29477394307231747,
+      "learning_rate": 6.731391585760519e-07,
+      "loss": 0.235,
+      "step": 8480
+    },
+    {
+      "epoch": 7.907312529110387,
+      "grad_norm": 0.28070352814466104,
+      "learning_rate": 6.407766990291262e-07,
+      "loss": 0.2242,
+      "step": 8485
+    },
+    {
+      "epoch": 7.911970190964136,
+      "grad_norm": 0.2929726293630473,
+      "learning_rate": 6.084142394822007e-07,
+      "loss": 0.2367,
+      "step": 8490
+    },
+    {
+      "epoch": 7.916627852817886,
+      "grad_norm": 0.30314735503554396,
+      "learning_rate": 5.760517799352751e-07,
+      "loss": 0.2249,
+      "step": 8495
+    },
+    {
+      "epoch": 7.921285514671635,
+      "grad_norm": 0.29427314766744256,
+      "learning_rate": 5.436893203883496e-07,
+      "loss": 0.2307,
+      "step": 8500
+    },
+    {
+      "epoch": 7.925943176525385,
+      "grad_norm": 0.2912139356277789,
+      "learning_rate": 5.11326860841424e-07,
+      "loss": 0.2319,
+      "step": 8505
+    },
+    {
+      "epoch": 7.930600838379133,
+      "grad_norm": 0.293854156431429,
+      "learning_rate": 4.789644012944983e-07,
+      "loss": 0.2365,
+      "step": 8510
+    },
+    {
+      "epoch": 7.935258500232883,
+      "grad_norm": 0.2918833196305414,
+      "learning_rate": 4.4660194174757285e-07,
+      "loss": 0.2346,
+      "step": 8515
+    },
+    {
+      "epoch": 7.939916162086632,
+      "grad_norm": 0.2923200670274681,
+      "learning_rate": 4.1423948220064724e-07,
+      "loss": 0.2291,
+      "step": 8520
+    },
+    {
+      "epoch": 7.944573823940382,
+      "grad_norm": 0.2924664374700889,
+      "learning_rate": 3.818770226537217e-07,
+      "loss": 0.235,
+      "step": 8525
+    },
+    {
+      "epoch": 7.949231485794131,
+      "grad_norm": 0.30212915417954544,
+      "learning_rate": 3.4951456310679613e-07,
+      "loss": 0.2322,
+      "step": 8530
+    },
+    {
+      "epoch": 7.953889147647881,
+      "grad_norm": 0.30164147260075297,
+      "learning_rate": 3.171521035598706e-07,
+      "loss": 0.2266,
+      "step": 8535
+    },
+    {
+      "epoch": 7.95854680950163,
+      "grad_norm": 0.2939683525337563,
+      "learning_rate": 2.8478964401294497e-07,
+      "loss": 0.2305,
+      "step": 8540
+    },
+    {
+      "epoch": 7.96320447135538,
+      "grad_norm": 0.2755637464638358,
+      "learning_rate": 2.524271844660194e-07,
+      "loss": 0.2358,
+      "step": 8545
+    },
+    {
+      "epoch": 7.967862133209129,
+      "grad_norm": 0.2901318305536487,
+      "learning_rate": 2.2006472491909384e-07,
+      "loss": 0.2309,
+      "step": 8550
+    },
+    {
+      "epoch": 7.972519795062879,
+      "grad_norm": 0.2944501952740433,
+      "learning_rate": 1.8770226537216828e-07,
+      "loss": 0.2308,
+      "step": 8555
+    },
+    {
+      "epoch": 7.977177456916628,
+      "grad_norm": 0.2849006022937681,
+      "learning_rate": 1.5533980582524273e-07,
+      "loss": 0.2329,
+      "step": 8560
+    },
+    {
+      "epoch": 7.981835118770377,
+      "grad_norm": 0.289904412851554,
+      "learning_rate": 1.2297734627831717e-07,
+      "loss": 0.2334,
+      "step": 8565
+    },
+    {
+      "epoch": 7.986492780624126,
+      "grad_norm": 0.2926599702016367,
+      "learning_rate": 9.061488673139159e-08,
+      "loss": 0.2296,
+      "step": 8570
+    },
+    {
+      "epoch": 7.991150442477876,
+      "grad_norm": 0.30510864446562197,
+      "learning_rate": 5.8252427184466026e-08,
+      "loss": 0.234,
+      "step": 8575
+    },
+    {
+      "epoch": 7.995808104331625,
+      "grad_norm": 0.30231794718044347,
+      "learning_rate": 2.5889967637540452e-08,
+      "loss": 0.2294,
+      "step": 8580
+    },
+    {
+      "epoch": 7.999534233814625,
+      "step": 8584,
+      "total_flos": 6105625239486464.0,
+      "train_loss": 0.028776329486367866,
+      "train_runtime": 7678.0521,
+      "train_samples_per_second": 17.891,
+      "train_steps_per_second": 1.118
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 8584,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 8,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6105625239486464.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}